Dynamic Peer Groups of Arbitrage Characteristics

We propose an asset pricing factor model constructed with semiparametric characteristics-based mispricing and loading functions. approximate the unknown functions by B-splines sieve where number of coefficients is diverging. estimate this test existence function a power enhanced hypothesis test. The solves low problem caused diverging coefficients, strengthened approaching one asymptotically. also investigate structure components through Hierarchical K-means Clusterings. apply our methodology to CRSP (Center for Research in Security Prices) Compustat data U.S. stock market one-year rolling windows during 1967–2017. This empirical study shows presence certain time blocks. find that distinct clusters same characteristics lead similar arbitrage returns, forming “peer group” characteristics.