Varieties of Capital Flows: What Do We Know?

CID Working Paper No. 296

Eduardo Levy Yeyati and Jimena Zuniga
May 2015


Capital flows have been the subject of key policy concern since the Brady plan launched the emerging markets asset class. Their massive volume, coupled with their volatile and procyclical nature, is often associated with a variety of financial and real risks: excess exchange rate volatility (gradual overvaluation and sharp corrections), dollar liquidity crunches, distressed asset sales, and crisis propensity.

These risks have changed over time. Emerging market crises in the 1990s and 2000s were inherently driven by financial dollarization and balance sheet effects, the latter were intimately related with capital inflows in the form of growing foreign liability positions. But, now that financial dollarization has receded in the emerging market word (either through debt deleveraging or international reserve accumulation), the focus shifted to the macroeconomic effects of cross market flows, including extended periods of exchange rate misalignment and the amplification of business cycles in a context of large and persistent terms-of-trade shocks and global liquidity swings. Hence, the difficulty of evaluating capital flows based on data mostly from the 1990s and early 2000s. Hence, also, the emphasis on the recent empirical literature that revisits the issue with fresh data and an open mind.

Capital flows cannot be addressed indistinctly or in isolation. Increasingly, academics and practitioners have flagged that different types of capital flows display different behaviors. Conventional wisdom tends to assume that, within portfolio flows, fixed income assets (bonds) are more harmful than equity in that they may introduce currency imbalances that may create deleterious balance sheet effects in the event of sharp exchange rate depreciation. By the same token, it is usually assumed that portfolio flows (including equity securities) are more volatile than foreign direct investment (FDI), because the latter is "sunk" in illiquid instruments that, precisely because of their illiquidity, are not prone to react to speculative motives or short-lived financial distress.

However, even this simple order of riskiness deserves some reassessment. Within debt liabilities, a distinction needs to be made between foreign and local currency denominated instruments, at a time when foreign-currency instruments still dominate local-currency ones as emerging market investments; duration is another critical aspect to consider. Is equity "safer" than a long domestic currency bond from a macro prudential perspective?