Comment

“The Feldstein‐Horioka Fact” by Domenico Giannone and Michele Lenza deploys a new methodology to address an old question: how to interpret the Feldstein‐Horioka puzzle that national saving and investment rates appear to be highly correlated for industrial countries, especially when looking at averages of annual observations. The authors use a new methodology popular in the business cycle literature, factor‐augmented panel regression, to attempt to isolate idiosyncratic sources of fluctuations. The contribution to existing studies is that countries are allowed to react to global shocks with specific sign and magnitude. The paper shows that the homogeneity restriction is rejected by the data and biases the estimation of the saving‐retention coefficient. Indeed, allowing for a heterogeneous propagation mechanism of global shocks, the saving‐retention coefficient drops significantly from the 1980s on, consistent with the increase in capital mobility across OECD countries. The structure of the model is given by