Futures Trading and the Excess Co-movement of Commodity Prices

Futures Trading and the Excess Co-movement of Commodity Prices Abstract We empirically reinvestigate the issue of the excess co-movement of commodity prices initially raised in Pindyck and Rotemberg (1990). Excess co-movement appears when commodity prices remain correlated even after adjusting for the impact of fundamentals. We use recent developments in large approximate factor models to consider a richer information set and adequately model these fundamentals. We consider a set of eight unrelated commodities along with 184 real and nominal macroeconomic variables, from developed and emerging economies, from which nine factors are extracted over the 1993–2013 period. Our estimates provide evidence of time-varying excess co-movement which is particularly high after 2007. We further show that speculative intensity is a driver of the estimated excess co-movement, as speculative trading is both correlated across the commodity futures markets and correlated with the futures prices. Our results can be taken as direct evidence of the significant impact of financialization on commodity-price cross-moments. 1. Introduction Commodity markets have undergone major changes over the past two decades. The popularity of commodity-related financial instruments, such as commodity indices, has led many observers to conclude that commodity markets are now more intimately connected to financial markets, and so may also co-move more significantly (Tang and Xiong, 2012; Cheng and Xiong, 2014a; Hamilton and Wu, 2015; Basak and Pavlova, 2016). While a greater number of participants in commodity markets may bring about improved risk sharing, the financialization process has been widely criticized as a potential source of excessive price volatility (Stoll and Whaley, 2010). This paper investigates whether the excess co-movement of commodity prices is related to the growing financial influence in commodity markets. The excess co-movement of commodity prices deserves analysis for at least two reasons. First, residual correlation (or “co-movement”) may mean that “[…] commodity demands and supplies are affected by unobserved forecasts of the economic variable.” [Pindyck and Rotemberg (PR hereafter), 1990, p. 1174], thereby indicating that the standard demand–supply model may not be able to explain commodity returns adequately. This conclusion, which is at odds with standard economic theory, suggests that further research is needed to uncover the new relevant fundamentals, or change the way in which these fundamentals are measured. Second, from a portfolio-management perspective, the presence of co-movement limits the diversification of investors who manage a portfolio containing a number of commodity futures.1 PR define excess co-movement as commodity prices remaining correlated even after adjusting for the impact of common macroeconomic variables. They select six variables: the US index of industrial production, the consumer price index, the effective $US exchange rate,2 the three-month Treasury bill interest rate (cf. Frankel and Rose, 2010), the M1 monetary measure (cf. Frankel, 2006), and the S&P 500 stock index, which are supposed to represent the fundamentals. Nevertheless, the authors recognize that: “[…] a major limitation of our approach is that we can never be sure we have included all relevant macroeconomic variables and latent variables.” (p. 1185).3 One major issue in filtering the returns from common factors is indeed the selection of the variables to be considered. To deal with the issue of omitted variables, we suggest relying on a large approximate factor model, along the lines of Stock and Watson (2002a, 2002b), which allows us to enlarge the information set significantly while preserving a sufficiently low dimension for the econometric estimation.4 We thus avoid the arbitrariness and computational difficulties of selecting relevant variables, in particular when the number of possible combinations is large. Borensztein and Reinhart (1994) underline the need to consider well-defined supply and demand variables in order to explain commodity prices. In particular, the authors advocate the inclusion of variables for Eastern Europe that are likely to be relevant for their sample period of 1970–92. In the same spirit, we consider a set of economic variables from developed and emerging countries (China, India, and Brazil, among others) that should allow us to filter out commodity returns more accurately, as these countries have played a central role in shaping commodity prices over recent years. While commodity prices are the product of transactions in one particular part of the world, they also reflect a great deal of information which has been generated throughout the world. For instance, the price of crude oil, say US West Texas Intermediate (WTI), is widely accepted as a world price, while being mainly traded in the USA (see Kilian, 2009). Following the idea of Ludvigson and Ng (2009) of grouping explanatory variables into meaningful categories, we uncover the sets of variables that best explain commodity returns. Our estimates show that monthly commodity returns over the last two decades are mainly correlated with the real aggregate variables in emerging countries, highlighting the important role played by these countries in shaping commodity prices over this period. The paper provides evidence of time-varying excess co-movement, which is particularly high after 2007.5 As such we extend PR’s analysis in two directions. First, we investigate the time-varying behavior of the phenomenon, thereby providing further insights into the analysis of excess co-movement. Second, we look at a recent period that includes both a pronounced increase in commodity prices around 2008 and the recent financial crisis. Last, we take heteroscedasticity into account as this can play a critical role in measures of correlation.6 Highlighting these stylized facts regarding excess co-movements in commodity prices in recent years is our first contribution. The main novelty in our paper, which constitutes our second contribution, is that we establish an empirical relationship between the notion of excess co-movement and speculative activity in commodity futures markets. Surprisingly, academic research has not yet investigated the potential determinants of excess co-movement in commodity prices. We suggest an explanation for this phenomenon following the intuition developed in Barberis and Shleifer (2003) that “investors categorize risky assets into different styles and move funds among these styles depending on their relative performance.” (p. 161). As such, if most commodities are classified into a “commodity style”, seemingly unrelated commodities are likely to co-move more than would be expected based on fundamental analysis.7 This is precisely what we demonstrate in our present work. Our results are also in line with the recent work by Basak and Pavlova (2016), who go beyond the behavioral approach in Barberis and Shleifer (2003) and develop a multi-asset, multi-good general equilibrium model with heterogeneous investors, some of whom are institutional investors, considering characteristics that are specific to commodities such as the presence of inventories. The model, in the tradition of Lucas-tree models, is solved in closed-form and provides a rich set of implications, among which an increase in the correlation between commodities following institutional positioning, and more so for commodities that are included in an index. Our results provide strong support for the outcome in Basak and Pavlova (2016), and may then be seen as an empirical validation of their model. Our empirical work makes use of data from the US Commodity Futures Trading Commission (CFTC) to estimate speculative intensity. While the categories in the publicly available data from the CFTC do not distinguish perfectly between the various categories of traders, as discussed previously in Bessembinder (1992) and Stoll and Whaley (2010) among many others, we here show that they are informative for the explanation of excess co-movement. Our measure of speculative activity in futures markets follows the recent work by Han (2008) on sentiments in financial markets but is reminiscent of the so-called Working’s T measure. Our empirical strategy provides direct evidence of the explanatory power of speculative intensity for excess co-movement: while the large number of fundamental variables have limited success in explaining the co-movement between commodities, we show that speculative activity is correlated across commodity futures markets and, at the same time, that speculative activity is correlated with futures prices. This last result is obtained from an instrumental-variable analysis to avoid endogeneity issues between returns and positions in futures markets. The empirical work closest to ours is Tang and Xiong (2012), which also considers the financialization of commodities as a potential source of the recent increase in co-movements between commodity returns. Their “analysis focuses on connecting the large inflow of commodity index investment to the large increase of commodity price co-movements in recent years by examining the difference in these co-movements between indexed and off-index commodities” (p. 55).8 The authors regress the S&P-GSCI on a measure of the net position change of different categories of traders and, as such, do not pick up the common factors that may affect the behavior of most, if not all, commodity prices. Tang and Xiong (2012) also investigate the relationship between economic activity in emerging countries and the co-movement of commodity prices using a novel time series of Chinese futures prices available since the late 1990s. While commodity prices are usually thought of as being global, the authors show that the picture is actually more complex. Interestingly, while US commodity prices exhibit a pronounced cycle, this is not the case for the Chinese prices of similar commodities, thereby raising “doubt about commodity demands from China as the driver of all commodity prices in the U.S.” (p. 63). Our regressions for commodity returns show that the demand from emerging economies does play a role in determining the prices of US non-agricultural commodity futures prices, while leaving a considerable role for other factors. Overall, while dealing with a research question similar to Tang and Xiong (2012), we adopt a very different empirical approach. In particular, we specifically consider fundamentals that are critical in the analysis of co-movements. The plan for the rest of the paper is as follows. In the next section, we present the data used for the empirical analysis. In Section 3, we very briefly review the factor-model methodology and calculate the factors used to filter the commodity returns. The excess co-movement is then estimated in Section 4, while Section 5 is dedicated to the analysis of the relationship between excess co-movement and speculation. Finally, Section 6 concludes by discussing some limits to and possible future extensions of our work. 2. Data We consider a set of eight commodity prices: wheat, copper, silver, soybeans, raw sugar, cotton, crude oil, and live cattle. These are representative of the main commodity classes and are assumed to be unrelated as defined in PR, in the sense that their supply or demand cross-elasticities are almost zero. All prices are cash prices except for crude oil, where the front-month contract price is taken as a proxy for the cash price, to avoid the distorting impact of delivery issues for this particular commodity. All prices are in nominal US$. Due to data limitations, in particular for macroeconomic variables from emerging countries, we consider monthly observations from February 1993 to November 2013. Data are from Datastream. The prices are displayed in Figure 1. They fluctuate around their mean level until 2005, except for oil and silver which have a rising trend. A first large price rise begins in 2005 and ends in 2008. Prices fell in 2009 but rise steeply in 2010; they stabilize or fell in 2012. Returns are log difference of prices.9 The descriptive statistics in Table I reveals evidence of skewness—negative in six cases out of eight—and excess kurtosis. The Jarque–Bera test consequently rejects the hypothesis of a Gaussian distribution for all returns. The presence of heteroscedasticity, which is a standard feature in financial price series, may lie behind this non-normality. Table I. Descriptive statistics for the eight commodities monthly returns—February 1993–November 2013 (i) Monthly returns are computed as price log differences. (ii) Commodity prices are cash prices except crude oil where the current month contract price is taken as a proxy for the cash price. (iii) ***, **, and *, respectively, denote rejection of the null hypothesis of a Gaussian distribution at 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Mean  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022  Max  0.3666  0.3266  0.2309  0.2013  0.2052  0.3855  0.2536  0.1178  Min  −0.2499  −0.3360  −0.3285  −0.4660  −0.3620  −0.2605  −0.3899  −0.2369  Std  0.0845  0.0796  0.0873  0.0835  0.0806  0.0903  0.1016  0.0460  Skewness  0.1660  −0.4043  −0.4079  −1.0658  −0.3611  0.2763  −0.5827  −0.5674  Kurtosis  4.3312  5.9263  4.0502  7.0722  5.1793  4.5636  3.9713  5.2846  Jarque-Bera  19.60***  96.01***  18.42***  220.06***  54.90***  28.64***  23.9759***  67.78***  Number of observations  250  250  250  250  250  250  250  250    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Mean  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022  Max  0.3666  0.3266  0.2309  0.2013  0.2052  0.3855  0.2536  0.1178  Min  −0.2499  −0.3360  −0.3285  −0.4660  −0.3620  −0.2605  −0.3899  −0.2369  Std  0.0845  0.0796  0.0873  0.0835  0.0806  0.0903  0.1016  0.0460  Skewness  0.1660  −0.4043  −0.4079  −1.0658  −0.3611  0.2763  −0.5827  −0.5674  Kurtosis  4.3312  5.9263  4.0502  7.0722  5.1793  4.5636  3.9713  5.2846  Jarque-Bera  19.60***  96.01***  18.42***  220.06***  54.90***  28.64***  23.9759***  67.78***  Number of observations  250  250  250  250  250  250  250  250  Figure 1. View largeDownload slide The eight commodity prices—January 1993–November 2013. Note: Prices are normalized to 100 at January 1993. Figure 1. View largeDownload slide The eight commodity prices—January 1993–November 2013. Note: Prices are normalized to 100 at January 1993. Table II presents the sample correlations between returns and their associated p-values. There are, respectively, 16, 15, and 11 significant correlations at the 10%, 5%, and 1% critical levels. All of the significant correlations are positive, ranging from 0.4789 (wheat and raw sugar) to 0.11 (raw sugar and crude oil), with an average figure of 0.2536. The figures in PR are a maximum of 0.322 and a minimum of 0.113, with an average of 0.161 for the significant correlations between 1960 and 1985. Therefore, it seems that the relationship between commodity prices has tightened over the last two decades. Table II. Correlation between the eight commodities monthly—February 1993–November 2013 The upper triangular matrix reports correlations while the lower reports their p-values. ***, **, and *, respectively, denote significance at 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.2988***  0.2256***  0.4789***  −0.0023  0.2763***  0.1569**  0.0352  Copper  0.0000  1  0.3799***  0.2345***  0.1258**  0.2352***  0.3496***  0.0077  Silver  0.0003  0.0000  1  0.2142***  0.1901**  0.0844  0.2141***  −0.0306  Soybeans  0.0000  0.0002  0.0006  1  −0.0767  0.3877***  0.1095*  −0.0898  Raw sugar  0.9714  0.0468  0.0025  0.2266  1  0.0543  0.1012  0.0530  Cotton  0.0000  0.0002  0.1834  0.0000  0.3930  1  0.1808**  −0.0288  Crude oil  0.0130  0.0000  0.0007  0.0839  0.1103  0.0041  1  0.0455  Live cattle  0.5791  0.9040  0.6302  0.1571  0.4045  0.6506  0.4739  1    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.2988***  0.2256***  0.4789***  −0.0023  0.2763***  0.1569**  0.0352  Copper  0.0000  1  0.3799***  0.2345***  0.1258**  0.2352***  0.3496***  0.0077  Silver  0.0003  0.0000  1  0.2142***  0.1901**  0.0844  0.2141***  −0.0306  Soybeans  0.0000  0.0002  0.0006  1  −0.0767  0.3877***  0.1095*  −0.0898  Raw sugar  0.9714  0.0468  0.0025  0.2266  1  0.0543  0.1012  0.0530  Cotton  0.0000  0.0002  0.1834  0.0000  0.3930  1  0.1808**  −0.0288  Crude oil  0.0130  0.0000  0.0007  0.0839  0.1103  0.0041  1  0.0455  Live cattle  0.5791  0.9040  0.6302  0.1571  0.4045  0.6506  0.4739  1  Our main aim is to analyze whether these correlations result from a common set of variables related to the fundamentals of commodities markets. If significant residual correlations remain, we would conclude in favor of excess co-movement. To model the eight commodity returns, we construct a set of 184 real and nominal macroeconomic variables. These variables, with a short description, are listed in Appendix A.10 Our data set contains variables from developed (118 variables from Australia, Canada, France, Germany, Japan, the UK, and the USA) and emerging countries (66 variables from China, Brazil, Korea, Taiwan, Mexico, etc.). In recent years, these countries have experienced high growth rates and their commodity demand has had a significant impact on commodity markets. Commonly used US databases such as those in Stock and Watson (2002b) and Ludvigson and Ng (2007, 2009) are thus not well-suited for our current purpose. We have the same classes of data for both developed and emerging countries. We include measures of the country’s aggregate activity level such as the industrial production index and manufacturing orders and capacity utilization. Other real variables are related to household expenditure: household consumption, housing starts, and car sales. We add variables related to the labor market (wages and unemployment) and international trade (exports, imports, and terms of trade). These real variables are assumed to be correlated with the world demand for commodities. The main categories of nominal variables that we include are monetary aggregates, stock indices, interest rates, exchange rates with the dollar, and producer and consumer price indices. These nominal variables help us to model the relationship between commodity returns and interest rate or the inflation rate. Finally, we add the Real Activity Index to the above, as developed in Kilian (2009). This is “based on dry cargo single voyage ocean freight rates and is explicitly designed to capture shifts in the demand for industrial commodities in global business markets” (p. 1055), following a long tradition of economists who have noted the correlation between economic activity and ocean-freight rates. 3. Filtering Commodity Returns Using Large Approximate Factors Models In this section, we first briefly review the large approximate factors method. Recent techniques to establish the optimal number of factors are presented in Appendix B; additional developments can be found in the survey by Bai and Ng (2008) of large approximate factors models. The remainder of the section is dedicated to the projection of commodity returns on the estimated factors. 3.1 Static Factors Calculation We use the static factor model of Stock and Watson (2002a). We do not consider the dynamic version of Forni et al. (2005), as recent work (Boivin and Ng, 2005) has shown that the dynamic and static factor models perform equally well, especially when the factors have unknown dynamics, which is often the case in empirical work. In addition, the dynamic factor model is best suited to forecasting, which is not the purpose of our work. We have a sample {xit} of i=1,…,N cross-section units and t=1,…,T time-series observations. Each xit is split into a component depending on a set of r<<N common factors Ft=(f1t,f2t,…,frt)′ and a specific component eit:   xit=λi′Ft+eit, where λi is the (r×1) factor loading. If we define the (N×1) vectors of observations and specific components at date t as Xt=(x1t,…,xNt)′, et=(e1t,…,eNt)′, and Λ=(λ1,…,λN)′ the (N×r) matrix of factor loadings, the factor decomposition is written as   Xt=ΛFt+et. Standard factor analysis makes the assumptions that Ft and et are serially and cross-sectionally uncorrelated, and the number of units of observation N is fixed. Stock and Watson’s (2002a, 2002b) “large dimensional approximate factor models” allow the specific errors to be “weakly correlated” across i and t11 and the sample size to tend to infinity in both directions. We assume k factors and use the principal components method to estimate the (T×k) factor matrix Fk and the corresponding (N×T) loading matrix Λk. The estimates solve the optimization problem:   min S(k)=(NT)−1∑i=1N∑t=1T(xit−λik′Ftk)2 subject to the normalization Λk′Λk/N=Ik.12 This classical principal component problem is solved by setting Λ^k equal to the eigenvectors of the largest k eigenvalues of X′X where X=(X1,X2,…,XT)′ is the (T×N) matrices of all observations.13 The principal components estimator of Fk is:   F^k=X′Λ^k/N. The consistency and asymptotic normal distribution of the principal component estimator as N,T→∞ have been, respectively, demonstrated by Stock and Watson (2002a), Bai and Ng (2002), and Bai (2003). The next step is to determine the optimal number of factors. The literature on this issue has not come to a clear consensus on how to select relevant factors and, as shown in Table BI in Appendix B, different methods lead to very different outcomes.14 We follow traditional practice in principal component analysis and choose the first nine factors, as the incremental explanatory power beyond these nine factors is only small. The nine factors explain 37% of the variability of the 184 macroeconomic variables. 3.2 Modeling Commodity Returns Our measure of excess co-movement makes use of commodity returns which have been filtered for common components. As such, once we have calculated the static factors, the second step of the empirical analysis consists in filtering the returns using these estimated factors. The first step is the linear regression of returns on the first three factors:   rit=αi+∑k=13βikF^k,t+uiti=1,…,8 t=1,…,T=αi+βi′F^t+uit, where rit represents the ith commodity return at date t, αi is the constant, βi the vector of factor coefficients for the ith commodity, and F^t=(F^1,t,F^2,t,F^3,t)′ the vector of the first three factors at date t. The results from seemingly unrelated regressions (SUR) appear in Table III. The R2 varies from 1.07% for soybeans to 28.58% for crude oil. The factors F^2 and F^1 are significant in most regressions. While the explanatory power figures for agricultural commodity returns are not substantially higher than those in PR, we do obtain a much higher R2 for metals and energy commodities.15 The ARCH-LM test shows that six out of the eight series of residuals have time-varying variance. Table III. Modeling the eight commodities returns: the three factors regressions—February 1993–November 2013 (i) This table reports the SUR estimates of the regression of the eight commodities monthly returns. The explanatory variables are reported in far-left column. A constant is always included in the regression and F^i denotes the ith factor. (ii) t-Statistics are reported in parenthesis under the estimates. ***, **, and *, respectively, denotes rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels. (iii) For the ARCH_LM, ***, **, and *, respectively, denote rejection of the null hypothesis of no ARCH effect at the 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022    (0.45)  (1.05)  (1.22)  (0.66)  (0.59)  (0.24)  (1.32)  (0.78)  F^1  −0.0254  −0.0951***  −0.0271  −0.0242  −0.0114  −0.0517**  −0.1200***  −0.0164**    (−1.19)  (−6.55)  (−1.40)  (−1.10)  (−0.87)  (−2.44)  (−7.10)  (−2.00)  F^2  0.0377**  0.0867***  0.0526***  0.0400**  0.0288*  0.0566***  0.1445***  0.0108    (2.28)  (6.32)  (2.77)  (2.15)  (1.73)  (2.93)  (8.34)  (1.04)  F^3  0.0020  0.0212  0.0060  0.0288  −0.0012  −0.0024  −0.0609**  −0.0062    (0.09)  (0.95)  (0.23)  (1.21)  (−0.05)  (−0.08)  (−2.34)  (−0.52)  R2  0.0222  0.2184  0.0343  0.0290  0.0107  0.0577  0.2858  0.0165  R¯2  0.0103  0.2089  0.0226  0.0172  −0.0014  0.0462  0.2771  0.0045  ARCH_LM(2)  13.10***  52.18***  7.31**  29.66***  3.54  5.84*  5.18*  1.63    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022    (0.45)  (1.05)  (1.22)  (0.66)  (0.59)  (0.24)  (1.32)  (0.78)  F^1  −0.0254  −0.0951***  −0.0271  −0.0242  −0.0114  −0.0517**  −0.1200***  −0.0164**    (−1.19)  (−6.55)  (−1.40)  (−1.10)  (−0.87)  (−2.44)  (−7.10)  (−2.00)  F^2  0.0377**  0.0867***  0.0526***  0.0400**  0.0288*  0.0566***  0.1445***  0.0108    (2.28)  (6.32)  (2.77)  (2.15)  (1.73)  (2.93)  (8.34)  (1.04)  F^3  0.0020  0.0212  0.0060  0.0288  −0.0012  −0.0024  −0.0609**  −0.0062    (0.09)  (0.95)  (0.23)  (1.21)  (−0.05)  (−0.08)  (−2.34)  (−0.52)  R2  0.0222  0.2184  0.0343  0.0290  0.0107  0.0577  0.2858  0.0165  R¯2  0.0103  0.2089  0.0226  0.0172  −0.0014  0.0462  0.2771  0.0045  ARCH_LM(2)  13.10***  52.18***  7.31**  29.66***  3.54  5.84*  5.18*  1.63  In a second approach, as in Stock and Watson (2002b) and Ludvigson and Ng (2009), we consider all possible combinations of the first nine estimated factors and, for each commodity, select the regression which minimizes the BIC criterion. Once each set of regressors has been selected, we jointly estimate the eight regressions via SUR. Our aim here is to identify the best model from a set of common regressors for each commodity. This approach aims to eliminate as much residual correlation as possible, and so strengthen our evidence for any excess co-movement. The SUR estimates appear in Table IV and show a significant increase in explanatory power for crude oil, while this figure remains low for the other commodities. Again, the F^1 and F^2 factors are significant for most of the eight commodities and the ARCH-LM test rejects the null hypothesis of constant variance for three series of residuals. Table IV. Modeling the eight commodities returns: the BIC minimizing regressions—February 1993–November 2013 This table reports the SUR estimates of the regression of the eight commodities monthly returns. The explanatory variables are reported in left column. A constant is always included in the regression and F^i denotes the ith factor. (ii) t-Statistics are reported in parenthesis under the estimates. ***, **, and *, respectively, denotes rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels. (iii) For the ARCH_LM, ***, **, and *, respectively, denote rejection of the null hypothesis of no ARCH effect at the 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022    (0.44)  (1.05)  (1.24)  (0.66)  (0.59)  (0.24)  (1.45)  (0.78)  F^1    −0.0844***        −0.0422**  −0.1194***        (−6.29)        (−2.51)  (−7.39)    F^2  0.0371*  0.0875***  0.0533***  0.0419**  0.0279  0.0577***  0.1449***      (1.84)  (5.17)  (2.64)  (2.12)  (1.45)  (2.74)  (7.72)    F^3              −0.0617***                  (−2.65)    F^6              0.1142***                  (4.00)    F^8      0.0967***        0.1858***          (3.01)        (5.74)    F^9                −0.0393**                  (−2.05)  R2  0.0138  0.2136  0.0728  0.0159  0.0088  0.0566  0.4079  0.0171  R¯2  0.0098  0.2072  0.0653  0.0119  0.0048  0.0489  0.3958  0.0131  ARCH_ LM(2)  7.65**  1.49  3.31  18.54***  2.83  5.50*  4.12  0.96    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022    (0.44)  (1.05)  (1.24)  (0.66)  (0.59)  (0.24)  (1.45)  (0.78)  F^1    −0.0844***        −0.0422**  −0.1194***        (−6.29)        (−2.51)  (−7.39)    F^2  0.0371*  0.0875***  0.0533***  0.0419**  0.0279  0.0577***  0.1449***      (1.84)  (5.17)  (2.64)  (2.12)  (1.45)  (2.74)  (7.72)    F^3              −0.0617***                  (−2.65)    F^6              0.1142***                  (4.00)    F^8      0.0967***        0.1858***          (3.01)        (5.74)    F^9                −0.0393**                  (−2.05)  R2  0.0138  0.2136  0.0728  0.0159  0.0088  0.0566  0.4079  0.0171  R¯2  0.0098  0.2072  0.0653  0.0119  0.0048  0.0489  0.3958  0.0131  ARCH_ LM(2)  7.65**  1.49  3.31  18.54***  2.83  5.50*  4.12  0.96  While the factors cannot be identified econometrically, it is very useful to identify the macroeconomic variables behind the factors affecting commodity returns. To interpret the factors, we follow Ludvigson and Ng (2009) and divide our 184 series into developed and emerging countries, and then real and nominal variables.16 Each of the 184 original variables is then regressed on one factor with the resulting R2 appearing on the horizontal axis. We can thus see which macroeconomic variables obtain the highest R2. The factor in question can then be thought to represent this set of variables. Figure 2 plots R2 for both F^1 (top panel) and F^2 (bottom panel) factors. F^1, which explains a significant part of crude oil and copper returns, is mostly correlated with real variables in emerging economies. This illustrates the importance of emerging countries in shaping commodity prices.17 This result corroborates recent work that also demonstrates that oil (e.g., Hamilton, 2009 or Kilian and Hicks, 2013) and agricultural prices (e.g., Hamilton and Wu, 2015) are partly driven by demand from emerging countries and that speculative activity only plays a minor role. Figure 2. View largeDownload slide Marginal R2 of macroeconomic and financial variables regressed on the first two estimated factors. Notes: Each panel shows the R2 from regressing the series number given on the x-axis on to each individual factor F^i. The series are detailed in Appendix A and sorted as they appear in the figure (real variables for developed countries, nominal variables for developed countries, real variables for emerging countries, and nominal variables for emerging countries). Figure 2. View largeDownload slide Marginal R2 of macroeconomic and financial variables regressed on the first two estimated factors. Notes: Each panel shows the R2 from regressing the series number given on the x-axis on to each individual factor F^i. The series are detailed in Appendix A and sorted as they appear in the figure (real variables for developed countries, nominal variables for developed countries, real variables for emerging countries, and nominal variables for emerging countries). The interpretation of factor F^2 is less obvious. It is highly correlated with a small number of real variables but its explanatory power with respect to interest rates, producer and consumer price indices, and monetary aggregates in both developed and emerging countries is greater than that of F^1. F^2 is likely to represent these nominal variables. Earlier contributions (Barsky and Kilian, 2002; Frankel and Rose, 2010) only provide mixed evidence on the relationship between interest rates and commodity prices. Our estimates give additional support to such a link. In this regard, price indices and monetary aggregates may pick up the impact of inflation on commodity prices. The activity index from Kilian (2009) brings no additional information as it attracts only an insignificant estimated coefficient—except for copper, at the 10% threshold only—indicating that F^1 does a better job of modeling commodity returns. This conclusion is of interest as this real-activity index is considered to be as a proxy for economic activity. We believe that this result confirms the ability of statistical factors to aggregate information from a large number of variables and capture high-frequency growth rates. To better understand the insignificance of Kilian’s index, Table V shows the estimates from univariate regressions of the nine empirical factors on Kilian’s index. The estimates are very significant but with little explanatory power. This is likely due to the low-frequency nature of the Kilian Index, and further demonstrates the benefit from using statistical factors in modeling monthly commodity returns. Table V. Regression of the Kilian real activity index on each of the nine factors Coefficient reports the estimated coefficient of each factor and t-stat its Student’s statistic. ***, **, and *, respectively, denote rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels. The real activity index is taken from Lutz Kilian’s homepage. See Kilian (2009) for a definition of this index.   