How do the daily price statistics of commodities differ, and how do they compare with those for equities? In their May 2011 paper entitled “The Dynamics of Commodity Prices”, Chris Brooks and Marcel Prokopczuk examine the daily price statistics for six major commodity markets (crude oil, gasoline, gold, silver, soybeans and wheat) individually and relative to each other and the equity market. Using daily spot prices for the commodities and daily levels of the S&P 500 Index for January 1985 through March 2010 (over 25 years), they find that:
- Mean daily return varies considerably across commodities (wheat lowest and gold highest), with all lower than that for the S&P 500 Index.
- Only gold has a lower standard deviation (volatility) of daily returns than the S&P 500 Index, with crude oil and gasoline the highest (annualized 42.8% and 44.1%, respectively).
- Relationships between returns and volatilities are: negative for crude oil and equities; close to zero for gasoline and wheat; and, positive for gold, silver and soybeans. A negative (positive) relationship implies a return distribution skewed to the left (right). These differences may arise from the relative influence of hedgers (whose activities tend to suppress volatility) and speculators (whose activities tend to elevate volatility).
- While daily returns for commodities exhibit lower kurtosis than those of the S&P 500 Index (with soybeans and wheat lowest), values are still higher than for a normal distribution.
- Daily price jumps (up or down) are more frequent on average for commodity markets than the 1.8 times per year for the equity market: about six times a year for crude oil, gasoline and wheat; 12 to 13 times a year for soybeans and gold; and, about 22 times a year for silver. Mean jumps are negative for all markets, but energy prices tend to jump up when equities jump down.
- Daily return correlations are high for pairs of related commodities (0.68 for gold/silver and 0.49 for crude oil/gasoline), but very low (never exceeding 0.11) for unrelated commodities. The likelihood of coincident price jumps follows a similar pattern.
In summary, evidence from daily prices indicates that it is inappropriate to lump different kinds of commodities into a single asset class and that commodities in general are a useful diversifier of both equity returns and equity volatility.
The study does not address the stability of the daily commodity price statistics over time. Results summarized in “Basic Equity Return Statistics” suggest that these statistics (and, hence, diversification power for equities) may vary considerably across extended subperiods.
Frequently Asked Questions
What does volatility of daily returns mean when comparing commodities to stocks?
Volatility of daily returns is the standard deviation of a market's daily price changes, often annualized to compare assets. In a study of six commodities (crude oil, gasoline, gold, silver, soybeans, wheat) versus the S&P 500 Index from January 1985 to March 2010, only gold was less volatile than the S&P 500, while crude oil and gasoline were highest at 42.8% and 44.1% annualized.
Are commodities a useful diversifier for stock portfolios?
Commodities in general are a useful diversifier of both equity returns and equity volatility, according to daily price data for six major commodities versus the S&P 500 Index from January 1985 to March 2010. Unrelated commodities had daily return correlations never exceeding 0.11, and energy prices tended to jump up when equities jumped down. The study does not address whether these statistics are stable over time, so diversification power may vary across subperiods.
How are price jumps and return distributions measured for commodity markets?
The study of six commodity markets (crude oil, gasoline, gold, silver, soybeans and wheat) used daily spot prices and daily S&P 500 Index levels from January 1985 to March 2010 to compute mean returns, standard deviations, skewness, kurtosis, correlations and the frequency of daily price jumps. Jumps occurred about six times a year for crude oil, gasoline and wheat, 12 to 13 times for soybeans and gold, and about 22 times for silver, versus 1.8 times a year for equities. Mean jumps were negative for all markets.
Should all commodities be treated as one asset class like equities?
Lumping different commodities into a single asset class is inappropriate, based on daily price statistics for six commodities from January 1985 to March 2010. Related pairs were highly correlated (0.68 for gold/silver, 0.49 for crude oil/gasoline) while unrelated commodities had correlations never exceeding 0.11. Return-volatility relationships also differed: negative for crude oil and equities, near zero for gasoline and wheat, and positive for gold, silver and soybeans.
What are the limitations of using historical commodity price statistics to judge diversification benefits?
Historical commodity price statistics may not remain stable, because the study of six commodities (January 1985 to March 2010) does not address how the statistics change over time. Related equity return statistics suggest they, and hence diversification power for equities, may vary considerably across extended subperiods. The results also rely on daily spot prices for only six markets, so they may not generalize to all commodities.