Working Paper: NBER ID: w26184
Authors: John D. Burger; Francis E. Warnock; Veronica Cacdac Warnock
Abstract: We put forward a theory-based time-varying supply-side measure of the natural level of capital flows, KF*. Out-of-sample empirical features of KF* are impressive. For countries that have quarterly time series data of capital flows, we show that KF* is a level to which flows converge in the medium term; greatly improves our ability to model notoriously volatile capital flows; and performs well against out-of-sample and in-sample filtering techniques. The gap between actual inflows and KF* also helps predict both 6-quarters ahead sudden stop episodes and medium-term equity returns. We close with how KF* and lessons from the global financial crisis help us think about capital flows during the Covid-19 shock.
Keywords: capital flows; economic indicators; portfolio flows
JEL Codes: F3; F4
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| natural level of capital flows (kf) (F32) | actual capital flows (F32) |
| deviations of actual flows from kf (F32) | sudden stop episodes (E32) |
| deviations of actual flows from kf (F32) | medium-term equity returns (G12) |
| kf gap (the difference between actual flows and kf) (F32) | probability of sudden stop (C69) |
| natural level of capital flows (kf) (F32) | future capital flows (F32) |
| natural level of capital flows (kf) (F32) | medium-run variation in flows (E32) |