F^1  F^2  F^3  F^4  F^5  F^6  F^7  F^8  F^9  Coefficient  −17.8231***  19.3546***  −5.6256  28.5721***  1.6348  −2.0679  4.7489  −9.6327  4.0327  t-stat  −3.78  3.504  −0.77  4.77  0.19  −0.21  0.48  −0.85  0.34  R2  0.0438  0.0387  0.0021  0.0434  0.0001  0.0002  0.0008  0.0032  0.0005  R¯2  0.0400  0.0349  −0.0019  0.0396  −0.0039  −0.0038  −0.0031  −0.0008  −0.0035    F^1  F^2  F^3  F^4  F^5  F^6  F^7  F^8  F^9  Coefficient  −17.8231***  19.3546***  −5.6256  28.5721***  1.6348  −2.0679  4.7489  −9.6327  4.0327  t-stat  −3.78  3.504  −0.77  4.77  0.19  −0.21  0.48  −0.85  0.34  R2  0.0438  0.0387  0.0021  0.0434  0.0001  0.0002  0.0008  0.0032  0.0005  R¯2  0.0400  0.0349  −0.0019  0.0396  −0.0039  −0.0038  −0.0031  −0.0008  −0.0035  Finally, the omission of inventory data in our analysis is worthy of mention. It is commonly thought that stock levels may help us to better model commodity returns, following Working’s Theory of Storage. For instance, Pindyck (2001) uses weekly inventory data from the US Department of Energy to model the convenience yield in the WTI crude oil market. Geman and Nguyen (2005) rely on a number of worldwide sources to construct their own inventory series for soybeans which they use to model this commodity’s forward curve. Baumeister and Kilian (2012) consider a number of oil-specific inventory series to forecast real-time monthly oil prices. We do not include inventory information in our empirical analysis as we wish to filter returns using fundamentals that are, at least partly, common to all commodities. By doing so, data related to commodity demands that we proxy via our factors are relevant as they represent common fundamentals. Conversely, data such as inventories are very particular to each commodity and so less likely to explain any correlation in commodity returns. As such, even if we recognize that inventories matter in particular cases such as, for instance, forecasting commodity prices [see Baumeister and Kilian (2012, 2014) for the case of oil], they do not do so here, where it is rather common factors that are our primary concern. 4. Testing for the Excess Co-movement of Commodity Returns 4.1 Testing for Residual Correlation The residuals from the regressions above reflect commodity returns after controlling for fundamentals. We first evaluate the correlation in residuals, as in PR. Tables VI and VII show the sample correlations (in the upper triangular part) and their p-values18 (in the lower triangular part) for the residuals from the three-factor and BIC linear filters. Table VI. Correlation between residuals from the three factors linear model The upper triangular matrix reports correlation while the lower reports the p-values. The p-value is computed by transforming the correlation ρ^ to create a t-statistic having T – 2 degrees of freedom, where T is the number of observations. ***, **, and *, respectively, denote significance at 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.3452***  0.2564***  0.4939***  0.0204  0.3133***  0.2330***  0.0660  Copper  0.0000  1  0.4317***  0.2817***  0.1698***  0.3386***  0.5312***  0.0913  Silver  0.0000  0.0000  1  0.2398***  0.2110***  0.1439**  0.3017***  0.0101  Soybeans  0.0000  0.0000  0.0001  1  −0.0568  0.4140***  0.1842***  −0.0574  Raw sugar  0.7478  0.0071  0.0008  0.3709  1  0.0884  0.1567**  0.0740  Cotton  0.0000  0.0000  0.0228  0.0000  0.1636  1  0.3102***  0.0225  Crude oil  0.0002  0.0000  0.0000  0.0035  0.0131  0.0000  1  0.1337  Live cattle  0.2986  0.1499  0.8739  0.3663  0.2438  0.7228  0.0346  1  Breusch–Pagan LM test  186.20  p-value  0    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.3452***  0.2564***  0.4939***  0.0204  0.3133***  0.2330***  0.0660  Copper  0.0000  1  0.4317***  0.2817***  0.1698***  0.3386***  0.5312***  0.0913  Silver  0.0000  0.0000  1  0.2398***  0.2110***  0.1439**  0.3017***  0.0101  Soybeans  0.0000  0.0000  0.0001  1  −0.0568  0.4140***  0.1842***  −0.0574  Raw sugar  0.7478  0.0071  0.0008  0.3709  1  0.0884  0.1567**  0.0740  Cotton  0.0000  0.0000  0.0228  0.0000  0.1636  1  0.3102***  0.0225  Crude oil  0.0002  0.0000  0.0000  0.0035  0.0131  0.0000  1  0.1337  Live cattle  0.2986  0.1499  0.8739  0.3663  0.2438  0.7228  0.0346  1  Breusch–Pagan LM test  186.20  p-value  0  Table VII. Correlation between residuals from the BIC minimizing regressions The upper triangular matrix reports correlation while the lower reports the p-values. The p-value is computed by transforming the correlation ρ^ to create a t-statistic having T – 2 degrees of freedom, where T is the number of observations. ***, **, and *, respectively, denote significance at 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.2674***  0.1990***  0.4711***  −0.0135  0.2528***  0.0609  0.0287  Copper  0.0000  1  0.3328***  0.1930***  0.0962  0.1458**  0.1068*  −0.0431  Silver  0.0016  0.0000  1  0.1869***  0.1733***  0.0375  0.0411  −0.0390  Soybeans  0.0000  0.0022  0.0030  1  −0.0897  0.3677***  −0.0087  −0.0990  Raw sugar  0.8324  0.1294  0.0060  0.1572  1  0.0337  0.0251  0.0450  Cotton  0.0001  0.0211  0.5552  0.0000  0.5957  1  0.0261  −0.0496  Crude oil  0.3375  0.0920  0.5179  0.8912  0.6929  0.6812  1  −0.0249  Live cattle  0.6512  0.4978  0.5389  0.1184  0.4783  0.4345  0.6956  1  Breusch– Pagan LM test  99.39  p-value  0    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.2674***  0.1990***  0.4711***  −0.0135  0.2528***  0.0609  0.0287  Copper  0.0000  1  0.3328***  0.1930***  0.0962  0.1458**  0.1068*  −0.0431  Silver  0.0016  0.0000  1  0.1869***  0.1733***  0.0375  0.0411  −0.0390  Soybeans  0.0000  0.0022  0.0030  1  −0.0897  0.3677***  −0.0087  −0.0990  Raw sugar  0.8324  0.1294  0.0060  0.1572  1  0.0337  0.0251  0.0450  Cotton  0.0001  0.0211  0.5552  0.0000  0.5957  1  0.0261  −0.0496  Crude oil  0.3375  0.0920  0.5179  0.8912  0.6929  0.6812  1  −0.0249  Live cattle  0.6512  0.4978  0.5389  0.1184  0.4783  0.4345  0.6956  1  Breusch– Pagan LM test  99.39  p-value  0  The results from both sets of regressions confirm the hypothesis of excess co-movement. We find 16 and 18 significant sample correlations at the 1% and 5% significance levels, respectively, for the three-factor regressions; the analogous numbers for the BIC-minimizing regressions are 9 and 10. Unsurprisingly, the Breusch–Pagan LM test rejects the null hypothesis of no residual correlation in both cases. In the BIC-minimizing regressions, five sample correlations are no longer significant, mostly related to crude oil.19 Filtering commodity returns therefore somewhat reduces the number of significant correlations. However, as the significant correlations range from 0.4711 (wheat and soybeans) to 0.1066 (copper and crude oil) the level of residual correlation remains quite substantial. 4.2 A Global, Unbiased, and Time-Varying Measure of Excess Co-movement One major limit of the use of sample correlation to gauge excess co-movement is the bias in the former when volatility is time-varying.20 This argument has been put forward in the contagion literature21 by Forbes and Rigobon (2002), among others.22 When there is a simultaneous rise in the respective volatility of two variables, the typical sample correlation measure overestimates the true correlation. Forbes and Rigobon (2002) propose an unbiased correlation estimator: as our residuals very often have time-varying volatility, this is the estimator we use to evaluate excess co-movement. We follow Kallberg and Pasquariello (2008), who apply the Forbes and Rigobon estimator on a moving-window basis to yield a more precise estimator of the true correlation. We end up with a global measure, as we treat all residual correlations equally, without focusing on the correlation of one particular commodity with another. We calculate a time-varying measure of excess co-movement which will show us whether excess co-movement is a permanent feature of commodity markets or if it is only occasional. Our global measure of excess co-movement is the average of all the squared unbiased correlations. We use squared correlation measures as some of the estimated correlations are negative. Our estimate is nonparametric and avoids the mean-reversion problem inherent in the parametric approach, such as in the dynamic conditional correlation (DCC) model (Engle, 2002). Indeed, in many instances (cf. Kallberg and Pasquariello, 2008, among others), methods based on rolling windows filters are very competitive with parametric methods when the object of interest is the estimation of correlations. In the following, we present the bias-corrected correlation estimator and the aggregation process to obtain our overall excess co-movement measure. For all pairs of non-redundant returns i≠j, we calculate the residual correlation:   ρ^ij,t=cov(u^i,t,u^j,t)[var(u^i,t)var(u^j,t)]1/2, where u^i,t is the residual from the ith commodity-return equation. As the sample correlation ρ^ij,t is biased in the case of heteroscedasticity, this is called the “conditional correlation”. The Forbes and Rigobon (2002) bias-corrected correlation estimator is   ρ^ij,t*=ρ^ij,t[1+δ^i,t(1−(ρ^ij,t2)]1/2, where the ratio δ^i,t=var(u^i,t)var(u^i,t)LT−1 corrects the conditional correlation ρ^ij,t for the change between the ith return’s short-term var(u^i,t) and long-term var(u^i,t)LT volatilities.23 ρ^ij,t* is called the unconditional correlation. As we do not make any ex ante assumption regarding the direction of the propagation of shocks from one commodity to another, we alternately assume that the source of these shocks is asset i (in ρ^ij,t*) or asset j (in ρ^ji,t*). We therefore have two unconditional and possibly different correlations, ρ^ij,t* and ρ^ji,t*. Our global excess co-movement measure is based on these unconditional correlations. As suggested in King, Sentana, and Wadhwani (1994) and Kallberg and Pasquariello (2008), we compute the arithmetic mean of the pairwise squared unbiased correlations for each commodity i. A non-null unconditional correlation ρ^ij,t*≠0 and ρ^ji,t*≠0, whatever its sign, is taken as evidence of excess co-movement between commodities i and j. A measure of excess co-movement between commodity i and the others is defined as:   ρ^i,t*=1K−1∑j=1,j≠iK(ρ^ij,t*)2 for all commodity returns i=1,…,K, where K = 8 is the number of commodities. Our global and time-varying measure of excess co-movement is then the mean of the excess squared unconditional correlations over all commodities:   ρ^t*=1K∑i=1Kρ^i,t*. We treat the covariance matrix of return residuals as observable, and construct a time series of rolling realized excess squared correlations for each commodity i. δ^i,t and ρ^i,t* are estimated over short- and long-term intervals of fixed length N [t−N+1,t] and gN (with g > 1) [t−gN+1,t], respectively. We use a rolling window of N = 30 monthly observations for short-term volatility and gN = 60 monthly observations for long-term volatility. 4.3 Estimation Results We compute three averages of squared correlations, all of which appear in Figure 3, to evaluate the importance of filtering returns and illustrate the time-variation in volatility. The first (dashed–dotted line) is the average of the squared unconditional correlations in returns: ρ^ret,t*=1K∑i=1Kρ^ret i,t*, where the unbiased correlations are calculated for non-filtered returns. The second (dashed line) is the average of the squared conditional correlations between residuals: ρ^t=1K∑i=1Kρ^i,t, where ρ^i,t=1K−1∑j=1,j≠iK(ρ^ij,t)2. We here use residual correlations that are not corrected for changes in volatility. The solid line is the average of the unconditional squared correlations ρ^t* as defined in the previous section, which is our estimate of excess co-movement. Figure 3. View largeDownload slide Mean excess squared correlation for commodity returns and residuals. Notes: (i) “av sq unc corr ret” is the average squared unconditional correlation of returns: ρret.t*. (ii) “av sq cond corr res fund” is the average squared conditional correlation of factors regression residual: ρt*. (iii) “av sq unc corr res all” is the average squared unconditional of factors regression residual. (iv) The confidence band is the minimal value above which squared correlation is significant at 5% level. It is computed from the t-squared ratio test t^ijt2=(ρ^ijt*)2[1−ρ^ijt*]−1(N−2)∼F(1.N−2) and is equal to 1.6990. Figure 3. View largeDownload slide Mean excess squared correlation for commodity returns and residuals. Notes: (i) “av sq unc corr ret” is the average squared unconditional correlation of returns: ρret.t*. (ii) “av sq cond corr res fund” is the average squared conditional correlation of factors regression residual: ρt*. (iii) “av sq unc corr res all” is the average squared unconditional of factors regression residual. (iv) The confidence band is the minimal value above which squared correlation is significant at 5% level. It is computed from the t-squared ratio test t^ijt2=(ρ^ijt*)2[1−ρ^ijt*]−1(N−2)∼F(1.N−2) and is equal to 1.6990. Table VIII shows the descriptive statistics for the returns and residual average squared correlations estimated over the full sample. We draw three main conclusions from this table. First, while the means of ρ^ret,t* and ρ^t* are very similar over the whole sample, there is a notable difference—almost equal to 10% in some months—between the two measures over the 2008–13 period. This emphasizes the importance of filtering returns using some measures of fundamentals and shows that the rise in commodity-returns correlation is partly due to common factors. Juvenal and Petrella (2015) find that the co-movements between the prices of oil and other commodities reflect global demand shocks. We are partially in line with them in that, once factors related to demand are taken into account, the residual correlation is lower, with this effect being stronger in recent years when demand shocks were larger. Table VIII. Descriptive statistics on returns and residuals average squared unconditional correlations (i) This table reports summary statistics on average squared unconditional return correlation ρ^ret.t* and average squared unconditional residual correlation ρ^t*. (ii) Fρ*2 is the mean percentage of average squared unconditional correlation significant at the 5% level using the t-squared ratio test t^ijt2=(ρ^ijt*)2[1−ρ^ijt*]−1(N−2)∼F(1.N−2). (iii) ***, **, and *, respectively, denote significance at 1%, 5%, and 10%, levels. (iv) Cρ is the correlation between ρ^ret.t* and ρ^t*.   ρ^ret.t*  ρ^t*  μ  0.1982**  0.1803**  σ  0.0455  0.0246  Fρ*2  0.6230  0.6440  Cρ  0.9319      ρ^ret.t*  ρ^t*  μ  0.1982**  0.1803**  σ  0.0455  0.0246  Fρ*2  0.6230  0.6440  Cρ  0.9319    Second, looking at both ρ^t* and ρ^t, taking time-variation in volatility into consideration only moderately affects the estimated correlation: the two lines are almost identical except in periods of high volatility, where there is a difference (although only small) between the two measures. Third, and most importantly, our measure of excess co-movement is significant at the 5% level only half of the time in the period under consideration.24 We thus conclude that the excess co-movement in commodity prices cannot be viewed as a general feature of commodity markets but is rather sample-dependent. As PR do not investigate time-variation in their excess co-movement measure, our results cannot be compared to theirs. There is a possibility, however, that the estimated excess co-movement over the 1960–85 period that PR find is insignificant over some sub-samples, thereby questioning the determinants of this phenomenon. In the same vein as the correlation plot in Tang and Xiong (2012), the chart of average squared correlations in Figure 1 provides a finer description of the estimated excess co-movement. This latter is mostly significant during periods of financial crisis: from mid-2000 to early 2003, and from 2008 onward. In their “convective risk flows” model, Cheng, Kirilenko, and Xiong (2015) show that financial traders cut their net long positions in response to market distress. A coordinated drop in the long positions of financial traders may help explain excess co-movement. Alternatively, excess co-movement may also reflect a “flight-to-quality” phenomenon, where investors decide to partly leave the stock market and invest heavily in commodities to diversify their positions. Moreover, the period starting in 2000 also corresponds to the growing financialization of commodity futures markets, as excellently surveyed in Cheng and Xiong (2014a). As such, excess co-movement might be related to speculative activity in commodity futures markets. Whether excess co-movement comes the changing nature of trading in commodity markets is a central question that we answer in the next section. 5. Explaining Excess Co-movement The literature on institutional investors and their possible impact on commodity prices has grown dramatically in recent years [see the nice surveys in Irwin and Sanders (2011), Cheng and Xiong (2014a), and Haase, Zimmermann, and Zimmermann (2016)]. Prior research has, however, produced mixed results. While some authors have produced evidence of a significant effect of index funds on commodity prices (Tang and Xiong, 2012; Singleton, 2013, among others), others have found evidence to the contrary (Rouwenhorst and Tang, 2012; Hamilton and Wu, 2015; Lehecka, 2015). Surprisingly, there is no work dealing with the impact of financialization on cross-market return linkages except for that in Tang and Xiong (2012).25 The latter attempt to explain the recent rise in the co-movement of a number of commodity prices via five hypotheses: (i) the financialization of commodities, (ii) the rapid growth of emerging economies, (iii) the recent world financial crisis, (iv) inflation, and (v) the adoption of biofuels. Our research question is linked to the arguments in Tang and Xiong (2012), in that we arguably jointly test their first and third hypotheses, and consider the second and fourth in Section 3 when we filter returns using common factors. In particular, we have shown that growth in emerging economies leads commodity prices (hypothesis (ii)), and that commodity returns are correlated with a “nominal variables” factor (hypothesis (iv)). Both of these effects likely contribute to excess co-movement and are expressly taken into account in our work. This section aims to show that speculation in commodity futures markets is a significant determinant of our estimated excess co-movement. The issue is new and challenging, as no significant evidence has been put forward in the literature to date. Our empirical approach is as follows. In a first step, we show that speculative activity and filtered futures returns are correlated for most of the commodities in our sample. Then, in a second step, we show that measures of speculative activity are correlated across commodities. Taken together, these results provide direct evidence of speculation as a driver of excess co-movement.26 5.1 Measure of Speculative Intensity Our measure of speculative intensity builds on the work in Han (2008), where a new speculative index is developed following the literature on investor sentiment (see Baker and Wurgler, 2007).27 The basic idea is to pick up the net view of speculators in a given futures market by comparing their long and short positions. Han’s index is given by the number of long non-commercial contracts minus the number of short non-commercial contracts, scaled by the total open interest in futures markets for the commodity of interest; as such this is a directional index of speculative activity in the futures market. We calculate Han’s index for our eight commodities using CFTC data. All traders who are considered as large enough—positions are above a specified level that is commodity-dependent—are required to provide the CFTC with their daily positions. The Commitments of Traders (COT) Report corresponds to the weekly aggregation of the daily positions and is released each Tuesday. CFTC differentiates between “commercial” and “non-commercial” traders and provides long and short positions for both categories.28 “Commercial” traders should be able to prove an involvement in the physical market and are thus considered as hedgers while “non-commercial” traders have no relation with the cash business: this latter group then consists of speculators. The usefulness of these CFTC data has previously been discussed in Bessembinder (1992) and Cheng and Xiong (2014b), as many traders carry out activities which cover both hedging and speculation. In particular, Cheng and Xiong (2014b) show that hedgers may react to changes in commodity futures prices in a number of US Agricultural markets, which is undoubtedly a form of speculation. In what follows, we show that CFTC data are informative for our purpose despite the potential bias in the definition of categories of traders. We also experiment with alternative measures of trading activity. The first of these is the Working’s T speculative index, as recently used in Byksahin and Robe (2014). A second measure of trading activity is hedging pressure, as defined in de Roon, Nijman, and Veld (2000), who showed that futures risk premia depend on both own-market and cross-market hedging pressure. Their measure of hedging pressure is calculated as the difference between the number of short and long hedge positions, divided by the total number of hedge positions. This measure focuses on the positions of traders who are hedgers, that is, who have a cash business for the commodity. It is different from the Han index, where the denominator is the total open interest and not the total number of speculative positions, but the idea is roughly similar as hedging pressure also picks up the difference between long and short positions.29 Results from using either Working’s T or hedging pressure are very similar to those presented here, and are not reported to save space (but available upon request). 5.2 Empirical Results To deal with the potential correlation of Han’s indices and the business cycle, we regress our speculative indices on the same set of factors ( F^1t,F^2t,F^3t,F^6t,F^8t) as was used to filter commodity returns. Then, to gauge the explanatory power of Han’s indices, we include them in univariate regressions of the form:   Resi,t=ai+∑j=18bi,jHj,t+ei,t, where Resi,t is the ith commodity return residual at time t and Hj,t is Han’s index for the jth commodity at time t adjusted for the factors. Our choice to use contemporaneous variables in the regressions is motivated by the monthly frequency and the efficient-market hypothesis stating that any impact of index funds should be instantaneously reflected in prices [see Gilbert and Pfuderer (2014) for further developments on this issue]. We also choose to consider all the Han indices in each univariate regression following existing research on cross-commodity trading and its potential impact on prices (e.g., de Roon, Nijman, and Veld, 2000).30 The estimated coefficients appear in Table IX. We observe that, with the exception of oil, copper, and silver, commodity returns are correlated with their own Han’s index. The R2 mostly ranges between 1% and 3%, but reaches 12.29% for raw sugar. We find a positive and significant impact of the Han index on its corresponding commodity for wheat, raw sugar, soybeans, and live cattle. For these commodities, speculative trading and returns move in the same direction. We also observe some cross-effects: the raw sugar and copper Han indices have an impact on wheat returns, and the cotton Han index has an effect on crude oil returns. In these cases, the estimated coefficient is negative but only weakly significant. The Han index for wheat has a positive effect on live cattle residual. The interpretation of these cross-effects is quite challenging as there is no link, such as substitutability or complementarity, between the commodities concerned. Table IX. OLS regressions of residual returns on speculative Han indices (i) This table reports OLS estimates of the regression of the eight commodities monthly residual returns on Han speculative indices. (ii) The Han indices are corrected for the factors {F^1t,F^2t,F^3t,F^6t,F^8t} to control for the effect of the business cycle. (iii) t-statistics are reported in parenthesis under the estimates. ***, **, and *, respectively, denote rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0009      −0.0002  −0.0013  −0.0001  0.0004  0.0000    (0.20)      (−0.04)  (−0.23)  (−0.03)  (0.10)  (0.02)  Han_Wheat  0.0778***              0.0429*    (2.67)              (1.88)  Han_Copper  −0.0597*                  (−1.92)                Han_Silver                  Han_Soybeans        0.0639*                  (1.79)          Han_Raw Sugar  −0.0514*        0.2063***          (−1.90)        (5.52)        Han_Cotton            0.0636***  −0.0396*                (2.98)  (−1.79)    Han_Crude Oil                  Han_Live Cattle                0.0816***                  (4.05)  R2  0.0286      0.0110  0.1229  0.0206  0.0101  0.0397  R¯  0.0166      0.0070  0.1194  0.0166  0.0061  0.0318    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0009      −0.0002  −0.0013  −0.0001  0.0004  0.0000    (0.20)      (−0.04)  (−0.23)  (−0.03)  (0.10)  (0.02)  Han_Wheat  0.0778***              0.0429*    (2.67)              (1.88)  Han_Copper  −0.0597*                  (−1.92)                Han_Silver                  Han_Soybeans        0.0639*                  (1.79)          Han_Raw Sugar  −0.0514*        0.2063***          (−1.90)        (5.52)        Han_Cotton            0.0636***  −0.0396*                (2.98)  (−1.79)    Han_Crude Oil                  Han_Live Cattle                0.0816***                  (4.05)  R2  0.0286      0.0110  0.1229  0.0206  0.0101  0.0397  R¯  0.0166      0.0070  0.1194  0.0166  0.0061  0.0318  One may rightly suspect that these OLS estimates are plagued by endogeneity. To assess the presence of endogeneity, we estimate all previous regressions via GMM and use one-period and two-period lagged Han indices as instruments.31 The results in Table X show that the test for exogeneity based on the difference in the J-test does not reject the exogeneity of the Han index for cotton, crude oil. We therefore consider the previous OLS estimates as valid: the Han index has a positive effect on its own commodity for cotton while that for cotton has a negative impact on oil. We reject the exogeneity of the Han index in the wheat, raw sugar, soybeans, and live cattle regressions. Hansen’s (1982)J-test for overidentifying restrictions validates our set of instruments. We also check that the instruments are not weak. With the exception of soybeans, the GMM estimates are in line with the previous OLS regressions for these four commodities. Wheat return is still negatively and significantly impacted by the Han index for raw sugar, although the indices for wheat and copper are no longer significant. The raw sugar Han index still has a positive impact on its return. The wheat Han indices have a positive and significant impact on live cattle residual return. Table X. GMM regressions of residual returns on speculative Han indices; Instruments {Hant−1,Hant−2} (i) This table reports GMM estimates of the regression of the eight commodities monthly residuals returns. The explanatory variables are reported in far-left column. (ii) The Han indices are corrected for the factors {F^1t,F^2t,F^3t,F^6t,F^8t} to control for the effect of the business cycle. (iii) t-statistics are reported in parenthesis under the estimates. ***, **, and *, respectively, denotes rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels. (iv) J-test is the Hansen (1982) test of overidentifying restrictions and the Diff in J-test is the test for endogeneity of regressors. P-values are reported under the test-statistics.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0010      0.0008  0.0009  −0.0008  0.0003  0.0005    (0.21)      (0.17)  (0.18)  (−0.16)  (0.09)  (0.20)  Han_Wheat  −0.0351              0.0555*    (−0.61)              (1.85)  Han_Copper  0.0010                  (0.02)                Han_Silver                  Han_Soybeans        −0.0252                  (−0.65)          Han_Raw Sugar  −0.1018**        0.0987**          (−2.57)        (2.33)        Han_Cotton            0.0439  −0.0362                (1.59)  (−1.47)    Han_Crude Oil                  Han_Live Cattle                0.0122                  (0.36)  Exogeneity test                  Diff in J-test  9.64**      7.75***  17.28***  1.83  0.0678  10.75***  pval  0.0219      0.0054  0.0000  0.1750  0.79  0.00  Overidentifying test  J-test  1.1770      1.3462  2.5290  0.4705  0.0077  1.028  pval  0.75      0.9966  0.11  0.2459  0.49  0.31    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0010      0.0008  0.0009  −0.0008  0.0003  0.0005    (0.21)      (0.17)  (0.18)  (−0.16)  (0.09)  (0.20)  Han_Wheat  −0.0351              0.0555*    (−0.61)              (1.85)  Han_Copper  0.0010                  (0.02)                Han_Silver                  Han_Soybeans        −0.0252                  (−0.65)          Han_Raw Sugar  −0.1018**        0.0987**          (−2.57)        (2.33)        Han_Cotton            0.0439  −0.0362                (1.59)  (−1.47)    Han_Crude Oil                  Han_Live Cattle                0.0122                  (0.36)  Exogeneity test                  Diff in J-test  9.64**      7.75***  17.28***  1.83  0.0678  10.75***  pval  0.0219      0.0054  0.0000  0.1750  0.79  0.00  Overidentifying test  J-test  1.1770      1.3462  2.5290  0.4705  0.0077  1.028  pval  0.75      0.9966  0.11  0.2459  0.49  0.31  Our approach through instrumental variables unambiguously shows that there is a significant impact of changes in the speculative index on contemporaneous returns, even after controlling for endogeneity for most commodities. This impact is positive when the Han index and the return pertain to the same commodity. We now focus on sample cross-correlations between speculative intensities in commodity futures markets. As expected from the “style investing” hypothesis developed in Barberis and Shleifer (2003) or the more general increase in non-commercial positions in commodity-futures markets in the last decades (see Cheng and Xiong, 2014a), the cross-correlations between speculative indices are mostly positive and significant as shown in Table XI. There are, respectively, 10, 15, and 19 significant cross-correlations at the 1%, 5%, and 10% significance levels. The cross-correlations are significantly negative in only three cases (silver and raw sugar, silver and soybeans, wheat and live cattle). We therefore have evidence that speculative indices move together, even for commodities of different classes such as, for instance, wheat and copper, cotton and crude oil, or raw sugar and live cattle. Table XI. Correlation between the eight Han indices (i) The upper triangular matrix reports correlations while the lower reports their p-values. ***, **, and *, respectively, denote significance at 1%, 5%, and 10% levels. (ii) The Han indices are corrected for the factors {F^1t,F^2t,F^3t,F^6t,F^8t} to control for the effect of the business cycle.