The Feldstein–Horioka puzzle of international finance refers to the concept that savings and investment are highly correlated at the country level. In the absence of transaction costs and other frictions, savings and investment shouldn’t correlate at the country level for a small country. This prediction implies that, assuming no restrictions between countries, capital should flow from lower-return regions to higher-return regions.

This idea makes sense, if you can think of yourself as an investor: You want to invest in a country with the highest return (assuming the same risk and maturity of the instrument).

The overwhelming empirical evidence indicates that savings and investments are highly correlated at the country level. But if capital flows freely between countries, a country’s investment shouldn’t correlate with its savings.

Domestic agents can get funds anywhere in the world and invest in the home country. The most important empirical evidence is obtained in studies that investigate capital flows within the OECD (Organization of Economic Cooperation and Development) countries that consist of mostly developed countries, with minimum restrictions on capital flows.

What makes this puzzle even more interesting is that it isn’t a short-term phenomenon. The data used in most empirical studies regarding this puzzle involve decade averages.

So why do domestic savings overwhelmingly finance domestic investment? As in the case of other puzzles, among the usual explanations are market imperfections. Costly information about international risk (including exchange rate risk), as well as restrictions in capital flows, may force domestic savings and investment to correlate.

Restrictions on capital flows may seem small, but their interaction with information costs may be enough to reject the predictions that savings and investment shouldn’t be highly correlated.