International finance and macroeconomics.

• Author: Rose, Andrew K.

It has been three years since the NBER's Program in International Finance and Macroeconomics was last reviewed in the NBER Reporter. During this period, many researchers have continued to tackle the traditional problems of international finance, including: 1) the large and persistent apparent deviations from uncovered interest parity; 2) insufficient global diversification of risk; and 3) the slow convergence of real exchange rates to equilibrium levels. But many researchers have been attracted by more contemporary problems. Most notably, in the wake of the dramatic events in Europe and Latin America, there has been a resurgence of interest in the analysis of speculative attacks on fixed exchange rates and open economy monetary policy. A common goal of much of this research has been to understand better the nature of international capital flows.

This report does not attempt to be comprehensive. Many researchers in the IFM program work in overlapping fields, and much of their work is covered most appropriately in other program reports. For this reason, and for the sake of brevity, this report omits four recent "hot" areas of IFM-related research: international aspects of long-run growth; political economy; regional trading blocks; and international aspects of fiscal policy.

Real Exchange Rates

One of the recent areas of resurgent research in open economy macro-economics has been the examination of real exchange rates. Much of this work is distinguished by the use of innovative datasets. The new datasets are long (in terms of time span), wide (in terms of the number of economic factors, commodities, or countries examined simultaneously), or clever (they involve the prices of Big Macs[R] or The Economist).

Perhaps three years ago a loose consensus had developed that deviations from purchasing power parity (PPP) have a half-life of around four years, as demonstrated by Froot and Rogoff.(1) Not only did they find convergence to PPP, but it seemed remarkably stable across different regimes, as exemplified by a dataset stretching back almost 700 years!(2) Parsley, Wei, Frankel and others have confirmed this using panels of data covering many countries in the postwar period, while Cumby shows even faster convergence using an imaginative panel of Big Mac[R] prices.(3) However, Engel shows that the statistical evidence for convergence to PPP is weaker than it seems, while Taylor argues that the evidence seems to depend on the era considered.(4)

Since PPP can be expected to hold only in narrow circumstances, it is not surprising that the long-run determinants of real exchange rates continue to be a subject of great interest. In a pure accounting sense, Engel shows that real American exchange rate changes are accounted for almost completely by changes in nominal exchange rates; prices (even the relative price of nontradables) account for almost none of the variance, even at low frequencies.(5) Chinn and Johnston find that government spending and productivity trends help in the analysis of real exchange rates; their finding is confirmed by Canzoneri, Cumby, and Diba; and by De Gregorio and Wolf.(6) On the other hand, Clarida and Gali find little evidence of important sup-ply-side determinants.(7)

International pricing per se remains a subject of interest to IFM researchers. Engel and Rogers show that price disparities within countries or regions are much more closely linked and likely to converge than price disparities across countries. This finding is confirmed by Parsley and Wei, while Ghosh and Wolf document the importance of price stickiness and menu costs using a dataset consisting of prices for The Economist magazine.(8)

The renewed interest in empirical analysis of real exchange rates has not been matched by a comparable interest in nominal exchange rates.(9) However, the promising examination of market microstructure in the foreign exchanges begun by Lyons, Goldberg and others continues, albeit at a somewhat slower clip.(10)

The intertemporal approach to macroeconomic fluctuations in the open economy also continues to be an area of interest to IFM researchers. Obstfeld and Rogoff are most closely identified with this area, which forms an integral part of their new book; Razin and Milesi-Ferretti have provided related empirical analysis.(11)

International Financial Markets

Perhaps the most important continuing mystery of international finance is the "forward discount premium puzzle"; countries with high interest rates tend to have appreciating rather than depreciating currencies. Thus investors who receive a high interest rate return also tend to experience capital gains from currency appreciation, a deviation from the uncovered interest parity condition. Backus, Foresi, and Telmer show just how hard it is to rationalize this in models based on the absence of arbitrage and a reasonable risk premium. On the other hand, Elliott and Ito argue that the profits apparently available from investments in high interest rate countries are small and variable.(12) Marston looks at a number of parity conditions jointly, while Lyons and I examine them during currency crises.(13) Flood and I find that interest differentials are linked more closely to exchange rate changes for fixed exchange rate regimes than flexible ones; Favero, Giavazzi, and Spaventa analyze European interest rate differentials directly.(14) This area has been surveyed separately by both Engel and Lewis.(15)

While the forward discount premium puzzle links different financial asset prices, the other great unsolved problem of international finance is "home market bias" in asset stocks. This is the fact that investors tend to hold too many domestic securities for typical portfolios to be well-diversified internationally. Baxter, King, and Jermann show that the existence of nontradables (such as human capital) makes the problem look even worse.(16) But Lewis, Taylor, Baxter and Crucini, show that even though risk appears not to be shared or smoothed adequately internationally, there are a number of reasons why there may be few important...