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.1744***  0.0107  0.2133***  −0.0334  0.1394**  −0.0226  −0.1194*  Copper  0.00  1  0.2146***  0.0029  0.1271**  0.1888***  0.1071*  0.2227***  Silver  0.86  0.00  1  −0.1408**  −0.1128*  −0.0935  −0.1024  −0.0925  Soybeans  0.00  0.96  0.02  1  0.1853***  0.4619***  0.2787***  0.0830  Raw sugar  0.59  0.04  0.07  0.00  1  0.0222  0.1060*  0.2280***  Cotton  0.02  0.00  0.14  0.00  0.72  1  0.3005***  0.1324**  Crude oil  0.72  0.09  0.10  0.00  0.09  0.00  1  0.1102*  Live cattle  0.06  0.00  0.14  0.19  0.00  0.03  0.08  1    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.1744***  0.0107  0.2133***  −0.0334  0.1394**  −0.0226  −0.1194*  Copper  0.00  1  0.2146***  0.0029  0.1271**  0.1888***  0.1071*  0.2227***  Silver  0.86  0.00  1  −0.1408**  −0.1128*  −0.0935  −0.1024  −0.0925  Soybeans  0.00  0.96  0.02  1  0.1853***  0.4619***  0.2787***  0.0830  Raw sugar  0.59  0.04  0.07  0.00  1  0.0222  0.1060*  0.2280***  Cotton  0.02  0.00  0.14  0.00  0.72  1  0.3005***  0.1324**  Crude oil  0.72  0.09  0.10  0.00  0.09  0.00  1  0.1102*  Live cattle  0.06  0.00  0.14  0.19  0.00  0.03  0.08  1  Overall, our empirical results demonstrate that speculative activity is a significant driver of excess co-movement. We thus confirm the implications in Basak and Pavlova (2016) that institutional investors do play a role in linking commodity futures prices. Our results are also in line with those in Tang and Xiong (2012), but provide stronger evidence of the impact of speculation on co-movements as we on purpose control for the impact of real variables on commodity prices. More generally, our results demonstrate the critical role of trading for price determination, and the overall importance of the “financialization of the commodity markets”, a concept that has attracted growing interest in academic and political spheres over recent years. 6. Concluding Remarks The aim of this paper was to reconsider the question of the excess co-movement of commodity prices and to provide an explanation of this phenomenon, if it was found to be present in the data. We believe that our paper offers new perspectives for the analysis of co-movement in commodity returns. First, as discussed above, we use the large approximate factor model method to uncover the relevant factors that allow us to explain commodity returns. To the best of our knowledge, this is the first time that this method has been used to filter out returns before looking for excess co-movement. The main advantage of factors is that they allow us to deal with a large number of variables, while retaining econometric tractability, thereby including a richer set of fundamentals. We thus avoid any artificial limit on the information set, which has been a major constraint in previous work. Our second contribution is to provide an explanation of the excess co-movement in commodity returns. Previous work has emphasized the methodological aspects of the assessment of the hypothesis of excess co-movements. Surprisingly, however, the issue of which variables are related to this phenomenon has not been analyzed to date. Our indicator of speculative activity, calculated using traders’ positions available from CFTC, is both correlated across commodities and with futures prices, thereby providing evidence of speculation as a driver of excess co-movement. The limits of our analysis are good topics for future research. First, we consider, as in most factor-models in the literature, the factors as if they were data rather than being estimated. Even if this may have only a small effect on our results, it would be useful to investigate the small-sample case using simulation techniques as in Ludvigson and Ng (2007, 2009). Second, mixed-data sampling (MIDAS) regressions may be used to include more information at different frequencies. Tang and Xiong (2012) consider daily and monthly regressions, and MIDAS may help to combine the two data sources, with daily market indices and monthly or quarterly macroeconomic variables. This is the setting in Karali and Power (2013), who mix high- and low-frequency variables to explain the volatility of commodity returns. Such a setting may allow us to consider volatility spillovers, as in the penultimate section in Tang and Xiong (2012). The analysis of commodity volatility co-movement may have interesting implications for financial risk management. Third, alternative measures of trading activity, such as liquidity measures, may help better explain excess co-movement. In this respect, the recent contributions of Marshall, Nguyen, and Visaltanachoti (2012, 2013) may aid in the selection of appropriate liquidity measures for commodities and the evaluation of the explanatory power of their common liquidity factor. These measures may additionally be calculated on a daily basis, thereby permitting the high-frequency analysis of the common evolution of commodity prices. Appendix A Table AI. List of the 184 variables considered in the computation of the common factors In the Trans column, we report the transformation used to ensure the stationarity of each variable. ln denotes the logarithm, Δln  and Δ2ln denote the first and second difference of the logarithm, lv denotes the level of the series, and Δlv denotes the first difference of the series. Developed countries   Series number  Short name  Mnemonic  Trans  Description  Industrial production  1  IP: USA  USIPMAN.G  Δln   US INDUSTRIAL PRODUCTION—MANUFACTURING (NAICS) VOLA  2  IP: France  FRIPMAN.G  Δln   FR INDUSTRIAL PRODUCTION—MANUFACTURING VOLA  3  IP: France  FRINDSYNQ  lv  FR SURVEY: MANUFACTURING—SYNTHETIC BUSINESS INDICATOR SADJ  4  IP: Germany  BDIPMAN.G  Δln   BD INDUSTRIAL PRODUCTION: MANUFACTURING VOLA  5  IP: UK  UKIPMAN.G  Δln   UK INDUSTRIAL PRODUCTION INDEX—MANUFACTURING VOLA  6  IP: Japan  JPIPMAN.G  Δln   JP INDUSTRIAL PRODUCTION—MANUFACTURING VOLAJ  7  IP: Japan  JPIPTOT.G  Δln   JP INDUSTRIAL PRODUCTION—MINING & MANUFACTURING VOLA (2005 = 100)  Orders and capacity utilization  8  Capacity utilization: USA  USCUMANUG  Δlv  US CAPACITY UTILIZATION—MANUFACTURING VOLA  9  Manufct. new ord.: USA  USNOCOGMC  Δln   US MANUFACTURERS NEW ORDERS—CONSUMER GOODS & MATERIALS CONN (BASE 1982)  10  Manufct. new ord.: USA  USNOMXTRB  Δln   US NEW ORDERS—MANUFACTURING. EXCLUDING TRANSPORTATION CURA  11  New orders: Canada  CNNEWORDB  Δln   CN NEW ORDERS: ALL MANUFACTURING INDUSTRIES (SA) CURA  12  Manufct. ord.: Germany  BDNEWORDE  Δln   BD MANUFACTURING ORDERS SADJ (2000 = 100)  13  Manufct. ord.: Japan  JPNEWORDB  Δln   JP MACHINERY ORDERS: DOM.DEMAND–PRIVATE DEMAND (EXCL. SHIP) CURA  14  Operating ratio: Japan  JPCAPUTLQ  Δlv  JP OPERATING RATIO—MANUFACTURING SADJ (2005 = 100)  15  Business failures: Japan  JPBNKRPTP  Δln   JP BUSINESS FAILURES VOLN  Housing start  16  Housing permits: USA  USHOUSE.O  Δln   US NEW PRIVATE HOUSING UNITS STARTED (AR) VOLA  17  Housing permits: Canada  CNHOUSE.O  Δln   CN HOUSING STARTS: ALL AREAS (SA. AR) VOLA  18  Housing permits: Germany  BDHOUSE.G  Δln   BD CONSTRUCTION ORDERS RECEIVED—RESIDENTIAL CONSTRUCTION VOLA  19  Housing permits: Australia  AUHOUSE_A  Δln   AU BUILDING APPROVALS: NEW HOUSES CURN  20  Housing permits: Japan  JPHOUSSTF  ln   JP NEW HOUSING CONSTRUCTION STARTED VOLN  Car sales  21  Car registration: USA  USCAR…P  Δln   US NEW PASSENGER CARS—TOTAL REGISTRATIONS VOLN  22  Car registration: France  FRCARREGP  Δln   FR NEW CAR REGISTRATIONS VOLN  23  Car registration: Germany  BDRVNCARP  ln   BD NEW REGISTRATIONS—CARS VOLN  24  Car registration: UK  UKCAR…P  Δln   UK CAR REGISTRATIONS VOLN  25  Car registration: Japan  JPCARREGF  ln   JP MOTOR VEHICLE NEW REGISTRATIONS: PASSENGER CARS EXCL. BELOW 66  Consumption  26  Consumer sentiment: USA  USUMCONEH  Δln   US UNIV OF MICHIGAN CONSUMER SENTIMENT—EXPECTATIONS VOLN (BASE 1966 = 100)  27  Pers. cons. exp.: USA  USPERCONB  Δln   US PERSONAL CONSUMPTION EXPENDITURES (AR) CURA  28  Pers. saving: USA  USPERSAVE  Δlv  US PERSONAL SAVING AS % OF DISPOSABLE PERSONAL INCOME SADJ  29  Retail sale: Canada  CNRETTOTB  Δln   CN RETAIL SALES: TOTAL (ADJUSTED) CURA  30  Household confidence: France  FRCNFCONQ  Δlv  FR SURVEY—HOUSEHOLD CONFIDENCE INDICATOR SADJ  31  Household confidence: Germany  BDCNFCONQ  lv  BD CONSUMER CONFIDENCE INDICATOR—GERMANY SADJ  32  Retail sales: UK  UKRETTOTB  Δln   UK RETAIL SALES (MONTHLY ESTIMATE. DS CALCULATED) CURA  33  Household confidence: UK  UKCNFCONQ  Δlv  UK CONSUMER CONFIDENCE INDICATOR—UK SADJ  34  Retail sales: Australia  AURETTOTT  Δln   AU RETAIL SALES (TREND) VOLA  35  Household confidence: Australia  AUCNFCONR  lv  AU MELBOURNE/WESTPAC CONSUMER SENTIMENT INDEX NADJ  36  Household expenditure: Japan  JPHLEXPWA  Δln   JP WORKERS HOUSEHOLD LIVING EXPENDITURE (INCL. AFF) CURN  37  Retail sales: Japan  JPRETTOTA  Δln   JP RETAIL SALES CURN  Wages and labor  38  Av. Hourly Real Earnings: USA  USWRIM.D  Δln   US AVG HOURLY REAL EARNINGS—MANUFACTURING CONA (BASE 82–84)  39  Av. overtime hours: USA  USOOL024Q  Δlv  US OVERTIME HOURS—MANUFACTURING. WEEKLY VOLA  40  Av. wkly hours: USA  USHKIM_O  Δlv  US AVG WKLY HOURS—MANUFACTURING VOLA  41  Av. hourly real earnings: Canada  CNWAGES.A  Δln   CN AVG. HOURLY EARN—INDUSTRIAL AGGREGATE EXCL. UNCLASSIFIED CURN  42  Labor productivity: Germany  BDPRODVTQ  Δln   BD PRODUCTIVITY: OUTPUT PER MAN-HOUR WORKED IN INDUSTRY SADJ (2005 = 100)  43  wages: Germany  BDWAGES.F  Δln   BD WAGE & SALARY.OVERALL ECONOMY—ON A MTHLY BASIS (PAN BD M0191)  44  wages index: Japan  JPWAGES_E  Δln   JP WAGE INDEX: CASH EARNINGS—ALL INDUSTRIES SADJ  Unemployment  45  U rate: USA  USUNEM15Q  Δlv  US UNEMPLOYMENT RATE—15 WEEKS & OVER SADJ  46  U rate: USA  USUNTOTQ_pc  Δlv  US UNEMPLOYMENT RATE SADJ  47  Employment: Canada  CNEMPTOTO  Δln   CN EMPLOYMENT—CANADA (15 YRS & OVER SA) VOLA  48  U all: Germany  BDUNPTOTP  Δln   BD UNEMPLOYMENT LEVEL (PAN BD FROM SEPT. 1990) VOLN  49  U rate: UK  UKUNTOTQ_pc  Δlv  UK UNEMPLOYMENT RATE SADJ  50  Emp: Australia  AUEMPTOTO  Δln   AU EMPLOYED: PERSONS VOLA  51  U all: Australia  AUUNPTOTO  Δln   AU UNEMPLOYMENT LEVEL VOLA  52  U rate: Japan  JPUNTOTQ_pc  Δlv  JP UNEMPLOYMENT RATE SADJ  International trade  53  Exports: USA  USI70_A  Δln   US EXPORTS CURN  54  Exports: France  FREXPGDSB  Δln   FR EXPORTS FOB CURA  55  Exports: Germany  BDEXPBOPB  Δln   BD EXPORTS FOB CURA  56  Exports: UK  UKI70_A  Δln   UK EXPORTS CURN  57  Exports: Australia  AUEXPG&SB  Δln   AU EXPORTS OF GOODS & SERVICES (BOP BASIS) CURA  58  Exports: Japan  JPEXPGDSB  Δln   JP EXPORTS OF GOODS—CUSTOMSBASIS CURA  59  Imports: USA  USIMPGDSB  Δln   US IMPORTS F.A.S. CURA  60  Imports: France  FRIMPGDSB  Δln   FR IMPORTS FOB CURA  61  Imports: Germany  BDIMPGDSB  Δln   BD IMPORTS CIF (PAN BD M0790) CURA  62  Imports: UK  UKIMPBOPB  Δln   UK IMPORTS—BALANCE OF PAYMENTS BASIS CURA  63  Imports: Australia  AUIMPG&SB  Δln   AU IMPORTS OF GOODS & SERVICES (BOP BASIS) CURA  64  Imports: Japan  JPOXT009B  Δln   JP IMPORTS CURA  65  Terms of trade: UK  UKTOTPRCF  Δln   UK TERMS OF TRADE—EXPORT/IMPORT PRICES (BOP BASIS) NADJ  66  Terms of trade: Japan  JPTOTPRCF  Δln   JP TERMS OF TRADE INDEX NADJ  Money and credit  67  Money supply: USA  USM0_A  Δln   US MONETARY BASE CURA  68  Money supply: USA  USM2_B  Δln   US MONEY SUPPLY M2 CURA  69  Money supply: France  FRM2_A  Δln   FR MONEY SUPPLY—M2 (NATIONAL CONTRIBUTION TO M2) CURN  70  Money supply: France  FRM3_A  Δln   FR MONEY SUPPLY—M3 (NATIONAL CONTRIBUTION TO M3) CURN  71  Money supply: Germany  BDM1_A  Δln   BD MONEY SUPPLY—GERMAN CONTRIBUTION TO EURO M1 (PAN BD M0790)  72  Money supply: Germany  BDM3_B  Δln   BD MONEY SUPPLY—M3 (CONTRIBUTION TO EURO BASIS FROM M0195) CURA  73  Money supply: UK  UKM1_B  Δln   UK MONEY SUPPLY M1 (ESTIMATE OF EMU AGGREGATE FOR THE UK) CURA  74  Money supply: UK  UKM3_B  Δln   UK UK MONEY SUPPLY M3 (ESTIMATE OF EMU AGGREGATE FORTHE UK) CURA  75  Money supply: Australia  AUM1_B  Δln   AU MONEY SUPPLY—M1 CURA  76  Money supply: Australia  AUM3_B  Δln   AU MONEY SUPPLY - M3 (SEE AUM3…OB) CURA  77  Money supply: Japan  JPM1_A  Δln   JP MONEY SUPPLY: M1 (METHO-BREAK. APR. 2003) CURN  78  Money supply: Japan  JPM2_A  Δln   JP MONEY SUPPLY: M2 (METHO-BREAK. APR. 2003) CURN  79  Credit: USA  USCOMILND  Δln   US COMMERCIAL & INDUSTRIAL LOANS OUTSTANDING (BCI 101) CONA (BASE 2005)  80  Credit: USA  USCILNNCB  lv  US COMMERCIAL & INDL LOANS. NET CHANGE (AR) (BCI 112) CURA  81  Credit: USA  USCRDNRVB  Δln   US NONREVOLVING CONSUMER CREDIT OUTSTANDING CURA  82  Credit: France  FRBANKLPA  Δln   FR MFI LOANS TO RESIDENT PRIVATE SECTOR CURN  83  Credit: Germany  BDBANKLPA  Δ2ln   BD LENDING TO ENTERPRISES & INDIVIDUALS CURN  84  Credit: Australia  AUCRDCONB  Δ2ln   AU FINANCIAL INTERMEDIARIES: NARROW CREDIT—PRIVATE SECTOR CURA  85  Credit: Japan  JPBANKLPA  Δ2ln   JP AGGREGATE BANK LENDING (EXCL. SHINKIN BANKS) CURN  Stock index  86  Stock index: USA  USSHRPRCF  Δln   US DOW JONES INDUSTRIALS SHARE PRICE INDEX (EP) NADJ  87  Stock index: France  FRSHRPRCF  Δln   FR SHARE PRICE INDEX—SBF 250 NADJ  88  Stock index: Germany  BDSHRPRCF  Δln   BD DAX SHARE PRICE INDEX. EP NADJ  89  Stock index: UK  UKOSP001F  Δln   UK FTSE 100 SHARE PRICE INDEXNADJ (2005 = 100)  90  Stock index: Japan  JPSHRPRCF  Δln   JP TOKYO STOCK EXCHANGE—TOPIX (EP) NADJ (1968 = 100)  Interest rate  91  Interest rate: USA  USFEDFUN  Δlv  US FEDERAL FUNDS RATE (AVG.)  92  Interest rate: USA  USCRBBAA  Δlv  US CORPORATE BOND YIELD—MOODY’S BAA. SEASONED ISSUES  93  Interest rate: USA  USGBOND  Δlv  US TREASURY YIELD ADJUSTED TO CONSTANT MATURITY—20 YEAR  94  Interest rate: France  FRPRATE  Δlv  FR AVERAGE COST OF FUNDS FOR BANKS/EURO REPO RATE  95  Interest rate: France  FRGBOND  Δlv  FR GOVERNMENT GUARANTEED BOND YIELD (EP) NADJ  96  Interest rate: Germany  BDPRATE  Δlv  BD DISCOUNT RATE/SHORT TERM EURO REPO RATE  97  Interest rate: Germany  BDGBOND  Δlv  BD LONG TERM GOVERNMENT BOND YIELD—9–10 YEARS  98  Interest rate: UK  UKPRATE  Δlv  UK BANK OF ENGLAND BASE RATE (EP)  99  Interest rate: UK  UKGBOND  Δlv  UK GROSS REDEMPTION YIELD ON 20 YEAR GILTS (PERIOD AVERAGE) NADJ  100  Interest rate: Australia  AUPRATE  Δlv  AU RBA CASH RATE TARGET  101  Interest rate: Australia  AUBOND  Δlv  AU COMMONWEALTH GOVERNMENT BOND YIELD 10 YEAR (EP)  102  Interest rate: Japan  JPGBOND  Δlv  JP INTEREST-BEARING GOVERNMENT BONDS—10-YEAR (EP)  Exchange rate  103  Exchange rate: DM to US$  BBDEMSP  Δln   GERMAN MARK TO US$ (BBI)—EXCHANGE RATE  104  Exchange rate: SK to US$  SDXRUSD  Δln   SD SWEDISH KRONOR TO US$ (BBI. EP)  105  Exchange rate: to $  UKDOLLR  Δln   UK TO US$ (WMR)—EXCHANGE RATE  106  Exchange rate: Yen to $  JPXRUSD  Δln   JP JAPANESE YEN TO US$  107  Exchange rate: Aus.$ to US$  AUXRUSD  Δln   AU AUSTRALIAN $ TO US$ (MTH. AVG.)  Producer price index  108  PPI: USA  USPROPRCE  Δln   US PPI—FINISHED GOODS SADJ  109  PPI: Canada  CNPROPRCF  Δln   CN INDUSTRIAL PRICE INDEX: ALL COMMODITIES NADJ  110  PPI: Germany  BDPROPRCF  Δln   BD PPI: INDL. PRODUCTS. TOTAL. SOLD ON THE DOMESTIC MARKET NADJ (2005 = 100)  111  PPI: UK  UKPROPRCF  Δln   UK PPI—OUTPUT OF MANUFACTURED PRODUCTS (HOME SALES) NADJ  112  PPI: Japan  JPPROPRCF  Δln   JP CORPORATE GOODS PRICE INDEX: DOMESTIC—ALL COMMODITIES NADJ  Consumer price index  113  CPI: USA  USCONPRCE  Δln   US CPI—ALL URBAN: ALL ITEMS SADJ  114  CPI: Canada  CNCONPRCF  Δln   CN CPI NADJ  115  CPI: France  FRCONPRCE  Δln   FR CPI SADJ  116  CPI: Germany  BDCONPRCE  Δln   BD CPI SADJ  117  CPI: UK  UKCONPRCF  Δln   UK CPI INDEX 00: ALL ITEMS- ESTIMATED PRE-97 2005 = 100 NADJ  118  CPI: Japan  JPCONPRCF  Δln   JP CPI: NATIONAL MEASURE NADJ    Developed countries   Series number  Short name  Mnemonic  Trans  Description  Industrial production  1  IP: USA  USIPMAN.G  Δln   US INDUSTRIAL PRODUCTION—MANUFACTURING (NAICS) VOLA  2  IP: France  FRIPMAN.G  Δln   FR INDUSTRIAL PRODUCTION—MANUFACTURING VOLA  3  IP: France  FRINDSYNQ  lv  FR SURVEY: MANUFACTURING—SYNTHETIC BUSINESS INDICATOR SADJ  4  IP: Germany  BDIPMAN.G  Δln   BD INDUSTRIAL PRODUCTION: MANUFACTURING VOLA  5  IP: UK  UKIPMAN.G  Δln   UK INDUSTRIAL PRODUCTION INDEX—MANUFACTURING VOLA  6  IP: Japan  JPIPMAN.G  Δln   JP INDUSTRIAL PRODUCTION—MANUFACTURING VOLAJ  7  IP: Japan  JPIPTOT.G  Δln   JP INDUSTRIAL PRODUCTION—MINING & MANUFACTURING VOLA (2005 = 100)  Orders and capacity utilization  8  Capacity utilization: USA  USCUMANUG  Δlv  US CAPACITY UTILIZATION—MANUFACTURING VOLA  9  Manufct. new ord.: USA  USNOCOGMC  Δln   US MANUFACTURERS NEW ORDERS—CONSUMER GOODS & MATERIALS CONN (BASE 1982)  10  Manufct. new ord.: USA  USNOMXTRB  Δln   US NEW ORDERS—MANUFACTURING. EXCLUDING TRANSPORTATION CURA  11  New orders: Canada  CNNEWORDB  Δln   CN NEW ORDERS: ALL MANUFACTURING INDUSTRIES (SA) CURA  12  Manufct. ord.: Germany  BDNEWORDE  Δln   BD MANUFACTURING ORDERS SADJ (2000 = 100)  13  Manufct. ord.: Japan  JPNEWORDB  Δln   JP MACHINERY ORDERS: DOM.DEMAND–PRIVATE DEMAND (EXCL. SHIP) CURA  14  Operating ratio: Japan  JPCAPUTLQ  Δlv  JP OPERATING RATIO—MANUFACTURING SADJ (2005 = 100)  15  Business failures: Japan  JPBNKRPTP  Δln   JP BUSINESS FAILURES VOLN  Housing start  16  Housing permits: USA  USHOUSE.O  Δln   US NEW PRIVATE HOUSING UNITS STARTED (AR) VOLA  17  Housing permits: Canada  CNHOUSE.O  Δln   CN HOUSING STARTS: ALL AREAS (SA. AR) VOLA  18  Housing permits: Germany  BDHOUSE.G  Δln   BD CONSTRUCTION ORDERS RECEIVED—RESIDENTIAL CONSTRUCTION VOLA  19  Housing permits: Australia  AUHOUSE_A  Δln   AU BUILDING APPROVALS: NEW HOUSES CURN  20  Housing permits: Japan  JPHOUSSTF  ln   JP NEW HOUSING CONSTRUCTION STARTED VOLN  Car sales  21  Car registration: USA  USCAR…P  Δln   US NEW PASSENGER CARS—TOTAL REGISTRATIONS VOLN  22  Car registration: France  FRCARREGP  Δln   FR NEW CAR REGISTRATIONS VOLN  23  Car registration: Germany  BDRVNCARP  ln   BD NEW REGISTRATIONS—CARS VOLN  24  Car registration: UK  UKCAR…P  Δln   UK CAR REGISTRATIONS VOLN  25  Car registration: Japan  JPCARREGF  ln   JP MOTOR VEHICLE NEW REGISTRATIONS: PASSENGER CARS EXCL. BELOW 66  Consumption  26  Consumer sentiment: USA  USUMCONEH  Δln   US UNIV OF MICHIGAN CONSUMER SENTIMENT—EXPECTATIONS VOLN (BASE 1966 = 100)  27  Pers. cons. exp.: USA  USPERCONB  Δln   US PERSONAL CONSUMPTION EXPENDITURES (AR) CURA  28  Pers. saving: USA  USPERSAVE  Δlv  US PERSONAL SAVING AS % OF DISPOSABLE PERSONAL INCOME SADJ  29  Retail sale: Canada  CNRETTOTB  Δln   CN RETAIL SALES: TOTAL (ADJUSTED) CURA  30  Household confidence: France  FRCNFCONQ  Δlv  FR SURVEY—HOUSEHOLD CONFIDENCE INDICATOR SADJ  31  Household confidence: Germany  BDCNFCONQ  lv  BD CONSUMER CONFIDENCE INDICATOR—GERMANY SADJ  32  Retail sales: UK  UKRETTOTB  Δln   UK RETAIL SALES (MONTHLY ESTIMATE. DS CALCULATED) CURA  33  Household confidence: UK  UKCNFCONQ  Δlv  UK CONSUMER CONFIDENCE INDICATOR—UK SADJ  34  Retail sales: Australia  AURETTOTT  Δln   AU RETAIL SALES (TREND) VOLA  35  Household confidence: Australia  AUCNFCONR  lv  AU MELBOURNE/WESTPAC CONSUMER SENTIMENT INDEX NADJ  36  Household expenditure: Japan  JPHLEXPWA  Δln   JP WORKERS HOUSEHOLD LIVING EXPENDITURE (INCL. AFF) CURN  37  Retail sales: Japan  JPRETTOTA  Δln   JP RETAIL SALES CURN  Wages and labor  38  Av. Hourly Real Earnings: USA  USWRIM.D  Δln   US AVG HOURLY REAL EARNINGS—MANUFACTURING CONA (BASE 82–84)  39  Av. overtime hours: USA  USOOL024Q  Δlv  US OVERTIME HOURS—MANUFACTURING. WEEKLY VOLA  40  Av. wkly hours: USA  USHKIM_O  Δlv  US AVG WKLY HOURS—MANUFACTURING VOLA  41  Av. hourly real earnings: Canada  CNWAGES.A  Δln   CN AVG. HOURLY EARN—INDUSTRIAL AGGREGATE EXCL. UNCLASSIFIED CURN  42  Labor productivity: Germany  BDPRODVTQ  Δln   BD PRODUCTIVITY: OUTPUT PER MAN-HOUR WORKED IN INDUSTRY SADJ (2005 = 100)  43  wages: Germany  BDWAGES.F  Δln   BD WAGE & SALARY.OVERALL ECONOMY—ON A MTHLY BASIS (PAN BD M0191)  44  wages index: Japan  JPWAGES_E  Δln   JP WAGE INDEX: CASH EARNINGS—ALL INDUSTRIES SADJ  Unemployment  45  U rate: USA  USUNEM15Q  Δlv  US UNEMPLOYMENT RATE—15 WEEKS & OVER SADJ  46  U rate: USA  USUNTOTQ_pc  Δlv  US UNEMPLOYMENT RATE SADJ  47  Employment: Canada  CNEMPTOTO  Δln   CN EMPLOYMENT—CANADA (15 YRS & OVER SA) VOLA  48  U all: Germany  BDUNPTOTP  Δln   BD UNEMPLOYMENT LEVEL (PAN BD FROM SEPT. 1990) VOLN  49  U rate: UK  UKUNTOTQ_pc  Δlv  UK UNEMPLOYMENT RATE SADJ  50  Emp: Australia  AUEMPTOTO  Δln   AU EMPLOYED: PERSONS VOLA  51  U all: Australia  AUUNPTOTO  Δln   AU UNEMPLOYMENT LEVEL VOLA  52  U rate: Japan  JPUNTOTQ_pc  Δlv  JP UNEMPLOYMENT RATE SADJ  International trade  53  Exports: USA  USI70_A  Δln   US EXPORTS CURN  54  Exports: France  FREXPGDSB  Δln   FR EXPORTS FOB CURA  55  Exports: Germany  BDEXPBOPB  Δln   BD EXPORTS FOB CURA  56  Exports: UK  UKI70_A  Δln   UK EXPORTS CURN  57  Exports: Australia  AUEXPG&SB  Δln   AU EXPORTS OF GOODS & SERVICES (BOP BASIS) CURA  58  Exports: Japan  JPEXPGDSB  Δln   JP EXPORTS OF GOODS—CUSTOMSBASIS CURA  59  Imports: USA  USIMPGDSB  Δln   US IMPORTS F.A.S. CURA  60  Imports: France  FRIMPGDSB  Δln   FR IMPORTS FOB CURA  61  Imports: Germany  BDIMPGDSB  Δln   BD IMPORTS CIF (PAN BD M0790) CURA  62  Imports: UK  UKIMPBOPB  Δln   UK IMPORTS—BALANCE OF PAYMENTS BASIS CURA  63  Imports: Australia  AUIMPG&SB  Δln   AU IMPORTS OF GOODS & SERVICES (BOP BASIS) CURA  64  Imports: Japan  JPOXT009B  Δln   JP IMPORTS CURA  65  Terms of trade: UK  UKTOTPRCF  Δln   UK TERMS OF TRADE—EXPORT/IMPORT PRICES (BOP BASIS) NADJ  66  Terms of trade: Japan  JPTOTPRCF  Δln   JP TERMS OF TRADE INDEX NADJ  Money and credit  67  Money supply: USA  USM0_A  Δln   US MONETARY BASE CURA  68  Money supply: USA  USM2_B  Δln   US MONEY SUPPLY M2 CURA  69  Money supply: France  FRM2_A  Δln   FR MONEY SUPPLY—M2 (NATIONAL CONTRIBUTION TO M2) CURN  70  Money supply: France  FRM3_A  Δln   FR MONEY SUPPLY—M3 (NATIONAL CONTRIBUTION TO M3) CURN  71  Money supply: Germany  BDM1_A  Δln   BD MONEY SUPPLY—GERMAN CONTRIBUTION TO EURO M1 (PAN BD M0790)  72  Money supply: Germany  BDM3_B  Δln   BD MONEY SUPPLY—M3 (CONTRIBUTION TO EURO BASIS FROM M0195) CURA  73  Money supply: UK  UKM1_B  Δln   UK MONEY SUPPLY M1 (ESTIMATE OF EMU AGGREGATE FOR THE UK) CURA  74  Money supply: UK  UKM3_B  Δln   UK UK MONEY SUPPLY M3 (ESTIMATE OF EMU AGGREGATE FORTHE UK) CURA  75  Money supply: Australia  AUM1_B  Δln   AU MONEY SUPPLY—M1 CURA  76  Money supply: Australia  AUM3_B  Δln   AU MONEY SUPPLY - M3 (SEE AUM3…OB) CURA  77  Money supply: Japan  JPM1_A  Δln   JP MONEY SUPPLY: M1 (METHO-BREAK. APR. 2003) CURN  78  Money supply: Japan  JPM2_A  Δln   JP MONEY SUPPLY: M2 (METHO-BREAK. APR. 2003) CURN  79  Credit: USA  USCOMILND  Δln   US COMMERCIAL & INDUSTRIAL LOANS OUTSTANDING (BCI 101) CONA (BASE 2005)  80  Credit: USA  USCILNNCB  lv  US COMMERCIAL & INDL LOANS. NET CHANGE (AR) (BCI 112) CURA  81  Credit: USA  USCRDNRVB  Δln   US NONREVOLVING CONSUMER CREDIT OUTSTANDING CURA  82  Credit: France  FRBANKLPA  Δln   FR MFI LOANS TO RESIDENT PRIVATE SECTOR CURN  83  Credit: Germany  BDBANKLPA  Δ2ln   BD LENDING TO ENTERPRISES & INDIVIDUALS CURN  84  Credit: Australia  AUCRDCONB  Δ2ln   AU FINANCIAL INTERMEDIARIES: NARROW CREDIT—PRIVATE SECTOR CURA  85  Credit: Japan  JPBANKLPA  Δ2ln   JP AGGREGATE BANK LENDING (EXCL. SHINKIN BANKS) CURN  Stock index  86  Stock index: USA  USSHRPRCF  Δln   US DOW JONES INDUSTRIALS SHARE PRICE INDEX (EP) NADJ  87  Stock index: France  FRSHRPRCF  Δln   FR SHARE PRICE INDEX—SBF 250 NADJ  88  Stock index: Germany  BDSHRPRCF  Δln   BD DAX SHARE PRICE INDEX. EP NADJ  89  Stock index: UK  UKOSP001F  Δln   UK FTSE 100 SHARE PRICE INDEXNADJ (2005 = 100)  90  Stock index: Japan  JPSHRPRCF  Δln   JP TOKYO STOCK EXCHANGE—TOPIX (EP) NADJ (1968 = 100)  Interest rate  91  Interest rate: USA  USFEDFUN  Δlv  US FEDERAL FUNDS RATE (AVG.)  92  Interest rate: USA  USCRBBAA  Δlv  US CORPORATE BOND YIELD—MOODY’S BAA. SEASONED ISSUES  93  Interest rate: USA  USGBOND  Δlv  US TREASURY YIELD ADJUSTED TO CONSTANT MATURITY—20 YEAR  94  Interest rate: France  FRPRATE  Δlv  FR AVERAGE COST OF FUNDS FOR BANKS/EURO REPO RATE  95  Interest rate: France  FRGBOND  Δlv  FR GOVERNMENT GUARANTEED BOND YIELD (EP) NADJ  96  Interest rate: Germany  BDPRATE  Δlv  BD DISCOUNT RATE/SHORT TERM EURO REPO RATE  97  Interest rate: Germany  BDGBOND  Δlv  BD LONG TERM GOVERNMENT BOND YIELD—9–10 YEARS  98  Interest rate: UK  UKPRATE  Δlv  UK BANK OF ENGLAND BASE RATE (EP)  99  Interest rate: UK  UKGBOND  Δlv  UK GROSS REDEMPTION YIELD ON 20 YEAR GILTS (PERIOD AVERAGE) NADJ  100  Interest rate: Australia  AUPRATE  Δlv  AU RBA CASH RATE TARGET  101  Interest rate: Australia  AUBOND  Δlv  AU COMMONWEALTH GOVERNMENT BOND YIELD 10 YEAR (EP)  102  Interest rate: Japan  JPGBOND  Δlv  JP INTEREST-BEARING GOVERNMENT BONDS—10-YEAR (EP)  Exchange rate  103  Exchange rate: DM to US$  BBDEMSP  Δln   GERMAN MARK TO US$ (BBI)—EXCHANGE RATE  104  Exchange rate: SK to US$  SDXRUSD  Δln   SD SWEDISH KRONOR TO US$ (BBI. EP)  105  Exchange rate: to $  UKDOLLR  Δln   UK TO US$ (WMR)—EXCHANGE RATE  106  Exchange rate: Yen to $  JPXRUSD  Δln   JP JAPANESE YEN TO US$  107  Exchange rate: Aus.$ to US$  AUXRUSD  Δln   AU AUSTRALIAN $ TO US$ (MTH. AVG.)  Producer price index  108  PPI: USA  USPROPRCE  Δln   US PPI—FINISHED GOODS SADJ  109  PPI: Canada  CNPROPRCF  Δln   CN INDUSTRIAL PRICE INDEX: ALL COMMODITIES NADJ  110  PPI: Germany  BDPROPRCF  Δln   BD PPI: INDL. PRODUCTS. TOTAL. SOLD ON THE DOMESTIC MARKET NADJ (2005 = 100)  111  PPI: UK  UKPROPRCF  Δln   UK PPI—OUTPUT OF MANUFACTURED PRODUCTS (HOME SALES) NADJ  112  PPI: Japan  JPPROPRCF  Δln   JP CORPORATE GOODS PRICE INDEX: DOMESTIC—ALL COMMODITIES NADJ  Consumer price index  113  CPI: USA  USCONPRCE  Δln   US CPI—ALL URBAN: ALL ITEMS SADJ  114  CPI: Canada  CNCONPRCF  Δln   CN CPI NADJ  115  CPI: France  FRCONPRCE  Δln   FR CPI SADJ  116  CPI: Germany  BDCONPRCE  Δln   BD CPI SADJ  117  CPI: UK  UKCONPRCF  Δln   UK CPI INDEX 00: ALL ITEMS- ESTIMATED PRE-97 2005 = 100 NADJ  118  CPI: Japan  JPCONPRCF  Δln   JP CPI: NATIONAL MEASURE NADJ    Emerging countries           Series number  Short name  Mnemonic  Trans  Description  Industrial production  119  IP: Argentina  AGIPTOT.G  Δln   AG INDUSTRIAL PRODUCTION VOLA  120  IP: Chile  CLIPMAN.H  Δln   CL INDUSTRIAL PRODUCTION INDEX BY INE—MANUFACTURING TOTAL VOLN  121  IP: Brazil  BRIPTOT_G  Δln   BR INDUSTRIAL PRODUCTION VOLA index 2002 = BASE  122  IP: Brazil  BRIPMAN.G  Δln   BR INDUSTRIAL PRODUCTION—MANUFACTURING VOLA  123  IP: China  CHPBRENTP (electricity)  Δln   CH INDUSTRIAL PRODUCTION: ELECTRICITY VOLN  124  IP: Korea  KOIPTOT.G  Δln   KO INDUSTRIAL PRODUCTION VOLA (2005 = 100)  125  IP: Korea  KOIPMAN.G  Δln   KO MANUFACTURING PRODUCTION INDEX VOLA  126  IP: Mexico  MXIPTOT_H  Δln   MX INDUSTRIAL PRODUCTION INDEX VOLN  127  IP: Philippines  PHIPMAN_F  Δln   PH MANUFACTURING PRODUCTION NADJ  128  IP: South Africa  SAIPMAN.G  Δln   SA INDUSTRIAL PRODUCTION (MANUFACTURING SECTOR) VOLA  129  IP: Taiwan  TWIPMAN.H  Δln   TW INDUSTRIAL PRODUCTION INDEX—MANUFACTURING VOLN  Orders and capacity utilization  130  Operating ratio: Brazil  BRCAPUTLR  Δlv  BR CAPACITY UTILIZATION—MANUFACTURING NADJ  131  Mach. ord.: Korea  KONEWORDA  Δln   KO MACHINERY ORDERS RECEIVEDCURN  132  Manufct. prod capa.: Korea  KOCAPUTLF  Δlv  KO MANUFACTURING PRODUCTION CAPACITY NADJ (2005 = 100)  Consumption  133  Car sales: Argentina  AGCARSLSP  Δln   AG SALES—NATIONAL CARS TO DEALERS VOLN  134  Retail sales: Chile  CLRETTOTH  Δln   CL RETAIL SALES AT SUPERMARKETS (REAL INDEX) VOLN  135  Gasoline consumption: Korea  KOOPCGSLP  Δln   KO OIL PRODUCTS CONSUMPTION—GASOLINE VOLN  136  Retail sales: Singapore  SPRETTOTG  Δlv  SP RETAIL SALES INDEX (CONSTANT) VOLA  137  Retail sales: Russia  RSRETTOTA  Δln   RS RETAIL TRADE TURNOVER—TOTAL CURN  Wages and labor  138  Labor cost: Brazil  BRLCOST.F  Δln   BR UNIT LABOR COST NADJ  Unemployment  139  Unemployment: Hong Kong  HKUNPTOTP  Δln   HK UNEMPLOYMENT (3 MONTHS ENDING) VOLN  140  U rate: Taiwan  TWUN%TOTQ  Δlv  TW UNEMPLOYMENT RATE SADJ  141  Unemployment: Russia  RSUNPTOTP  Δln   RS ECONOMICALLY ACTIVE POPULATION—UNEMPLOYED VOLN  International trade  142  Exports: Brazil  BREXPBOPA  Δln   BR BOP: CURRENT ACCOUNT—GOODS (CREDIT) CURN  143  Exports: China  CHEXPGDSA  Δln   CH EXPORTS CURN  144  Exports: India  INEXPGDSA  Δln   IN EXPORTS FOB CURN  145  Exports: Indonesia  IDEXPGDSA  Δln   ID EXPORTS FOB CURN  146  Exports: Korea  KOEXPGDSA  Δln   KO EXPORTS FOB (CUSTOMS CLEARANCE BASIS) CURN  147  Exports: Philippines  PHEXPGDSA  Δln   PH EXPORTS CURN  148  Exports: Singapore  SPEXPGDSA  Δln   SP EXPORTS CURN  149  Exports: Taiwan  TWEXPGDSA  Δln   TW EXPORTS CURN  150  Imports: Brazil  BRIMPBOPA  Δln   BR BOP: CURRENT ACCOUNT—GOODS (DEBIT) CURN  151  Imports: China  CHIMPGDSA  Δln   CH IMPORTS CURN  152  Imports: Korea  KOIMPGDSA  Δln   KO IMPORTS CIF (CUSTOMS CLEARANCE BASIS) CURN  153  Imports: Singapore  SPIMPGDSA  Δln   SP IMPORTS CURN  154  Imports: Taiwan  TWIMPGDSA  Δln   TW IMPORTS CURN  155  Terms of trade: Brazil  BRTOTPRCF  Δln   BR TERMS OF TRADE NADJ (2006 = 100)  Money and credit  156  Money supply: Brazil  BRM1_A  Δln   BR MONEY SUPPLY—M1 (EP) CURN  157  Money supply: Brazil  BRM3_A  Δln   BR MONEY SUPPLY—M3 (EP) CURN  158  Money supply: China  CHM0_A  Δln   CH MONEY SUPPLY—CURRENCY IN CIRCULATION CURN  159  Money supply: China  CHM1_A  Δln   CH MONEY SUPPLY—M1 CURN  160  Money supply: India  INM1_A  Δ2ln   IN MONEY SUPPLY: M1 (EP) CURN  161  Money supply: India  INM3_A  Δ2ln   IN MONEY SUPPLY: M3 (EP) CURN  162  Money supply: Indonesia  IDM1_A  Δln   ID MONEY SUPPLY: M1 CURN  163  Money supply: Indonesia  IDM2_A  Δln   ID MONEY SUPPLY—M2 CURN  164  Money supply: Korea  KOM2_B  Δln   KO MONEY SUPPLY—M2 (EP) CURA  165  Money supply: Mexico  MXM1_A  Δln   MX MONEY SUPPLY: M1 (EP) CURN BASE = END OF PERIOD  166  Money supply: Mexico  MXM3_A  Δ2ln   MX MONEY SUPPLY: M3 (EP) CURN  167  Money supply: Philippines  PHM1_A  Δln   PH MONEY SUPPLY—M1 (METHO BREAK AT 12/03) CURN  168  Money supply: Philippines  PHM3_A  Δln   PH MONEY SUPPLY—M3 (METHO BREAK AT 12/03) CURN  169  Money supply: Russia  RSM 0_A  Δln   RS MONEY SUPPLY—M0 CURN  Stock index  170  Stock index: Brazil  BRSHRPRCF  Δln   BR BOVESPA SHARE PRICE INDEX (EP) NADJ  171  Stock index: Hong Kong  HKSHRPRCF  Δln   HK HANG SENG SHARE PRICE INDEX (EP) NADJ (31 JULY 1964 =100)  Exchange rate  172  Exchange rate: Br.R. to US$  BRXRUSD  Δln   BR BRAZILIAN REAIS TO US$ (AVG)  173  Exchange rate: Ch.Y. to US$  CHXRUSD  Δln   CH CHINESE YUAN TO US$ (AVERAGE AMOUNT)  174  Exchange rate: In.R. to US$  INXRUSD  Δln   IN INDIAN RUPEES PER US$ (RBI)  175  Exchange rate: Id.R. to US$  IDXRUSD  Δln   ID INDONESIAN RUPIAHS TO US$  176  Exchange rate: Mx.P. to US$  MXXRUSD  Δln   MX MEXICAN PESOS TO US$—CENTRAL BANK SETTLEMENT RATE (AVG)  177  Exchange rate: Rs.R. to US$  RSXRUSD  Δln   RS RUSSIAN ROUBLES TO US$ NADJ  Consumer price index  178  CPI: Brazil  BRCPIGENF  Δln   BR CPI—GENERAL NADJ  179  CPI: China  CHCONPRCF  Δln   CH CPI NADJ  180  CPI: India  INCONPRCF  Δln   IN CPI: INDUSTRIAL LABOURERS(DS CALCULATED) NADJ (2001 = 100)  181  CPI: Korea  KOCONPRCF  Δln   KO CPI NADJ (2005 = 100)  182  CPI: Mexico  MXCONPRCF  Δln   MX CPI NADJ (JUN 2002 = 100)  183  CPI: Philippines  PHCONPRCF  Δln   PH CPI NADJ  184  CPI: Russia  RSCONPRCF  Δln   RS CPI NADJ    Emerging countries           Series number  Short name  