During this period, the PPP tests are mainly using the ordinary least squares method to simply estimate parameters of PPP. The general result was significantly rejecting PPP hypothesis except some countries with hyperinflation. Frenkel (1981) summarized the results in an article entitled “The collapse of purchasing power parities during the 1970s “. It is quite well-known that the Froot and Rogoff (1995) so-called “the first phase test” does not take into account the non-stationary of the time series of relative prices and exchange rates, and result in spurious regressions. If the nominal exchange rate and relative prices are non-stationary variables and when there is no cointegration relationship, then this equation is a spurious regression. Therefore using the traditional ordinary least squares method to do the statistical inference is invalid. Even error term is stationary; there is a strong long-term linear relationship between exchange rates and relative prices. If there is an error in estimated standard deviation, the traditional statistical estimation is still invalid. Therefore, when building a model testing PPP, we need to explicitly point out the stationarity of each variable. If the real exchange rate itself is stationary, then it is able to effectively support the long-term PPP a hypothesis, if the subordinate unit root process, then that will not converge in the long-term equilibrium level.
在这期间，PPP测试主要采用普通最小二乘法估计参数的简单的PPP。总的结果是显着抑制PPP假设除了一些国家恶性通货膨胀。弗兰克尔（1981）总结的结果，在一篇题为“崩溃的购买力平价在70年代”。这是相当众所周知，Froot和罗戈夫（1995）所谓的“第一阶段测试”没有考虑到相对价格和汇率的时间序列的非平稳，并导致伪回归。如果名义汇率和相对价格都是不固定的变量，当不存在协整关系，那么这是一个虚假的回归方程。因此，采用传统的普通最小二乘法做统计推断是无效的。即使误差项是固定的；强烈的长期的线性关系，汇率和相对价格有。如果有一个错误的估计的标准偏差，传统的统计估计仍然是无效的。因此，在建设一个模型试验的PPP，我们需要明确指出各变量的平稳性。如果实际汇率本身是静止的，那么它是能够有效地支持长期的PPP假说，如果下属单位根过程，然后，将不会在长期均衡水平收敛。
