Capital flows at risk: Taming the ebbs and flows

We propose a new quantile regression framework to predict the entire future probability distribution of capital flows to emerging markets, based on changes in global financial conditions, domestic structural characteristics, and policies. The approach allows us to differentiate between the impact on the median versus the tails of the future predicted density of flows, and between short- and medium-term effects. We find that FX- and macroprudential interventions are effective in mitigating downside risks to portfolio flows stemming from adverse global shocks, while tightening of capital controls in response appears to be counterproductive. Good institutional frameworks are not able to shield countries from the increased volatility of portfolio flows in the immediate aftermath of global shocks, but can contribute to a more rapid bounce-back of foreign flows over the medium term. Our results highlight the limitations of the standard approach, which focuses only on the short-term behavior of average flows. •A new approach allows to predict the entire future probability distribution of capital flows to emerging markets.•The method allows to identify tail risks and to distinguish between short- and medium-term effects.•FX- and macroprudential interventions are effective in reducing outflow risks following adverse global shocks, but capital controls are not.•Good institutional frameworks do not to help in the short term but do so in the medium term.