Mnemonic  Trans  Description  Industrial production  119  IP: Argentina  AGIPTOT.G  Δln   AG INDUSTRIAL PRODUCTION VOLA  120  IP: Chile  CLIPMAN.H  Δln   CL INDUSTRIAL PRODUCTION INDEX BY INE—MANUFACTURING TOTAL VOLN  121  IP: Brazil  BRIPTOT_G  Δln   BR INDUSTRIAL PRODUCTION VOLA index 2002 = BASE  122  IP: Brazil  BRIPMAN.G  Δln   BR INDUSTRIAL PRODUCTION—MANUFACTURING VOLA  123  IP: China  CHPBRENTP (electricity)  Δln   CH INDUSTRIAL PRODUCTION: ELECTRICITY VOLN  124  IP: Korea  KOIPTOT.G  Δln   KO INDUSTRIAL PRODUCTION VOLA (2005 = 100)  125  IP: Korea  KOIPMAN.G  Δln   KO MANUFACTURING PRODUCTION INDEX VOLA  126  IP: Mexico  MXIPTOT_H  Δln   MX INDUSTRIAL PRODUCTION INDEX VOLN  127  IP: Philippines  PHIPMAN_F  Δln   PH MANUFACTURING PRODUCTION NADJ  128  IP: South Africa  SAIPMAN.G  Δln   SA INDUSTRIAL PRODUCTION (MANUFACTURING SECTOR) VOLA  129  IP: Taiwan  TWIPMAN.H  Δln   TW INDUSTRIAL PRODUCTION INDEX—MANUFACTURING VOLN  Orders and capacity utilization  130  Operating ratio: Brazil  BRCAPUTLR  Δlv  BR CAPACITY UTILIZATION—MANUFACTURING NADJ  131  Mach. ord.: Korea  KONEWORDA  Δln   KO MACHINERY ORDERS RECEIVEDCURN  132  Manufct. prod capa.: Korea  KOCAPUTLF  Δlv  KO MANUFACTURING PRODUCTION CAPACITY NADJ (2005 = 100)  Consumption  133  Car sales: Argentina  AGCARSLSP  Δln   AG SALES—NATIONAL CARS TO DEALERS VOLN  134  Retail sales: Chile  CLRETTOTH  Δln   CL RETAIL SALES AT SUPERMARKETS (REAL INDEX) VOLN  135  Gasoline consumption: Korea  KOOPCGSLP  Δln   KO OIL PRODUCTS CONSUMPTION—GASOLINE VOLN  136  Retail sales: Singapore  SPRETTOTG  Δlv  SP RETAIL SALES INDEX (CONSTANT) VOLA  137  Retail sales: Russia  RSRETTOTA  Δln   RS RETAIL TRADE TURNOVER—TOTAL CURN  Wages and labor  138  Labor cost: Brazil  BRLCOST.F  Δln   BR UNIT LABOR COST NADJ  Unemployment  139  Unemployment: Hong Kong  HKUNPTOTP  Δln   HK UNEMPLOYMENT (3 MONTHS ENDING) VOLN  140  U rate: Taiwan  TWUN%TOTQ  Δlv  TW UNEMPLOYMENT RATE SADJ  141  Unemployment: Russia  RSUNPTOTP  Δln   RS ECONOMICALLY ACTIVE POPULATION—UNEMPLOYED VOLN  International trade  142  Exports: Brazil  BREXPBOPA  Δln   BR BOP: CURRENT ACCOUNT—GOODS (CREDIT) CURN  143  Exports: China  CHEXPGDSA  Δln   CH EXPORTS CURN  144  Exports: India  INEXPGDSA  Δln   IN EXPORTS FOB CURN  145  Exports: Indonesia  IDEXPGDSA  Δln   ID EXPORTS FOB CURN  146  Exports: Korea  KOEXPGDSA  Δln   KO EXPORTS FOB (CUSTOMS CLEARANCE BASIS) CURN  147  Exports: Philippines  PHEXPGDSA  Δln   PH EXPORTS CURN  148  Exports: Singapore  SPEXPGDSA  Δln   SP EXPORTS CURN  149  Exports: Taiwan  TWEXPGDSA  Δln   TW EXPORTS CURN  150  Imports: Brazil  BRIMPBOPA  Δln   BR BOP: CURRENT ACCOUNT—GOODS (DEBIT) CURN  151  Imports: China  CHIMPGDSA  Δln   CH IMPORTS CURN  152  Imports: Korea  KOIMPGDSA  Δln   KO IMPORTS CIF (CUSTOMS CLEARANCE BASIS) CURN  153  Imports: Singapore  SPIMPGDSA  Δln   SP IMPORTS CURN  154  Imports: Taiwan  TWIMPGDSA  Δln   TW IMPORTS CURN  155  Terms of trade: Brazil  BRTOTPRCF  Δln   BR TERMS OF TRADE NADJ (2006 = 100)  Money and credit  156  Money supply: Brazil  BRM1_A  Δln   BR MONEY SUPPLY—M1 (EP) CURN  157  Money supply: Brazil  BRM3_A  Δln   BR MONEY SUPPLY—M3 (EP) CURN  158  Money supply: China  CHM0_A  Δln   CH MONEY SUPPLY—CURRENCY IN CIRCULATION CURN  159  Money supply: China  CHM1_A  Δln   CH MONEY SUPPLY—M1 CURN  160  Money supply: India  INM1_A  Δ2ln   IN MONEY SUPPLY: M1 (EP) CURN  161  Money supply: India  INM3_A  Δ2ln   IN MONEY SUPPLY: M3 (EP) CURN  162  Money supply: Indonesia  IDM1_A  Δln   ID MONEY SUPPLY: M1 CURN  163  Money supply: Indonesia  IDM2_A  Δln   ID MONEY SUPPLY—M2 CURN  164  Money supply: Korea  KOM2_B  Δln   KO MONEY SUPPLY—M2 (EP) CURA  165  Money supply: Mexico  MXM1_A  Δln   MX MONEY SUPPLY: M1 (EP) CURN BASE = END OF PERIOD  166  Money supply: Mexico  MXM3_A  Δ2ln   MX MONEY SUPPLY: M3 (EP) CURN  167  Money supply: Philippines  PHM1_A  Δln   PH MONEY SUPPLY—M1 (METHO BREAK AT 12/03) CURN  168  Money supply: Philippines  PHM3_A  Δln   PH MONEY SUPPLY—M3 (METHO BREAK AT 12/03) CURN  169  Money supply: Russia  RSM 0_A  Δln   RS MONEY SUPPLY—M0 CURN  Stock index  170  Stock index: Brazil  BRSHRPRCF  Δln   BR BOVESPA SHARE PRICE INDEX (EP) NADJ  171  Stock index: Hong Kong  HKSHRPRCF  Δln   HK HANG SENG SHARE PRICE INDEX (EP) NADJ (31 JULY 1964 =100)  Exchange rate  172  Exchange rate: Br.R. to US$  BRXRUSD  Δln   BR BRAZILIAN REAIS TO US$ (AVG)  173  Exchange rate: Ch.Y. to US$  CHXRUSD  Δln   CH CHINESE YUAN TO US$ (AVERAGE AMOUNT)  174  Exchange rate: In.R. to US$  INXRUSD  Δln   IN INDIAN RUPEES PER US$ (RBI)  175  Exchange rate: Id.R. to US$  IDXRUSD  Δln   ID INDONESIAN RUPIAHS TO US$  176  Exchange rate: Mx.P. to US$  MXXRUSD  Δln   MX MEXICAN PESOS TO US$—CENTRAL BANK SETTLEMENT RATE (AVG)  177  Exchange rate: Rs.R. to US$  RSXRUSD  Δln   RS RUSSIAN ROUBLES TO US$ NADJ  Consumer price index  178  CPI: Brazil  BRCPIGENF  Δln   BR CPI—GENERAL NADJ  179  CPI: China  CHCONPRCF  Δln   CH CPI NADJ  180  CPI: India  INCONPRCF  Δln   IN CPI: INDUSTRIAL LABOURERS(DS CALCULATED) NADJ (2001 = 100)  181  CPI: Korea  KOCONPRCF  Δln   KO CPI NADJ (2005 = 100)  182  CPI: Mexico  MXCONPRCF  Δln   MX CPI NADJ (JUN 2002 = 100)  183  CPI: Philippines  PHCONPRCF  Δln   PH CPI NADJ  184  CPI: Russia  RSCONPRCF  Δln   RS CPI NADJ    Appendix B: Estimating the Number of Factors Bai and Ng (2002) propose to select the number of common factors which minimize the following information criteria:   PCPi(k)=S(k)+kσ¯2gi(N.T),  ICi(k)=ln(S(k))+kgi(N.T), where k is the number of factors, S(k)=(NT)−1∑i=1N∑t=1T(xit−λ^ik′F^tk)2 is the sum of squared residuals (divided by NT), g(N.T) is a penalty function,32 and σ¯2 equals S(kmax) for a pre-specified value of kmax. The optimal number of factors k^ minimizes these information criteria. Kapetanios (2010) proposes a sequential test to determine the number of factors. When the true number of factors is k0, under some regularity condition, the first k0 eigenvalues of the population covariance matrix Σ increase at rate N while the others are bounded. Let’s note λ^k, k=1,…,N, the N eigenvalues (in decreasing order) of the sample covariance matrix X′X and kmax a finite number such that k0<kmax. The difference λ^k−λ^kmax+1 will go to infinity for k=1,…,k0, but is bounded for k=k0+1,⋯,kmax. λ^k−λ^kmax+1 is then used as a the test statistics to discriminate the null hypothesis that the true number of factors k0 equals k ( H0.k:k0=k) against the alternative hypothesis ( H1.k:k0>k). When there is no factor structure, λ^k−λ^kmax+1, appropriately normalized, converge to a law limit, but tend to infinity in the presence of factors. We begin by testing ( H0.k:k0=k=0) against ( H1.k:k0>0). If we reject the null hypothesis, then we consider the null ( H0.k:k0=k+1=1). We stop once we cannot reject the null hypothesis. Kapetanios (2010) called this algorithm the maximal eigenvalue distribution (MED) algorithm. As shown in Table BI, there is no agreement on the optimal number of factors.33Bai and Ng (2002) information criteria select between two and nine factors while Kapetanios (2010) sequential test suggests two. Previous empirical work also reveals considerable variance in estimates of the correct number of factors.34 The factors autocorrelation of factors F^t are displayed in Table BII. They show that most factors are persistent. Statistics on their explanatory power reveals that only 20% of the variance in the 184 time series is explained by the first three factors. This figure is equal to 36% for the first nine factors, which leads us to keep the first nine factors as potential regressors for modeling commodity returns. Table BI. Static factor selection results MED denotes the number of factors given by the maximum eigenvalue distribution algorithm. ICi and PCPi denote, respectively, the number of factors given by the information criteria IC and PCP estimated with the penalty function gi(N.T). Method  Number of static factors  MED  2  IC1  4  IC2  3  IC3  12  IC4  20  PCP1  9  PCP2  8  PCP3  18  PCP4  20  Method  Number of static factors  MED  2  IC1  4  IC2  3  IC3  12  IC4  20  PCP1  9  PCP2  8  PCP3  18  PCP4  20  Table BII. Summary statistics for estimated static factors F^t.i for i=1,…,9 For i = 1, … , 9. F^it is estimated by the method of principal components using a panel of data with 184 indicators of economic activity from 1993:03 to 2010:03 (205 time-series observations). The data are transformed (taking logs and differenced where appropriate) and standardized prior to estimation. ρi denotes the ith autocorrelation. The 95% confidence bounds are ±0.1397. The relative importance of the common component. Ri2 is calculated as the fraction of total variance in the data explained by factors 1 to i. Factor i  ρ1  ρ2  ρ3  Ri2  1  0.1614  0.1256  0.3176  0.0930  2  0.1357  0.0805  0.3110  0.1623  3  −0.0748  0.0145  −0.0294  0.2066  4  −0.0285  −0.0694  0.1866  0.2424  5  −0.1439  −0.0966  0.0950  0.2740  6  0.2546  0.0328  −0.0091  0.3035  7  0.1012  0.3234  0.3844  0.3288  8  0.3405  0.4066  0.1768  0.3518  9  −0.0065  −0.0413  −0.1447  0.3739  Factor i  ρ1  ρ2  ρ3  Ri2  1  0.1614  0.1256  0.3176  0.0930  2  0.1357  0.0805  0.3110  0.1623  3  −0.0748  0.0145  −0.0294  0.2066  4  −0.0285  −0.0694  0.1866  0.2424  5  −0.1439  −0.0966  0.0950  0.2740  6  0.2546  0.0328  −0.0091  0.3035  7  0.1012  0.3234  0.3844  0.3288  8  0.3405  0.4066  0.1768  0.3518  9  −0.0065  −0.0413  −0.1447  0.3739  Footnotes 1 Investment in commodity markets from a portfolio perspective is discussed in Gorton and Rouwenhorst (2006); Erb and Harvey (2006); Rouwenhorst and Tang (2012); and Gorton, Hayashi, and Rouwenhorst (2013), among many others. 2 The early contribution by Gilbert (1989) emphasizes the relevance of the exchange rate as an explanatory variable for commodity prices; see also the recent papers by Chen et al. (2010) and Ferraro, Rogoff, and Rossi (2015). 3 The same variables are used in Deb, Trivedi, and Varangis (1996). Leybourne, Lloyd, and Reed (1994) further discuss the issue of omitted variables. 4 Recent economic research on the determination of commodity prices occasionally makes use of factor models. Examples of this growing literature are Byrne, Fazio, and Fiess (2013); Gospodinov and Ng (2013); West and Wong (2014); and Christoffersen, Lunde, and Olesen (2014). While these papers investigate more or less directly the issue of the co-movement of commodity prices, they all extract principal components from a set of commodity prices to explain the evolution of commodity prices, only considering a few additional macroeconomic variables—such as interest rates, exchange rates, and inflation—to analyze the link between these variables and their estimated factors. As such, their approach is very different from ours. 5 As will be made clear in the empirical sections, we adopt a measure of excess co-movement that is similar to that used in Kallberg and Pasquariello (2008), in that we consider the average of the squared residual correlations between all pairs of commodities. We hence allow both positive and negative correlations to contribute to the excess co-movement estimate. 6 As shown by Forbes and Rigobon (2002), the usual sample correlation is a biased measure of the true correlation when volatility is time-varying, which is a well-known stylized fact regarding financial series. As most of our commodity returns are characterized by time-varying volatility, we use the correlation coefficient corrected for heteroscedasticity of Forbes and Rigobon (2002). 7 Interestingly, Barberis and Shleifer (2003) cite the empirical results in PR as exemplifying their model. Conversely, PR observe that: “[…] traders are alternatively bullish or bearish on all commodities for no plausible reason” (p. 1173), which is behind our basic idea to measure the impact of speculation in commodity-futures markets on commodity price co-movements. 8 Their research question builds on Barberis, Shleifer, and Wurgler (2005), who theoretically and empirically analyze the behavior of newly included stocks in a stock index. It is shown that the price co-movement between the stock and the index significantly increases after this inclusion. 9 Part of the existing literature (Palaskas and Varangis, 1991; Leybourne, Lloyd, and Reed, 1994) considers excess co-movement of nominal or real price rather than return, and relies on a co-integration analysis. We think that return is more appealing when dealing with risk management issues, and thus consider the excess co-movement of returns as in the seminal work of PR. Returns rather than prices have also been considered more recently in Ai, Chatrath, and Song (2006) and Malliaris and Urritia (1996) for the main agricultural commodities. 10 Each variable is rendered stationary in an appropriate manner: the chosen transformation appears in the penultimate column of the table in Appendix A. 11 Although Stock and Watson (2002a) use different sets of assumptions to characterize “weak correlations”, the main idea is that the cross-correlations and serial correlations have an upper bound. 12 As the factors Ft and the loading matrix Λ are not separately identifiable [see Bai and Ng (2008) for more details], constraints are imposed to obtain a unique estimate. 13 When N > T, a computationally simpler approach is to use the T × T matrix XX′. 14 Methods based on information criteria and Kapetanios (2010) are described in Appendix B. 15 To further improve the explanatory power, we also considered potential nonlinearities with quadratic or cubic factors. We choose the specification with the highest adjusted R¯2. The set of factors is now F¯tnl=(F^1,t,F^2,t,F^3,t,F^4,t,F^2,t3,F^4,t3)′ and the regressions become: rit=ωi+∑k=14γikF^k,t+ωi,5F^2,t3+ωi,6F^4,t3+uiti=1,…,8 t=1,…,T=ωi+γi′F¯tnl+uit. The best specifications results are not shown here (but are available upon request). We do not find any notable increase in the R2 for any commodity. We therefore retain linear factors in the returns equation. 16 The classification in Ludvigson and Ng (2009) is finer but is applied to US variables only. Their classification is likely not applicable when a number of economies are considered for reasons of interpretability. 17 China imports 30% of all the copper traded in the world. 18 The p-value is calculated by transforming the correlation ρ^ to create a t-statistic with T – 2 degrees of freedom, where T is the number of observations. 19 One possible explanation is that the oil-return filtering is more successful than that for other commodities. 20 This is an issue in the contribution of PR which was further considered in Deb, Trivedi, and Varangis (1996) by means of the multivariate GARCH model in its BEKK form (Engle and Kroner, 1995). Multivariate GARCH models deal with standardized returns and no further correction for heteroscedasticity is needed [see Brenner, Pasquariello, and Subrahmanyam (2009) for a recent application using standardized returns for the analysis of co-movements in US financial markets around scheduled macroeconomic announcements]. 21 Co-movement is a concept which may at first sight be confused with contagion. However, there is a significant difference between the two concepts. While excess co-movement is defined as significant residual correlation once common factors are considered, contagion is defined as a significant increase in correlation following a shock in one market. At this point, two remarks are in order. First, most of the literature on contagion either uses very simple common factors or ignores them entirely. This is quite different from the excess co-movement literature where “excess” means “beyond common factors”, and the determination of common factors strongly affects the estimated co-movement. Second, we do not need to observe an increase in correlation to confirm excess co-movement, but rather a significant correlation most of the time or on average over a given period. 22 Similar results appear in Boyer, Gibson, and Loretan (1999) and Loretan and English (2000). Tang and Xiong (2012) also correct the correlation for time-varying volatility using the method in Forbes and Rigobon (2002), which has only a small insignificant effect on their estimates. 23 This correction is valid if we assume no omitted variables or endogeneity. 24 The significance threshold is 0.1669 and is plotted as a horizontal dotted line in Figure 3. 25 See also Bruno, Büyüksahin, and Robe (2016), who consider co-movement across food commodities along with financialization, but whose main focus is rather on the commodity–equity relationship. 26 We wish to thank a referee for suggesting this methodological approach. 27 The analysis in Han (2008) deals with S&P 500 futures contracts. The author also relies on a proxy base on the Investors Intelligence’s weekly that is not relevant for commodity markets. 28 Since 2006, the CFTC has also released a weekly Disaggregated Commitments of Traders (DCOT) report each Friday. This complements the COT report by providing more detailed categories of traders such as Index Traders who have played a significant role in recent years. We do not use the DCOT data here, as it would considerably restrict the analysis sample period. 29 The sample correlation between Han’s index and hedging pressure ranges from −0.78 for live cattle to − 0.98 for cotton. 30 Singleton (2014) implicitly considers the role of cross-positions, as his measure of index funds in oil markets is derived from index funds positions in grain markets. 31 Our methodology resembles the approach in Raman, Robe, and Yadav (2016), who gauge the effect of the participation of financial traders in oil futures post-electronification using two-stage least squares, or the method in Gilbert and Pfuderer (2014), who investigate the causal role of index trading on grain markets using instrumental variables. As changes in futures positions and futures returns are simultaneously determined, these methods are naturally relevant when analyzing financialization. 32 The penalty functions suggested by Bai and Ng (2002) are: g1(N.T)=N+TNTln(NTN+T), g2(N.T)=N+TNTln(CNT2), g3(N.T)=ln(CNT2)CNT2, and g4(N.T)=(N+T−k)ln(NT)NT. 33 We use the Matlab routine provided by Ng for the estimation and information criteria. The Kapetanios test routine is ours. 34 See for example the empirical applications in Kapetanios (2010) which show that there may be some variation in the estimation of the number of factors under different selection criteria. References Ai C., Chatrath A., Song F. ( 2006): On the comovement of commodity prices, American Journal of Agricultural Economics  88, 574– 588. Google Scholar CrossRef Search ADS   Bai J. ( 2003): Inferential theory for factor models of large dimensions, Econometrica  71, 135– 172. Google Scholar CrossRef Search ADS   Bai J., Ng S. ( 2002): Determining the number of factors in approximate factor models, Econometrica  70, 191– 221. Google Scholar CrossRef Search ADS   Bai J., Ng S. ( 2008): Large dimensional factor analysis, Foundations and Trends in Econometrics  3, 89– 163. Google Scholar CrossRef Search ADS   Baker M., Wurgler J. ( 2007): Investor sentiment in the stock market, Journal of Economic Perspectives  21, 129– 151. Google Scholar CrossRef Search ADS   Barberis N., Shleifer A. ( 2003): Style investing, Journal of Financial Economics  68, 161– 199. Google Scholar CrossRef Search ADS   Barberis N., Shleifer A., Wurgler J. ( 2005): Comovement, Journal of Financial Economics  75, 283– 317. Google Scholar CrossRef Search ADS   Barsky R. B., Kilian L. ( 2002): Do we really know that oil caused the great stagflation? A monetary alternative, in: Bernanke B. S., Rogoff K. (eds.), NBER Macroeconomics Annual 2001 , Vol. 16. MIT Press, Cambridge, Massachusetts, pp. 137– 183. Basak S., Pavlova A. ( 2016): A model of fiinancialization of commodities, Journal of Finance  71, 1511– 1556. Google Scholar CrossRef Search ADS   Baumeister C., Kilian L. ( 2012): Real-time forecasts of the real price of oil, Journal of Business & Economic Statistics  30, 326– 336. Google Scholar CrossRef Search ADS   Baumeister C., Kilian L. ( 2014): What central bankers need to know about forecasting oil prices, International Economic Review  55, 869– 889. Google Scholar CrossRef Search ADS   Bessembinder H. ( 1992): Systematic risk, hedging pressure, and risk premiums in futures markets, Review of Financial Studies  5, 637– 667. Google Scholar CrossRef Search ADS   Boivin J., Ng S. ( 2005): Understanding and comparing factor based forecasts, International Journal of Central Banking  1, 117– 152. Borensztein E., Reinhart C. M. ( 1994): The macroeconomic determinants of commodity prices, IMF Staff Papers  41, 236– 261. Google Scholar CrossRef Search ADS   Boyer B., Gibson M., Loretan M. ( 1999): Pitfalls in tests for changes in correlations. Unpublished working Paper, Vol. 597R, Federal Reserve Board. Brenner M., Pasquariello P., Subrahmanyam M. ( 2009): On the volatility and comovement of U.S. financial markets around macroeconomic news announcements, Journal of Financial and Quantitative Analysis  44, 1265– 1289. Google Scholar CrossRef Search ADS   Bruno V. G., Byksahin B., Robe M. A. ( 2016): The financialization of food?, American Journal of Agricultural Economics  99, 243– 264. Google Scholar CrossRef Search ADS   Byksahin B., Robe M. A. ( 2014): Speculators, commodities and cross-market linkages, Journal of International Money and Finance  42, 38– 70. Google Scholar CrossRef Search ADS   Byrne J. P., Fazio G., Fiess N. ( 2013): Primary commodity prices, Co-movements, common factors and fundamentals, Journal of Development Economics  101, 16– 26. Google Scholar CrossRef Search ADS   Cheng I.-H., Kirilenko A., Xiong W. ( 2015): Convective risk flows in commodity futures markets, Review of Finance  19, 1733– 1781. Google Scholar CrossRef Search ADS   Cheng I.-H., Xiong W. ( 2014a): The financialization of commodity markets, Annual Review of Financial Economics  6, 419– 441. Google Scholar CrossRef Search ADS   Cheng I.-H., Xiong W. ( 2014b): Why do hedgers trade so much?, Journal of Legal Studies  43, S183– S207. Google Scholar CrossRef Search ADS   Chen S.-L., Jackson J. D., Kim H., Resiandini P. ( 2014): What drives commodity prices?, American Journal of Agricultural Economics  96, 1455– 1468. Google Scholar CrossRef Search ADS   Chen Y.-C., Rogoff K. S., Rossi B. ( 2010): Can exchange rates forecast commodity prices?, Quarterly Journal of Economics  125, 1145– 1194. Google Scholar CrossRef Search ADS   Christoffersen P., Lunde A., Olesen K. V. ( 2014): Factor structure in commodity futures return and volatility. Unpublished working paper No. 2495779, Rotman School of Management. Deb P., Trivedi P. K., Varangis P. ( 1996): The excess comovement of commodity prices reconsidered, Journal of Applied Econometrics  11, 275– 291. Google Scholar CrossRef Search ADS   Engle R. F. ( 2002): Dynamic conditional correlation: a simple class of multivariate generalized autoregressive conditional heteroskedasticity models, Journal of Business & Economic Statistics  20, 339– 350. Google Scholar CrossRef Search ADS   Engle R. F., Kroner K. F. ( 1995): Multivariate simultaneous generalized ARCH, Econometric Theory  11, 122– 150. Google Scholar CrossRef Search ADS   Erb C., Harvey C. ( 2006): The strategic and tactical value of commodity futures, Financial Analysts Journal  62, 69– 97. Google Scholar CrossRef Search ADS   Ferraro D., Rogoff K. S., Rossi B. ( 2015): Can oil prices forecast exchange rates? An empirical analysis of the relationship between commodity prices and exchange rates, Journal of International Money and Finance  54, 116– 141. Google Scholar CrossRef Search ADS   Forbes K., Rigobon R. ( 2002): No contagion, only interdependence: measuring stock market co-movements, Journal of Finance  57, 2223– 2261. Google Scholar CrossRef Search ADS   Forni M., Hallin M., Lippi M., Reichlin L. ( 2005): The generalized dynamic factor model, one sided estimation and forecasting, Journal of the American Statistical Association  100, 830– 840. Google Scholar CrossRef Search ADS   Frankel J. A. ( 2006): The effect of monetary policy on real commodity prices. Unpublished working paper No. w12713, National Bureau of Economic Research. Frankel J. A., Rose A. K. ( 2010): Determinants of agricultural and mineral commodity prices. Unpublished working paper No. 10-038, Harvard Kennedy School. Geman H., Nguyen V.-N. ( 2005): Soybean inventory and forward curve dynamics, Management Science  51, 1076– 1091. Google Scholar CrossRef Search ADS   Gilbert C. L. ( 1989): The impact of exchange rates and developing country debt on commodity prices, Economic Journal  99, 773– 784. Google Scholar CrossRef Search ADS   Gilbert C. L., Pfuderer S. ( 2014): The role of index trading in price formation in the grains and oilseeds markets, Journal of Agricultural Economics  65, 303– 322. Google Scholar CrossRef Search ADS   Gorton G., Hayashi F., Rouwenhorst K. G. ( 2013): The fundamentals of commodity futures returns, Review of Finance  17, 35– 105. Google Scholar CrossRef Search ADS   Gorton G., Rouwenhorst K. G. ( 2006): Facts and fantasies about commodity returns, Financial Analysts Journal  62, 47– 68. Google Scholar CrossRef Search ADS   Gospodinov N., Ng S. ( 2013): Commodity prices, convenience yields, and inflation, Review of Economics and Statistics  95, 206– 219. Google Scholar CrossRef Search ADS   Haase M., Zimmermann Y. S., Zimmermann H. ( 2016): The impact of speculation on commodity futures markets—a review of the findings of 100 empirical studies, Journal of Commodity Markets  3, 1– 15. Google Scholar CrossRef Search ADS   Hamilton J. D. ( 2009): Understanding crude oil prices, Energy Journal  30, 179– 206. Google Scholar CrossRef Search ADS   Hamilton J. D., Wu J. C. ( 2015): Effects of index-fund investing on commodity futures prices, International Economic Review  56, 187– 205. Google Scholar CrossRef Search ADS   Han B. ( 2008): Investor sentiment and option prices, Review of Financial Studies  21, 387– 414. Google Scholar CrossRef Search ADS   Hansen L. ( 1982): Large sample properties of the generalized method of moments estimator, Econometrica  50, 1029– 1054. Google Scholar CrossRef Search ADS   Irwin S. H., Sanders D. R. ( 2011): Index funds, financialization, and commodity futures markets, Applied Economic Perspectives and Policy  33, 1– 31. Google Scholar CrossRef Search ADS   Juvenal L., Petrella I. ( 2015): Speculation in the oil market, Journal of Applied Econometrics  30, 621– 649. Google Scholar CrossRef Search ADS   Kallberg J., Pasquariello P. ( 2008): Time-series and cross-sectional excess comovement in stock indexes, Journal of Empirical Finance  15, 481– 502. Google Scholar CrossRef Search ADS   Kapetanios G. ( 2010): A testing procedure for determining the number of factors in approximate factor models with large datasets, Journal of Business & Economic Statistics  28, 397– 409. Google Scholar CrossRef Search ADS   Karali B., Power G. J. ( 2013): Short and long-run determinants of commodity price volatility, American Journal of Agricultural Economics  95, 724– 738. Google Scholar CrossRef Search ADS   Kilian L. ( 2009): Not all oil price shocks are alike: disentangling demand and supply shocks in the crude oil market, American Economic Review  99, 1053– 1069. Google Scholar CrossRef Search ADS   Kilian L., Hicks B. ( 2013): Did unexpectedly strong economic growth cause the oil price shock of 2003–2008?, Journal of Forecasting  32, 385– 394. Google Scholar CrossRef Search ADS   King M., Sentana E., Wadhwani S. ( 1994): Volatility and links between national markets, Econometrica  62, 901– 933. Google Scholar CrossRef Search ADS   Lehecka G. V. ( 2015): Do hedging and speculative pressures drive commodity prices, or the other way round?, Empirical Economics  49, 575– 603. Google Scholar CrossRef Search ADS   Leybourne S. J., Lloyd T. A., Reed G. V. ( 1994): The excess comovement of commodity prices revisited, World Economy  22, 1747– 1758. Loretan M., English W. ( 2000): Evaluation “correlation breakdowns” during periods of market volatility. BIS Conference Papers 8, Bank for International Settlements. Ludvigson S. C., Ng S. ( 2007): The empirical risk return relation: a factor analysis approach, Journal of Financial Economics  83, 171– 222. Google Scholar CrossRef Search ADS   Ludvigson S. C., Ng S. ( 2009): Macro factors in bond risk premia, Review of Financial Studies  22, 5027– 5067. Google Scholar CrossRef Search ADS   Malliaris A. G., Urritia J. L. ( 1996): Linkages between agricultural commodity futures contracts, Journal of Futures Markets  16, 595–610. Google Scholar CrossRef Search ADS   Marshall B. R., Nguyen N. H., Visaltanachoti N. ( 2012): Commodity liquidity measurement and transaction costs, Review of Financial Studies  25, 599– 638. Google Scholar CrossRef Search ADS   Marshall B. R., Nguyen N. H., Visaltanachoti N. ( 2013): Liquidity commonality in commodities, Journal of Banking & Finance  37, 11– 20. Google Scholar CrossRef Search ADS   Palaskas T. B., Varangis P. N. ( 1991): Is there excess co-movement of primary commodity prices? A co-integration test. Unpublished working paper series No. 758, World Bank , Washington, DC, International Economics Department. Pindyck R. S. ( 2001): The dynamics of commodity spot and futures markets: a primer, Energy Journal  22, 1– 29. Google Scholar CrossRef Search ADS   Pindyck R. S., Rotemberg J. J. ( 1990): The excess co-movement of commodity prices, Economic Journal  100, 1173– 1189. Google Scholar CrossRef Search ADS   Raman V., Robe M. A., Yadav P. K. ( 2016): Financialization, intraday institutional trading, and commodity market quality. Unpublished manuscript. de Roon F. A., Nijman T. E., Veld C. ( 2000): Hedging pressure effects in futures markets, Journal of Finance  55, 1437– 1456. Google Scholar CrossRef Search ADS   Rouwenhorst K. G., Tang K. ( 2012): Commodity investing. Unpublished working paper No. 06-12, Yale ICF. Singleton K. J. ( 2014): Investor flows and the 2008 boom/bust in oil prices, Management Science  60, 300– 318. Google Scholar CrossRef Search ADS   Stock J. H., Watson M. W. ( 2002a): Forecasting using principal components from a large number of predictors, Journal of the American Statistical Association  97, 1167– 1179. Google Scholar CrossRef Search ADS   Stock J. H., Watson M. W. ( 2002b): Macroeconomic forecasting using diffusion indexes, Journal of Business & Economic Statistics  20, 147– 162. Google Scholar CrossRef Search ADS   Stoll H. R., Whaley R. E. ( 2010): Commodity index investing and commodity futures prices, Journal of Applied Finance  20, 1– 40. Tang K., Xiong W. ( 2012): Index investment and financialization of commodities, Financial Analysts Journal  68, 54– 74. Google Scholar CrossRef Search ADS   West K. D., Wong K.-F. ( 2014): A factor model for co-movements of commodity prices, Journal of International Money and Finance  42, 289– 309. Google Scholar CrossRef Search ADS   © The Authors 2017. Published by Oxford University Press on behalf of the European Finance Association. All rights reserved. For Permissions, please email: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Review of Finance Oxford University Press

Futures Trading and the Excess Co-movement of Commodity Prices

Loading next page...
 
/lp/ou_press/futures-trading-and-the-excess-co-movement-of-commodity-prices-deA1Kxmt64
Publisher
Oxford University Press
Copyright
© The Authors 2017. Published by Oxford University Press on behalf of the European Finance Association. All rights reserved. For Permissions, please email: journals.permissions@oup.com
ISSN
1572-3097
eISSN
1573-692X
D.O.I.
