31.7 CONCLUSION

To understand the risk-return properties of CTAs, one needs to apply special measures designed to account for special characteristics of CTAs. The fact that CTAs allow investors to gain exposure to a variety of markets with minimal investments and that the margin requirement for each contract is closely linked to the riskiness of the underlying market means investors should carefully examine the amount of margin that a CTA is carrying. In addition to the margin-to-equity ratio, investors should use other metrics, such as volatility, VaR, and CaR, to gain a better understanding of a CTA's risk profile.

One important issue that investors need to address is the lack of proper benchmarks for CTAs in general, and discretionary CTAs in particular. Although passive strategies have been developed to mimic the risk-return properties of trend-following CTAs, no such passive strategies have been developed for discretionary CTAs. The MLM Index represents one such attempt to produce a passive and investable benchmark for systematic CTAs. The performance of this index can be used to estimate the returns on CTAs due to their beta exposures to a passive trend-following strategy. The results presented in this chapter showed that less than 50% of the return earned by trend-following CTAs is due to their beta exposure to a passive trend-following index.

This chapter has also examined the role of CTAs in providing downside protection for traditional asset classes, such as equities and high-yield bonds, as well as alternative assets, such as hedge funds. It was argued that though CTA return profiles give the appearance that they are long volatility, it is more appropriate to characterize CTAs as being long gamma. This gives the appearance that CTAs are long volatility while what trend-following CTAs do is increase the delta of their positions as prices move in their favor, the basic characteristic of a long gamma position.

1 For example, using Microsoft Excel, one can obtain an estimate of α using the following function: NORM.S.INV(1–Confidence Level).

2 See CAIA Level II: Current and Integrated Topics.

3 See Shadwick and Keating (2002) and Kazemi, Schneeweis, and Gupta (2003).

4 It is important to realize that many of the reported properties are time specific and may change significantly over short periods of time, especially when market conditions change.

5 See McCarthy, Schneeweis, and Spurgin (1996); Edwards and Park (1996); and Schneeweis, Spurgin, and Potter (1997).

6 The case for creating passive indices and the associated methodology are described in Spurgin, Schneeweis, and Georgiev (2001).

7 See www.ingarm.org.

8 See Schneeweis (2009) and Kat (2002).

9 See Gregoriou et al. (2004).

10 For example, see Fung and Hsieh (1997).

11 This section follows closely the arguments set forth in Malek and Dobrovolsky (2009).