Feldstein–Horioka puzzles

Abstract

The high correlation between national saving and investment rates in advanced economies—the Feldstein–Horioka puzzle—has been referred to as the “mother of all puzzles.” Perhaps more puzzling is that for emerging economies saving–investment correlations tend to be significantly lower, though still positive. This deepens the Feldstein–Horioka puzzle because the mobility of capital is generally believed to be much lower in emerging economies than in advanced economies, and a country with less mobile capital should have a tighter relationship between local saving and investment rates. This paper develops a DSGE model that, without resorting to any real or financial friction, simultaneously explains these two aspects of the Feldstein–Horioka puzzle: positive saving–investment correlations in both advanced and emerging economies and significantly lower saving–investment correlations in emerging economies than in advanced economies. The main features of the model include long-run risk, an endogenous world interest rate, and cross-correlations of national and global shocks. The findings hold for both quarterly time series and long-run averages.