Empirical Tests of Exchange Rate Theory

There have been very few direct applications of experimental techniques in macroeconomics for two main reasons. First, macroeconomics is about the in­ teraction between markets, and it is not easy to design an elegant treatment that gets at the essence of how a national economy functions. Second, there is a vestigial prejudice that favors econometrics over putatively unorthodox em­ pirical approaches. Still, the Nobel Prize-winning economist Edward Prescott is alleged to have said, “Don’t regress; progress!” This paper takes that methodological exhortation seriously, and it pays homage two recent Nobel Prizes in economics. Following the pioneering work of Vernon Smith and others, I use the laboratory to study some foundations of international finance. Following in the footsteps of Fynn Kydland, Edward Prescott, and others, I eschew esoteric econometric techniques and design direct empirical tests of the three most basic elements of exchange rate theory: purchasing power parity, covered interest parity, and uncovered interest parity. My research establishes two points. First, each of these three elements of exchange rate theory fares well in the laboratory. Second, not-traded goods and non-stationary domestic prices do cause deviations from simple theoretical predictions, but a non-stationary environment has a more significant effect than does not-traded goods in explaining why simple predictions of exchange rate theories break down. Perhaps the most celebrated article in the literature on exchange rate econometrics is Meese and Rogoff [9]. Those authors set the standard for a vast literature in empirical international finance. The entire literature on