Attention-Driven Buying: Why Retail Investors Chase High-Visibility Stocks
Attention-driven buying is individual investors' documented tendency to become net buyers of stocks that have recently attracted notice. A stock appears in headlines or posts an extreme one-day gain or loss, and retail investors buy it regardless of fundamentals. In "All That Glitters," Barber and Odean (2008) examined 78,000 households at a large U.S. discount brokerage from January 1991 to November 1996. Buy-sell imbalances rose from −18.15% for the lowest-volume decile by 29.5 percentage points to the highest-volume 5% of stocks.
What the Study Found
Individual investors at the large discount brokerage had a buy-sell imbalance of −18.15% for stocks in the lowest abnormal-volume decile. For the highest-volume vingtile (top 5%), imbalances rose by 29.5 percentage points from the lowest decile. Return sorts produced a U-shaped curve: 29.4% for worst-return stocks and 24% for best-return stocks. By value of trades, the worst-return vingtile showed a 29.1% imbalance versus −8.6% in the eighth return decile. At the large retail brokerage (665,533 accounts), stocks in the news carried a 16.17% imbalance versus −1.84% for no-news stocks.
Methodology
Trading records from a large discount brokerage covered 78,000 U.S. households from January 1991 to November 1996. A large retail brokerage added 665,533 accounts (30 months ending June 1999). A small discount brokerage added 14,667 accounts (January 1996–June 1999) and 43 Plexus Group institutional money managers spanned January 1993–March 1996. Stocks were sorted daily into deciles and vingtiles by abnormal trading volume, previous-day return, and Dow Jones News Service coverage. Buy-sell imbalances equaled purchases minus sales divided by total trades, with Newey-West corrections for serial dependence.
Key Statistics
| Metric | Finding | Context |
|---|---|---|
| Buy-sell imbalance, lowest volume decile | −18.15% | Large discount brokerage, by number of trades |
| Buy-sell imbalance spread: lowest to highest volume | +29.5 percentage points (from −18.15%) | Large discount brokerage, by number of trades |
| Buy-sell imbalance, lowest volume decile (by value) | −16.28% | Large discount brokerage, by value of trades |
| Buy-sell imbalance, highest volume vingtile (by value) | +17.67% | Large discount brokerage, by value of trades |
| Buy-sell imbalance, worst return vingtile | 29.4% | Large discount brokerage, by number of trades |
| Buy-sell imbalance, eighth return decile | 1.8% | Large discount brokerage, by number of trades |
| Buy-sell imbalance, best return vingtile | 24% | Large discount brokerage, by number of trades |
| Buy-sell imbalance, stocks in the news | 9.35% vs. 2.70% | Large discount brokerage: in-news vs. no-news, by number of trades |
| Buy-sell imbalance, stocks in the news | 16.17% vs. −1.84% | Large retail brokerage: in-news vs. no-news, by number of trades |
| Mean stocks held per household | 4.3 stocks ($47,334) | Mean household, large discount brokerage |
| Abnormal volume formula | AV_it = V_it / V̄_it | V̄_it = mean daily dollar volume over prior 252 trading days (Eq. 1) |
| Buy-sell imbalance formula | BSI_pt = (ΣN_Bit − ΣN_Sit) / (ΣN_Bit + ΣN_Sit) | Across all stocks in partition p on day t (Eq. 3) |
Why This Matters
Attention narrows the buy-side choice set before preferences enter, making individual investors prone to purchasing already-noticed and potentially overpriced securities. Because every attention-driven buyer requires a willing seller, retail attention-buying systematically transfers wealth to less attention-constrained counterparties. The model predicts that noise trader losses increase with attention intensity. Retail portfolios concentrated in high-attention stocks therefore face a structural performance drag.