New framework exposes flaws in strategy capacity estimates
Research shows that standard methods for measuring how much capital a trading strategy can absorb often ignore the compounding effects of crowding.
Trading strategy capacity is usually measured using observational proxies that rely on incompatible assumptions. This research establishes a formal experimental design to determine exactly how much capital a strategy can deploy before its edge disappears, identifying a fundamental conflict between statistical robustness and the ability to detect crowding.
Most capacity tests fail because they treat parallel implementations of a strategy as independent units. In reality, these implementations trade the same securities, creating a shared crowding effect. When researchers compare implementations on the same date to remove market-wide shocks, they inadvertently remove the very signal they are looking for: the impact of the strategy's own accumulated position.
This design flaw means that a fixed-length trial typically understates the eventual erosion of an edge. Because deployed capital erodes returns gradually, a short-term experiment misses the long-term decay caused by aggregate positioning. The paper provides a correction method to account for this underestimation and defines the specific costs associated with identifying the true aggregate effect.
For firms managing large quantitative portfolios, this suggests that current capital allocation models may be overestimating the ceiling of their strategies. If the capacity is measured via a same-date design, the firm is seeing a private response to a prevailing level of positioning rather than the total impact of their own scale.
To get an accurate reading, the research argues that managers must either use implementations with deliberately different exposure levels or track variation in positions over time. This requirement increases the operational cost and time needed to validate a strategy's scalability.
Asset managers who rely on rapid scaling of quantitative signals face a hidden risk where the perceived capacity is a mirage created by the experimental design. The ability to price the cost of a proper capacity study allows firms to decide if the precision of a capacity estimate is worth the deployment cost.