Trend-followers and managers futures need a better definition of their value-added than "crisis alpha"

Trend-followers have focused on their strategy's ability to provide "crisis alpha", yet there are fundamental issues with the definition of this concept that generates a mixed message for potential investors. 

Crisis alpha has been defined as the excess return of managed futures relative to the market during periods of extreme negative stock returns, a market sell-off. Right away, there is an issue between the traditional or common definition of alpha, the return unrelated to market risk, and the definition of crisis alpha. Crisis alpha is the relative return versus the market conditional on a level of negative market risk. This alpha can vary based on an arbitrary condition used to measure a downturn. It is often set at -20 percent for the equity benchmark, but there is no reason for any special conditional number. 

The definition used in much of the trend-following research and with many marketing presentations focuses on what happens to managed futures trend-following if there is an equity sell-off. In fact, this equity sell-off story has become the standard for funds who discuss the diversification benefit of managed futures. Find periods when there has been a sharp equity declines and then show the returns for a managed futures fund for that defined period. The manager can change the scale or time to form the best conditional relative returns. This crisis definition is related to stock performance and not any exogenous factors that may be associated with a crisis. 

The meaning of crisis alpha becomes more perplexing when you look at the beta of a managed futures fund. The beta for many managed futures funds is often close to zero when measured against the equity market portfolio. By definition and measurement, if there is a significant decline in equities and managed futures have a zero beta, the trend-following strategy will do better in a down market. There is no "crisis alpha". There is just low beta. There may be skill creating a zero beta portfolio and there may be skill with finding trends, but there is no skill in the traditional definition of alpha. 

By forming this relative return concept versus the market condition on a large market decline, any asset that has a stable beta less than one will have some crisis alpha. If there is a lower beta, there will likely be more crisis alpha. At the extreme, an asset that does not move with the market and has no traditional alpha will have positive crisis alpha in a down market equal to the market decline. Trend-followers may produce significant value-added; however, it may not be effectively described through the wording of crisis alpha. 

The definition of crisis alpha as just relative performance during a defined sell-off can apply to many strategies. All of these forms of "crisis alpha" may not employ special skill during a downturn. Any fund well-diversified across a number of asset classes and with the ability to sell short a declining asset should have crisis alpha diversification benefit. The question becomes how well trend-following performs versus other strategies, or what is the overall return versus periods of conditional market stress. The value of the crisis risk strategy should only be judged by accounting for the opportunity cost of holding the offset over time.

• Any asset uncorrelated to equity markets may have "crisis alpha".
• Any diversified portfolio may have "crisis alpha".
• Any strategy that can short an asset class can have "crisis  alpha".
• Any equity put option can have "crisis alpha"

There is nothing wrong with the analysis provided by managers. In fact, even if the research work stops with the traditional crisis alpha charts, the benefit from managed futures is still clear; however, the naming and simple analysis may not provide a deeper understanding of when trend-following will work or why managed futures may be a better strategy than just holding safe assets at critical times. In a perfect world, investors should know the ex-ante conditions that will be good for trend-following. This is a much harder question.