The mispricing factors model, proposed by Stambaugh and Yuan (2017), is a four-factor model that constructs factors by aggregating information across multiple anomalies rather than relying on a single characteristic per factor.
Model specification
The model combines market and size factors with two “mispricing” factors:
- MGMT (Management): averages stock rankings across anomalies related to management decisions (net stock issues, composite equity issuance, accruals, net operating assets, asset growth, investment-to-assets)
- PERF (Performance): averages rankings across anomalies related to performance (financial distress, O-score, momentum, gross profitability, return on assets)
The two clusters are identified by grouping the 11 anomalies from Stambaugh, Yu, and Yuan (2012, 2014, 2015) by their return co-movement.
Key results
- The four-factor model accommodates a large set of anomalies better than the Fama-French five-factor, Carhart four-factor, and q-factor models
- The size factor in this model reveals a small-firm premium nearly twice usual estimates (because mispricing factors absorb the junk-in-small-cap effect)
- Investor sentiment predicts the mispricing factors, especially their short legs, consistent with a behavioral mispricing interpretation
Theoretical motivation
Unlike FF5 (dividend discount model) or q-theory (production-side equilibrium), this model is explicitly motivated by mispricing. The asymmetry between long and short legs (short legs are more sentiment-sensitive, reflecting short-sale constraints) supports a behavioral rather than risk-based interpretation.
Relationship to other models
The model differs from alternatives in factor construction philosophy: rather than one anomaly per factor, it averages across anomalies for a less noisy mispricing signal. This addresses the factor-zoo concern about proliferating factors by compressing information from 11 anomalies into two factors.