Detecting Crowded Trades in Currency Funds

Investors and regulators suspect that crowded trades may pose a special risk. The authors propose a methodology to measure crowded trades and apply it to currency managers. This methodology offers useful insights regarding the popularity of certain trades among hedge funds and provides regulators with another tool for monitoring markets.The financial crisis of 2008 highlighted the importance of detecting crowded trades because of the risks they pose to the stability of both the global financial system and the global economy. Crowded trades, however, are perceived as difficult to identify. To date, no single measure that captures the crowdedness of a trade or trading style has been developed.We proposed a methodology to measure crowded trades and applied it to professional currency managers. We investigated three trading strategies: carry, trend, and value. Earlier research has shown that returns on an index of currency hedge funds, as well as returns earned by individual fund managers, are closely related to returns on indices that represent these three trading strategies. Our measure of crowdedness attempts to capture the popularity of a strategy by counting the number of funds whose returns have significant style betas versus the three basic strategies. Specifically, we focused on a measure of net crowdedness defined as the percentage of funds in our sample with positive style betas (with respect to a trading strategy) minus the percentage of funds with negative style betas (“contrarians”).We used daily data from the Deutsche Bank FXSelect currency trading platform to form weekly returns for 107 managers listed on the platform. The data allowed us to control for backfill bias and survivorship bias.Our results show that our measure of crowdedness varied considerably over time. For example, from September 2005 to June 2010, carry crowdedness ranged from a low of –10.5 percent to a high of 32.1 percent, trend crowdedness moved from –3.4 percent to 33.9 percent, and value crowdedness varied between –28.3 percent and 12.2 percent. Moreover, each trading strategy experienced at least two high and two low points of crowdedness as managers alternately adopted and then abandoned a strategy. Our results show that a trading strategy becomes crowded not only because existing managers switch to that strategy but also because newcomers join the database and adopt the strategy. A trading strategy becomes less crowded as managers close their positions and as other managers exit the platform.Specifically, we found that carry became a crowded trading strategy toward the end of the first quarter of 2008, shortly before a massive liquidation of carry trades. The timing suggests a possible adverse relationship between our measure of style crowdedness and the future performance of that trading style. Crowdedness in both the trend and value strategies support this hypothesis.Our sample period covered 63 months, of which 27 were effectively an out-of-sample period. The out-of-sample results confirm the usefulness of our measure of crowdedness. After a period when carry returns were very favorable, the carry strategy became crowded again in the fall of 2009 and then experienced a sharp reversal during the European sovereign debt crises in the spring of 2010 and after the “flash crash” of May 2010.Although we applied our approach to currencies, the methodology could be used to measure the popularity or crowdedness of any trade with an identifiable time-series return. Our methodology may offer helpful insights regarding the popularity of certain trades—in currencies, gold, or other assets—among hedge funds. Further research in this area could yield useful findings for investors, managers, and regulators.