Media Coverage and the Cross-section of Stock Returns
Source: Fang, L. H. & Peress, J. (2009). Journal of Finance 64(5), 2023–2052.
TL;DR
Stocks with no media coverage earn higher returns than otherwise-similar stocks with high
coverage. A no-coverage-minus-high-coverage portfolio outperforms by about 3% per year raw, and
the risk-adjusted "no-media premium" ranges from 8% to 12% per year depending on the model (Fama-
French three factors + momentum + Pastor-Stambaugh liquidity). The effect fits Merton's (1987)
investor-recognition hypothesis: neglected stocks need a higher expected return to attract under-
diversified, attention-constrained investors. Sample: Jan 1, 1993 – Dec 31, 2002.
What anomaly it documents
Predictor: media coverage — the number of newspaper articles about a firm (low/zero coverage
predicts high returns).
Direction: negative — no/low coverage → higher future returns ("no-media premium").
Shape: concentrated in no-coverage stocks; the premium is remarkably stable for at least 12
months after formation and is largest among small, high-individual-ownership, low-analyst,
high-volatility stocks.
OSAP predictor: media coverage / firm visibility (recognition channel).
How to construct it
Universe: all NYSE firms plus 500 randomly selected NASDAQ firms, 1993–2002.
Coverage measure: count LexisNexis articles in four major U.S. national dailies — **New York Times,
USA Today, Wall Street Journal, Washington Post** — keeping articles with a LexisNexis relevance
score ≥ 90%; aggregate to a monthly count per firm.
Sort stocks into coverage portfolios (including a no-coverage group); go long no-coverage, short
high-coverage; measure risk-adjusted return differences controlling for size, BM, momentum, and
liquidity.
Evidence and replication (IS/OOS)
In-sample (1993–2002): no-coverage portfolio beats high-coverage by ~3%/yr raw following
formation; 8%–12%/yr after risk adjustment. Premium stable for ≥12 months.
Strongest among small, low-analyst-coverage, high-individual-ownership, high-idiosyncratic-volatility
stocks — where impediments to trade and recognition frictions are largest.
The authors weigh two interpretations: an impediments-to-trade / arbitrage (mispricing) story and
the Merton recognition story; the cross-sectional pattern supports recognition with trade frictions
sustaining the premium.
Why it might work
Merton (1987) recognition: investors only hold stocks they know about; low-visibility stocks are
held by fewer, less-diversified investors and must offer higher expected returns.
Impediments to trade: limits to arbitrage let the premium persist in hard-to-trade names.
Distinct from the short-horizon attention-buying pressure on high-coverage stocks (Barber-Odean
2008, "All That Glitters").
Limitations and risks
Premium concentrates in small, illiquid, hard-to-arbitrage stocks; trading costs limit capture.
Coverage measured only via four print dailies over 1993–2002; the media landscape (online/social)
has changed enormously since, raising external-validity concerns.
Recognition vs mispricing interpretation is not fully resolved.
Key references
Fang, L. & Peress, J. (2009) — Media Coverage and the Cross-section of Stock Returns — Journal of Finance
Merton, R. (1987) — A Simple Model of Capital Market Equilibrium with Incomplete Information — Journal of Finance
Barber, B. & Odean, T. (2008) — All That Glitters — Review of Financial Studies
Provenance: verified/generated from the paper's full text.
