This paper evaluates the performance of a pairs trading strategy using daily data from 1962 to 1997. The strategy involves identifying pairs of stocks with historically similar price movements and taking long-short positions when the spread between them widens. The authors test several trading rules and find that top pairs generate average annualized excess returns of up to 12%. These profits may be partly due to market microstructure effects, but historical trading profits exceed conservative estimates of transaction costs. The authors use bootstrapping to distinguish pairs trading from pure mean-reversion strategies, finding that the pairs effect differs from previously documented mean reversion profits. The study examines the risk and return characteristics of pairs trading using daily data from 1962 to 1997. A simple algorithm is used to select pairs based on historical price similarity. The authors find that top pairs generate average annualized excess returns of about 12%. They argue that part of this may be due to institutional factors, such as asynchronous trading and bid-ask bounce. After accounting for these factors, pairs trading still yields positive and significant returns. The study explores the statistical characteristics of pairs trading portfolios and helps understand one mechanism of market efficiency - relative pricing of close substitutes. The authors find that pairs trading is a profitable strategy, but its profitability is not sufficient to support large amounts of investor capital. The study also finds that pairs trading is not limited to a particular sector, as it is profitable in every broad sector category. The authors also examine the effects of transaction costs on pairs trading profits. They find that the average excess return of unrestricted pairs strategies falls from 5.98% to 3.99% per month when trades are postponed to the day following the crossing. This suggests that transaction costs are a significant factor in pairs trading profits. The authors estimate that the average transaction cost per pair is 83 basis points, which is consistent with other estimates of transaction costs. The study finds that pairs trading is a profitable strategy, but its profitability is not sufficient to support large amounts of investor capital. The authors conclude that pairs trading is an important mechanism of relative price equilibration by market participants. However, competition in the industry and the price impact of large trades may be important factors limiting the scale of pairs trading.This paper evaluates the performance of a pairs trading strategy using daily data from 1962 to 1997. The strategy involves identifying pairs of stocks with historically similar price movements and taking long-short positions when the spread between them widens. The authors test several trading rules and find that top pairs generate average annualized excess returns of up to 12%. These profits may be partly due to market microstructure effects, but historical trading profits exceed conservative estimates of transaction costs. The authors use bootstrapping to distinguish pairs trading from pure mean-reversion strategies, finding that the pairs effect differs from previously documented mean reversion profits. The study examines the risk and return characteristics of pairs trading using daily data from 1962 to 1997. A simple algorithm is used to select pairs based on historical price similarity. The authors find that top pairs generate average annualized excess returns of about 12%. They argue that part of this may be due to institutional factors, such as asynchronous trading and bid-ask bounce. After accounting for these factors, pairs trading still yields positive and significant returns. The study explores the statistical characteristics of pairs trading portfolios and helps understand one mechanism of market efficiency - relative pricing of close substitutes. The authors find that pairs trading is a profitable strategy, but its profitability is not sufficient to support large amounts of investor capital. The study also finds that pairs trading is not limited to a particular sector, as it is profitable in every broad sector category. The authors also examine the effects of transaction costs on pairs trading profits. They find that the average excess return of unrestricted pairs strategies falls from 5.98% to 3.99% per month when trades are postponed to the day following the crossing. This suggests that transaction costs are a significant factor in pairs trading profits. The authors estimate that the average transaction cost per pair is 83 basis points, which is consistent with other estimates of transaction costs. The study finds that pairs trading is a profitable strategy, but its profitability is not sufficient to support large amounts of investor capital. The authors conclude that pairs trading is an important mechanism of relative price equilibration by market participants. However, competition in the industry and the price impact of large trades may be important factors limiting the scale of pairs trading.