Decomposing the persistence of international equity flows

The portfolio flows of institutional investors are widely known to be persistent. What is less well-known, however, is the source of this persistence. One possibility is the ‘informed trading hypothesis:’ that persistence arises from autocorrelated trades of individual investors who believe they have information about value and who face an imperfectly liquid market. Another possibility is that there are asynchroneities with respect to investment decisions across funds, across investments, or both. These asynchroneities could be due to wealth effects (across investments for a single fund), investor herding (across funds for a single investment), or generalized contagion (across funds and across investments). We use daily data on institutional flows into 21 developed countries by 471 funds to measure and decompose aggregate flow persistence. We find that the informed trading hypothesis explains about 75% of total persistence, and that the remaining amount is attributable entirely to cross-fund own-country persistence. While asynchroneities across funds investing in the same country are important, asynchroneities across countries, either within a given fund, or across funds, are not important. The cross-fund flow lags we identify might result from different fund investment processes, or from some funds mimicking others' decisions. We reject the hypothesis that wealth effects explain persistence.