ommodity futures funds are an important part of today's futures markets. Orga-C nized as limited partnerships, these funds range in size from a few hundred thousand up to several hundred million dollars in capital. With this capital, Brorsen and Irwin (1987) estimated funds held an average of 23% of open interest (one side of the market) in the nearby contract in ten important futures markets in 1984.
Most of the trading by these funds is based on technical trading systems (Irwin and Brorsen (1985)). Lukac et al. (1988), Irwin and Uhrig (1984), and others have shown that these technical systems can provide returns greater than zero. Murphy (1986), Irwin and Brorsen (1985), and Irwin and Landa (1987) showed technically traded futures funds would enter an investor's optimal portfolio that included stocks, Treasury Bills, and other assets. However, not all of the research is unanimous. Elton et al. argued technically traded commodity funds' returns were poor and that commodity funds did not belong in an optimal portfolio.
Managing risks is an important factor in the success or failure of a futures fund. The risk of technical trading systems was demonstrated by Lukac et al. (1988) and by Murphy (1986). They showed the standard deviation of return was quite large and, in fact, over double that of a naive portfolio of equally weighted long positions. Elton et al. showed that the commodity funds' standard deviation was over twice that of common stocks, long-term corporate bonds, long-term government bonds, and Treasury Bills. To be a more competitive investment, futures funds need to increase their return relative to their risk.
The purpose of this article is to suggest a means of reducing the riskiness of futures funds by finding an optimal allocation of capital across commodities in a futures fund portfolio. Previous simulation studies and actual futures funds allocate trading capital by varied methods.
In Brorsen and Irwin's (1987) survey, commodity fund managers responded that margin is not a deciding factor in their allocation decision. Eighty-five percent of managers responding said they hold numbers of contracts according to their volatility. Half of these held positions based on variances only, -by holding more contracts of less volatile commodities. Thirty-five percent based their allocation on a tradeoff between risk
The authors thank David C. Wilson and Paul Preckel for computational assistance and Scott Irwin, Bill Uhrig, and Deb Brown for helpful comments.