10.1093/rof/rfx039
Publisher site
See Article on Publisher Site

Abstract

Abstract We empirically reinvestigate the issue of the excess co-movement of commodity prices initially raised in Pindyck and Rotemberg (1990). Excess co-movement appears when commodity prices remain correlated even after adjusting for the impact of fundamentals. We use recent developments in large approximate factor models to consider a richer information set and adequately model these fundamentals. We consider a set of eight unrelated commodities along with 184 real and nominal macroeconomic variables, from developed and emerging economies, from which nine factors are extracted over the 1993–2013 period. Our estimates provide evidence of time-varying excess co-movement which is particularly high after 2007. We further show that speculative intensity is a driver of the estimated excess co-movement, as speculative trading is both correlated across the commodity futures markets and correlated with the futures prices. Our results can be taken as direct evidence of the significant impact of financialization on commodity-price cross-moments. 1. Introduction Commodity markets have undergone major changes over the past two decades. The popularity of commodity-related financial instruments, such as commodity indices, has led many observers to conclude that commodity markets are now more intimately connected to financial markets, and so may also co-move more significantly (Tang and Xiong, 2012; Cheng and Xiong, 2014a; Hamilton and Wu, 2015; Basak and Pavlova, 2016). While a greater number of participants in commodity markets may bring about improved risk sharing, the financialization process has been widely criticized as a potential source of excessive price volatility (Stoll and Whaley, 2010). This paper investigates whether the excess co-movement of commodity prices is related to the growing financial influence in commodity markets. The excess co-movement of commodity prices deserves analysis for at least two reasons. First, residual correlation (or “co-movement”) may mean that “[…] commodity demands and supplies are affected by unobserved forecasts of the economic variable.” [Pindyck and Rotemberg (PR hereafter), 1990, p. 1174], thereby indicating that the standard demand–supply model may not be able to explain commodity returns adequately. This conclusion, which is at odds with standard economic theory, suggests that further research is needed to uncover the new relevant fundamentals, or change the way in which these fundamentals are measured. Second, from a portfolio-management perspective, the presence of co-movement limits the diversification of investors who manage a portfolio containing a number of commodity futures.1 PR define excess co-movement as commodity prices remaining correlated even after adjusting for the impact of common macroeconomic variables. They select six variables: the US index of industrial production, the consumer price index, the effective $US exchange rate,2 the three-month Treasury bill interest rate (cf. Frankel and Rose, 2010), the M1 monetary measure (cf. Frankel, 2006), and the S&P 500 stock index, which are supposed to represent the fundamentals. Nevertheless, the authors recognize that: “[…] a major limitation of our approach is that we can never be sure we have included all relevant macroeconomic variables and latent variables.” (p. 1185).3 One major issue in filtering the returns from common factors is indeed the selection of the variables to be considered. To deal with the issue of omitted variables, we suggest relying on a large approximate factor model, along the lines of Stock and Watson (2002a, 2002b), which allows us to enlarge the information set significantly while preserving a sufficiently low dimension for the econometric estimation.4 We thus avoid the arbitrariness and computational difficulties of selecting relevant variables, in particular when the number of possible combinations is large. Borensztein and Reinhart (1994) underline the need to consider well-defined supply and demand variables in order to explain commodity prices. In particular, the authors advocate the inclusion of variables for Eastern Europe that are likely to be relevant for their sample period of 1970–92. In the same spirit, we consider a set of economic variables from developed and emerging countries (China, India, and Brazil, among others) that should allow us to filter out commodity returns more accurately, as these countries have played a central role in shaping commodity prices over recent years. While commodity prices are the product of transactions in one particular part of the world, they also reflect a great deal of information which has been generated throughout the world. For instance, the price of crude oil, say US West Texas Intermediate (WTI), is widely accepted as a world price, while being mainly traded in the USA (see Kilian, 2009). Following the idea of Ludvigson and Ng (2009) of grouping explanatory variables into meaningful categories, we uncover the sets of variables that best explain commodity returns. Our estimates show that monthly commodity returns over the last two decades are mainly correlated with the real aggregate variables in emerging countries, highlighting the important role played by these countries in shaping commodity prices over this period. The paper provides evidence of time-varying excess co-movement, which is particularly high after 2007.5 As such we extend PR’s analysis in two directions. First, we investigate the time-varying behavior of the phenomenon, thereby providing further insights into the analysis of excess co-movement. Second, we look at a recent period that includes both a pronounced increase in commodity prices around 2008 and the recent financial crisis. Last, we take heteroscedasticity into account as this can play a critical role in measures of correlation.6 Highlighting these stylized facts regarding excess co-movements in commodity prices in recent years is our first contribution. The main novelty in our paper, which constitutes our second contribution, is that we establish an empirical relationship between the notion of excess co-movement and speculative activity in commodity futures markets. Surprisingly, academic research has not yet investigated the potential determinants of excess co-movement in commodity prices. We suggest an explanation for this phenomenon following the intuition developed in Barberis and Shleifer (2003) that “investors categorize risky assets into different styles and move funds among these styles depending on their relative performance.” (p. 161). As such, if most commodities are classified into a “commodity style”, seemingly unrelated commodities are likely to co-move more than would be expected based on fundamental analysis.7 This is precisely what we demonstrate in our present work. Our results are also in line with the recent work by Basak and Pavlova (2016), who go beyond the behavioral approach in Barberis and Shleifer (2003) and develop a multi-asset, multi-good general equilibrium model with heterogeneous investors, some of whom are institutional investors, considering characteristics that are specific to commodities such as the presence of inventories. The model, in the tradition of Lucas-tree models, is solved in closed-form and provides a rich set of implications, among which an increase in the correlation between commodities following institutional positioning, and more so for commodities that are included in an index. Our results provide strong support for the outcome in Basak and Pavlova (2016), and may then be seen as an empirical validation of their model. Our empirical work makes use of data from the US Commodity Futures Trading Commission (CFTC) to estimate speculative intensity. While the categories in the publicly available data from the CFTC do not distinguish perfectly between the various categories of traders, as discussed previously in Bessembinder (1992) and Stoll and Whaley (2010) among many others, we here show that they are informative for the explanation of excess co-movement. Our measure of speculative activity in futures markets follows the recent work by Han (2008) on sentiments in financial markets but is reminiscent of the so-called Working’s T measure. Our empirical strategy provides direct evidence of the explanatory power of speculative intensity for excess co-movement: while the large number of fundamental variables have limited success in explaining the co-movement between commodities, we show that speculative activity is correlated across commodity futures markets and, at the same time, that speculative activity is correlated with futures prices. This last result is obtained from an instrumental-variable analysis to avoid endogeneity issues between returns and positions in futures markets. The empirical work closest to ours is Tang and Xiong (2012), which also considers the financialization of commodities as a potential source of the recent increase in co-movements between commodity returns. Their “analysis focuses on connecting the large inflow of commodity index investment to the large increase of commodity price co-movements in recent years by examining the difference in these co-movements between indexed and off-index commodities” (p. 55).8 The authors regress the S&P-GSCI on a measure of the net position change of different categories of traders and, as such, do not pick up the common factors that may affect the behavior of most, if not all, commodity prices. Tang and Xiong (2012) also investigate the relationship between economic activity in emerging countries and the co-movement of commodity prices using a novel time series of Chinese futures prices available since the late 1990s. While commodity prices are usually thought of as being global, the authors show that the picture is actually more complex. Interestingly, while US commodity prices exhibit a pronounced cycle, this is not the case for the Chinese prices of similar commodities, thereby raising “doubt about commodity demands from China as the driver of all commodity prices in the U.S.” (p. 63). Our regressions for commodity returns show that the demand from emerging economies does play a role in determining the prices of US non-agricultural commodity futures prices, while leaving a considerable role for other factors. Overall, while dealing with a research question similar to Tang and Xiong (2012), we adopt a very different empirical approach. In particular, we specifically consider fundamentals that are critical in the analysis of co-movements. The plan for the rest of the paper is as follows. In the next section, we present the data used for the empirical analysis. In Section 3, we very briefly review the factor-model methodology and calculate the factors used to filter the commodity returns. The excess co-movement is then estimated in Section 4, while Section 5 is dedicated to the analysis of the relationship between excess co-movement and speculation. Finally, Section 6 concludes by discussing some limits to and possible future extensions of our work. 2. Data We consider a set of eight commodity prices: wheat, copper, silver, soybeans, raw sugar, cotton, crude oil, and live cattle. These are representative of the main commodity classes and are assumed to be unrelated as defined in PR, in the sense that their supply or demand cross-elasticities are almost zero. All prices are cash prices except for crude oil, where the front-month contract price is taken as a proxy for the cash price, to avoid the distorting impact of delivery issues for this particular commodity. All prices are in nominal US$. Due to data limitations, in particular for macroeconomic variables from emerging countries, we consider monthly observations from February 1993 to November 2013. Data are from Datastream. The prices are displayed in Figure 1. They fluctuate around their mean level until 2005, except for oil and silver which have a rising trend. A first large price rise begins in 2005 and ends in 2008. Prices fell in 2009 but rise steeply in 2010; they stabilize or fell in 2012. Returns are log difference of prices.9 The descriptive statistics in Table I reveals evidence of skewness—negative in six cases out of eight—and excess kurtosis. The Jarque–Bera test consequently rejects the hypothesis of a Gaussian distribution for all returns. The presence of heteroscedasticity, which is a standard feature in financial price series, may lie behind this non-normality. Table I. Descriptive statistics for the eight commodities monthly returns—February 1993–November 2013 (i) Monthly returns are computed as price log differences. (ii) Commodity prices are cash prices except crude oil where the current month contract price is taken as a proxy for the cash price. (iii) ***, **, and *, respectively, denote rejection of the null hypothesis of a Gaussian distribution at 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Mean  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022  Max  0.3666  0.3266  0.2309  0.2013  0.2052  0.3855  0.2536  0.1178  Min  −0.2499  −0.3360  −0.3285  −0.4660  −0.3620  −0.2605  −0.3899  −0.2369  Std  0.0845  0.0796  0.0873  0.0835  0.0806  0.0903  0.1016  0.0460  Skewness  0.1660  −0.4043  −0.4079  −1.0658  −0.3611  0.2763  −0.5827  −0.5674  Kurtosis  4.3312  5.9263  4.0502  7.0722  5.1793  4.5636  3.9713  5.2846  Jarque-Bera  19.60***  96.01***  18.42***  220.06***  54.90***  28.64***  23.9759***  67.78***  Number of observations  250  250  250  250  250  250  250  250    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Mean  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022  Max  0.3666  0.3266  0.2309  0.2013  0.2052  0.3855  0.2536  0.1178  Min  −0.2499  −0.3360  −0.3285  −0.4660  −0.3620  −0.2605  −0.3899  −0.2369  Std  0.0845  0.0796  0.0873  0.0835  0.0806  0.0903  0.1016  0.0460  Skewness  0.1660  −0.4043  −0.4079  −1.0658  −0.3611  0.2763  −0.5827  −0.5674  Kurtosis  4.3312  5.9263  4.0502  7.0722  5.1793  4.5636  3.9713  5.2846  Jarque-Bera  19.60***  96.01***  18.42***  220.06***  54.90***  28.64***  23.9759***  67.78***  Number of observations  250  250  250  250  250  250  250  250  Figure 1. View largeDownload slide The eight commodity prices—January 1993–November 2013. Note: Prices are normalized to 100 at January 1993. Figure 1. View largeDownload slide The eight commodity prices—January 1993–November 2013. Note: Prices are normalized to 100 at January 1993. Table II presents the sample correlations between returns and their associated p-values. There are, respectively, 16, 15, and 11 significant correlations at the 10%, 5%, and 1% critical levels. All of the significant correlations are positive, ranging from 0.4789 (wheat and raw sugar) to 0.11 (raw sugar and crude oil), with an average figure of 0.2536. The figures in PR are a maximum of 0.322 and a minimum of 0.113, with an average of 0.161 for the significant correlations between 1960 and 1985. Therefore, it seems that the relationship between commodity prices has tightened over the last two decades. Table II. Correlation between the eight commodities monthly—February 1993–November 2013 The upper triangular matrix reports correlations while the lower reports their p-values. ***, **, and *, respectively, denote significance at 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.2988***  0.2256***  0.4789***  −0.0023  0.2763***  0.1569**  0.0352  Copper  0.0000  1  0.3799***  0.2345***  0.1258**  0.2352***  0.3496***  0.0077  Silver  0.0003  0.0000  1  0.2142***  0.1901**  0.0844  0.2141***  −0.0306  Soybeans  0.0000  0.0002  0.0006  1  −0.0767  0.3877***  0.1095*  −0.0898  Raw sugar  0.9714  0.0468  0.0025  0.2266  1  0.0543  0.1012  0.0530  Cotton  0.0000  0.0002  0.1834  0.0000  0.3930  1  0.1808**  −0.0288  Crude oil  0.0130  0.0000  0.0007  0.0839  0.1103  0.0041  1  0.0455  Live cattle  0.5791  0.9040  0.6302  0.1571  0.4045  0.6506  0.4739  1    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.2988***  0.2256***  0.4789***  −0.0023  0.2763***  0.1569**  0.0352  Copper  0.0000  1  0.3799***  0.2345***  0.1258**  0.2352***  0.3496***  0.0077  Silver  0.0003  0.0000  1  0.2142***  0.1901**  0.0844  0.2141***  −0.0306  Soybeans  0.0000  0.0002  0.0006  1  −0.0767  0.3877***  0.1095*  −0.0898  Raw sugar  0.9714  0.0468  0.0025  0.2266  1  0.0543  0.1012  0.0530  Cotton  0.0000  0.0002  0.1834  0.0000  0.3930  1  0.1808**  −0.0288  Crude oil  0.0130  0.0000  0.0007  0.0839  0.1103  0.0041  1  0.0455  Live cattle  0.5791  0.9040  0.6302  0.1571  0.4045  0.6506  0.4739  1  Our main aim is to analyze whether these correlations result from a common set of variables related to the fundamentals of commodities markets. If significant residual correlations remain, we would conclude in favor of excess co-movement. To model the eight commodity returns, we construct a set of 184 real and nominal macroeconomic variables. These variables, with a short description, are listed in Appendix A.10 Our data set contains variables from developed (118 variables from Australia, Canada, France, Germany, Japan, the UK, and the USA) and emerging countries (66 variables from China, Brazil, Korea, Taiwan, Mexico, etc.). In recent years, these countries have experienced high growth rates and their commodity demand has had a significant impact on commodity markets. Commonly used US databases such as those in Stock and Watson (2002b) and Ludvigson and Ng (2007, 2009) are thus not well-suited for our current purpose. We have the same classes of data for both developed and emerging countries. We include measures of the country’s aggregate activity level such as the industrial production index and manufacturing orders and capacity utilization. Other real variables are related to household expenditure: household consumption, housing starts, and car sales. We add variables related to the labor market (wages and unemployment) and international trade (exports, imports, and terms of trade). These real variables are assumed to be correlated with the world demand for commodities. The main categories of nominal variables that we include are monetary aggregates, stock indices, interest rates, exchange rates with the dollar, and producer and consumer price indices. These nominal variables help us to model the relationship between commodity returns and interest rate or the inflation rate. Finally, we add the Real Activity Index to the above, as developed in Kilian (2009). This is “based on dry cargo single voyage ocean freight rates and is explicitly designed to capture shifts in the demand for industrial commodities in global business markets” (p. 1055), following a long tradition of economists who have noted the correlation between economic activity and ocean-freight rates. 3. Filtering Commodity Returns Using Large Approximate Factors Models In this section, we first briefly review the large approximate factors method. Recent techniques to establish the optimal number of factors are presented in Appendix B; additional developments can be found in the survey by Bai and Ng (2008) of large approximate factors models. The remainder of the section is dedicated to the projection of commodity returns on the estimated factors. 3.1 Static Factors Calculation We use the static factor model of Stock and Watson (2002a). We do not consider the dynamic version of Forni et al. (2005), as recent work (Boivin and Ng, 2005) has shown that the dynamic and static factor models perform equally well, especially when the factors have unknown dynamics, which is often the case in empirical work. In addition, the dynamic factor model is best suited to forecasting, which is not the purpose of our work. We have a sample {xit} of i=1,…,N cross-section units and t=1,…,T time-series observations. Each xit is split into a component depending on a set of r<<N common factors Ft=(f1t,f2t,…,frt)′ and a specific component eit:   xit=λi′Ft+eit, where λi is the (r×1) factor loading. If we define the (N×1) vectors of observations and specific components at date t as Xt=(x1t,…,xNt)′, et=(e1t,…,eNt)′, and Λ=(λ1,…,λN)′ the (N×r) matrix of factor loadings, the factor decomposition is written as   Xt=ΛFt+et. Standard factor analysis makes the assumptions that Ft and et are serially and cross-sectionally uncorrelated, and the number of units of observation N is fixed. Stock and Watson’s (2002a, 2002b) “large dimensional approximate factor models” allow the specific errors to be “weakly correlated” across i and t11 and the sample size to tend to infinity in both directions. We assume k factors and use the principal components method to estimate the (T×k) factor matrix Fk and the corresponding (N×T) loading matrix Λk. The estimates solve the optimization problem:   min S(k)=(NT)−1∑i=1N∑t=1T(xit−λik′Ftk)2 subject to the normalization Λk′Λk/N=Ik.12 This classical principal component problem is solved by setting Λ^k equal to the eigenvectors of the largest k eigenvalues of X′X where X=(X1,X2,…,XT)′ is the (T×N) matrices of all observations.13 The principal components estimator of Fk is:   F^k=X′Λ^k/N. The consistency and asymptotic normal distribution of the principal component estimator as N,T→∞ have been, respectively, demonstrated by Stock and Watson (2002a), Bai and Ng (2002), and Bai (2003). The next step is to determine the optimal number of factors. The literature on this issue has not come to a clear consensus on how to select relevant factors and, as shown in Table BI in Appendix B, different methods lead to very different outcomes.14 We follow traditional practice in principal component analysis and choose the first nine factors, as the incremental explanatory power beyond these nine factors is only small. The nine factors explain 37% of the variability of the 184 macroeconomic variables. 3.2 Modeling Commodity Returns Our measure of excess co-movement makes use of commodity returns which have been filtered for common components. As such, once we have calculated the static factors, the second step of the empirical analysis consists in filtering the returns using these estimated factors. The first step is the linear regression of returns on the first three factors:   rit=αi+∑k=13βikF^k,t+uiti=1,…,8 t=1,…,T=αi+βi′F^t+uit, where rit represents the ith commodity return at date t, αi is the constant, βi the vector of factor coefficients for the ith commodity, and F^t=(F^1,t,F^2,t,F^3,t)′ the vector of the first three factors at date t. The results from seemingly unrelated regressions (SUR) appear in Table III. The R2 varies from 1.07% for soybeans to 28.58% for crude oil. The factors F^2 and F^1 are significant in most regressions. While the explanatory power figures for agricultural commodity returns are not substantially higher than those in PR, we do obtain a much higher R2 for metals and energy commodities.15 The ARCH-LM test shows that six out of the eight series of residuals have time-varying variance. Table III. Modeling the eight commodities returns: the three factors regressions—February 1993–November 2013 (i) This table reports the SUR estimates of the regression of the eight commodities monthly returns. The explanatory variables are reported in far-left column. A constant is always included in the regression and F^i denotes the ith factor. (ii) t-Statistics are reported in parenthesis under the estimates. ***, **, and *, respectively, denotes rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels. (iii) For the ARCH_LM, ***, **, and *, respectively, denote rejection of the null hypothesis of no ARCH effect at the 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022    (0.45)  (1.05)  (1.22)  (0.66)  (0.59)  (0.24)  (1.32)  (0.78)  F^1  −0.0254  −0.0951***  −0.0271  −0.0242  −0.0114  −0.0517**  −0.1200***  −0.0164**    (−1.19)  (−6.55)  (−1.40)  (−1.10)  (−0.87)  (−2.44)  (−7.10)  (−2.00)  F^2  0.0377**  0.0867***  0.0526***  0.0400**  0.0288*  0.0566***  0.1445***  0.0108    (2.28)  (6.32)  (2.77)  (2.15)  (1.73)  (2.93)  (8.34)  (1.04)  F^3  0.0020  0.0212  0.0060  0.0288  −0.0012  −0.0024  −0.0609**  −0.0062    (0.09)  (0.95)  (0.23)  (1.21)  (−0.05)  (−0.08)  (−2.34)  (−0.52)  R2  0.0222  0.2184  0.0343  0.0290  0.0107  0.0577  0.2858  0.0165  R¯2  0.0103  0.2089  0.0226  0.0172  −0.0014  0.0462  0.2771  0.0045  ARCH_LM(2)  13.10***  52.18***  7.31**  29.66***  3.54  5.84*  5.18*  1.63    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022    (0.45)  (1.05)  (1.22)  (0.66)  (0.59)  (0.24)  (1.32)  (0.78)  F^1  −0.0254  −0.0951***  −0.0271  −0.0242  −0.0114  −0.0517**  −0.1200***  −0.0164**    (−1.19)  (−6.55)  (−1.40)  (−1.10)  (−0.87)  (−2.44)  (−7.10)  (−2.00)  F^2  0.0377**  0.0867***  0.0526***  0.0400**  0.0288*  0.0566***  0.1445***  0.0108    (2.28)  (6.32)  (2.77)  (2.15)  (1.73)  (2.93)  (8.34)  (1.04)  F^3  0.0020  0.0212  0.0060  0.0288  −0.0012  −0.0024  −0.0609**  −0.0062    (0.09)  (0.95)  (0.23)  (1.21)  (−0.05)  (−0.08)  (−2.34)  (−0.52)  R2  0.0222  0.2184  0.0343  0.0290  0.0107  0.0577  0.2858  0.0165  R¯2  0.0103  0.2089  0.0226  0.0172  −0.0014  0.0462  0.2771  0.0045  ARCH_LM(2)  13.10***  52.18***  7.31**  29.66***  3.54  5.84*  5.18*  1.63  In a second approach, as in Stock and Watson (2002b) and Ludvigson and Ng (2009), we consider all possible combinations of the first nine estimated factors and, for each commodity, select the regression which minimizes the BIC criterion. Once each set of regressors has been selected, we jointly estimate the eight regressions via SUR. Our aim here is to identify the best model from a set of common regressors for each commodity. This approach aims to eliminate as much residual correlation as possible, and so strengthen our evidence for any excess co-movement. The SUR estimates appear in Table IV and show a significant increase in explanatory power for crude oil, while this figure remains low for the other commodities. Again, the F^1 and F^2 factors are significant for most of the eight commodities and the ARCH-LM test rejects the null hypothesis of constant variance for three series of residuals. Table IV. Modeling the eight commodities returns: the BIC minimizing regressions—February 1993–November 2013 This table reports the SUR estimates of the regression of the eight commodities monthly returns. The explanatory variables are reported in left column. A constant is always included in the regression and F^i denotes the ith factor. (ii) t-Statistics are reported in parenthesis under the estimates. ***, **, and *, respectively, denotes rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels. (iii) For the ARCH_LM, ***, **, and *, respectively, denote rejection of the null hypothesis of no ARCH effect at the 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022    (0.44)  (1.05)  (1.24)  (0.66)  (0.59)  (0.24)  (1.45)  (0.78)  F^1    −0.0844***        −0.0422**  −0.1194***        (−6.29)        (−2.51)  (−7.39)    F^2  0.0371*  0.0875***  0.0533***  0.0419**  0.0279  0.0577***  0.1449***      (1.84)  (5.17)  (2.64)  (2.12)  (1.45)  (2.74)  (7.72)    F^3              −0.0617***                  (−2.65)    F^6              0.1142***                  (4.00)    F^8      0.0967***        0.1858***          (3.01)        (5.74)    F^9                −0.0393**                  (−2.05)  R2  0.0138  0.2136  0.0728  0.0159  0.0088  0.0566  0.4079  0.0171  R¯2  0.0098  0.2072  0.0653  0.0119  0.0048  0.0489  0.3958  0.0131  ARCH_ LM(2)  7.65**  1.49  3.31  18.54***  2.83  5.50*  4.12  0.96    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0024  0.0047  0.0066  0.0034  0.0030  0.0013  0.0072  0.0022    (0.44)  (1.05)  (1.24)  (0.66)  (0.59)  (0.24)  (1.45)  (0.78)  F^1    −0.0844***        −0.0422**  −0.1194***        (−6.29)        (−2.51)  (−7.39)    F^2  0.0371*  0.0875***  0.0533***  0.0419**  0.0279  0.0577***  0.1449***      (1.84)  (5.17)  (2.64)  (2.12)  (1.45)  (2.74)  (7.72)    F^3              −0.0617***                  (−2.65)    F^6              0.1142***                  (4.00)    F^8      0.0967***        0.1858***          (3.01)        (5.74)    F^9                −0.0393**                  (−2.05)  R2  0.0138  0.2136  0.0728  0.0159  0.0088  0.0566  0.4079  0.0171  R¯2  0.0098  0.2072  0.0653  0.0119  0.0048  0.0489  0.3958  0.0131  ARCH_ LM(2)  7.65**  1.49  3.31  18.54***  2.83  5.50*  4.12  0.96  While the factors cannot be identified econometrically, it is very useful to identify the macroeconomic variables behind the factors affecting commodity returns. To interpret the factors, we follow Ludvigson and Ng (2009) and divide our 184 series into developed and emerging countries, and then real and nominal variables.16 Each of the 184 original variables is then regressed on one factor with the resulting R2 appearing on the horizontal axis. We can thus see which macroeconomic variables obtain the highest R2. The factor in question can then be thought to represent this set of variables. Figure 2 plots R2 for both F^1 (top panel) and F^2 (bottom panel) factors. F^1, which explains a significant part of crude oil and copper returns, is mostly correlated with real variables in emerging economies. This illustrates the importance of emerging countries in shaping commodity prices.17 This result corroborates recent work that also demonstrates that oil (e.g., Hamilton, 2009 or Kilian and Hicks, 2013) and agricultural prices (e.g., Hamilton and Wu, 2015) are partly driven by demand from emerging countries and that speculative activity only plays a minor role. Figure 2. View largeDownload slide Marginal R2 of macroeconomic and financial variables regressed on the first two estimated factors. Notes: Each panel shows the R2 from regressing the series number given on the x-axis on to each individual factor F^i. The series are detailed in Appendix A and sorted as they appear in the figure (real variables for developed countries, nominal variables for developed countries, real variables for emerging countries, and nominal variables for emerging countries). Figure 2. View largeDownload slide Marginal R2 of macroeconomic and financial variables regressed on the first two estimated factors. Notes: Each panel shows the R2 from regressing the series number given on the x-axis on to each individual factor F^i. The series are detailed in Appendix A and sorted as they appear in the figure (real variables for developed countries, nominal variables for developed countries, real variables for emerging countries, and nominal variables for emerging countries). The interpretation of factor F^2 is less obvious. It is highly correlated with a small number of real variables but its explanatory power with respect to interest rates, producer and consumer price indices, and monetary aggregates in both developed and emerging countries is greater than that of F^1. F^2 is likely to represent these nominal variables. Earlier contributions (Barsky and Kilian, 2002; Frankel and Rose, 2010) only provide mixed evidence on the relationship between interest rates and commodity prices. Our estimates give additional support to such a link. In this regard, price indices and monetary aggregates may pick up the impact of inflation on commodity prices. The activity index from Kilian (2009) brings no additional information as it attracts only an insignificant estimated coefficient—except for copper, at the 10% threshold only—indicating that F^1 does a better job of modeling commodity returns. This conclusion is of interest as this real-activity index is considered to be as a proxy for economic activity. We believe that this result confirms the ability of statistical factors to aggregate information from a large number of variables and capture high-frequency growth rates. To better understand the insignificance of Kilian’s index, Table V shows the estimates from univariate regressions of the nine empirical factors on Kilian’s index. The estimates are very significant but with little explanatory power. This is likely due to the low-frequency nature of the Kilian Index, and further demonstrates the benefit from using statistical factors in modeling monthly commodity returns. Table V. Regression of the Kilian real activity index on each of the nine factors Coefficient reports the estimated coefficient of each factor and t-stat its Student’s statistic. ***, **, and *, respectively, denote rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels. The real activity index is taken from Lutz Kilian’s homepage. See Kilian (2009) for a definition of this index.   