The Capacity of Trading Strategies.

Summary Due to non-linear transaction costs, the financial performance of a trading strategy decreases with portfolio size. Using a dynamic trading model a la Garleanu and Pedersen (2013), the authors derive closed-form formulas for the performance-to-scale frontier reached by a trader endowed with a signal predicting stock returns. The decay with scale of the realized Sharpe ratio is slower for strategies that (1) trade more liquid stocks (2) are based on signals that do not fade away quickly and (3) have strong frictionless performance. For an investor ready to accept a Sharpe reduction by 30%, portfolio scale (measured in dollar volatility) is given by a simple formula that is a function of the frictionless Sharpe, a measure of price impact, and a measure of the speed at which the signal fades away. They apply the framework to four well-known strategies. Because stocks have become more liquid, the capacity of strategies has increased in the 2000s compared to the 1990s. Due to high signal persistence, the capacity of a "quality" strategy is an order of magnitude larger than the others and is the only one highly scalable in the mid-cap range.