# Why Investors Buy 'Lottery Stocks' (And Why They Underperform)

Cite as: Gorak, R. (2026). Why Investors Buy 'Lottery Stocks' (And Why They Underperform). Tradicted. https://www.tradicted.com/research/kumar-gambling-2009/
Paper: Alok Kumar — *Who Gambles in the Stock Market?*
Published in: Journal of Finance (2009)
Original: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=686022

Key finding: Individual investors allocate an average of 3.74% of their aggregate portfolio weight to lottery-type stocks, compared with 1.25% in the aggregate market portfolio and only 0.76% in the aggregate institutional portfolio, during the 1991-1996 sample period.

---
Lottery-type stocks are low-priced, highly volatile stocks with a small chance of a very large positive return, similar to a state lottery ticket. A speculative, low-priced biotech or mining stock that rarely pays off but occasionally posts a huge one-day gain fits this profile. Kumar's (2009) paper, Who Gambles in the Stock Market?, studied 77,995 individual investors at a U.S. discount brokerage house from 1991 to 1996. Individual investors allocated an average of 3.74% of their portfolios to lottery-type stocks, compared with 1.25% in the market portfolio and 0.76% in institutional portfolios.

## What the Study Found

Individual investors allocated an average of 3.74% of their portfolio weight to lottery-type stocks, versus 1.25% in the market portfolio and 0.76% in institutional portfolios. A 65-year-old investor allocated 2.19% lower weight to lottery-type stocks than he would have at age 30, a three standard deviation change in age. Average annual risk-adjusted under-performance from lottery-type stock investment was 1.10%, rising over 2.50% for investors with at least one-third of their portfolio in lottery stocks. A one-percentage-point increase in the unemployment rate corresponded to a 1.95% increase in the excess buy-sell imbalance for lottery-type stocks. Lottery-type stocks posted a -7.10% annualized risk-adjusted performance differential relative to non-lottery-type stocks.

## Methodology

Kumar used six years of monthly brokerage portfolio holdings and trades, combined with 13F institutional holdings, CRSP, and COMPUSTAT data. The brokerage sample included 77,995 individual investors, of whom 62,387 traded common stocks. The data span January 1991 to November 1996, with an ISSM/TAQ small-trades robustness sample covering 1983 to 2000. Stock-level Fama-MacBeth and panel regressions controlled for market beta, firm size, book-to-market ratio, past return, coskewness, turnover, and dividend status.

## Key Statistics

| Metric | Finding | Context |
|---|---|---|
| Aggregate portfolio weight in lottery-type stocks | 3.74% (individual) vs. 1.25% (market) vs. 0.76% (institutional) | 1991-1996 brokerage and 13F data |
| Annual risk-adjusted under-performance from lottery investing | 1.10% average; over 2.50% for investors with ≥1/3 portfolio in lottery stocks | Investor-specific hypothetical portfolios |
| Lottery-type stock under-performance vs. non-lottery stocks | -7.10% annualized, risk-adjusted | Value-weighted portfolio comparison |
| Effect of unemployment on lottery-stock demand | 1.95% increase in EBSI per 1-percentage-point rise in unemployment | Time-series regression, 1991-1996 |
| Aggregate Investor Preference formula | EW_ipt = ((w_ipt − w_imt) / w_imt) × 100 | Excess portfolio weight in a stock relative to its market-capitalization weight |

## Why This Matters

Retail investors appear to treat certain individual stocks as substitutes for state lottery tickets rather than as claims on discounted future cash flows. Because this behavior concentrates among lower-income households, the resulting losses function similarly to the regressive cost structure documented in state lottery research. Advisors working with clients drawn to speculative, low-priced names can use this framework to identify gambling-driven trades rather than research-driven theses. Aggregate retail appetite for speculative stocks may also shift predictably with macroeconomic conditions such as the unemployment rate. Lottery-stock preference shows up as a pattern across many trades rather than in any single one. A [trading journal](/trading-journal/) makes the distribution visible.

## FAQ

### What percentage of their portfolio do individual investors allocate to lottery-type stocks?

3.74% of the aggregate individual investor portfolio was allocated to lottery-type stocks. The aggregate market portfolio held 1.25%, and the aggregate institutional portfolio held 0.76%. Kumar (2009) measured this using monthly brokerage and 13F holdings data from 1991 to 1996.

### How does the research define a lottery-type stock?

Three characteristics jointly define a lottery-type stock: low price, high idiosyncratic volatility, and high idiosyncratic skewness. Each characteristic is measured in the lowest or highest 50th percentile of CRSP stocks. Kumar (2009) computed volatility and skewness using the prior six months of daily stock returns.

### Do institutional investors also gamble on lottery-type stocks?

0.76% was the average weight institutions allocated to lottery-type stocks, versus 3.74% for individual investors. Institutions exhibited a relative aversion to lottery-type stocks and instead overweighted stocks with higher coskewness, the opposite preference shown by individual investors. Kumar (2009) identified this pattern using stock-level Fama-MacBeth and panel regressions.

### How much do gambling-motivated investments cost individual investors?

1.10% was the average annual risk-adjusted under-performance attributable to lottery-type stock holdings. Under-performance rose to over 2.50% annually for investors allocating at least one-third of their portfolio to lottery-type stocks. A typical lottery investor could have earned 2.84% more annually by holding only non-lottery stocks.

## Related Research

Coming soon

## Source

Kumar, A. (2009). Who Gambles in the Stock Market? Journal of Finance.

[Read the full paper →](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=686022)