Adjusting toward long-run purchasing power parity
使用1870-2020年16国面板数据，通过系统混合均值组估计检验长期购买力平价，发现汇率和价格会向购买力平价调整，并测量了调整速度。
• We use the System Pooled Mean Group model to test relative PPP. • We use long-span data from the Jordà-Schularick-Taylor microhistory database. • This approach addresses episodic cointegration and cross-section dependence. • The pooled long-run elasticity and bootstrapped confidence interval support PPP. • We measure the speed of adjustment of exchange rates and prices toward relative PPP. • We detect consensus for adjusting toward PPP and by considering exchange rate regimes. Under purchasing power parity (PPP) exchange rates and relative prices adjust to maintain a constant real exchange rate in the long run. Its empirical validity continues to be questioned. We use data on exchange rates and prices relative to the U.S. for a long-span (1870–2020) panel of 16 countries to examine (a) whether the long-run elasticity is one; (b) whether there is adjustment by exchange rates of prices to maintain a constant real exchange rate and (c) the time taken to adjust. We use four estimators, which increasingly restrict the model. These are country-specific vector error correction model in exchange rates and relative prices; the Johansen estimator, which has the cross-equation restriction that the long-run coefficient in the two equations is the same; the system pooled mean group estimator, which has a homogeneous long-run coefficient over countries and heterogeneous short-run dynamics, and a univariate real exchange rate equation used to obtain median unbiased estimates of the half-life.