Jeffrey Wurgler of New York University and Ekaterina Zhuravskaya of the Centre for Economic and Financial Research (Moscow) and Centre for Economic Policy Research examine whether arbitrage flattens demand curves for stocks. In textbook theory, arbitrage between perfect substitutes keeps demand curves flat. However, in reality, individual stocks do not have perfect substitutes. The authors develop a model showing that risk inherent in arbitrage between imperfect substitutes deters risk-averse arbitrageurs from flattening demand curves. Stocks without close substitutes experience higher price jumps upon inclusion into the S&P 500 Index. The results suggest that arbitrage is weaker and mispricing is more frequent and severe among such stocks. The authors test predictions of their model using the cross-section of price responses of stocks added to the S&P 500 Index. They find that high-arbitrage-risk stocks experience higher price jumps, controlling for demand shock size. Stocks hit by large index fund demand shocks also experience higher price jumps, controlling for arbitrage risk. The model predicts that the price response to an excess demand shock increases with the shock's size and the stock's arbitrage risk. The authors also find that the demand curves for stocks are not perfectly flat, as arbitrage risk prevents them from being so. The results suggest that risk prevents arbitrage from completely flattening demand curves for stocks. The study also shows that arbitrage risk can be added to a growing list of forces that inhibit market efficiency, including informational and transaction costs, heterogeneous beliefs about fundamental value, noise trader sentiment risk, short-sales constraints, and agency costs in delegated fund management. The study concludes that the results are consistent with the hypothesis that risk prevents arbitrage from completely flattening demand curves for stocks. The results suggest that typical between-stock arbitrage trades are too risky to be worth a couple of percentage points, which is what arbitrageurs stand to gain from reversion of the average S&P effect in their sample. However, the arbitrage argument does explain some of the cross-section: stocks that have closer substitutes have flatter demand curves. The study also finds that the addition effect is at least partially permanent, while other studies suggest it may completely reverse within a short period. The results indicate that arbitrage fails to flatten demand curves for stocks and that this happens in part because perfect substitutes are not available.Jeffrey Wurgler of New York University and Ekaterina Zhuravskaya of the Centre for Economic and Financial Research (Moscow) and Centre for Economic Policy Research examine whether arbitrage flattens demand curves for stocks. In textbook theory, arbitrage between perfect substitutes keeps demand curves flat. However, in reality, individual stocks do not have perfect substitutes. The authors develop a model showing that risk inherent in arbitrage between imperfect substitutes deters risk-averse arbitrageurs from flattening demand curves. Stocks without close substitutes experience higher price jumps upon inclusion into the S&P 500 Index. The results suggest that arbitrage is weaker and mispricing is more frequent and severe among such stocks. The authors test predictions of their model using the cross-section of price responses of stocks added to the S&P 500 Index. They find that high-arbitrage-risk stocks experience higher price jumps, controlling for demand shock size. Stocks hit by large index fund demand shocks also experience higher price jumps, controlling for arbitrage risk. The model predicts that the price response to an excess demand shock increases with the shock's size and the stock's arbitrage risk. The authors also find that the demand curves for stocks are not perfectly flat, as arbitrage risk prevents them from being so. The results suggest that risk prevents arbitrage from completely flattening demand curves for stocks. The study also shows that arbitrage risk can be added to a growing list of forces that inhibit market efficiency, including informational and transaction costs, heterogeneous beliefs about fundamental value, noise trader sentiment risk, short-sales constraints, and agency costs in delegated fund management. The study concludes that the results are consistent with the hypothesis that risk prevents arbitrage from completely flattening demand curves for stocks. The results suggest that typical between-stock arbitrage trades are too risky to be worth a couple of percentage points, which is what arbitrageurs stand to gain from reversion of the average S&P effect in their sample. However, the arbitrage argument does explain some of the cross-section: stocks that have closer substitutes have flatter demand curves. The study also finds that the addition effect is at least partially permanent, while other studies suggest it may completely reverse within a short period. The results indicate that arbitrage fails to flatten demand curves for stocks and that this happens in part because perfect substitutes are not available.