The widely used Fama-French three-factor model explains stock returns based on aggregate market return, firm size (small versus large) and firm valuation (value versus growth). Since the Fama-French model does not explain the stock price momentum effect, researchers and investors often add momentum as a fourth factor to predict future stock returns. Might some other small set of factors (three) outperform the Fama-French model in explaining stock returns, obviating the need for a momentum factor and accounting for other stock return anomalies as well? In their June 2009 paper entitled “A Better Three-Factor Model That Explains More Anomalies”, Long Chen and Lu Zhang argue that a three-factor model based on aggregate market return, level of firm investment relative to assets (low versus high) and return on assets (high versus low) substantially outperforms the Fama-French model in explaining stock returns. Using a wide range of firm and stock data for a broad sample of stocks over the period 1972-2006 to test this model, they conclude that:
- The investment premium (return for the lowest 30% investment firms minus return for the highest 30% investment firms) averages about 0.43% per month over the entire sample period.
- The return on assets (ROA) premium (return for highest 30% ROA firms minus return for lowest 30% ROA firms) averages about 0.96% per month over the entire sample period.
- This new three-factor model:
- Fully explains the effects of firm size, stock price momentum (such that no momentum factor is needed) and firm financial distress on stock returns.
- Outperforms the Fama-French model in explaining the effects of net stock issuance, asset growth and earnings surprises on stock returns.
- Roughly matches the Fama-French model in explaining the effect of book-to-market value on stock returns.
- This new model does not interpret factor effects in terms of reward for risk and is not directly amenable to behavioral mispricing explanations.
The authors propose that researchers and investors view this new market-investment-ROA model, which economically resolves many known anomalies, as a workhorse for estimating future stock returns.
In summary, investors who pick stocks may want to focus on two firm characteristics: (1) investment-to-assets ratio; and, (2) return on assets. Evidence indicates that these two characteristics are foundational to a broad range of stock return anomalies.
