Review of Quantitative Finance and Accounting, 15 (2000): 127±135
# 2000 Kluwer Academic Publishers. Manufactured in The Netherlands.
Future Stock Performance of Oil and Gas Firms
Conditional on the Imputed Value of Reserves
STEVEN L. HENNING AND WAYNE H. SHAW
Southern Methodist University
Abstract. Harris and Ohlson (1990) provide evidence suggesting market inef®ciencies in the pricing of oil and
gas ®rms in the 1979±1984 period. This paper examines three possible explanations for their results. First, are
differences in oil and gas market values (IVR) explained by risk differences. Second, is the trading rule sensitive
to changes in oil and gas prices? Third, can the results be replicated in a later period?
We provide evidence in support of the risk-based explanation by demonstrating that approximately 75% of the
trading rule return can be explained by adjusting for a stock's covariance with market and energy price
movements. In addition, we demonstrate some time speci®c element to the results since the trading rule performs
poorly in a subsequent time period.
Key words: oil and gas, valuation, organizational structure
JEL Classi®cation: G14, G20
Harris and Ohlson (1990) provide evidence suggesting market inef®ciencies in the pricing
of oil and gas ®rms during the early 1980s. Speci®cally, they demonstrate that positive
returns could have been earned from zero-investment portfolios created by investing long
in ®rms with low imputed values of oil and gas properties per equivalent barrel and selling
short in the opposite case (denoted as Trading Rule #1).
However, Tinic (1990), in a
review of the paper, suggested that their results (1) might be due to risk differences in the
portfolios, (2) might be explained by changes in oil and gas prices, and/or (3) might be
The purpose of this paper is to examine the extent to which the success of Trading Rule
#1 is explained by these factors. The study proceeds in the following manner. First, we
replicate the results of Trading Rule #1 during the 1980±1985 time period investigated by
Harris and Ohlson (1990). Second, we estimate market and oil and gas price betas for each
®rm in the sample. Using these estimates, Trading Rule #1 is implemented using residual
returns from an expanded market model to determine whether the success of the trading
rule is explained by the covariance of individual stocks with these risk measures. Finally,
we perform the same tests for ®rms in a later period, 1986±1990, to determine whether the
Harris and Ohlson (1990) ®ndings are time speci®c.
In summary, we were able to replicate the results of Harris and Ohlson (1990) in their
1980±1985 sample period, documenting an average one-year return on a zero-investment
portfolio of 17.5%. However, after adjusting for market-based risk using a standard market
model, the one-year return declined to nine percent. Adjustment for each ®rm's sensitivity