Factor Investing
Value, momentum, and quality — the systematic edge · 17 min
Why Factors?
In the 1990s, Eugene Fama and Kenneth French discovered that two characteristics predicted stock returns better than CAPM: company size (small-cap stocks outperformed large-cap) and company value (cheap stocks — low price-to-book — outperformed expensive stocks). These persistent, systematic return differences are called factors.
The Fama-French 3-factor model extended CAPM:
E[R_i] = R_f + β × MKT + s × SMB + h × HML
- MKT: Market factor (as in CAPM)
- SMB (Small Minus Big): Return of small-cap stocks minus large-cap stocks
- HML (High Minus Low): Return of high book-to-market (value) stocks minus low book-to-market (growth) stocks
The Major Factors
1. Value: Cheap stocks (high earnings yield, high book/price, low P/E) historically outperform expensive (growth) stocks. The intuition: cheap stocks are often out-of-favor businesses that are undervalued. Investors overpay for growth and underpay for recovery stories.
2. Momentum: Stocks that performed well over the past 3–12 months tend to continue outperforming for the next 3–12 months. Jegadeesh and Titman (1993) documented this. The intuition: investors underreact to news; prices trend as information is gradually priced in. Note: momentum reverses over very short (1-month) and very long (3–5 year) horizons.
3. Quality: Companies with high profitability, low debt, stable earnings, and strong free cash flow generation tend to outperform. Buffett's investing style is largely quality investing — "wonderful companies at fair prices."
4. Low Volatility: Lower-risk stocks have historically produced competitive risk-adjusted returns despite having lower raw returns. This "low-vol anomaly" contradicts CAPM and is attributed to investor preference for lottery-like high-beta stocks, creating mispricing on the low-risk end.
5. Size: Small-cap stocks have outperformed large-cap historically, but this premium has been elusive in recent decades and may partly reflect higher transaction costs and liquidity risk rather than true mispricing.
Factor Portfolios: Long-Short Construction
A factor portfolio is typically constructed as a long-short portfolio: go long the stocks with the highest factor scores and short the stocks with the lowest factor scores. This isolates the factor return independently of market direction.
For a momentum factor portfolio:
- Rank all S&P 500 stocks by 12-1 momentum (12-month return excluding the most recent month)
- Go long the top quintile (20%) — the winners
- Go short the bottom quintile — the losers
- Equal-weight within each quintile
- Rebalance monthly
The return of this portfolio is the "momentum factor premium." Factor premia can be earned without a directional bet on the market — they have low beta to MKT.
Factor Decay and "Factor Zoo"
A cautionary note: researchers have published hundreds of "factors" that appear to predict returns. Most don't survive out-of-sample testing, post-publication decay, or simple data-mining corrections. Harvey, Liu, and Zhu (2016) argue that due to multiple-testing bias, a t-statistic of 3.0 (not the usual 2.0) should be the bar for factor publication.
After controlling for data mining and transaction costs, robust factors likely number fewer than ten. The ones with the strongest theoretical and empirical support: market beta, value, momentum, profitability/quality, and low volatility.
Smart Beta ETFs: Factor Investing for Everyone
Factor investing was once available only to institutional investors. Today, "smart beta" or "factor ETFs" provide low-cost factor exposure:
- VLUE (iShares MSCI USA Value Factor): Value exposure
- MTUM (iShares MSCI USA Momentum Factor): Momentum exposure
- QUAL (iShares MSCI USA Quality Factor): Quality exposure
- USMV (iShares MSCI USA Min Vol): Low volatility exposure