F^1  F^2  F^3  F^4  F^5  F^6  F^7  F^8  F^9  Coefficient  −17.8231***  19.3546***  −5.6256  28.5721***  1.6348  −2.0679  4.7489  −9.6327  4.0327  t-stat  −3.78  3.504  −0.77  4.77  0.19  −0.21  0.48  −0.85  0.34  R2  0.0438  0.0387  0.0021  0.0434  0.0001  0.0002  0.0008  0.0032  0.0005  R¯2  0.0400  0.0349  −0.0019  0.0396  −0.0039  −0.0038  −0.0031  −0.0008  −0.0035    F^1  F^2  F^3  F^4  F^5  F^6  F^7  F^8  F^9  Coefficient  −17.8231***  19.3546***  −5.6256  28.5721***  1.6348  −2.0679  4.7489  −9.6327  4.0327  t-stat  −3.78  3.504  −0.77  4.77  0.19  −0.21  0.48  −0.85  0.34  R2  0.0438  0.0387  0.0021  0.0434  0.0001  0.0002  0.0008  0.0032  0.0005  R¯2  0.0400  0.0349  −0.0019  0.0396  −0.0039  −0.0038  −0.0031  −0.0008  −0.0035  Finally, the omission of inventory data in our analysis is worthy of mention. It is commonly thought that stock levels may help us to better model commodity returns, following Working’s Theory of Storage. For instance, Pindyck (2001) uses weekly inventory data from the US Department of Energy to model the convenience yield in the WTI crude oil market. Geman and Nguyen (2005) rely on a number of worldwide sources to construct their own inventory series for soybeans which they use to model this commodity’s forward curve. Baumeister and Kilian (2012) consider a number of oil-specific inventory series to forecast real-time monthly oil prices. We do not include inventory information in our empirical analysis as we wish to filter returns using fundamentals that are, at least partly, common to all commodities. By doing so, data related to commodity demands that we proxy via our factors are relevant as they represent common fundamentals. Conversely, data such as inventories are very particular to each commodity and so less likely to explain any correlation in commodity returns. As such, even if we recognize that inventories matter in particular cases such as, for instance, forecasting commodity prices [see Baumeister and Kilian (2012, 2014) for the case of oil], they do not do so here, where it is rather common factors that are our primary concern. 4. Testing for the Excess Co-movement of Commodity Returns 4.1 Testing for Residual Correlation The residuals from the regressions above reflect commodity returns after controlling for fundamentals. We first evaluate the correlation in residuals, as in PR. Tables VI and VII show the sample correlations (in the upper triangular part) and their p-values18 (in the lower triangular part) for the residuals from the three-factor and BIC linear filters. Table VI. Correlation between residuals from the three factors linear model The upper triangular matrix reports correlation while the lower reports the p-values. The p-value is computed by transforming the correlation ρ^ to create a t-statistic having T – 2 degrees of freedom, where T is the number of observations. ***, **, and *, respectively, denote significance at 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.3452***  0.2564***  0.4939***  0.0204  0.3133***  0.2330***  0.0660  Copper  0.0000  1  0.4317***  0.2817***  0.1698***  0.3386***  0.5312***  0.0913  Silver  0.0000  0.0000  1  0.2398***  0.2110***  0.1439**  0.3017***  0.0101  Soybeans  0.0000  0.0000  0.0001  1  −0.0568  0.4140***  0.1842***  −0.0574  Raw sugar  0.7478  0.0071  0.0008  0.3709  1  0.0884  0.1567**  0.0740  Cotton  0.0000  0.0000  0.0228  0.0000  0.1636  1  0.3102***  0.0225  Crude oil  0.0002  0.0000  0.0000  0.0035  0.0131  0.0000  1  0.1337  Live cattle  0.2986  0.1499  0.8739  0.3663  0.2438  0.7228  0.0346  1  Breusch–Pagan LM test  186.20  p-value  0    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.3452***  0.2564***  0.4939***  0.0204  0.3133***  0.2330***  0.0660  Copper  0.0000  1  0.4317***  0.2817***  0.1698***  0.3386***  0.5312***  0.0913  Silver  0.0000  0.0000  1  0.2398***  0.2110***  0.1439**  0.3017***  0.0101  Soybeans  0.0000  0.0000  0.0001  1  −0.0568  0.4140***  0.1842***  −0.0574  Raw sugar  0.7478  0.0071  0.0008  0.3709  1  0.0884  0.1567**  0.0740  Cotton  0.0000  0.0000  0.0228  0.0000  0.1636  1  0.3102***  0.0225  Crude oil  0.0002  0.0000  0.0000  0.0035  0.0131  0.0000  1  0.1337  Live cattle  0.2986  0.1499  0.8739  0.3663  0.2438  0.7228  0.0346  1  Breusch–Pagan LM test  186.20  p-value  0  Table VII. Correlation between residuals from the BIC minimizing regressions The upper triangular matrix reports correlation while the lower reports the p-values. The p-value is computed by transforming the correlation ρ^ to create a t-statistic having T – 2 degrees of freedom, where T is the number of observations. ***, **, and *, respectively, denote significance at 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.2674***  0.1990***  0.4711***  −0.0135  0.2528***  0.0609  0.0287  Copper  0.0000  1  0.3328***  0.1930***  0.0962  0.1458**  0.1068*  −0.0431  Silver  0.0016  0.0000  1  0.1869***  0.1733***  0.0375  0.0411  −0.0390  Soybeans  0.0000  0.0022  0.0030  1  −0.0897  0.3677***  −0.0087  −0.0990  Raw sugar  0.8324  0.1294  0.0060  0.1572  1  0.0337  0.0251  0.0450  Cotton  0.0001  0.0211  0.5552  0.0000  0.5957  1  0.0261  −0.0496  Crude oil  0.3375  0.0920  0.5179  0.8912  0.6929  0.6812  1  −0.0249  Live cattle  0.6512  0.4978  0.5389  0.1184  0.4783  0.4345  0.6956  1  Breusch– Pagan LM test  99.39  p-value  0    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.2674***  0.1990***  0.4711***  −0.0135  0.2528***  0.0609  0.0287  Copper  0.0000  1  0.3328***  0.1930***  0.0962  0.1458**  0.1068*  −0.0431  Silver  0.0016  0.0000  1  0.1869***  0.1733***  0.0375  0.0411  −0.0390  Soybeans  0.0000  0.0022  0.0030  1  −0.0897  0.3677***  −0.0087  −0.0990  Raw sugar  0.8324  0.1294  0.0060  0.1572  1  0.0337  0.0251  0.0450  Cotton  0.0001  0.0211  0.5552  0.0000  0.5957  1  0.0261  −0.0496  Crude oil  0.3375  0.0920  0.5179  0.8912  0.6929  0.6812  1  −0.0249  Live cattle  0.6512  0.4978  0.5389  0.1184  0.4783  0.4345  0.6956  1  Breusch– Pagan LM test  99.39  p-value  0  The results from both sets of regressions confirm the hypothesis of excess co-movement. We find 16 and 18 significant sample correlations at the 1% and 5% significance levels, respectively, for the three-factor regressions; the analogous numbers for the BIC-minimizing regressions are 9 and 10. Unsurprisingly, the Breusch–Pagan LM test rejects the null hypothesis of no residual correlation in both cases. In the BIC-minimizing regressions, five sample correlations are no longer significant, mostly related to crude oil.19 Filtering commodity returns therefore somewhat reduces the number of significant correlations. However, as the significant correlations range from 0.4711 (wheat and soybeans) to 0.1066 (copper and crude oil) the level of residual correlation remains quite substantial. 4.2 A Global, Unbiased, and Time-Varying Measure of Excess Co-movement One major limit of the use of sample correlation to gauge excess co-movement is the bias in the former when volatility is time-varying.20 This argument has been put forward in the contagion literature21 by Forbes and Rigobon (2002), among others.22 When there is a simultaneous rise in the respective volatility of two variables, the typical sample correlation measure overestimates the true correlation. Forbes and Rigobon (2002) propose an unbiased correlation estimator: as our residuals very often have time-varying volatility, this is the estimator we use to evaluate excess co-movement. We follow Kallberg and Pasquariello (2008), who apply the Forbes and Rigobon estimator on a moving-window basis to yield a more precise estimator of the true correlation. We end up with a global measure, as we treat all residual correlations equally, without focusing on the correlation of one particular commodity with another. We calculate a time-varying measure of excess co-movement which will show us whether excess co-movement is a permanent feature of commodity markets or if it is only occasional. Our global measure of excess co-movement is the average of all the squared unbiased correlations. We use squared correlation measures as some of the estimated correlations are negative. Our estimate is nonparametric and avoids the mean-reversion problem inherent in the parametric approach, such as in the dynamic conditional correlation (DCC) model (Engle, 2002). Indeed, in many instances (cf. Kallberg and Pasquariello, 2008, among others), methods based on rolling windows filters are very competitive with parametric methods when the object of interest is the estimation of correlations. In the following, we present the bias-corrected correlation estimator and the aggregation process to obtain our overall excess co-movement measure. For all pairs of non-redundant returns i≠j, we calculate the residual correlation:   ρ^ij,t=cov(u^i,t,u^j,t)[var(u^i,t)var(u^j,t)]1/2, where u^i,t is the residual from the ith commodity-return equation. As the sample correlation ρ^ij,t is biased in the case of heteroscedasticity, this is called the “conditional correlation”. The Forbes and Rigobon (2002) bias-corrected correlation estimator is   ρ^ij,t*=ρ^ij,t[1+δ^i,t(1−(ρ^ij,t2)]1/2, where the ratio δ^i,t=var(u^i,t)var(u^i,t)LT−1 corrects the conditional correlation ρ^ij,t for the change between the ith return’s short-term var(u^i,t) and long-term var(u^i,t)LT volatilities.23 ρ^ij,t* is called the unconditional correlation. As we do not make any ex ante assumption regarding the direction of the propagation of shocks from one commodity to another, we alternately assume that the source of these shocks is asset i (in ρ^ij,t*) or asset j (in ρ^ji,t*). We therefore have two unconditional and possibly different correlations, ρ^ij,t* and ρ^ji,t*. Our global excess co-movement measure is based on these unconditional correlations. As suggested in King, Sentana, and Wadhwani (1994) and Kallberg and Pasquariello (2008), we compute the arithmetic mean of the pairwise squared unbiased correlations for each commodity i. A non-null unconditional correlation ρ^ij,t*≠0 and ρ^ji,t*≠0, whatever its sign, is taken as evidence of excess co-movement between commodities i and j. A measure of excess co-movement between commodity i and the others is defined as:   ρ^i,t*=1K−1∑j=1,j≠iK(ρ^ij,t*)2 for all commodity returns i=1,…,K, where K = 8 is the number of commodities. Our global and time-varying measure of excess co-movement is then the mean of the excess squared unconditional correlations over all commodities:   ρ^t*=1K∑i=1Kρ^i,t*. We treat the covariance matrix of return residuals as observable, and construct a time series of rolling realized excess squared correlations for each commodity i. δ^i,t and ρ^i,t* are estimated over short- and long-term intervals of fixed length N [t−N+1,t] and gN (with g > 1) [t−gN+1,t], respectively. We use a rolling window of N = 30 monthly observations for short-term volatility and gN = 60 monthly observations for long-term volatility. 4.3 Estimation Results We compute three averages of squared correlations, all of which appear in Figure 3, to evaluate the importance of filtering returns and illustrate the time-variation in volatility. The first (dashed–dotted line) is the average of the squared unconditional correlations in returns: ρ^ret,t*=1K∑i=1Kρ^ret i,t*, where the unbiased correlations are calculated for non-filtered returns. The second (dashed line) is the average of the squared conditional correlations between residuals: ρ^t=1K∑i=1Kρ^i,t, where ρ^i,t=1K−1∑j=1,j≠iK(ρ^ij,t)2. We here use residual correlations that are not corrected for changes in volatility. The solid line is the average of the unconditional squared correlations ρ^t* as defined in the previous section, which is our estimate of excess co-movement. Figure 3. View largeDownload slide Mean excess squared correlation for commodity returns and residuals. Notes: (i) “av sq unc corr ret” is the average squared unconditional correlation of returns: ρret.t*. (ii) “av sq cond corr res fund” is the average squared conditional correlation of factors regression residual: ρt*. (iii) “av sq unc corr res all” is the average squared unconditional of factors regression residual. (iv) The confidence band is the minimal value above which squared correlation is significant at 5% level. It is computed from the t-squared ratio test t^ijt2=(ρ^ijt*)2[1−ρ^ijt*]−1(N−2)∼F(1.N−2) and is equal to 1.6990. Figure 3. View largeDownload slide Mean excess squared correlation for commodity returns and residuals. Notes: (i) “av sq unc corr ret” is the average squared unconditional correlation of returns: ρret.t*. (ii) “av sq cond corr res fund” is the average squared conditional correlation of factors regression residual: ρt*. (iii) “av sq unc corr res all” is the average squared unconditional of factors regression residual. (iv) The confidence band is the minimal value above which squared correlation is significant at 5% level. It is computed from the t-squared ratio test t^ijt2=(ρ^ijt*)2[1−ρ^ijt*]−1(N−2)∼F(1.N−2) and is equal to 1.6990. Table VIII shows the descriptive statistics for the returns and residual average squared correlations estimated over the full sample. We draw three main conclusions from this table. First, while the means of ρ^ret,t* and ρ^t* are very similar over the whole sample, there is a notable difference—almost equal to 10% in some months—between the two measures over the 2008–13 period. This emphasizes the importance of filtering returns using some measures of fundamentals and shows that the rise in commodity-returns correlation is partly due to common factors. Juvenal and Petrella (2015) find that the co-movements between the prices of oil and other commodities reflect global demand shocks. We are partially in line with them in that, once factors related to demand are taken into account, the residual correlation is lower, with this effect being stronger in recent years when demand shocks were larger. Table VIII. Descriptive statistics on returns and residuals average squared unconditional correlations (i) This table reports summary statistics on average squared unconditional return correlation ρ^ret.t* and average squared unconditional residual correlation ρ^t*. (ii) Fρ*2 is the mean percentage of average squared unconditional correlation significant at the 5% level using the t-squared ratio test t^ijt2=(ρ^ijt*)2[1−ρ^ijt*]−1(N−2)∼F(1.N−2). (iii) ***, **, and *, respectively, denote significance at 1%, 5%, and 10%, levels. (iv) Cρ is the correlation between ρ^ret.t* and ρ^t*.   ρ^ret.t*  ρ^t*  μ  0.1982**  0.1803**  σ  0.0455  0.0246  Fρ*2  0.6230  0.6440  Cρ  0.9319      ρ^ret.t*  ρ^t*  μ  0.1982**  0.1803**  σ  0.0455  0.0246  Fρ*2  0.6230  0.6440  Cρ  0.9319    Second, looking at both ρ^t* and ρ^t, taking time-variation in volatility into consideration only moderately affects the estimated correlation: the two lines are almost identical except in periods of high volatility, where there is a difference (although only small) between the two measures. Third, and most importantly, our measure of excess co-movement is significant at the 5% level only half of the time in the period under consideration.24 We thus conclude that the excess co-movement in commodity prices cannot be viewed as a general feature of commodity markets but is rather sample-dependent. As PR do not investigate time-variation in their excess co-movement measure, our results cannot be compared to theirs. There is a possibility, however, that the estimated excess co-movement over the 1960–85 period that PR find is insignificant over some sub-samples, thereby questioning the determinants of this phenomenon. In the same vein as the correlation plot in Tang and Xiong (2012), the chart of average squared correlations in Figure 1 provides a finer description of the estimated excess co-movement. This latter is mostly significant during periods of financial crisis: from mid-2000 to early 2003, and from 2008 onward. In their “convective risk flows” model, Cheng, Kirilenko, and Xiong (2015) show that financial traders cut their net long positions in response to market distress. A coordinated drop in the long positions of financial traders may help explain excess co-movement. Alternatively, excess co-movement may also reflect a “flight-to-quality” phenomenon, where investors decide to partly leave the stock market and invest heavily in commodities to diversify their positions. Moreover, the period starting in 2000 also corresponds to the growing financialization of commodity futures markets, as excellently surveyed in Cheng and Xiong (2014a). As such, excess co-movement might be related to speculative activity in commodity futures markets. Whether excess co-movement comes the changing nature of trading in commodity markets is a central question that we answer in the next section. 5. Explaining Excess Co-movement The literature on institutional investors and their possible impact on commodity prices has grown dramatically in recent years [see the nice surveys in Irwin and Sanders (2011), Cheng and Xiong (2014a), and Haase, Zimmermann, and Zimmermann (2016)]. Prior research has, however, produced mixed results. While some authors have produced evidence of a significant effect of index funds on commodity prices (Tang and Xiong, 2012; Singleton, 2013, among others), others have found evidence to the contrary (Rouwenhorst and Tang, 2012; Hamilton and Wu, 2015; Lehecka, 2015). Surprisingly, there is no work dealing with the impact of financialization on cross-market return linkages except for that in Tang and Xiong (2012).25 The latter attempt to explain the recent rise in the co-movement of a number of commodity prices via five hypotheses: (i) the financialization of commodities, (ii) the rapid growth of emerging economies, (iii) the recent world financial crisis, (iv) inflation, and (v) the adoption of biofuels. Our research question is linked to the arguments in Tang and Xiong (2012), in that we arguably jointly test their first and third hypotheses, and consider the second and fourth in Section 3 when we filter returns using common factors. In particular, we have shown that growth in emerging economies leads commodity prices (hypothesis (ii)), and that commodity returns are correlated with a “nominal variables” factor (hypothesis (iv)). Both of these effects likely contribute to excess co-movement and are expressly taken into account in our work. This section aims to show that speculation in commodity futures markets is a significant determinant of our estimated excess co-movement. The issue is new and challenging, as no significant evidence has been put forward in the literature to date. Our empirical approach is as follows. In a first step, we show that speculative activity and filtered futures returns are correlated for most of the commodities in our sample. Then, in a second step, we show that measures of speculative activity are correlated across commodities. Taken together, these results provide direct evidence of speculation as a driver of excess co-movement.26 5.1 Measure of Speculative Intensity Our measure of speculative intensity builds on the work in Han (2008), where a new speculative index is developed following the literature on investor sentiment (see Baker and Wurgler, 2007).27 The basic idea is to pick up the net view of speculators in a given futures market by comparing their long and short positions. Han’s index is given by the number of long non-commercial contracts minus the number of short non-commercial contracts, scaled by the total open interest in futures markets for the commodity of interest; as such this is a directional index of speculative activity in the futures market. We calculate Han’s index for our eight commodities using CFTC data. All traders who are considered as large enough—positions are above a specified level that is commodity-dependent—are required to provide the CFTC with their daily positions. The Commitments of Traders (COT) Report corresponds to the weekly aggregation of the daily positions and is released each Tuesday. CFTC differentiates between “commercial” and “non-commercial” traders and provides long and short positions for both categories.28 “Commercial” traders should be able to prove an involvement in the physical market and are thus considered as hedgers while “non-commercial” traders have no relation with the cash business: this latter group then consists of speculators. The usefulness of these CFTC data has previously been discussed in Bessembinder (1992) and Cheng and Xiong (2014b), as many traders carry out activities which cover both hedging and speculation. In particular, Cheng and Xiong (2014b) show that hedgers may react to changes in commodity futures prices in a number of US Agricultural markets, which is undoubtedly a form of speculation. In what follows, we show that CFTC data are informative for our purpose despite the potential bias in the definition of categories of traders. We also experiment with alternative measures of trading activity. The first of these is the Working’s T speculative index, as recently used in Byksahin and Robe (2014). A second measure of trading activity is hedging pressure, as defined in de Roon, Nijman, and Veld (2000), who showed that futures risk premia depend on both own-market and cross-market hedging pressure. Their measure of hedging pressure is calculated as the difference between the number of short and long hedge positions, divided by the total number of hedge positions. This measure focuses on the positions of traders who are hedgers, that is, who have a cash business for the commodity. It is different from the Han index, where the denominator is the total open interest and not the total number of speculative positions, but the idea is roughly similar as hedging pressure also picks up the difference between long and short positions.29 Results from using either Working’s T or hedging pressure are very similar to those presented here, and are not reported to save space (but available upon request). 5.2 Empirical Results To deal with the potential correlation of Han’s indices and the business cycle, we regress our speculative indices on the same set of factors ( F^1t,F^2t,F^3t,F^6t,F^8t) as was used to filter commodity returns. Then, to gauge the explanatory power of Han’s indices, we include them in univariate regressions of the form:   Resi,t=ai+∑j=18bi,jHj,t+ei,t, where Resi,t is the ith commodity return residual at time t and Hj,t is Han’s index for the jth commodity at time t adjusted for the factors. Our choice to use contemporaneous variables in the regressions is motivated by the monthly frequency and the efficient-market hypothesis stating that any impact of index funds should be instantaneously reflected in prices [see Gilbert and Pfuderer (2014) for further developments on this issue]. We also choose to consider all the Han indices in each univariate regression following existing research on cross-commodity trading and its potential impact on prices (e.g., de Roon, Nijman, and Veld, 2000).30 The estimated coefficients appear in Table IX. We observe that, with the exception of oil, copper, and silver, commodity returns are correlated with their own Han’s index. The R2 mostly ranges between 1% and 3%, but reaches 12.29% for raw sugar. We find a positive and significant impact of the Han index on its corresponding commodity for wheat, raw sugar, soybeans, and live cattle. For these commodities, speculative trading and returns move in the same direction. We also observe some cross-effects: the raw sugar and copper Han indices have an impact on wheat returns, and the cotton Han index has an effect on crude oil returns. In these cases, the estimated coefficient is negative but only weakly significant. The Han index for wheat has a positive effect on live cattle residual. The interpretation of these cross-effects is quite challenging as there is no link, such as substitutability or complementarity, between the commodities concerned. Table IX. OLS regressions of residual returns on speculative Han indices (i) This table reports OLS estimates of the regression of the eight commodities monthly residual returns on Han speculative indices. (ii) The Han indices are corrected for the factors {F^1t,F^2t,F^3t,F^6t,F^8t} to control for the effect of the business cycle. (iii) t-statistics are reported in parenthesis under the estimates. ***, **, and *, respectively, denote rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0009      −0.0002  −0.0013  −0.0001  0.0004  0.0000    (0.20)      (−0.04)  (−0.23)  (−0.03)  (0.10)  (0.02)  Han_Wheat  0.0778***              0.0429*    (2.67)              (1.88)  Han_Copper  −0.0597*                  (−1.92)                Han_Silver                  Han_Soybeans        0.0639*                  (1.79)          Han_Raw Sugar  −0.0514*        0.2063***          (−1.90)        (5.52)        Han_Cotton            0.0636***  −0.0396*                (2.98)  (−1.79)    Han_Crude Oil                  Han_Live Cattle                0.0816***                  (4.05)  R2  0.0286      0.0110  0.1229  0.0206  0.0101  0.0397  R¯  0.0166      0.0070  0.1194  0.0166  0.0061  0.0318    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0009      −0.0002  −0.0013  −0.0001  0.0004  0.0000    (0.20)      (−0.04)  (−0.23)  (−0.03)  (0.10)  (0.02)  Han_Wheat  0.0778***              0.0429*    (2.67)              (1.88)  Han_Copper  −0.0597*                  (−1.92)                Han_Silver                  Han_Soybeans        0.0639*                  (1.79)          Han_Raw Sugar  −0.0514*        0.2063***          (−1.90)        (5.52)        Han_Cotton            0.0636***  −0.0396*                (2.98)  (−1.79)    Han_Crude Oil                  Han_Live Cattle                0.0816***                  (4.05)  R2  0.0286      0.0110  0.1229  0.0206  0.0101  0.0397  R¯  0.0166      0.0070  0.1194  0.0166  0.0061  0.0318  One may rightly suspect that these OLS estimates are plagued by endogeneity. To assess the presence of endogeneity, we estimate all previous regressions via GMM and use one-period and two-period lagged Han indices as instruments.31 The results in Table X show that the test for exogeneity based on the difference in the J-test does not reject the exogeneity of the Han index for cotton, crude oil. We therefore consider the previous OLS estimates as valid: the Han index has a positive effect on its own commodity for cotton while that for cotton has a negative impact on oil. We reject the exogeneity of the Han index in the wheat, raw sugar, soybeans, and live cattle regressions. Hansen’s (1982)J-test for overidentifying restrictions validates our set of instruments. We also check that the instruments are not weak. With the exception of soybeans, the GMM estimates are in line with the previous OLS regressions for these four commodities. Wheat return is still negatively and significantly impacted by the Han index for raw sugar, although the indices for wheat and copper are no longer significant. The raw sugar Han index still has a positive impact on its return. The wheat Han indices have a positive and significant impact on live cattle residual return. Table X. GMM regressions of residual returns on speculative Han indices; Instruments {Hant−1,Hant−2} (i) This table reports GMM estimates of the regression of the eight commodities monthly residuals returns. The explanatory variables are reported in far-left column. (ii) The Han indices are corrected for the factors {F^1t,F^2t,F^3t,F^6t,F^8t} to control for the effect of the business cycle. (iii) t-statistics are reported in parenthesis under the estimates. ***, **, and *, respectively, denotes rejection of the null hypothesis of no significance at the 1%, 5%, and 10% levels. (iv) J-test is the Hansen (1982) test of overidentifying restrictions and the Diff in J-test is the test for endogeneity of regressors. P-values are reported under the test-statistics.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0010      0.0008  0.0009  −0.0008  0.0003  0.0005    (0.21)      (0.17)  (0.18)  (−0.16)  (0.09)  (0.20)  Han_Wheat  −0.0351              0.0555*    (−0.61)              (1.85)  Han_Copper  0.0010                  (0.02)                Han_Silver                  Han_Soybeans        −0.0252                  (−0.65)          Han_Raw Sugar  −0.1018**        0.0987**          (−2.57)        (2.33)        Han_Cotton            0.0439  −0.0362                (1.59)  (−1.47)    Han_Crude Oil                  Han_Live Cattle                0.0122                  (0.36)  Exogeneity test                  Diff in J-test  9.64**      7.75***  17.28***  1.83  0.0678  10.75***  pval  0.0219      0.0054  0.0000  0.1750  0.79  0.00  Overidentifying test  J-test  1.1770      1.3462  2.5290  0.4705  0.0077  1.028  pval  0.75      0.9966  0.11  0.2459  0.49  0.31    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Intercept  0.0010      0.0008  0.0009  −0.0008  0.0003  0.0005    (0.21)      (0.17)  (0.18)  (−0.16)  (0.09)  (0.20)  Han_Wheat  −0.0351              0.0555*    (−0.61)              (1.85)  Han_Copper  0.0010                  (0.02)                Han_Silver                  Han_Soybeans        −0.0252                  (−0.65)          Han_Raw Sugar  −0.1018**        0.0987**          (−2.57)        (2.33)        Han_Cotton            0.0439  −0.0362                (1.59)  (−1.47)    Han_Crude Oil                  Han_Live Cattle                0.0122                  (0.36)  Exogeneity test                  Diff in J-test  9.64**      7.75***  17.28***  1.83  0.0678  10.75***  pval  0.0219      0.0054  0.0000  0.1750  0.79  0.00  Overidentifying test  J-test  1.1770      1.3462  2.5290  0.4705  0.0077  1.028  pval  0.75      0.9966  0.11  0.2459  0.49  0.31  Our approach through instrumental variables unambiguously shows that there is a significant impact of changes in the speculative index on contemporaneous returns, even after controlling for endogeneity for most commodities. This impact is positive when the Han index and the return pertain to the same commodity. We now focus on sample cross-correlations between speculative intensities in commodity futures markets. As expected from the “style investing” hypothesis developed in Barberis and Shleifer (2003) or the more general increase in non-commercial positions in commodity-futures markets in the last decades (see Cheng and Xiong, 2014a), the cross-correlations between speculative indices are mostly positive and significant as shown in Table XI. There are, respectively, 10, 15, and 19 significant cross-correlations at the 1%, 5%, and 10% significance levels. The cross-correlations are significantly negative in only three cases (silver and raw sugar, silver and soybeans, wheat and live cattle). We therefore have evidence that speculative indices move together, even for commodities of different classes such as, for instance, wheat and copper, cotton and crude oil, or raw sugar and live cattle. Table XI. Correlation between the eight Han indices (i) The upper triangular matrix reports correlations while the lower reports their p-values. ***, **, and *, respectively, denote significance at 1%, 5%, and 10% levels. (ii) The Han indices are corrected for the factors {F^1t,F^2t,F^3t,F^6t,F^8t} to control for the effect of the business cycle.   Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.1744***  0.0107  0.2133***  −0.0334  0.1394**  −0.0226  −0.1194*  Copper  0.00  1  0.2146***  0.0029  0.1271**  0.1888***  0.1071*  0.2227***  Silver  0.86  0.00  1  −0.1408**  −0.1128*  −0.0935  −0.1024  −0.0925  Soybeans  0.00  0.96  0.02  1  0.1853***  0.4619***  0.2787***  0.0830  Raw sugar  0.59  0.04  0.07  0.00  1  0.0222  0.1060*  0.2280***  Cotton  0.02  0.00  0.14  0.00  0.72  1  0.3005***  0.1324**  Crude oil  0.72  0.09  0.10  0.00  0.09  0.00  1  0.1102*  Live cattle  0.06  0.00  0.14  0.19  0.00  0.03  0.08  1    Wheat  Copper  Silver  Soybeans  Raw sugar  Cotton  Crude oil  Live cattle  Wheat  1  0.1744***  0.0107  0.2133***  −0.0334  0.1394**  −0.0226  −0.1194*  Copper  0.00  1  0.2146***  0.0029  0.1271**  0.1888***  0.1071*  0.2227***  Silver  0.86  0.00  1  −0.1408**  −0.1128*  −0.0935  −0.1024  −0.0925  Soybeans  0.00  0.96  0.02  1  0.1853***  0.4619***  0.2787***  0.0830  Raw sugar  0.59  0.04  0.07  0.00  1  0.0222  0.1060*  0.2280***  Cotton  0.02  0.00  0.14  0.00  0.72  1  0.3005***  0.1324**  Crude oil  0.72  0.09  0.10  0.00  0.09  0.00  1  0.1102*  Live cattle  0.06  0.00  0.14  0.19  0.00  0.03  0.08  1  Overall, our empirical results demonstrate that speculative activity is a significant driver of excess co-movement. We thus confirm the implications in Basak and Pavlova (2016) that institutional investors do play a role in linking commodity futures prices. Our results are also in line with those in Tang and Xiong (2012), but provide stronger evidence of the impact of speculation on co-movements as we on purpose control for the impact of real variables on commodity prices. More generally, our results demonstrate the critical role of trading for price determination, and the overall importance of the “financialization of the commodity markets”, a concept that has attracted growing interest in academic and political spheres over recent years. 6. Concluding Remarks The aim of this paper was to reconsider the question of the excess co-movement of commodity prices and to provide an explanation of this phenomenon, if it was found to be present in the data. We believe that our paper offers new perspectives for the analysis of co-movement in commodity returns. First, as discussed above, we use the large approximate factor model method to uncover the relevant factors that allow us to explain commodity returns. To the best of our knowledge, this is the first time that this method has been used to filter out returns before looking for excess co-movement. The main advantage of factors is that they allow us to deal with a large number of variables, while retaining econometric tractability, thereby including a richer set of fundamentals. We thus avoid any artificial limit on the information set, which has been a major constraint in previous work. Our second contribution is to provide an explanation of the excess co-movement in commodity returns. Previous work has emphasized the methodological aspects of the assessment of the hypothesis of excess co-movements. Surprisingly, however, the issue of which variables are related to this phenomenon has not been analyzed to date. Our indicator of speculative activity, calculated using traders’ positions available from CFTC, is both correlated across commodities and with futures prices, thereby providing evidence of speculation as a driver of excess co-movement. The limits of our analysis are good topics for future research. First, we consider, as in most factor-models in the literature, the factors as if they were data rather than being estimated. Even if this may have only a small effect on our results, it would be useful to investigate the small-sample case using simulation techniques as in Ludvigson and Ng (2007, 2009). Second, mixed-data sampling (MIDAS) regressions may be used to include more information at different frequencies. Tang and Xiong (2012) consider daily and monthly regressions, and MIDAS may help to combine the two data sources, with daily market indices and monthly or quarterly macroeconomic variables. This is the setting in Karali and Power (2013), who mix high- and low-frequency variables to explain the volatility of commodity returns. Such a setting may allow us to consider volatility spillovers, as in the penultimate section in Tang and Xiong (2012). The analysis of commodity volatility co-movement may have interesting implications for financial risk management. Third, alternative measures of trading activity, such as liquidity measures, may help better explain excess co-movement. In this respect, the recent contributions of Marshall, Nguyen, and Visaltanachoti (2012, 2013) may aid in the selection of appropriate liquidity measures for commodities and the evaluation of the explanatory power of their common liquidity factor. These measures may additionally be calculated on a daily basis, thereby permitting the high-frequency analysis of the common evolution of commodity prices. Appendix A Table AI. List of the 184 variables considered in the computation of the common factors In the Trans column, we report the transformation used to ensure the stationarity of each variable. ln denotes the logarithm, Δln  and Δ2ln denote the first and second difference of the logarithm, lv denotes the level of the series, and Δlv denotes the first difference of the series. Developed countries   Series number  Short name  Mnemonic  Trans  Description  Industrial production  1  IP: USA  USIPMAN.G  Δln   US INDUSTRIAL PRODUCTION—MANUFACTURING (NAICS) VOLA  2  IP: France  FRIPMAN.G  Δln   FR INDUSTRIAL PRODUCTION—MANUFACTURING VOLA  3  IP: France  FRINDSYNQ  lv  FR SURVEY: MANUFACTURING—SYNTHETIC BUSINESS INDICATOR SADJ  4  IP: Germany  BDIPMAN.G  Δln   BD INDUSTRIAL PRODUCTION: MANUFACTURING VOLA  5  IP: UK  UKIPMAN.G  Δln   UK INDUSTRIAL PRODUCTION INDEX—MANUFACTURING VOLA  6  IP: Japan  JPIPMAN.G  Δln   JP INDUSTRIAL PRODUCTION—MANUFACTURING VOLAJ  7  IP: Japan  JPIPTOT.G  Δln   JP INDUSTRIAL PRODUCTION—MINING & MANUFACTURING VOLA (2005 = 100)  Orders and capacity utilization  8  Capacity utilization: USA  USCUMANUG  Δlv  US CAPACITY UTILIZATION—MANUFACTURING VOLA  9  Manufct. new ord.: USA  USNOCOGMC  Δln   US MANUFACTURERS NEW ORDERS—CONSUMER GOODS & MATERIALS CONN (BASE 1982)  10  Manufct. new ord.: USA  USNOMXTRB  Δln   US NEW ORDERS—MANUFACTURING. EXCLUDING TRANSPORTATION CURA  11  New orders: Canada  CNNEWORDB  Δln   CN NEW ORDERS: ALL MANUFACTURING INDUSTRIES (SA) CURA  12  Manufct. ord.: Germany  BDNEWORDE  Δln   BD MANUFACTURING ORDERS SADJ (2000 = 100)  13  Manufct. ord.: Japan  JPNEWORDB  Δln   JP MACHINERY ORDERS: DOM.DEMAND–PRIVATE DEMAND (EXCL. SHIP) CURA  14  Operating ratio: Japan  JPCAPUTLQ  Δlv  JP OPERATING RATIO—MANUFACTURING SADJ (2005 = 100)  15  Business failures: Japan  JPBNKRPTP  Δln   JP BUSINESS FAILURES VOLN  Housing start  16  Housing permits: USA  USHOUSE.O  Δln   US NEW PRIVATE HOUSING UNITS STARTED (AR) VOLA  17  Housing permits: Canada  CNHOUSE.O  Δln   CN HOUSING STARTS: ALL AREAS (SA. AR) VOLA  18  Housing permits: Germany  BDHOUSE.G  Δln   BD CONSTRUCTION ORDERS RECEIVED—RESIDENTIAL CONSTRUCTION VOLA  19  Housing permits: Australia  AUHOUSE_A  Δln   AU BUILDING APPROVALS: NEW HOUSES CURN  20  Housing permits: Japan  JPHOUSSTF  ln   JP NEW HOUSING CONSTRUCTION STARTED VOLN  Car sales  21  Car registration: USA  USCAR…P  Δln   US NEW PASSENGER CARS—TOTAL REGISTRATIONS VOLN  22  Car registration: France  FRCARREGP  Δln   FR NEW CAR REGISTRATIONS VOLN  23  Car registration: Germany  BDRVNCARP  ln   BD NEW REGISTRATIONS—CARS VOLN  24  Car registration: UK  UKCAR…P  Δln   UK CAR REGISTRATIONS VOLN  25  Car registration: Japan  JPCARREGF  ln   JP MOTOR VEHICLE NEW REGISTRATIONS: PASSENGER CARS EXCL. BELOW 66  Consumption  26  Consumer sentiment: USA  USUMCONEH  Δln   US UNIV OF MICHIGAN CONSUMER SENTIMENT—EXPECTATIONS VOLN (BASE 1966 = 100)  27  Pers. cons. exp.: USA  USPERCONB  Δln   US PERSONAL CONSUMPTION EXPENDITURES (AR) CURA  28  Pers. saving: USA  USPERSAVE  Δlv  US PERSONAL SAVING AS % OF DISPOSABLE PERSONAL INCOME SADJ  29  Retail sale: Canada  CNRETTOTB  Δln   CN RETAIL SALES: TOTAL (ADJUSTED) CURA  30  Household confidence: France  FRCNFCONQ  Δlv  FR SURVEY—HOUSEHOLD CONFIDENCE INDICATOR SADJ  31  Household confidence: Germany  BDCNFCONQ  lv  BD CONSUMER CONFIDENCE INDICATOR—GERMANY SADJ  32  Retail sales: UK  UKRETTOTB  Δln   UK RETAIL SALES (MONTHLY ESTIMATE. DS CALCULATED) CURA  33  Household confidence: UK  UKCNFCONQ  Δlv  UK CONSUMER CONFIDENCE INDICATOR—UK SADJ  34  Retail sales: Australia  AURETTOTT  Δln   AU RETAIL SALES (TREND) VOLA  35  Household confidence: Australia  AUCNFCONR  lv  AU MELBOURNE/WESTPAC CONSUMER SENTIMENT INDEX NADJ  36  Household expenditure: Japan  JPHLEXPWA  Δln   JP WORKERS HOUSEHOLD LIVING EXPENDITURE (INCL. AFF) CURN  37  Retail sales: Japan  JPRETTOTA  Δln   JP RETAIL SALES CURN  Wages and labor  38  Av. Hourly Real Earnings: USA  USWRIM.D  Δln   US AVG HOURLY REAL EARNINGS—MANUFACTURING CONA (BASE 82–84)  39  Av. overtime hours: USA  USOOL024Q  Δlv  US OVERTIME HOURS—MANUFACTURING. WEEKLY VOLA  40  Av. wkly hours: USA  USHKIM_O  Δlv  US AVG WKLY HOURS—MANUFACTURING VOLA  41  Av. hourly real earnings: Canada  CNWAGES.A  Δln   CN AVG. HOURLY EARN—INDUSTRIAL AGGREGATE EXCL. UNCLASSIFIED CURN  42  Labor productivity: Germany  BDPRODVTQ  Δln   BD PRODUCTIVITY: OUTPUT PER MAN-HOUR WORKED IN INDUSTRY SADJ (2005 = 100)  43  wages: Germany  BDWAGES.F  Δln   BD WAGE & SALARY.OVERALL ECONOMY—ON A MTHLY BASIS (PAN BD M0191)  44  wages index: Japan  JPWAGES_E  Δln   JP WAGE INDEX: CASH EARNINGS—ALL INDUSTRIES SADJ  Unemployment  45  U rate: USA  USUNEM15Q  Δlv  US UNEMPLOYMENT RATE—15 WEEKS & OVER SADJ  46  U rate: USA  USUNTOTQ_pc  Δlv  US UNEMPLOYMENT RATE SADJ  47  Employment: Canada  CNEMPTOTO  Δln   CN EMPLOYMENT—CANADA (15 YRS & OVER SA) VOLA  48  U all: Germany  BDUNPTOTP  Δln   BD UNEMPLOYMENT LEVEL (PAN BD FROM SEPT. 1990) VOLN  49  U rate: UK  UKUNTOTQ_pc  Δlv  UK UNEMPLOYMENT RATE SADJ  50  Emp: Australia  AUEMPTOTO  Δln   AU EMPLOYED: PERSONS VOLA  51  U all: Australia  AUUNPTOTO  Δln   AU UNEMPLOYMENT LEVEL VOLA  52  U rate: Japan  JPUNTOTQ_pc  Δlv  JP UNEMPLOYMENT RATE SADJ  International trade  53  Exports: USA  USI70_A  Δln   US EXPORTS CURN  54  Exports: France  FREXPGDSB  Δln   FR EXPORTS FOB CURA  55  Exports: Germany  BDEXPBOPB  Δln   BD EXPORTS FOB CURA  56  Exports: UK  UKI70_A  Δln   UK EXPORTS CURN  57  Exports: Australia  AUEXPG&SB  Δln   AU EXPORTS OF GOODS & SERVICES (BOP BASIS) CURA  58  Exports: Japan  JPEXPGDSB  Δln   JP EXPORTS OF GOODS—CUSTOMSBASIS CURA  59  Imports: USA  USIMPGDSB  Δln   US IMPORTS F.A.S. CURA  60  Imports: France  FRIMPGDSB  Δln   FR IMPORTS FOB CURA  61  Imports: Germany  BDIMPGDSB  Δln   BD IMPORTS CIF (PAN BD M0790) CURA  62  Imports: UK  UKIMPBOPB  Δln   UK IMPORTS—BALANCE OF PAYMENTS BASIS CURA  63  Imports: Australia  AUIMPG&SB  Δln   AU IMPORTS OF GOODS & SERVICES (BOP BASIS) CURA  64  Imports: Japan  JPOXT009B  Δln   JP IMPORTS CURA  65  Terms of trade: UK  UKTOTPRCF  Δln   UK TERMS OF TRADE—EXPORT/IMPORT PRICES (BOP BASIS) NADJ  66  Terms of trade: Japan  JPTOTPRCF  Δln   JP TERMS OF TRADE INDEX NADJ  Money and credit  67  Money supply: USA  USM0_A  Δln   US MONETARY BASE CURA  68  Money supply: USA  USM2_B  Δln   US MONEY SUPPLY M2 CURA  69  Money supply: France  FRM2_A  Δln   FR MONEY SUPPLY—M2 (NATIONAL CONTRIBUTION TO M2) CURN  70  Money supply: France  FRM3_A  Δln   FR MONEY SUPPLY—M3 (NATIONAL CONTRIBUTION TO M3) CURN  71  Money supply: Germany  BDM1_A  Δln   BD MONEY SUPPLY—GERMAN CONTRIBUTION TO EURO M1 (PAN BD M0790)  72  Money supply: Germany  BDM3_B  Δln   BD MONEY SUPPLY—M3 (CONTRIBUTION TO EURO BASIS FROM M0195) CURA  73  Money supply: UK  UKM1_B  Δln   UK MONEY SUPPLY M1 (ESTIMATE OF EMU AGGREGATE FOR THE UK) CURA  74  Money supply: UK  UKM3_B  Δln   UK UK MONEY SUPPLY M3 (ESTIMATE OF EMU AGGREGATE FORTHE UK) CURA  75  Money supply: Australia  AUM1_B  Δln   AU MONEY SUPPLY—M1 CURA  76  Money supply: Australia  AUM3_B  Δln   AU MONEY SUPPLY - M3 (SEE AUM3…OB) CURA  77  Money supply: Japan  JPM1_A  Δln   JP MONEY SUPPLY: M1 (METHO-BREAK. APR. 2003) CURN  78  Money supply: Japan  JPM2_A  Δln   JP MONEY SUPPLY: M2 (METHO-BREAK. APR. 2003) CURN  79  Credit: USA  USCOMILND  Δln   US COMMERCIAL & INDUSTRIAL LOANS OUTSTANDING (BCI 101) CONA (BASE 2005)  80  Credit: USA  USCILNNCB  lv  US COMMERCIAL & INDL LOANS. NET CHANGE (AR) (BCI 112) CURA  81  Credit: USA  USCRDNRVB  Δln   US NONREVOLVING CONSUMER CREDIT OUTSTANDING CURA  82  Credit: France  FRBANKLPA  Δln   FR MFI LOANS TO RESIDENT PRIVATE SECTOR CURN  83  Credit: Germany  BDBANKLPA  Δ2ln   BD LENDING TO ENTERPRISES & INDIVIDUALS CURN  84  Credit: Australia  AUCRDCONB  Δ2ln   AU FINANCIAL INTERMEDIARIES: NARROW CREDIT—PRIVATE SECTOR CURA  85  Credit: Japan  JPBANKLPA  Δ2ln   JP AGGREGATE BANK LENDING (EXCL. SHINKIN BANKS) CURN  Stock index  86  Stock index: USA  USSHRPRCF  Δln   US DOW JONES INDUSTRIALS SHARE PRICE INDEX (EP) NADJ  87  Stock index: France  FRSHRPRCF  Δln   FR SHARE PRICE INDEX—SBF 250 NADJ  88  Stock index: Germany  BDSHRPRCF  Δln   BD DAX SHARE PRICE INDEX. EP NADJ  89  Stock index: UK  UKOSP001F  Δln   UK FTSE 100 SHARE PRICE INDEXNADJ (2005 = 100)  90  Stock index: Japan  JPSHRPRCF  Δln   JP TOKYO STOCK EXCHANGE—TOPIX (EP) NADJ (1968 = 100)  Interest rate  91  Interest rate: USA  USFEDFUN  Δlv  US FEDERAL FUNDS RATE (AVG.)  92  Interest rate: USA  USCRBBAA  Δlv  US CORPORATE BOND YIELD—MOODY’S BAA. SEASONED ISSUES  93  Interest rate: USA  USGBOND  Δlv  US TREASURY YIELD ADJUSTED TO CONSTANT MATURITY—20 YEAR  94  Interest rate: France  FRPRATE  Δlv  FR AVERAGE COST OF FUNDS FOR BANKS/EURO REPO RATE  95  Interest rate: France  FRGBOND  Δlv  FR GOVERNMENT GUARANTEED BOND YIELD (EP) NADJ  96  Interest rate: Germany  BDPRATE  Δlv  BD DISCOUNT RATE/SHORT TERM EURO REPO RATE  97  Interest rate: Germany  BDGBOND  Δlv  BD LONG TERM GOVERNMENT BOND YIELD—9–10 YEARS  98  Interest rate: UK  UKPRATE  Δlv  UK BANK OF ENGLAND BASE RATE (EP)  99  Interest rate: UK  UKGBOND  Δlv  UK GROSS REDEMPTION YIELD ON 20 YEAR GILTS (PERIOD AVERAGE) NADJ  100  Interest rate: Australia  AUPRATE  Δlv  AU RBA CASH RATE TARGET  101  Interest rate: Australia  AUBOND  Δlv  AU COMMONWEALTH GOVERNMENT BOND YIELD 10 YEAR (EP)  102  Interest rate: Japan  JPGBOND  Δlv  JP INTEREST-BEARING GOVERNMENT BONDS—10-YEAR (EP)  Exchange rate  103  Exchange rate: DM to US$  BBDEMSP  Δln   GERMAN MARK TO US$ (BBI)—EXCHANGE RATE  104  Exchange rate: SK to US$  SDXRUSD  Δln   SD SWEDISH KRONOR TO US$ (BBI. EP)  105  Exchange rate: to $  UKDOLLR  Δln   UK TO US$ (WMR)—EXCHANGE RATE  106  Exchange rate: Yen to $  JPXRUSD  Δln   JP JAPANESE YEN TO US$  107  Exchange rate: Aus.$ to US$  AUXRUSD  Δln   AU AUSTRALIAN $ TO US$ (MTH. AVG.)  Producer price index  108  PPI: USA  USPROPRCE  Δln   US PPI—FINISHED GOODS SADJ  109  PPI: Canada  CNPROPRCF  Δln   CN INDUSTRIAL PRICE INDEX: ALL COMMODITIES NADJ  110  PPI: Germany  BDPROPRCF  Δln   BD PPI: INDL. PRODUCTS. TOTAL. SOLD ON THE DOMESTIC MARKET NADJ (2005 = 100)  111  PPI: UK  UKPROPRCF  Δln   UK PPI—OUTPUT OF MANUFACTURED PRODUCTS (HOME SALES) NADJ  112  PPI: Japan  JPPROPRCF  Δln   JP CORPORATE GOODS PRICE INDEX: DOMESTIC—ALL COMMODITIES NADJ  Consumer price index  113  CPI: USA  USCONPRCE  Δln   US CPI—ALL URBAN: ALL ITEMS SADJ  114  CPI: Canada  CNCONPRCF  Δln   CN CPI NADJ  115  CPI: France  FRCONPRCE  Δln   FR CPI SADJ  116  CPI: Germany  BDCONPRCE  Δln   BD CPI SADJ  117  CPI: UK  UKCONPRCF  Δln   UK CPI INDEX 00: ALL ITEMS- ESTIMATED PRE-97 2005 = 100 NADJ  118  CPI: Japan  JPCONPRCF  Δln   JP CPI: NATIONAL MEASURE NADJ    Developed countries   Series number  Short name  Mnemonic  Trans  Description  Industrial production  1  IP: USA  USIPMAN.G  Δln   US INDUSTRIAL PRODUCTION—MANUFACTURING (NAICS) VOLA  2  IP: France  FRIPMAN.G  Δln   FR INDUSTRIAL PRODUCTION—MANUFACTURING VOLA  3  IP: France  FRINDSYNQ  lv  FR SURVEY: MANUFACTURING—SYNTHETIC BUSINESS INDICATOR SADJ  4  IP: Germany  BDIPMAN.G  Δln   BD INDUSTRIAL PRODUCTION: MANUFACTURING VOLA  5  IP: UK  UKIPMAN.G  Δln   UK INDUSTRIAL PRODUCTION INDEX—MANUFACTURING VOLA  6  IP: Japan  JPIPMAN.G  Δln   JP INDUSTRIAL PRODUCTION—MANUFACTURING VOLAJ  7  IP: Japan  JPIPTOT.G  Δln   JP INDUSTRIAL PRODUCTION—MINING & MANUFACTURING VOLA (2005 = 100)  Orders and capacity utilization  8  Capacity utilization: USA  USCUMANUG  Δlv  US CAPACITY UTILIZATION—MANUFACTURING VOLA  9  Manufct. new ord.: USA  USNOCOGMC  Δln   US MANUFACTURERS NEW ORDERS—CONSUMER GOODS & MATERIALS CONN (BASE 1982)  10  Manufct. new ord.: USA  USNOMXTRB  Δln   US NEW ORDERS—MANUFACTURING. EXCLUDING TRANSPORTATION CURA  11  New orders: Canada  CNNEWORDB  Δln   CN NEW ORDERS: ALL MANUFACTURING INDUSTRIES (SA) CURA  12  Manufct. ord.: Germany  BDNEWORDE  Δln   BD MANUFACTURING ORDERS SADJ (2000 = 100)  13  Manufct. ord.: Japan  JPNEWORDB  Δln   JP MACHINERY ORDERS: DOM.DEMAND–PRIVATE DEMAND (EXCL. SHIP) CURA  14  Operating ratio: Japan  JPCAPUTLQ  Δlv  JP OPERATING RATIO—MANUFACTURING SADJ (2005 = 100)  15  Business failures: Japan  JPBNKRPTP  Δln   JP BUSINESS FAILURES VOLN  Housing start  16  Housing permits: USA  USHOUSE.O  Δln   US NEW PRIVATE HOUSING UNITS STARTED (AR) VOLA  17  Housing permits: Canada  CNHOUSE.O  Δln   CN HOUSING STARTS: ALL AREAS (SA. AR) VOLA  18  Housing permits: Germany  BDHOUSE.G  Δln   BD CONSTRUCTION ORDERS RECEIVED—RESIDENTIAL CONSTRUCTION VOLA  19  Housing permits: Australia  AUHOUSE_A  Δln   AU BUILDING APPROVALS: NEW HOUSES CURN  20  Housing permits: Japan  JPHOUSSTF  ln   JP NEW HOUSING CONSTRUCTION STARTED VOLN  Car sales  21  Car registration: USA  USCAR…P  Δln   US NEW PASSENGER CARS—TOTAL REGISTRATIONS VOLN  22  Car registration: France  FRCARREGP  Δln   FR NEW CAR REGISTRATIONS VOLN  23  Car registration: Germany  BDRVNCARP  ln   BD NEW REGISTRATIONS—CARS VOLN  24  Car registration: UK  UKCAR…P  Δln   UK CAR REGISTRATIONS VOLN  25  Car registration: Japan  JPCARREGF  ln   JP MOTOR VEHICLE NEW REGISTRATIONS: PASSENGER CARS EXCL. BELOW 66  Consumption  26  Consumer sentiment: USA  USUMCONEH  Δln   US UNIV OF MICHIGAN CONSUMER SENTIMENT—EXPECTATIONS VOLN (BASE 1966 = 100)  27  Pers. cons. exp.: USA  USPERCONB  Δln   US PERSONAL CONSUMPTION EXPENDITURES (AR) CURA  28  Pers. saving: USA  USPERSAVE  Δlv  US PERSONAL SAVING AS % OF DISPOSABLE PERSONAL INCOME SADJ  29  Retail sale: Canada  CNRETTOTB  Δln   CN RETAIL SALES: TOTAL (ADJUSTED) CURA  30  Household confidence: France  FRCNFCONQ  Δlv  FR SURVEY—HOUSEHOLD CONFIDENCE INDICATOR SADJ  31  Household confidence: Germany  BDCNFCONQ  lv  BD CONSUMER CONFIDENCE INDICATOR—GERMANY SADJ  32  Retail sales: UK  UKRETTOTB  Δln   UK RETAIL SALES (MONTHLY ESTIMATE. DS CALCULATED) CURA  33  Household confidence: UK  UKCNFCONQ  Δlv  UK CONSUMER CONFIDENCE INDICATOR—UK SADJ  34  Retail sales: Australia  AURETTOTT  Δln   AU RETAIL SALES (TREND) VOLA  35  Household confidence: Australia  AUCNFCONR  lv  AU MELBOURNE/WESTPAC CONSUMER SENTIMENT INDEX NADJ  36  Household expenditure: Japan  JPHLEXPWA  Δln   JP WORKERS HOUSEHOLD LIVING EXPENDITURE (INCL. AFF) CURN  37  Retail sales: Japan  JPRETTOTA  Δln   JP RETAIL SALES CURN  Wages and labor  38  Av. Hourly Real Earnings: USA  USWRIM.D  Δln   US AVG HOURLY REAL EARNINGS—MANUFACTURING CONA (BASE 82–84)  39  Av. overtime hours: USA  USOOL024Q  Δlv  US OVERTIME HOURS—MANUFACTURING. WEEKLY VOLA  40  Av. wkly hours: USA  USHKIM_O  Δlv  US AVG WKLY HOURS—MANUFACTURING VOLA  41  Av. hourly real earnings: Canada  CNWAGES.A  Δln   CN AVG. HOURLY EARN—INDUSTRIAL AGGREGATE EXCL. UNCLASSIFIED CURN  42  Labor productivity: Germany  BDPRODVTQ  Δln   BD PRODUCTIVITY: OUTPUT PER MAN-HOUR WORKED IN INDUSTRY SADJ (2005 = 100)  43  wages: Germany  BDWAGES.F  Δln   BD WAGE & SALARY.OVERALL ECONOMY—ON A MTHLY BASIS (PAN BD M0191)  44  wages index: Japan  JPWAGES_E  Δln   JP WAGE INDEX: CASH EARNINGS—ALL INDUSTRIES SADJ  Unemployment  45  U rate: USA  USUNEM15Q  Δlv  US UNEMPLOYMENT RATE—15 WEEKS & OVER SADJ  46  U rate: USA  USUNTOTQ_pc  Δlv  US UNEMPLOYMENT RATE SADJ  47  Employment: Canada  CNEMPTOTO  Δln   CN EMPLOYMENT—CANADA (15 YRS & OVER SA) VOLA  48  U all: Germany  BDUNPTOTP  Δln   BD UNEMPLOYMENT LEVEL (PAN BD FROM SEPT. 1990) VOLN  49  U rate: UK  UKUNTOTQ_pc  Δlv  UK UNEMPLOYMENT RATE SADJ  50  Emp: Australia  AUEMPTOTO  Δln   AU EMPLOYED: PERSONS VOLA  51  U all: Australia  AUUNPTOTO  Δln   AU UNEMPLOYMENT LEVEL VOLA  52  U rate: Japan  JPUNTOTQ_pc  Δlv  JP UNEMPLOYMENT RATE SADJ  International trade  53  Exports: USA  USI70_A  Δln   US EXPORTS CURN  54  Exports: France  FREXPGDSB  Δln   FR EXPORTS FOB CURA  55  Exports: Germany  BDEXPBOPB  Δln   BD EXPORTS FOB CURA  56  Exports: UK  UKI70_A  Δln   UK EXPORTS CURN  57  Exports: Australia  AUEXPG&SB  Δln   AU EXPORTS OF GOODS & SERVICES (BOP BASIS) CURA  58  Exports: Japan  JPEXPGDSB  Δln   JP EXPORTS OF GOODS—CUSTOMSBASIS CURA  59  Imports: USA  USIMPGDSB  Δln   US IMPORTS F.A.S. CURA  60  Imports: France  FRIMPGDSB  Δln   FR IMPORTS FOB CURA  61  Imports: Germany  BDIMPGDSB  Δln   BD IMPORTS CIF (PAN BD M0790) CURA  62  Imports: UK  UKIMPBOPB  Δln   UK IMPORTS—BALANCE OF PAYMENTS BASIS CURA  63  Imports: Australia  AUIMPG&SB  Δln   AU IMPORTS OF GOODS & SERVICES (BOP BASIS) CURA  64  Imports: Japan  JPOXT009B  Δln   JP IMPORTS CURA  65  Terms of trade: UK  UKTOTPRCF  Δln   UK TERMS OF TRADE—EXPORT/IMPORT PRICES (BOP BASIS) NADJ  66  Terms of trade: Japan  JPTOTPRCF  Δln   JP TERMS OF TRADE INDEX NADJ  Money and credit  67  Money supply: USA  USM0_A  Δln   US MONETARY BASE CURA  68  Money supply: USA  USM2_B  Δln   US MONEY SUPPLY M2 CURA  69  Money supply: France  FRM2_A  Δln   FR MONEY SUPPLY—M2 (NATIONAL CONTRIBUTION TO M2) CURN  70  Money supply: France  FRM3_A  Δln   FR MONEY SUPPLY—M3 (NATIONAL CONTRIBUTION TO M3) CURN  71  Money supply: Germany  BDM1_A  Δln   BD MONEY SUPPLY—GERMAN CONTRIBUTION TO EURO M1 (PAN BD M0790)  72  Money supply: Germany  BDM3_B  Δln   BD MONEY SUPPLY—M3 (CONTRIBUTION TO EURO BASIS FROM M0195) CURA  73  Money supply: UK  UKM1_B  Δln   UK MONEY SUPPLY M1 (ESTIMATE OF EMU AGGREGATE FOR THE UK) CURA  74  Money supply: UK  UKM3_B  Δln   UK UK MONEY SUPPLY M3 (ESTIMATE OF EMU AGGREGATE FORTHE UK) CURA  75  Money supply: Australia  AUM1_B  Δln   AU MONEY SUPPLY—M1 CURA  76  Money supply: Australia  AUM3_B  Δln   AU MONEY SUPPLY - M3 (SEE AUM3…OB) CURA  77  Money supply: Japan  JPM1_A  Δln   JP MONEY SUPPLY: M1 (METHO-BREAK. APR. 2003) CURN  78  Money supply: Japan  JPM2_A  Δln   JP MONEY SUPPLY: M2 (METHO-BREAK. APR. 2003) CURN  79  Credit: USA  USCOMILND  Δln   US COMMERCIAL & INDUSTRIAL LOANS OUTSTANDING (BCI 101) CONA (BASE 2005)  80  Credit: USA  USCILNNCB  lv  US COMMERCIAL & INDL LOANS. NET CHANGE (AR) (BCI 112) CURA  81  Credit: USA  USCRDNRVB  Δln   US NONREVOLVING CONSUMER CREDIT OUTSTANDING CURA  82  Credit: France  FRBANKLPA  Δln   FR MFI LOANS TO RESIDENT PRIVATE SECTOR CURN  83  Credit: Germany  BDBANKLPA  Δ2ln   BD LENDING TO ENTERPRISES & INDIVIDUALS CURN  84  Credit: Australia  AUCRDCONB  Δ2ln   AU FINANCIAL INTERMEDIARIES: NARROW CREDIT—PRIVATE SECTOR CURA  85  Credit: Japan  JPBANKLPA  Δ2ln   JP AGGREGATE BANK LENDING (EXCL. SHINKIN BANKS) CURN  Stock index  86  Stock index: USA  USSHRPRCF  Δln   US DOW JONES INDUSTRIALS SHARE PRICE INDEX (EP) NADJ  87  Stock index: France  FRSHRPRCF  Δln   FR SHARE PRICE INDEX—SBF 250 NADJ  88  Stock index: Germany  BDSHRPRCF  Δln   BD DAX SHARE PRICE INDEX. EP NADJ  89  Stock index: UK  UKOSP001F  Δln   UK FTSE 100 SHARE PRICE INDEXNADJ (2005 = 100)  90  Stock index: Japan  JPSHRPRCF  Δln   JP TOKYO STOCK EXCHANGE—TOPIX (EP) NADJ (1968 = 100)  Interest rate  91  Interest rate: USA  USFEDFUN  Δlv  US FEDERAL FUNDS RATE (AVG.)  92  Interest rate: USA  USCRBBAA  Δlv  US CORPORATE BOND YIELD—MOODY’S BAA. SEASONED ISSUES  93  Interest rate: USA  USGBOND  Δlv  US TREASURY YIELD ADJUSTED TO CONSTANT MATURITY—20 YEAR  94  Interest rate: France  FRPRATE  Δlv  FR AVERAGE COST OF FUNDS FOR BANKS/EURO REPO RATE  95  Interest rate: France  FRGBOND  Δlv  FR GOVERNMENT GUARANTEED BOND YIELD (EP) NADJ  96  Interest rate: Germany  BDPRATE  Δlv  BD DISCOUNT RATE/SHORT TERM EURO REPO RATE  97  Interest rate: Germany  BDGBOND  Δlv  BD LONG TERM GOVERNMENT BOND YIELD—9–10 YEARS  98  Interest rate: UK  UKPRATE  Δlv  UK BANK OF ENGLAND BASE RATE (EP)  99  Interest rate: UK  UKGBOND  Δlv  UK GROSS REDEMPTION YIELD ON 20 YEAR GILTS (PERIOD AVERAGE) NADJ  100  Interest rate: Australia  AUPRATE  Δlv  AU RBA CASH RATE TARGET  101  Interest rate: Australia  AUBOND  Δlv  AU COMMONWEALTH GOVERNMENT BOND YIELD 10 YEAR (EP)  102  Interest rate: Japan  JPGBOND  Δlv  JP INTEREST-BEARING GOVERNMENT BONDS—10-YEAR (EP)  Exchange rate  103  Exchange rate: DM to US$  BBDEMSP  Δln   GERMAN MARK TO US$ (BBI)—EXCHANGE RATE  104  Exchange rate: SK to US$  SDXRUSD  Δln   SD SWEDISH KRONOR TO US$ (BBI. EP)  105  Exchange rate: to $  UKDOLLR  Δln   UK TO US$ (WMR)—EXCHANGE RATE  106  Exchange rate: Yen to $  JPXRUSD  Δln   JP JAPANESE YEN TO US$  107  Exchange rate: Aus.$ to US$  AUXRUSD  Δln   AU AUSTRALIAN $ TO US$ (MTH. AVG.)  Producer price index  108  PPI: USA  USPROPRCE  Δln   US PPI—FINISHED GOODS SADJ  109  PPI: Canada  CNPROPRCF  Δln   CN INDUSTRIAL PRICE INDEX: ALL COMMODITIES NADJ  110  PPI: Germany  BDPROPRCF  Δln   BD PPI: INDL. PRODUCTS. TOTAL. SOLD ON THE DOMESTIC MARKET NADJ (2005 = 100)  111  PPI: UK  UKPROPRCF  Δln   UK PPI—OUTPUT OF MANUFACTURED PRODUCTS (HOME SALES) NADJ  112  PPI: Japan  JPPROPRCF  Δln   JP CORPORATE GOODS PRICE INDEX: DOMESTIC—ALL COMMODITIES NADJ  Consumer price index  113  CPI: USA  USCONPRCE  Δln   US CPI—ALL URBAN: ALL ITEMS SADJ  114  CPI: Canada  CNCONPRCF  Δln   CN CPI NADJ  115  CPI: France  FRCONPRCE  Δln   FR CPI SADJ  116  CPI: Germany  BDCONPRCE  Δln   BD CPI SADJ  117  CPI: UK  UKCONPRCF  Δln   UK CPI INDEX 00: ALL ITEMS- ESTIMATED PRE-97 2005 = 100 NADJ  118  CPI: Japan  JPCONPRCF  Δln   JP CPI: NATIONAL MEASURE NADJ    Emerging countries           Series number  Short name  Mnemonic  Trans  Description  Industrial production  119  IP: Argentina  AGIPTOT.G  Δln   AG INDUSTRIAL PRODUCTION VOLA  120  IP: Chile  CLIPMAN.H  Δln   CL INDUSTRIAL PRODUCTION INDEX BY INE—MANUFACTURING TOTAL VOLN  121  IP: Brazil  BRIPTOT_G  Δln   BR INDUSTRIAL PRODUCTION VOLA index 2002 = BASE  122  IP: Brazil  BRIPMAN.G  Δln   BR INDUSTRIAL PRODUCTION—MANUFACTURING VOLA  123  IP: China  CHPBRENTP (electricity)  Δln   CH INDUSTRIAL PRODUCTION: ELECTRICITY VOLN  124  IP: Korea  KOIPTOT.G  Δln   KO INDUSTRIAL PRODUCTION VOLA (2005 = 100)  125  IP: Korea  KOIPMAN.G  Δln   KO MANUFACTURING PRODUCTION INDEX VOLA  126  IP: Mexico  MXIPTOT_H  Δln   MX INDUSTRIAL PRODUCTION INDEX VOLN  127  IP: Philippines  PHIPMAN_F  Δln   PH MANUFACTURING PRODUCTION NADJ  128  IP: South Africa  SAIPMAN.G  Δln   SA INDUSTRIAL PRODUCTION (MANUFACTURING SECTOR) VOLA  129  IP: Taiwan  TWIPMAN.H  Δln   TW INDUSTRIAL PRODUCTION INDEX—MANUFACTURING VOLN  Orders and capacity utilization  130  Operating ratio: Brazil  BRCAPUTLR  Δlv  BR CAPACITY UTILIZATION—MANUFACTURING NADJ  131  Mach. ord.: Korea  KONEWORDA  Δln   KO MACHINERY ORDERS RECEIVEDCURN  132  Manufct. prod capa.: Korea  KOCAPUTLF  Δlv  KO MANUFACTURING PRODUCTION CAPACITY NADJ (2005 = 100)  Consumption  133  Car sales: Argentina  AGCARSLSP  Δln   AG SALES—NATIONAL CARS TO DEALERS VOLN  134  Retail sales: Chile  CLRETTOTH  Δln   CL RETAIL SALES AT SUPERMARKETS (REAL INDEX) VOLN  135  Gasoline consumption: Korea  KOOPCGSLP  Δln   KO OIL PRODUCTS CONSUMPTION—GASOLINE VOLN  136  Retail sales: Singapore  SPRETTOTG  Δlv  SP RETAIL SALES INDEX (CONSTANT) VOLA  137  Retail sales: Russia  RSRETTOTA  Δln   RS RETAIL TRADE TURNOVER—TOTAL CURN  Wages and labor  138  Labor cost: Brazil  BRLCOST.F  Δln   BR UNIT LABOR COST NADJ  Unemployment  139  Unemployment: Hong Kong  HKUNPTOTP  Δln   HK UNEMPLOYMENT (3 MONTHS ENDING) VOLN  140  U rate: Taiwan  TWUN%TOTQ  Δlv  TW UNEMPLOYMENT RATE SADJ  141  Unemployment: Russia  RSUNPTOTP  Δln   RS ECONOMICALLY ACTIVE POPULATION—UNEMPLOYED VOLN  International trade  142  Exports: Brazil  BREXPBOPA  Δln   BR BOP: CURRENT ACCOUNT—GOODS (CREDIT) CURN  143  Exports: China  CHEXPGDSA  Δln   CH EXPORTS CURN  144  Exports: India  INEXPGDSA  Δln   IN EXPORTS FOB CURN  145  Exports: Indonesia  IDEXPGDSA  Δln   ID EXPORTS FOB CURN  146  Exports: Korea  KOEXPGDSA  Δln   KO EXPORTS FOB (CUSTOMS CLEARANCE BASIS) CURN  147  Exports: Philippines  PHEXPGDSA  Δln   PH EXPORTS CURN  148  Exports: Singapore  SPEXPGDSA  Δln   SP EXPORTS CURN  149  Exports: Taiwan  TWEXPGDSA  Δln   TW EXPORTS CURN  150  Imports: Brazil  BRIMPBOPA  Δln   BR BOP: CURRENT ACCOUNT—GOODS (DEBIT) CURN  151  Imports: China  CHIMPGDSA  Δln   CH IMPORTS CURN  152  Imports: Korea  KOIMPGDSA  Δln   KO IMPORTS CIF (CUSTOMS CLEARANCE BASIS) CURN  153  Imports: Singapore  SPIMPGDSA  Δln   SP IMPORTS CURN  154  Imports: Taiwan  TWIMPGDSA  Δln   TW IMPORTS CURN  155  Terms of trade: Brazil  BRTOTPRCF  Δln   BR TERMS OF TRADE NADJ (2006 = 100)  Money and credit  156  Money supply: Brazil  BRM1_A  Δln   BR MONEY SUPPLY—M1 (EP) CURN  157  Money supply: Brazil  BRM3_A  Δln   BR MONEY SUPPLY—M3 (EP) CURN  158  Money supply: China  CHM0_A  Δln   CH MONEY SUPPLY—CURRENCY IN CIRCULATION CURN  159  Money supply: China  CHM1_A  Δln   CH MONEY SUPPLY—M1 CURN  160  Money supply: India  INM1_A  Δ2ln   IN MONEY SUPPLY: M1 (EP) CURN  161  Money supply: India  INM3_A  Δ2ln   IN MONEY SUPPLY: M3 (EP) CURN  162  Money supply: Indonesia  IDM1_A  Δln   ID MONEY SUPPLY: M1 CURN  163  Money supply: Indonesia  IDM2_A  Δln   ID MONEY SUPPLY—M2 CURN  164  Money supply: Korea  KOM2_B  Δln   KO MONEY SUPPLY—M2 (EP) CURA  165  Money supply: Mexico  MXM1_A  Δln   MX MONEY SUPPLY: M1 (EP) CURN BASE = END OF PERIOD  166  Money supply: Mexico  MXM3_A  Δ2ln   MX MONEY SUPPLY: M3 (EP) CURN  167  Money supply: Philippines  PHM1_A  Δln   PH MONEY SUPPLY—M1 (METHO BREAK AT 12/03) CURN  168  Money supply: Philippines  PHM3_A  Δln   PH MONEY SUPPLY—M3 (METHO BREAK AT 12/03) CURN  169  Money supply: Russia  RSM 0_A  Δln   RS MONEY SUPPLY—M0 CURN  Stock index  170  Stock index: Brazil  BRSHRPRCF  Δln   BR BOVESPA SHARE PRICE INDEX (EP) NADJ  171  Stock index: Hong Kong  HKSHRPRCF  Δln   HK HANG SENG SHARE PRICE INDEX (EP) NADJ (31 JULY 1964 =100)  Exchange rate  172  Exchange rate: Br.R. to US$  BRXRUSD  Δln   BR BRAZILIAN REAIS TO US$ (AVG)  173  Exchange rate: Ch.Y. to US$  CHXRUSD  Δln   CH CHINESE YUAN TO US$ (AVERAGE AMOUNT)  174  Exchange rate: In.R. to US$  INXRUSD  Δln   IN INDIAN RUPEES PER US$ (RBI)  175  Exchange rate: Id.R. to US$  IDXRUSD  Δln   ID INDONESIAN RUPIAHS TO US$  176  Exchange rate: Mx.P. to US$  MXXRUSD  Δln   MX MEXICAN PESOS TO US$—CENTRAL BANK SETTLEMENT RATE (AVG)  177  Exchange rate: Rs.R. to US$  RSXRUSD  Δln   RS RUSSIAN ROUBLES TO US$ NADJ  Consumer price index  178  CPI: Brazil  BRCPIGENF  Δln   BR CPI—GENERAL NADJ  179  CPI: China  CHCONPRCF  Δln   CH CPI NADJ  180  CPI: India  INCONPRCF  Δln   IN CPI: INDUSTRIAL LABOURERS(DS