Lesson 5 of 7 · 15 min
Market anomalies I: calendar, momentum, size and value
A market anomaly is a price change that cannot be linked to relevant information; to count it must persist over long periods, and the best-known time-series and cross-sectional anomalies have mostly faded, been explained by risk or turned out to be products of data mining.
In short
- An anomaly is a price change that cannot be tied to current relevant information or to newly released information. If persistent, it is an exception to efficiency.
- Evidence must hold over reasonably long periods; otherwise the anomaly may just reflect the sample period. Data mining (data snooping) reverses proper research: it searches the data first and builds the hypothesis afterwards.
- Anomalies are grouped by research method: time-series, cross-sectional and other (e.g. event studies).
- January effect: higher returns in early January, mostly in small caps. Tax-loss selling and window dressing explain only part, and after risk adjustment it no longer produces abnormal returns. Other calendar effects have mostly vanished.
- Overreaction: past 3–5-year losers later beat past winners. Momentum: short-term winners keep winning, which contradicts the weak form, though it may be a rational response to growth shocks.
- Size effect: small caps beat large caps on a risk-adjusted basis, but this was not confirmed later. Value effect: low P/E and M/B stocks beat growth stocks, which contradicts the semi-strong form, but it disappears in the Fama–French three-factor model.
Unlock this lesson free for 7 days
Create a free account to get 7 days of full access — every lesson, video, flashcard, mock and the question bank. No card needed.