The authors test the performance of a norm-constrained institutional portfolio when the constraint is lifted in the presence of reliable information about abnormal returns from investing in constrained stocks (e.g., “sin” stocks, lottery stocks) versus the performance of unconstrained institutional portfolios. The results corroborate abnormal returns to institutions that abandon the norm. They also show that institutional investors seem to have better access to information and send trade signals to the market.

Social norms significantly influence institutional holdings. There is a social norm against funding operations that promote human vice, but institutional investors pay a financial cost when abstaining from these stocks. Sin stocks are from publicly traded companies involved in alcohol, tobacco, or gambling. Sin stocks are held less by norm-constrained institutions, such as pension plans, than by mutual funds or hedge funds, which are natural arbitrageurs. These stocks also receive less coverage from analysts than stocks with otherwise comparable characteristics. Sin stocks have higher expected returns than other comparable stocks, which is consistent with their being neglected by norm-constrained investors and having to face greater litigation risk heightened by social norms. Another example is value stocks, whose source of value comes from being neglected because of social norms.

The authors examine the performance of norm-constrained institutional portfolios at times when they deviate from norms to reap abnormal returns. They find that these institutions deviate from the norm only when they have compelling information about a sin stock to act on. And by acting on this superior information, these norm-constrained institutions earn returns that are higher than those earned by norm-tolerant institutions. The research is useful for portfolio managers because it provides insight on an institution’s ability to earn superior returns versus a benchmark portfolio because of access to strong information.

The dataset includes quarterly stock holdings for institutions that filed Form 13F with the SEC from 1980 to 2008. The authors identify lottery-type stocks as those with high positive skewness and sin stocks as those in the industry groups of alcohol, tobacco, and gambling. The average percentage holding in sin stocks and lottery stocks is determined for all institutions and portfolio data are then divided into five parts. The quintile with the highest percentage allocation is norm tolerant and the one with the lowest is norm constrained. The ratio of Catholic adherents to Protestant adherents of the country where the institution is located is also used to estimate lottery or sin stock preferences. The value-weighted return on institutions’ holdings of stocks in each quintile is then estimated.

The authors tabulate key institutional characteristics for all five quintiles for both lottery and sin stocks in a cross section. This dataset is then compared with a fresh stock holdings dataset. A t-statistic is assigned to each value to assess its significance. To determine concrete implications, the dataset is further regrouped based on level of concentration, portfolio size, and different subperiods.

Research has been done on informed trading to identify reasons for portfolios being overweighted in a particular group of stocks versus benchmark portfolios. Such studies confirm that portfolio managers deviate from standards only when they have superior information. The authors try to confirm informed trades based on investors’ social preferences (i.e., aversion to lottery or sin stocks) by focusing on institutional portfolios’ stock positions. Further research in this area could apply the same logical analysis to portfolios with a preference for stocks that adhere to tenets of corporate social responsibility.