Managed futures vs. stocks

From the Issue of Managed Futures Today
Section: Performance
Diversification among both markets and strategies allows managed futures to generate profits when other investments are headed south.

30 years of managed futures” touched on the long-term performance of managed futures relative to stocks, and the role they can play in a larger investment portfolio. One of the advantages offered by managed futures is their historical lack of correlation to other assets, particularly the stock and bond markets. This characteristic is driven partly by the variety among futures markets, which comprise everything from financial assets to energy (oil and gas), precious metals (gold and silver), grains (wheat, corn, and soybeans), and livestock (hogs  and cattle). A second factor is the ease with which futures fund managers can go long or short, apply different trading strategies, and operate on different time frames.

In the markets, “positive correlation” simply refers to the tendency of one asset to move in the same direction, and at the same time, as another. Corn and wheat, for example, typically move higher or lower in tandem, although not always to the same degree. Similarly, stocks in the same industry or group typically have a high degree of positive correlation. Negative correlation occurs when two instruments have a tendency to move in opposite directions. Markets that are uncorrelated have no discernible relationship.

Correlation is commonly expressed mathematically by the “correlation coefficient” (see “Measuring correlation” below), which ranges from +1.00 (perfect positive correlation) to -1.00 (perfect negative correlation); the closer the value is to zero (positive or negative), the less correlation there is between the two instruments.

Diversification is only possible between assets with low correlation. Numerous studies (see “Additional resources” below) have shown the ability of managed futures to reduce portfolio volatility precisely because of their lack of correlation with other assets. Since 1980, for example, the BarclayHedge (www.barclayhedge.com) CTA Index has a correlation of +0.01 to the S&P 500, +0.11 to U.S. bonds, and -0.01 to world bonds — all values that reflect no meaningful connection between managed futures and these markets. This is what allows managed futures to reduce the volatility of a larger investment portfolio that includes stocks, bonds, and other assets.

Another advantage of managed futures is their ability to generate returns precisely when markets are experiencing extreme drawdowns. This tendency was clear during the 2008 financial panic.
 

A closer look at recent history
It’s one thing to know a certain managed futures index is historically uncorrelated to another asset, but it’s quite another to see how these relationships play out in specific instances. To get a different perspective on the performance of managed futures vs. stocks, we compared the performance of the S&P 500 and the BarclayHedge CTA Index to the returns of two randomly selected CTA sub-groups.

First, we looked at CTAs who were among the top-10 performers in February 2010, were managing at least $10 million, and had been in business since at least April 2005. Five of the top-10 CTAs in Barclay Hedge’s February rankings fulfilled this criteria. Second, we looked at the CTAs that fell into the opposite camp in February — the five with the worst percentage returns for that month. We had no other criteria and no other information about any of the funds.

We then compared the performance of the S&P 500, the CTA Index, and the two CTA sub-groups from 2005 through March 2010. Table 1 shows the annual percentage returns for the S&P 500 and the overall Barclay CTA index, along with the median (see “Key Concepts” for more on median) annual returns for the February top-five and bottom-five CTA portfolios.

Table 1: Managed Futures vs. Stocks

Figure 1 shows how these investments compared from the beginning of 2005 through March 2010 using the Value-Added Monthly Index (VAMI). This measurement reflects the performance of a hypothetical $1,000 initial investment based on each fund or index’s percentage returns (in this case, the annual returns from Table 1).

Figure 1: Long-term trends in the S&P 500 Index

Measuring correlation

The correlation coefficient, sometimes referred to simply as “correlation,” measures the degree of similarity between two data series. In the markets, correlation is typically used to measure how close the relationship is between the prices of two assets (e.g., two different futures markets or stocks), between an individual stock (or trading fund) and an index, or between different indices or investment benchmarks.

Correlation coefficients range between +1.00 and -1.00, with +1.00 representing perfect positive correlation (i.e., two series moving precisely in tandem), -1.00 representing perfect negative correlation (i.e., two series moving exactly opposite to one another), and zero meaning the two variables have no discernible relation.

The “coefficient of determination,” or “r-squared” (r2) is the square of the correlation coefficient. R-squared is always a positive number, ranging from zero to +1.The site — http://davidmlane.com/hyperstat/index.html — offers relatively easy-to-digest explanations of these and other statistical terms.

The results are interesting. We already knew the Barclay CTA Index outperformed the S&P 500 (2008 was the best year for the index since 1990), but more noteworthy are the even higher returns posted by the February top-five CTA portfolio and, especially, the February bottom-five CTA portfolio. The February top-five portfolio boasted a nearly 93-percent total return, but this was almost doubled by the February bottom-five portfolio’s 179-percent total return. (Apparently, February was a bad month for many CTAs who have performed quite well in recent years.)

Of course, we could have very well selected funds that had sub-par performance over this period, not to mention funds that might have failed  between 2005 and 2010. Figure 2 shows the VAMI lines for the two CTAs from the February top-five portfolio that posted the largest and smallest total returns for April 2005-March 2010. Although both funds ended up outperforming the S&P 500, the chart shows how the performance of any individual managed futures fund can vary greatly from the industry’s average performance. The worst-performing fund actually outperformed the S&P 500 and the best-performing fund until 2008, at which point it suffered an extremely large loss that dropped it below the S&P 500. Nonetheless, even this fund ended the analysis window in positive territory. (Potential investors can analyze a fund’s correlation to an overall CTA index to see how far it has deviated from the benchmark historically.)
 

Shining during the stock market’s darkest hour
The other salient point about the managed futures performance in Table 1 and Figures 1 and 2 is the funds’ dramatic outperformance in 2008, the year the S&P 500 lost nearly 40 percent. Not only did both five-CTA sub-groups (and the overall CTA Index) profit handsomely that year, the worst-performing fund highlighted in Figure 2 was, in fact, the only fund out of the 10 funds analyzed here that lost money in 2008. (The best individual 2008 return among the 10 CTAs was 109 percent.)

Figure 2: Individual CTA Perspective

A big reason CTAs performed so well as a group in 2008 — in stark contrast to the stock market, as well as hedge funds — is the flexibility offered by the futures markets and the prevalence of trend-following strategies in the managed futures space. Although at the height of the 2008-2009 financial crisis virtually all markets (with the exception of U.S. Treasuries and the dollar) entered a free fall, this actually worked to the advantage of many managed futures programs because the ease of shorting futures helped trend-following futures traders reap exceptionally large profits as massive downtrends emerged across all assets. As a result, even though most futures markets were highly correlated to the stock market during a large portion of the crisis (as shown in Figure 3), managed futures returns were not — because of ability of fund managers to profit from the unique conditions. (To learn more about trend-following techniques, see “Following the trend-followers”)

Figure 3: Trends Can Be Exploited

Nor was 2008 an isolated experience. During the last major stock bear market in 2000-2002 the S&P 500 index posted annual losses of -10 percent, -13 percent, and -23 percent. In contrast, the overall CTA Index gained 9 percent, 1 percent, and 12 percent during those years.  

Additional resources
The CME Group Web site has several studies regarding the ability of managed futures to reduce overall investment-portfolio risk, boost portfolio returns, and profit in different economic environments. Go to www.cmegroup.com/education/managed-futures-resource-center.html.