Investor Sentiment and the Cross-Section of Stock Returns

Source: Baker, M. & Wurgler, J. (2006). Journal of Finance 61(4), 1645–1680.

TL;DR

Builds a market-wide investor-sentiment index and shows it conditions the cross-section of returns:

when sentiment is high, stocks that are hardest to value and hardest to arbitrage — small, young,

volatile, unprofitable, non-dividend-paying, distressed, or extreme-growth firms — subsequently earn

low returns; when sentiment is low, the pattern reverses. Sentiment is not noise to be averaged

away; it predictably mis-prices a recognisable set of stocks.

A conditioning variable rather than a single tradable factor: the sign and size of many

cross-sectional return spreads depend on the prevailing level of sentiment. The affected

characteristics are precisely those that make a stock speculative and costly to short. Several effects

(age, return volatility) have no unconditional spread at all — they only appear once you condition

on sentiment, and they sign-flip across high vs. low sentiment.

Sentiment index: first principal component of six proxies — the closed-end fund discount, NYSE

share turnover, the number and average first-day returns of IPOs, the equity share in new issues,

and the dividend premium — each orthogonalised to a set of macroeconomic conditions. Proxies are

measured annually, 1962–2001.

Sorting: form decile portfolios on characteristics linked to valuation difficulty / arbitrage costs

(size, age, return volatility, profitability, dividend policy, growth/distress).

Test: sort firm-months first by beginning-of-period sentiment (above/below median), then by

characteristic decile; compare subsequent monthly returns conditional on sentiment.

Monthly stock returns, 1963–2001 (accounting data matched July t to June t+1, à la Fama-French).

Conditional return spreads from Table 3 (per month):

Characteristic	After LOW sentiment	After HIGH sentiment
Size (small − large)	small earn 2.33% vs large 0.91% (Banz size effect present)	size effect vanishes
Age (old − young, dec10−dec1)	old earn 0.47% more	old earn 0.72% less (sign flip; no unconditional effect)
Profitability (profitable − unprofitable)	profitable 0.85% lower	profitable 0.32% higher (sign flip)
Dividends (payers − nonpayers)	payers 0.77% lower	payers 0.45% higher (sign flip)

Returns are higher across the board after low sentiment. The result is conditional and statistical,

not a clean always-on long-short factor; its out-of-sample strength is debated and depends on how

sentiment is measured.

Limits to arbitrage: hard-to-value, hard-to-short stocks can stay mispriced longest.

Sentiment-driven demand from noise traders pushes speculative stocks above fundamentals when

optimism is high, depressing their subsequent returns.

Index construction: proxy choice, orthogonalisation, and look-ahead in building the index.

Regime dependence and non-stationarity of the sentiment-return relation.

Tradability: the short leg concentrates in expensive-to-short small caps.

Baker, M. & Wurgler, J. (2006) — Investor Sentiment and the Cross-Section of Stock Returns — Journal of Finance

Baker, M. & Wurgler, J. (2007) — Investor Sentiment in the Stock Market — Journal of Economic Perspectives

De Long, J. B. et al. (1990) — Noise Trader Risk in Financial Markets — Journal of Political Economy

Stambaugh, R., Yu, J. & Yuan, Y. (2012) — The Short of It: Investor Sentiment and Anomalies — Journal of Financial Economics

Provenance: verified/generated from the paper's full text.