MARKET MANIPULATION, MOMENTUM CRASHES, AND THE LIMITS OF PRICE-VOLUME SCREENING

M. Božović

Abstract

Momentum trading strategies earn persistent premia but face severe crashes during market recoveries. We investigate whether these crashes result from unwinding price manipulation. We intersect a forensic manipulation score (FMS) of abnormal turnover and illiquidity, calculated from CRSP data, with 489 SEC manipulation episodes (2000–2018). Only 14 episodes (2.9%) enter the investable momentum universe, and 94% of unmatched cases are OTC securities. In Firth penalized logit regressions, FMS loses predictive power once size is controlled, whereas Sloan operating accruals remain significant, although based on 32 events. In conditional crash regressions across 37 rebound months between 1980 and 2024, high-FMS stocks exhibit no detectable loser crash amplification (minimum detectable effect is 1.45). We therefore find no evidence that manipulation drives momentum crashes. Observable enforcement should not be treated as representative of investable equities for forensic screening, audit risk assessment, or market surveillance.

Source: semanticscholar · PDF

Read the AI summary and key takeaways for traders on WOBR Quant Research.