In our last article, we discussed the potential benefits of adding alternative assets to a strategically allocated portfolio, along with some of the difficulties that advisors may face in the process. In this posting, we will discuss one specific approach to constructing portfolios using alternatives.

As a reminder, we need to address the following set of questions: First, how much to allocate to alternatives in the overall portfolio? In addition, which sub-strategies should we include within the alternative sleeve, how should we combine them, and which managers should we use?

Analyzing Different Types of Alternative Strategies

To address these complex problems, we first need to group the investment strategies into similar approaches, or clusters, so we may compare and contrast them, and so we can meaningfully compare the strategies within each cluster. One of the more appealing techniques for grouping strategies is cluster analysis, a statistical technique that requires no preconceived notion of what each strategy does. Instead, this technique “clusters” similar strategies together based on their past performance.

Clustering allows us to examine and logically assign names to different types of alternative strategies. Three major clusters that tend to be the most effective in managing downside risk in a portfolio are market neutral, hedged equity/merger arbitrage and managed futures. They have the following key characteristics:

  • Market neutral strategies, which seek to profit from both increasing and decreasing prices in a single market or numerous markets, often by investing both long and short;
  • Hedged equity strategies, which attempt to capture potential stock upside while using active risk management techniques, such as short selling, to reduce the effect of market swings on long-term performance; and
  • Managed futures strategies, which employ a trend-following approach implemented through futures trading, which in turn provides risk management tools that may not be available through direct equity investments.

Portfolio Construction

The ideal portfolio construction process for including alternatives allows advisors to make informed choices in terms of trading off a portfolio’s downside protection in bad


times against achieving the same upside as the market in good times.

Our ideal process chooses the amount of the overall alternative slice and the weighting to each cluster so as to minimize the downside risk of the combined portfolio. Daily market data should be used to model each cluster, in order to better reflect the fact that distributions of alternatives tend to have large negative returns that are not matched by equal-size positive returns.

This optimization should be performed separately for each asset allocation risk-level portfolio. For example, the alternative sleeve for a conservative portfolio will have a different ratio of market neutral, hedged equity, and managed futures strategies than the alternative sleeve for an aggressive portfolio.

Once the optimal ratios of sub-strategies are determined for each risk level, the final step is to populate the allocations by selecting individual mutual funds or ETFs for each cluster. This can be done via either quantitative or qualitative means, so long as the strategies are all chosen from the correct cluster.

Testing the Results

To help determine the validity of this approach, extensive quantitative testing has been performed using historical data. Out-of-sample backtests have been conducted using this methodology for the market crash of 2008, and then the recovery from mid-2009 to the present. These backtests show what might have happened had these strategies been managed through the specified backtesting period, because at any given time period, the analysis only uses data known to that point to construct the portfolios for each period.

The results show that the portfolios created using this approach are very efficient in terms of limiting shortfall risk during the 2008 and 2009 period, since they significantly reduce not only the traditional measures of risk, such as overall portfolio standard deviation, but also such risk measures as drawdown, max daily loss and other commonly used downside risk measures. In addition, the sample portfolios captured a sizeable portion of the market upside during the 2009-2010 market recovery.

Over all, although adding alternatives to a long-only portfolio is a daunting process, with the right techniques these strategies can be included in portfolios with potentially favorable results.