Milton Ezrati's letter to the editor, "Ibbotson study defenders confuse collateral issue" (P&I, July 24), contained one kernel of truth: an analysis of futures returns assuming collateralization with Treasury bills does bias returns.
What Mr. Ezrati seems not to have realized is that the collateralization biases returns downward. Any argument that requires the removal of collateral returns from the numerator of the return equation must also require the removal of the cash outlay to purchase that collateral from the denominator of the equation.
Ten percent funding would meet exchange requirements for most diversified futures programs or indexes. While the returns would exclude the interest from the T-bills one might post to remove the implicit leverage of commodity futures, the resulting return on equity would have been significantly higher over the last several decades because the net returns would be earned on a 90% lower base. During a year when a fully collateralized investment of $1 million would have generated $50,000 of yield from collateral and $70,000 from the commodities futures for a total return of 12%, a portfolio that excluded the collateral in excess of that required by exchange rules would have earned only $5,000 of yield from collateral. Combined with the $70,000 of return from the futures, that $5,000 would have yielded a return of 75% on the investment of $100,000.
Academics and practitioners incorporate the collateral into both the numerator and denominator of the return equation for three reasons:
1. The volatility of an asset class leveraged 10:1 would grow linearly with the leverage. While the Sharpe ratio would remain the same in either case, most investors can not accept the level of volatility of a portfolio with such high leverage.
2. Incorporation of collateral allows direct comparisons of returns between assets like futures that are implicitly leveraged and those like equities that are implicitly unleveraged. There is no free lunch: returns of implicitly leveraged assets bear implicit costs of leverage. Comparability requires interest-bearing collateral to defease those costs or, alternately, the application of leverage and its requisite costs to the unleveraged assets.
3. Practical considerations compel most institutional investors to fully pay for their investments. In some cases, leverage is precluded by charter; in others it is rightly shunned for its contribution to portfolio risk. Mr. Ezrati's observation that the incorporation of collateral yield precludes direct comparison to fully paid assets could not be further from the truth. It is the exclusion of collateral from both the numerator and denominator of the return equation that effectively precludes comparison to such assets.
Douglas J. Hepworth
director of research
Gresham Investment Management LLC