Our data, improve on this, but also are not perfect. A US mutual fund will show up as the investor in the securities ultimately purchased. If the securities happened to be, say, Thai stocks, then the data will record a US inflow into Thailand. But clearly, if the mutual fund is a “Thai equity” fund, and if the purchase came from deposit made by a Thai resident into that fund, then our data would miss the round-trip nature of the cross-border flow. One can regain some meaning by emphasizing that the investment decision is guided by the investment manager rather than the ultimate beneficiary (if there is distinction between the two). But clearly, any data on cross-border flows will have flaws.13
It is also worth noting the consequences of the fact that we observe settlement date rather than trade date. Since the price for trades is set on trade date, we approximate trade date using the settlement date and the settlement conventions of each country.14 Occasionally trades may settle in more or less time than normal. This adds measurement error to our dating of trades. We try to allow for this by using longer intervals (every-other-day, weekly, etc.) in addition to daily intervals. Our interpretation of the results is that trade-date measurement error cannot explain our findings.
To scale the flows, denoted by Fjt, we divide by local market capitalization15, Mu, so scaled flows are denoted by fit =Fit /Mn. While we observe separate variables for purchases of local equity, sales of local equity, purchases of local-debt, and sales of local debt, we focus primarily on net equity transactions, i.e., purchases less sales.
We chose broad, well-known equity indices from each country. For instance, many of the indices used to calculate equity returns are the same ones listed in the “Financial Indicators” and “Emerging Market Indicators” sections of the Economist magazine. A complete list of the equity indices used is given in Appendix 1.