Active investment management’s weekly magazine for fee-based advisors

What does alpha have to do with the weather?

by Sep 18, 2014UpClose

What does alpha have to do with the weather?

by Sep 18, 2014UpClose

Understanding the “seasonal performance” of actively managed strategies using market type.

Imagine you like to hike and you live in a region where the weather can change quickly. When you go out in early summer, how do you know what to take with you? Do you pack a parka, snow shoes, and heavy gloves just because the weather might shift?

The answer, of course, is no. Each season has certain characteristics that tend to repeat. So while you may not know exactly what to wear, when you hike in early summer, you know you are better off using limited backpack space for a light raincoat and mosquito repellent than using the space for winter gear.

So what does that have to do with investing? More than you might think. As consultants, my partners and I are challenged with identifying ways to improve our clients’ investment strategies, some of which already have reasonably good performance compared to relevant benchmarks. One go-to method of strategy enhancement is assessing “seasonal performance.”

In investment lingo, a “season” is often referred to as either a market type, or, for the more academically inclined, a market regime. Adherents of using market type contend something that seems obvious to most market participants: no strategy works effectively all the time.

Market type is any systematic, macro-level means of categorizing investment strategy performance using information about market dynamics with a tendency to persist.

While definitions vary from source to source, in this article we’ll view a market type as any systematic, macro-level means of categorizing investment strategy performance using information about market dynamics with a tendency to persist. Sound complicated? Believe me, it’s not—an example will simplify things.

Let’s say you are an investment advisor and one of the strategies you frequently recommend to clients is a momentum-based system, which we’ll call Strategy X. Each month, Strategy X buys the three best-performing instruments from a list of exchange-traded funds (ETFs). The strategy takes advantage of a well-researched historical market edge (i.e., momentum) and meets long-run expectations, but experiences short periods of underperformance. Is there a way we can we improve Strategy X without significantly reducing the likelihood it will continue to perform well in the future? Exploring alternatives based on market type is one answer.

To consider improvements to Strategy X using market type, first we need a systematic means of segmenting historical system performance. To keep things straightforward, we will use the 200-day simple moving average (SMA) to define market type since it is commonly used by traders. More specifically, when the slope of the S&P 500 200-day SMA is rising, we place our historical strategy results into a bucket called Market Type A, and when the slope is falling, the historical results go into a bucket called Market Type B.

Figure 1: Knowing that all of the positive system performance occurs in Market Type A enables a new dimension of diversification,capable of delivering higher returns and lower drawdowns.

¹For illustrative purposes only. A 3% risk-free rate was used over the 22-year simulation period. For brevity, only 15 of the 22 years are displayed in the chart. Because “Strategy X” employs a rules-based system for entering and exiting trades, results “During Both Market Types” are not perfectly weighted averages of Market Type A plus B.

The numbered circles in Figure 1 show changes in market type based on this definition. Note that market type shifts happen infrequently—there were only seven changes in the roughly 15 years of S&P 500 data shown—which is often the case with effective market type definitions.

Aggregate “Strategy X” results across both market types are shown in the table: 12.7% compound annual return, drawdown roughly equivalent to the S&P 500, and a Sharpe ratio of 0.39, metrics which collectively represent noteworthy outperformance versus the benchmark. But the table also clearly shows all of the alpha—a 75% improvement in compound annual return and a 50% reduction in maximum drawdown—comes during what we’ve defined as Market Type A. So while the hypothetical “Strategy X” outperforms the S&P 500 overall through both up and down markets, is there room for strategy refinement to further boost alpha?

An often-overlooked feature of market type is its ability to open up a new, potentially more effective dimension of diversification. Market type allows you to employ each strategy only during “seasons” when they are likely to be profitable. By doing so, it enables you to seamlessly rotate capital between strategies specifically designed to complement one another. In the example above, choosing to only use Strategy X when in Market Type A allows you to diversify into strategies designed to work more effectively in Market Type B. When constructed correctly, this underused diversification method allows you to enhance returns and reduce drawdowns.

In the following figure, we’ve chosen a market type methodology with only two possible values: Market Type A (rising 200-day SMA) and Market Type B (falling 200-day SMA). Note, however, there are a number of ways to define market type, and the table shows a small sample of options. Each technique could be combined with one or more of the others resulting in a matrix of types.

Example of Market Typing Methods

So how do you choose which market typing method to use? As a general rule, a simple solution like using the direction of a key, long-term moving average of price or a relative level of historical volatility is often a good starting point. Even so, the most accurate answer is market types should be designed to support broad categories of client return objectives and risk tolerance, and need to be carefully matched to strategies. Just as investing a large portion of capital in equities is inappropriate for some categories of clients, so is choosing a universal definition for market type.

Clearly, effective market type diversification has the potential to provide your clients higher returns with the same, or even lower, risk. It allows active diversification not just across asset classes, but also across market types within a given asset class.

Diversification is often primarily viewed as a means of reducing risk, but that doesn’t mean it has to come at the cost of returns. For advisors and the third-party asset managers they partner with, market type-based strategy diversification can be a sophisticated way to boost alpha without increasing drawdowns.

The opinions expressed in this article are those of the author and do not necessarily represent the views of Proactive Advisor Magazine. These opinions are presented for educational purposes only.


Dave Witkin was formerly a partner in StatisTrade, a trading strategy modeling and consulting company. Mr. Witkin has traded stocks, commodities, and options over the last 20 years, and in 2011 was featured in Dr. Van Tharp's book, "Trading Beyond the Matrix."


Manage investment risk better than ever.
Get started – It’s free

About Us