Crowded futures market trade measurement and momentum, value and basis factor strategies - Watch speculators who are adding positions

 

No one wants to be in a crowded trade. You want to get in early, but when others rush in, you want to head to the exits. Yet, measuring crowdedness is not easy. Crowdedness, the excess capital or interest in specific trades, can vary through time and create time varying returns. 

Recent research has focused on the potential crowdedness of factor strategies for momentum, carry, and value in futures markets through tracking net trade flows for specific reporting groups. This offers a direct way of measuring speculative interest through the information collected in the CFTC's Commitment of Traders reports. See "Crowding and Factor Returns". 

The researchers find that when speculative non-commercial interest is high there is a negative predictive impact on expected factor strategy returns. Your speculative factor returns are better when there is less competitive from other speculators. 

The speculative positions are measured as the net non-commercial positions relative to open interest against the average net speculative positions over the last year. Higher crowding leads to lower subsequent excess futures returns. 

Variation in crowding can also be measured and is related to past returns, performance chasing, higher volatility, and the cost of arbitrage capital as measured by increases in repo or the TED spread. Speculators follow performance and will change positions with capital costs. 

Factor strategies can be created by ranking momentum returns for 26 commodity markets, value based on deviations from long-term mean prices, and carry (basis) constructed from the difference between the first and second nearby futures. Once factor returns have been calculated, the crowdedness measure across all markets can be used to sort factors between periods of high and low crowdedness. From this sort, a measure of conditional crowdedness can be constructed. Excess returns are higher during periods of low crowding. We show the momentum results below. Similar relationships are found for value and the basis. 

The cumulative return for a factor strategy conditional on crowding is compelling for momentum, value and basis trading. If you are going to trade even simple factor strategies, it is important to focus on whether other speculators are jumping into the markets.