CALCULATED) NADJ (2001 = 100)  181  CPI: Korea  KOCONPRCF  Δln   KO CPI NADJ (2005 = 100)  182  CPI: Mexico  MXCONPRCF  Δln   MX CPI NADJ (JUN 2002 = 100)  183  CPI: Philippines  PHCONPRCF  Δln   PH CPI NADJ  184  CPI: Russia  RSCONPRCF  Δln   RS CPI NADJ    Emerging countries           Series number  Short name  Mnemonic  Trans  Description  Industrial production  119  IP: Argentina  AGIPTOT.G  Δln   AG INDUSTRIAL PRODUCTION VOLA  120  IP: Chile  CLIPMAN.H  Δln   CL INDUSTRIAL PRODUCTION INDEX BY INE—MANUFACTURING TOTAL VOLN  121  IP: Brazil  BRIPTOT_G  Δln   BR INDUSTRIAL PRODUCTION VOLA index 2002 = BASE  122  IP: Brazil  BRIPMAN.G  Δln   BR INDUSTRIAL PRODUCTION—MANUFACTURING VOLA  123  IP: China  CHPBRENTP (electricity)  Δln   CH INDUSTRIAL PRODUCTION: ELECTRICITY VOLN  124  IP: Korea  KOIPTOT.G  Δln   KO INDUSTRIAL PRODUCTION VOLA (2005 = 100)  125  IP: Korea  KOIPMAN.G  Δln   KO MANUFACTURING PRODUCTION INDEX VOLA  126  IP: Mexico  MXIPTOT_H  Δln   MX INDUSTRIAL PRODUCTION INDEX VOLN  127  IP: Philippines  PHIPMAN_F  Δln   PH MANUFACTURING PRODUCTION NADJ  128  IP: South Africa  SAIPMAN.G  Δln   SA INDUSTRIAL PRODUCTION (MANUFACTURING SECTOR) VOLA  129  IP: Taiwan  TWIPMAN.H  Δln   TW INDUSTRIAL PRODUCTION INDEX—MANUFACTURING VOLN  Orders and capacity utilization  130  Operating ratio: Brazil  BRCAPUTLR  Δlv  BR CAPACITY UTILIZATION—MANUFACTURING NADJ  131  Mach. ord.: Korea  KONEWORDA  Δln   KO MACHINERY ORDERS RECEIVEDCURN  132  Manufct. prod capa.: Korea  KOCAPUTLF  Δlv  KO MANUFACTURING PRODUCTION CAPACITY NADJ (2005 = 100)  Consumption  133  Car sales: Argentina  AGCARSLSP  Δln   AG SALES—NATIONAL CARS TO DEALERS VOLN  134  Retail sales: Chile  CLRETTOTH  Δln   CL RETAIL SALES AT SUPERMARKETS (REAL INDEX) VOLN  135  Gasoline consumption: Korea  KOOPCGSLP  Δln   KO OIL PRODUCTS CONSUMPTION—GASOLINE VOLN  136  Retail sales: Singapore  SPRETTOTG  Δlv  SP RETAIL SALES INDEX (CONSTANT) VOLA  137  Retail sales: Russia  RSRETTOTA  Δln   RS RETAIL TRADE TURNOVER—TOTAL CURN  Wages and labor  138  Labor cost: Brazil  BRLCOST.F  Δln   BR UNIT LABOR COST NADJ  Unemployment  139  Unemployment: Hong Kong  HKUNPTOTP  Δln   HK UNEMPLOYMENT (3 MONTHS ENDING) VOLN  140  U rate: Taiwan  TWUN%TOTQ  Δlv  TW UNEMPLOYMENT RATE SADJ  141  Unemployment: Russia  RSUNPTOTP  Δln   RS ECONOMICALLY ACTIVE POPULATION—UNEMPLOYED VOLN  International trade  142  Exports: Brazil  BREXPBOPA  Δln   BR BOP: CURRENT ACCOUNT—GOODS (CREDIT) CURN  143  Exports: China  CHEXPGDSA  Δln   CH EXPORTS CURN  144  Exports: India  INEXPGDSA  Δln   IN EXPORTS FOB CURN  145  Exports: Indonesia  IDEXPGDSA  Δln   ID EXPORTS FOB CURN  146  Exports: Korea  KOEXPGDSA  Δln   KO EXPORTS FOB (CUSTOMS CLEARANCE BASIS) CURN  147  Exports: Philippines  PHEXPGDSA  Δln   PH EXPORTS CURN  148  Exports: Singapore  SPEXPGDSA  Δln   SP EXPORTS CURN  149  Exports: Taiwan  TWEXPGDSA  Δln   TW EXPORTS CURN  150  Imports: Brazil  BRIMPBOPA  Δln   BR BOP: CURRENT ACCOUNT—GOODS (DEBIT) CURN  151  Imports: China  CHIMPGDSA  Δln   CH IMPORTS CURN  152  Imports: Korea  KOIMPGDSA  Δln   KO IMPORTS CIF (CUSTOMS CLEARANCE BASIS) CURN  153  Imports: Singapore  SPIMPGDSA  Δln   SP IMPORTS CURN  154  Imports: Taiwan  TWIMPGDSA  Δln   TW IMPORTS CURN  155  Terms of trade: Brazil  BRTOTPRCF  Δln   BR TERMS OF TRADE NADJ (2006 = 100)  Money and credit  156  Money supply: Brazil  BRM1_A  Δln   BR MONEY SUPPLY—M1 (EP) CURN  157  Money supply: Brazil  BRM3_A  Δln   BR MONEY SUPPLY—M3 (EP) CURN  158  Money supply: China  CHM0_A  Δln   CH MONEY SUPPLY—CURRENCY IN CIRCULATION CURN  159  Money supply: China  CHM1_A  Δln   CH MONEY SUPPLY—M1 CURN  160  Money supply: India  INM1_A  Δ2ln   IN MONEY SUPPLY: M1 (EP) CURN  161  Money supply: India  INM3_A  Δ2ln   IN MONEY SUPPLY: M3 (EP) CURN  162  Money supply: Indonesia  IDM1_A  Δln   ID MONEY SUPPLY: M1 CURN  163  Money supply: Indonesia  IDM2_A  Δln   ID MONEY SUPPLY—M2 CURN  164  Money supply: Korea  KOM2_B  Δln   KO MONEY SUPPLY—M2 (EP) CURA  165  Money supply: Mexico  MXM1_A  Δln   MX MONEY SUPPLY: M1 (EP) CURN BASE = END OF PERIOD  166  Money supply: Mexico  MXM3_A  Δ2ln   MX MONEY SUPPLY: M3 (EP) CURN  167  Money supply: Philippines  PHM1_A  Δln   PH MONEY SUPPLY—M1 (METHO BREAK AT 12/03) CURN  168  Money supply: Philippines  PHM3_A  Δln   PH MONEY SUPPLY—M3 (METHO BREAK AT 12/03) CURN  169  Money supply: Russia  RSM 0_A  Δln   RS MONEY SUPPLY—M0 CURN  Stock index  170  Stock index: Brazil  BRSHRPRCF  Δln   BR BOVESPA SHARE PRICE INDEX (EP) NADJ  171  Stock index: Hong Kong  HKSHRPRCF  Δln   HK HANG SENG SHARE PRICE INDEX (EP) NADJ (31 JULY 1964 =100)  Exchange rate  172  Exchange rate: Br.R. to US$  BRXRUSD  Δln   BR BRAZILIAN REAIS TO US$ (AVG)  173  Exchange rate: Ch.Y. to US$  CHXRUSD  Δln   CH CHINESE YUAN TO US$ (AVERAGE AMOUNT)  174  Exchange rate: In.R. to US$  INXRUSD  Δln   IN INDIAN RUPEES PER US$ (RBI)  175  Exchange rate: Id.R. to US$  IDXRUSD  Δln   ID INDONESIAN RUPIAHS TO US$  176  Exchange rate: Mx.P. to US$  MXXRUSD  Δln   MX MEXICAN PESOS TO US$—CENTRAL BANK SETTLEMENT RATE (AVG)  177  Exchange rate: Rs.R. to US$  RSXRUSD  Δln   RS RUSSIAN ROUBLES TO US$ NADJ  Consumer price index  178  CPI: Brazil  BRCPIGENF  Δln   BR CPI—GENERAL NADJ  179  CPI: China  CHCONPRCF  Δln   CH CPI NADJ  180  CPI: India  INCONPRCF  Δln   IN CPI: INDUSTRIAL LABOURERS(DS CALCULATED) NADJ (2001 = 100)  181  CPI: Korea  KOCONPRCF  Δln   KO CPI NADJ (2005 = 100)  182  CPI: Mexico  MXCONPRCF  Δln   MX CPI NADJ (JUN 2002 = 100)  183  CPI: Philippines  PHCONPRCF  Δln   PH CPI NADJ  184  CPI: Russia  RSCONPRCF  Δln   RS CPI NADJ    Appendix B: Estimating the Number of Factors Bai and Ng (2002) propose to select the number of common factors which minimize the following information criteria:   PCPi(k)=S(k)+kσ¯2gi(N.T),  ICi(k)=ln(S(k))+kgi(N.T), where k is the number of factors, S(k)=(NT)−1∑i=1N∑t=1T(xit−λ^ik′F^tk)2 is the sum of squared residuals (divided by NT), g(N.T) is a penalty function,32 and σ¯2 equals S(kmax) for a pre-specified value of kmax. The optimal number of factors k^ minimizes these information criteria. Kapetanios (2010) proposes a sequential test to determine the number of factors. When the true number of factors is k0, under some regularity condition, the first k0 eigenvalues of the population covariance matrix Σ increase at rate N while the others are bounded. Let’s note λ^k, k=1,…,N, the N eigenvalues (in decreasing order) of the sample covariance matrix X′X and kmax a finite number such that k0<kmax. The difference λ^k−λ^kmax+1 will go to infinity for k=1,…,k0, but is bounded for k=k0+1,⋯,kmax. λ^k−λ^kmax+1 is then used as a the test statistics to discriminate the null hypothesis that the true number of factors k0 equals k ( H0.k:k0=k) against the alternative hypothesis ( H1.k:k0>k). When there is no factor structure, λ^k−λ^kmax+1, appropriately normalized, converge to a law limit, but tend to infinity in the presence of factors. We begin by testing ( H0.k:k0=k=0) against ( H1.k:k0>0). If we reject the null hypothesis, then we consider the null ( H0.k:k0=k+1=1). We stop once we cannot reject the null hypothesis. Kapetanios (2010) called this algorithm the maximal eigenvalue distribution (MED) algorithm. As shown in Table BI, there is no agreement on the optimal number of factors.33Bai and Ng (2002) information criteria select between two and nine factors while Kapetanios (2010) sequential test suggests two. Previous empirical work also reveals considerable variance in estimates of the correct number of factors.34 The factors autocorrelation of factors F^t are displayed in Table BII. They show that most factors are persistent. Statistics on their explanatory power reveals that only 20% of the variance in the 184 time series is explained by the first three factors. This figure is equal to 36% for the first nine factors, which leads us to keep the first nine factors as potential regressors for modeling commodity returns. Table BI. Static factor selection results MED denotes the number of factors given by the maximum eigenvalue distribution algorithm. ICi and PCPi denote, respectively, the number of factors given by the information criteria IC and PCP estimated with the penalty function gi(N.T). Method  Number of static factors  MED  2  IC1  4  IC2  3  IC3  12  IC4  20  PCP1  9  PCP2  8  PCP3  18  PCP4  20  Method  Number of static factors  MED  2  IC1  4  IC2  3  IC3  12  IC4  20  PCP1  9  PCP2  8  PCP3  18  PCP4  20  Table BII. Summary statistics for estimated static factors F^t.i for i=1,…,9 For i = 1, … , 9. F^it is estimated by the method of principal components using a panel of data with 184 indicators of economic activity from 1993:03 to 2010:03 (205 time-series observations). The data are transformed (taking logs and differenced where appropriate) and standardized prior to estimation. ρi denotes the ith autocorrelation. The 95% confidence bounds are ±0.1397. The relative importance of the common component. Ri2 is calculated as the fraction of total variance in the data explained by factors 1 to i. Factor i  ρ1  ρ2  ρ3  Ri2  1  0.1614  0.1256  0.3176  0.0930  2  0.1357  0.0805  0.3110  0.1623  3  −0.0748  0.0145  −0.0294  0.2066  4  −0.0285  −0.0694  0.1866  0.2424  5  −0.1439  −0.0966  0.0950  0.2740  6  0.2546  0.0328  −0.0091  0.3035  7  0.1012  0.3234  0.3844  0.3288  8  0.3405  0.4066  0.1768  0.3518  9  −0.0065  −0.0413  −0.1447  0.3739  Factor i  ρ1  ρ2  ρ3  Ri2  1  0.1614  0.1256  0.3176  0.0930  2  0.1357  0.0805  0.3110  0.1623  3  −0.0748  0.0145  −0.0294  0.2066  4  −0.0285  −0.0694  0.1866  0.2424  5  −0.1439  −0.0966  0.0950  0.2740  6  0.2546  0.0328  −0.0091  0.3035  7  0.1012  0.3234  0.3844  0.3288  8  0.3405  0.4066  0.1768  0.3518  9  −0.0065  −0.0413  −0.1447  0.3739  Footnotes 1 Investment in commodity markets from a portfolio perspective is discussed in Gorton and Rouwenhorst (2006); Erb and Harvey (2006); Rouwenhorst and Tang (2012); and Gorton, Hayashi, and Rouwenhorst (2013), among many others. 2 The early contribution by Gilbert (1989) emphasizes the relevance of the exchange rate as an explanatory variable for commodity prices; see also the recent papers by Chen et al. (2010) and Ferraro, Rogoff, and Rossi (2015). 3 The same variables are used in Deb, Trivedi, and Varangis (1996). Leybourne, Lloyd, and Reed (1994) further discuss the issue of omitted variables. 4 Recent economic research on the determination of commodity prices occasionally makes use of factor models. Examples of this growing literature are Byrne, Fazio, and Fiess (2013); Gospodinov and Ng (2013); West and Wong (2014); and Christoffersen, Lunde, and Olesen (2014). While these papers investigate more or less directly the issue of the co-movement of commodity prices, they all extract principal components from a set of commodity prices to explain the evolution of commodity prices, only considering a few additional macroeconomic variables—such as interest rates, exchange rates, and inflation—to analyze the link between these variables and their estimated factors. As such, their approach is very different from ours. 5 As will be made clear in the empirical sections, we adopt a measure of excess co-movement that is similar to that used in Kallberg and Pasquariello (2008), in that we consider the average of the squared residual correlations between all pairs of commodities. We hence allow both positive and negative correlations to contribute to the excess co-movement estimate. 6 As shown by Forbes and Rigobon (2002), the usual sample correlation is a biased measure of the true correlation when volatility is time-varying, which is a well-known stylized fact regarding financial series. As most of our commodity returns are characterized by time-varying volatility, we use the correlation coefficient corrected for heteroscedasticity of Forbes and Rigobon (2002). 7 Interestingly, Barberis and Shleifer (2003) cite the empirical results in PR as exemplifying their model. Conversely, PR observe that: “[…] traders are alternatively bullish or bearish on all commodities for no plausible reason” (p. 1173), which is behind our basic idea to measure the impact of speculation in commodity-futures markets on commodity price co-movements. 8 Their research question builds on Barberis, Shleifer, and Wurgler (2005), who theoretically and empirically analyze the behavior of newly included stocks in a stock index. It is shown that the price co-movement between the stock and the index significantly increases after this inclusion. 9 Part of the existing literature (Palaskas and Varangis, 1991; Leybourne, Lloyd, and Reed, 1994) considers excess co-movement of nominal or real price rather than return, and relies on a co-integration analysis. We think that return is more appealing when dealing with risk management issues, and thus consider the excess co-movement of returns as in the seminal work of PR. Returns rather than prices have also been considered more recently in Ai, Chatrath, and Song (2006) and Malliaris and Urritia (1996) for the main agricultural commodities. 10 Each variable is rendered stationary in an appropriate manner: the chosen transformation appears in the penultimate column of the table in Appendix A. 11 Although Stock and Watson (2002a) use different sets of assumptions to characterize “weak correlations”, the main idea is that the cross-correlations and serial correlations have an upper bound. 12 As the factors Ft and the loading matrix Λ are not separately identifiable [see Bai and Ng (2008) for more details], constraints are imposed to obtain a unique estimate. 13 When N > T, a computationally simpler approach is to use the T × T matrix XX′. 14 Methods based on information criteria and Kapetanios (2010) are described in Appendix B. 15 To further improve the explanatory power, we also considered potential nonlinearities with quadratic or cubic factors. We choose the specification with the highest adjusted R¯2. The set of factors is now F¯tnl=(F^1,t,F^2,t,F^3,t,F^4,t,F^2,t3,F^4,t3)′ and the regressions become: rit=ωi+∑k=14γikF^k,t+ωi,5F^2,t3+ωi,6F^4,t3+uiti=1,…,8 t=1,…,T=ωi+γi′F¯tnl+uit. The best specifications results are not shown here (but are available upon request). We do not find any notable increase in the R2 for any commodity. We therefore retain linear factors in the returns equation. 16 The classification in Ludvigson and Ng (2009) is finer but is applied to US variables only. Their classification is likely not applicable when a number of economies are considered for reasons of interpretability. 17 China imports 30% of all the copper traded in the world. 18 The p-value is calculated by transforming the correlation ρ^ to create a t-statistic with T – 2 degrees of freedom, where T is the number of observations. 19 One possible explanation is that the oil-return filtering is more successful than that for other commodities. 20 This is an issue in the contribution of PR which was further considered in Deb, Trivedi, and Varangis (1996) by means of the multivariate GARCH model in its BEKK form (Engle and Kroner, 1995). Multivariate GARCH models deal with standardized returns and no further correction for heteroscedasticity is needed [see Brenner, Pasquariello, and Subrahmanyam (2009) for a recent application using standardized returns for the analysis of co-movements in US financial markets around scheduled macroeconomic announcements]. 21 Co-movement is a concept which may at first sight be confused with contagion. However, there is a significant difference between the two concepts. While excess co-movement is defined as significant residual correlation once common factors are considered, contagion is defined as a significant increase in correlation following a shock in one market. At this point, two remarks are in order. First, most of the literature on contagion either uses very simple common factors or ignores them entirely. This is quite different from the excess co-movement literature where “excess” means “beyond common factors”, and the determination of common factors strongly affects the estimated co-movement. Second, we do not need to observe an increase in correlation to confirm excess co-movement, but rather a significant correlation most of the time or on average over a given period. 22 Similar results appear in Boyer, Gibson, and Loretan (1999) and Loretan and English (2000). Tang and Xiong (2012) also correct the correlation for time-varying volatility using the method in Forbes and Rigobon (2002), which has only a small insignificant effect on their estimates. 23 This correction is valid if we assume no omitted variables or endogeneity. 24 The significance threshold is 0.1669 and is plotted as a horizontal dotted line in Figure 3. 25 See also Bruno, Büyüksahin, and Robe (2016), who consider co-movement across food commodities along with financialization, but whose main focus is rather on the commodity–equity relationship. 26 We wish to thank a referee for suggesting this methodological approach. 27 The analysis in Han (2008) deals with S&P 500 futures contracts. The author also relies on a proxy base on the Investors Intelligence’s weekly that is not relevant for commodity markets. 28 Since 2006, the CFTC has also released a weekly Disaggregated Commitments of Traders (DCOT) report each Friday. This complements the COT report by providing more detailed categories of traders such as Index Traders who have played a significant role in recent years. We do not use the DCOT data here, as it would considerably restrict the analysis sample period. 29 The sample correlation between Han’s index and hedging pressure ranges from −0.78 for live cattle to − 0.98 for cotton. 30 Singleton (2014) implicitly considers the role of cross-positions, as his measure of index funds in oil markets is derived from index funds positions in grain markets. 31 Our methodology resembles the approach in Raman, Robe, and Yadav (2016), who gauge the effect of the participation of financial traders in oil futures post-electronification using two-stage least squares, or the method in Gilbert and Pfuderer (2014), who investigate the causal role of index trading on grain markets using instrumental variables. As changes in futures positions and futures returns are simultaneously determined, these methods are naturally relevant when analyzing financialization. 32 The penalty functions suggested by Bai and Ng (2002) are: g1(N.T)=N+TNTln(NTN+T), g2(N.T)=N+TNTln(CNT2), g3(N.T)=ln(CNT2)CNT2, and g4(N.T)=(N+T−k)ln(NT)NT. 33 We use the Matlab routine provided by Ng for the estimation and information criteria. The Kapetanios test routine is ours. 34 See for example the empirical applications in Kapetanios (2010) which show that there may be some variation in the estimation of the number of factors under different selection criteria. References Ai C., Chatrath A., Song F. ( 2006): On the comovement of commodity prices, American Journal of Agricultural Economics  88, 574– 588. Google Scholar CrossRef Search ADS   Bai J. ( 2003): Inferential theory for factor models of large dimensions, Econometrica  71, 135– 172. Google Scholar CrossRef Search ADS   Bai J., Ng S. ( 2002): Determining the number of factors in approximate factor models, Econometrica  70, 191– 221. Google Scholar CrossRef Search ADS   Bai J., Ng S. ( 2008): Large dimensional factor analysis, Foundations and Trends in Econometrics  3, 89– 163. Google Scholar CrossRef Search ADS   Baker M., Wurgler J. ( 2007): Investor sentiment in the stock market, Journal of Economic Perspectives  21, 129– 151. Google Scholar CrossRef Search ADS   Barberis N., Shleifer A. ( 2003): Style investing, Journal of Financial Economics  68, 161– 199. Google Scholar CrossRef Search ADS   Barberis N., Shleifer A., Wurgler J. ( 2005): Comovement, Journal of Financial Economics  75, 283– 317. Google Scholar CrossRef Search ADS   Barsky R. B., Kilian L. ( 2002): Do we really know that oil caused the great stagflation? A monetary alternative, in: Bernanke B. S., Rogoff K. (eds.), NBER Macroeconomics Annual 2001 , Vol. 16. MIT Press, Cambridge, Massachusetts, pp. 137– 183. Basak S., Pavlova A. ( 2016): A model of fiinancialization of commodities, Journal of Finance  71, 1511– 1556. Google Scholar CrossRef Search ADS   Baumeister C., Kilian L. ( 2012): Real-time forecasts of the real price of oil, Journal of Business & Economic Statistics  30, 326– 336. Google Scholar CrossRef Search ADS   Baumeister C., Kilian L. ( 2014): What central bankers need to know about forecasting oil prices, International Economic Review  55, 869– 889. Google Scholar CrossRef Search ADS   Bessembinder H. ( 1992): Systematic risk, hedging pressure, and risk premiums in futures markets, Review of Financial Studies  5, 637– 667. Google Scholar CrossRef Search ADS   Boivin J., Ng S. ( 2005): Understanding and comparing factor based forecasts, International Journal of Central Banking  1, 117– 152. Borensztein E., Reinhart C. M. ( 1994): The macroeconomic determinants of commodity prices, IMF Staff Papers  41, 236– 261. Google Scholar CrossRef Search ADS   Boyer B., Gibson M., Loretan M. ( 1999): Pitfalls in tests for changes in correlations. Unpublished working Paper, Vol. 597R, Federal Reserve Board. Brenner M., Pasquariello P., Subrahmanyam M. ( 2009): On the volatility and comovement of U.S. financial markets around macroeconomic news announcements, Journal of Financial and Quantitative Analysis  44, 1265– 1289. Google Scholar CrossRef Search ADS   Bruno V. G., Byksahin B., Robe M. A. ( 2016): The financialization of food?, American Journal of Agricultural Economics  99, 243– 264. Google Scholar CrossRef Search ADS   Byksahin B., Robe M. A. ( 2014): Speculators, commodities and cross-market linkages, Journal of International Money and Finance  42, 38– 70. Google Scholar CrossRef Search ADS   Byrne J. P., Fazio G., Fiess N. ( 2013): Primary commodity prices, Co-movements, common factors and fundamentals, Journal of Development Economics  101, 16– 26. Google Scholar CrossRef Search ADS   Cheng I.-H., Kirilenko A., Xiong W. ( 2015): Convective risk flows in commodity futures markets, Review of Finance  19, 1733– 1781. Google Scholar CrossRef Search ADS   Cheng I.-H., Xiong W. ( 2014a): The financialization of commodity markets, Annual Review of Financial Economics  6, 419– 441. Google Scholar CrossRef Search ADS   Cheng I.-H., Xiong W. ( 2014b): Why do hedgers trade so much?, Journal of Legal Studies  43, S183– S207. Google Scholar CrossRef Search ADS   Chen S.-L., Jackson J. D., Kim H., Resiandini P. ( 2014): What drives commodity prices?, American Journal of Agricultural Economics  96, 1455– 1468. Google Scholar CrossRef Search ADS   Chen Y.-C., Rogoff K. S., Rossi B. ( 2010): Can exchange rates forecast commodity prices?, Quarterly Journal of Economics  125, 1145– 1194. Google Scholar CrossRef Search ADS   Christoffersen P., Lunde A., Olesen K. V. ( 2014): Factor structure in commodity futures return and volatility. Unpublished working paper No. 2495779, Rotman School of Management. Deb P., Trivedi P. K., Varangis P. ( 1996): The excess comovement of commodity prices reconsidered, Journal of Applied Econometrics  11, 275– 291. Google Scholar CrossRef Search ADS   Engle R. F. ( 2002): Dynamic conditional correlation: a simple class of multivariate generalized autoregressive conditional heteroskedasticity models, Journal of Business & Economic Statistics  20, 339– 350. Google Scholar CrossRef Search ADS   Engle R. F., Kroner K. F. ( 1995): Multivariate simultaneous generalized ARCH, Econometric Theory  11, 122– 150. Google Scholar CrossRef Search ADS   Erb C., Harvey C. ( 2006): The strategic and tactical value of commodity futures, Financial Analysts Journal  62, 69– 97. Google Scholar CrossRef Search ADS   Ferraro D., Rogoff K. S., Rossi B. ( 2015): Can oil prices forecast exchange rates? An empirical analysis of the relationship between commodity prices and exchange rates, Journal of International Money and Finance  54, 116– 141. Google Scholar CrossRef Search ADS   Forbes K., Rigobon R. ( 2002): No contagion, only interdependence: measuring stock market co-movements, Journal of Finance  57, 2223– 2261. Google Scholar CrossRef Search ADS   Forni M., Hallin M., Lippi M., Reichlin L. ( 2005): The generalized dynamic factor model, one sided estimation and forecasting, Journal of the American Statistical Association  100, 830– 840. Google Scholar CrossRef Search ADS   Frankel J. A. ( 2006): The effect of monetary policy on real commodity prices. Unpublished working paper No. w12713, National Bureau of Economic Research. Frankel J. A., Rose A. K. ( 2010): Determinants of agricultural and mineral commodity prices. Unpublished working paper No. 10-038, Harvard Kennedy School. Geman H., Nguyen V.-N. ( 2005): Soybean inventory and forward curve dynamics, Management Science  51, 1076– 1091. Google Scholar CrossRef Search ADS   Gilbert C. L. ( 1989): The impact of exchange rates and developing country debt on commodity prices, Economic Journal  99, 773– 784. Google Scholar CrossRef Search ADS   Gilbert C. L., Pfuderer S. ( 2014): The role of index trading in price formation in the grains and oilseeds markets, Journal of Agricultural Economics  65, 303– 322. Google Scholar CrossRef Search ADS   Gorton G., Hayashi F., Rouwenhorst K. G. ( 2013): The fundamentals of commodity futures returns, Review of Finance  17, 35– 105. Google Scholar CrossRef Search ADS   Gorton G., Rouwenhorst K. G. ( 2006): Facts and fantasies about commodity returns, Financial Analysts Journal  62, 47– 68. Google Scholar CrossRef Search ADS   Gospodinov N., Ng S. ( 2013): Commodity prices, convenience yields, and inflation, Review of Economics and Statistics  95, 206– 219. Google Scholar CrossRef Search ADS   Haase M., Zimmermann Y. S., Zimmermann H. ( 2016): The impact of speculation on commodity futures markets—a review of the findings of 100 empirical studies, Journal of Commodity Markets  3, 1– 15. Google Scholar CrossRef Search ADS   Hamilton J. D. ( 2009): Understanding crude oil prices, Energy Journal  30, 179– 206. Google Scholar CrossRef Search ADS   Hamilton J. D., Wu J. C. ( 2015): Effects of index-fund investing on commodity futures prices, International Economic Review  56, 187– 205. Google Scholar CrossRef Search ADS   Han B. ( 2008): Investor sentiment and option prices, Review of Financial Studies  21, 387– 414. Google Scholar CrossRef Search ADS   Hansen L. ( 1982): Large sample properties of the generalized method of moments estimator, Econometrica  50, 1029– 1054. Google Scholar CrossRef Search ADS   Irwin S. H., Sanders D. R. ( 2011): Index funds, financialization, and commodity futures markets, Applied Economic Perspectives and Policy  33, 1– 31. Google Scholar CrossRef Search ADS   Juvenal L., Petrella I. ( 2015): Speculation in the oil market, Journal of Applied Econometrics  30, 621– 649. Google Scholar CrossRef Search ADS   Kallberg J., Pasquariello P. ( 2008): Time-series and cross-sectional excess comovement in stock indexes, Journal of Empirical Finance  15, 481– 502. Google Scholar CrossRef Search ADS   Kapetanios G. ( 2010): A testing procedure for determining the number of factors in approximate factor models with large datasets, Journal of Business & Economic Statistics  28, 397– 409. Google Scholar CrossRef Search ADS   Karali B., Power G. J. ( 2013): Short and long-run determinants of commodity price volatility, American Journal of Agricultural Economics  95, 724– 738. Google Scholar CrossRef Search ADS   Kilian L. ( 2009): Not all oil price shocks are alike: disentangling demand and supply shocks in the crude oil market, American Economic Review  99, 1053– 1069. Google Scholar CrossRef Search ADS   Kilian L., Hicks B. ( 2013): Did unexpectedly strong economic growth cause the oil price shock of 2003–2008?, Journal of Forecasting  32, 385– 394. Google Scholar CrossRef Search ADS   King M., Sentana E., Wadhwani S. ( 1994): Volatility and links between national markets, Econometrica  62, 901– 933. Google Scholar CrossRef Search ADS   Lehecka G. V. ( 2015): Do hedging and speculative pressures drive commodity prices, or the other way round?, Empirical Economics  49, 575– 603. Google Scholar CrossRef Search ADS   Leybourne S. J., Lloyd T. A., Reed G. V. ( 1994): The excess comovement of commodity prices revisited, World Economy  22, 1747– 1758. Loretan M., English W. ( 2000): Evaluation “correlation breakdowns” during periods of market volatility. BIS Conference Papers 8, Bank for International Settlements. Ludvigson S. C., Ng S. ( 2007): The empirical risk return relation: a factor analysis approach, Journal of Financial Economics  83, 171– 222. Google Scholar CrossRef Search ADS   Ludvigson S. C., Ng S. ( 2009): Macro factors in bond risk premia, Review of Financial Studies  22, 5027– 5067. Google Scholar CrossRef Search ADS   Malliaris A. G., Urritia J. L. ( 1996): Linkages between agricultural commodity futures contracts, Journal of Futures Markets  16, 595–610. Google Scholar CrossRef Search ADS   Marshall B. R., Nguyen N. H., Visaltanachoti N. ( 2012): Commodity liquidity measurement and transaction costs, Review of Financial Studies  25, 599– 638. Google Scholar CrossRef Search ADS   Marshall B. R., Nguyen N. H., Visaltanachoti N. ( 2013): Liquidity commonality in commodities, Journal of Banking & Finance  37, 11– 20. Google Scholar CrossRef Search ADS   Palaskas T. B., Varangis P. N. ( 1991): Is there excess co-movement of primary commodity prices? A co-integration test. Unpublished working paper series No. 758, World Bank , Washington, DC, International Economics Department. Pindyck R. S. ( 2001): The dynamics of commodity spot and futures markets: a primer, Energy Journal  22, 1– 29. Google Scholar CrossRef Search ADS   Pindyck R. S., Rotemberg J. J. ( 1990): The excess co-movement of commodity prices, Economic Journal  100, 1173– 1189. Google Scholar CrossRef Search ADS   Raman V., Robe M. A., Yadav P. K. ( 2016): Financialization, intraday institutional trading, and commodity market quality. Unpublished manuscript. de Roon F. A., Nijman T. E., Veld C. ( 2000): Hedging pressure effects in futures markets, Journal of Finance  55, 1437– 1456. Google Scholar CrossRef Search ADS   Rouwenhorst K. G., Tang K. ( 2012): Commodity investing. Unpublished working paper No. 06-12, Yale ICF. Singleton K. J. ( 2014): Investor flows and the 2008 boom/bust in oil prices, Management Science  60, 300– 318. Google Scholar CrossRef Search ADS   Stock J. H., Watson M. W. ( 2002a): Forecasting using principal components from a large number of predictors, Journal of the American Statistical Association  97, 1167– 1179. Google Scholar CrossRef Search ADS   Stock J. H., Watson M. W. ( 2002b): Macroeconomic forecasting using diffusion indexes, Journal of Business & Economic Statistics  20, 147– 162. Google Scholar CrossRef Search ADS   Stoll H. R., Whaley R. E. ( 2010): Commodity index investing and commodity futures prices, Journal of Applied Finance  20, 1– 40. Tang K., Xiong W. ( 2012): Index investment and financialization of commodities, Financial Analysts Journal  68, 54– 74. Google Scholar CrossRef Search ADS   West K. D., Wong K.-F. ( 2014): A factor model for co-movements of commodity prices, Journal of International Money and Finance  42, 289– 309. Google Scholar CrossRef Search ADS   © The Authors 2017. Published by Oxford University Press on behalf of the European Finance Association. All rights reserved. For Permissions, please email: journals.permissions@oup.com

Journal

Review of FinanceOxford University Press

Published: Feb 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off