Research Review

A Closer Look Beyond the Styles of Commodity Trading Advisors

Abstract

In a recent articleⁱ, “Look Beyond the Styles of Commodity Trading Advisors” Pauline Lam lays out common criteria in assessing CTAs for allocation purposes. In this review we examine some of the implications of the results of the article and provide updated estimates. Our results include the period March 2004 – September 2004 when the industry experienced some of its largest drawdowns in history.

Introduction

Under the Commodity Exchange Act, all individuals and firms, with certain exceptions, that intend to do business as futures professionals must be registered with the Commodity Futures Trading Commissionⁱⁱ (CFTC). The Commission has authorized the National Futures Association (NFA), a self-regulatory organization, to receive and review applications and grant registrations. Two such categories of futures professionals that must be registered include Commodity Pool Operators (CPOs) and Commodity Trading Advisors (CTAs). The terms Commodity Pool Operators (CPOs) and Commodity Trading Advisors (CTAs) are defined by the CFTC as follows:

• Commodity Trading Advisor - Any person who, for compensation or profit, directly or indirectly advises others as to the advisability of buying or selling commodity futures or option contracts.

• Commodity Pool Operator - An individual or firm which operates a commodity pool. (For example: If a pool is organized as a limited partnership, its general partner typically is its CPO.) A commodity pool is an investment trust, syndicate, or similar form of enterprise operated for the purpose of trading commodity futures or option contracts

It needs to be noted at the outset that there are many hedge funds that are registered as CTAs and even CPOs with the NFA. This often happens in the case of hedge funds that are classified as global macro. Lam’s suggestion that CTAs may trade financial futures as part of their portfolios is however true. CTAs indeed, also trade currency, stock index and commodity futures. The vast majority of CTAs in the CISDMⁱⁱⁱ Database however are classified as Diversified (ones that trade several markets) and a somewhat fewer number of CTAs are classified as Physicals (ones that trade markets such as energy, metals, softs etc.). The following better known hedge funds^iv or fund of funds are also registered either as CTAs or CPOs or both.

1. Caxton Associates, LLC
2. Kingdon Capital Management, LLC
3. Moore Capital Management, LLC
4. Renaissance Technologies Corp.
5. SAC Capital Advisors, LLC
6. Tudor Investment Corporation
7. Pacific Alternative Asset Management Co., LLC

The classifications in the CISDM database^v are manager reported. A fund is assigned to the CTA database if it trades futures and options markets and is generally invested through managed accounts. A fund is assigned to the CPO database if it pools funds from multiple participants. Note that CPOs may be single advisor (engaging a single CTA) or multi-advisor (engaging several CTAs). Funds may also be registered as both CTAs and CPOs as is often the case. Global macro funds on the other hand may trade equity markets as well and be registered as both CTAs and CPOs.

In this review we will reexamine the performance figures of Lam and provide updated results. In particular, we examine the period January 1992 – August 2005 which includes the period March 2004 through September 2004 when the managed futures industry experienced some of its largest drawdowns in history.

Diversification Benefits

Lam notes that CTAs exhibit low-to-negative correlation with traditional assets like stocks and bonds. The following exhibit presents correlation statistics between CISDM CTA sub-indices^vi, CISDM Global Macro Index, Stocks and Bonds over various periods. The period considered by Lam was January 1992 through May 2004. The CISDM indices underwent certain structural changes in October 2005. For purposes of this comparison we will use the CISDM Asset Weighted CTA Discretionary and Systematic indexes as well as the CISDM Global Macro Index. We also use the S&P 500 Total Return Index and the Merrill Lynch U.S. Treasuries, 10+ Yrs Total Return Index to represent stock and bond returns respectively.

The results for the period January 1992 – August 2005 generally follow the same patterns as the period January 1992 – May 2004. The correlations between large-cap stocks and the systematic and discretionary indexes are either low or negative. As is the case with Lam, bonds exhibit higher correlations with systematic and discretionary CTA indices. There are however, certain differences between the results above and the results reported in Lam. We find that the correlations between systematic and discretionary CTAs are around 0.5. A majority of the systematic CTAs are trend-followers whereas the trading decisions at discretionary CTAs are more fundamentally driven. The correlation between CISDM CTA Discretionary Index and the CISDM Global Macro Index is around 0.5 as well which is higher than the figures reported in Lam.

Performance Statistics

Lam computes and compares performance statistics for the period January 1990 through May 2004 as well as for five years preceding May 2004. In the exhibit below, we present performance statistics for the CTA Asset and Equal Weighted Indices, CPO Asset Weighted and Equal Weighted Indices and the CISDM Global Macro Index.

Our performance statistics in general are similar to the results reported by Lam. The average of the global macro strategy is higher than the averages of the asset and equally weighted CTA and CPO indices over the period January 1990 – September 2005. However over the period January 2000 – September 2005 the average of the global macro strategy is lower than the averages of the asset and equally weighted CTA and CPO indices. This difference in performance is attributable to the year 2002 when the CISDM Global Macro Index performed poorly.

Time Frames

Lam notes certain distinctions between commodity trading advisors and global macro hedge funds. These include:

a. GM trades are discretionary in nature with traders anticipating trends in their development.
b. GM has more flexibility in formulating its trades: they allow for faster reaction time for getting in and out of positions.

The CISDM database provides information on whether a CTA is systematic or discretionary as well as the time frame. We form six portfolios with both active and dead CTAs. These include long term discretionary CTAs, medium term discretionary CTAs and short term discretionary CTAs, long term trend followers, medium term trend followers and short term trend followers. Correlations of these portfolios are presented in Exhibit 3.

Lams observations are in general supported by the results. The CISDM Global Macro Index has low correlations with all six portfolios. As expected the long and medium term trend following portfolios are highly correlated. The performance of these portfolios over the period January 1990 – September 2005 is given in Exhibit 4.

__________________________________________________

i Published in the Journal of Wealth Management, Fall 2004 issue.
ii www.cftc.gov
iii The CISDM Database currently reports information on more than 700 Commodity Trading Advisors and Commodity Pool Operators.
iv Institutional Investor annually publishes rankings on the top 100 hedge funds in terms of assets.
v The CISDM Database is currently managed by Alternative Investments Research and Technologies (AIRT), LLC. vi The CISDM Hedge Fund and CTA Indices were restructured in October 2005.
vi The CISDM Hedge Fund and CTA Indices were restructured in October 2005.