A new test for stock market rationality
The purpose of this dissertation is to develop a new volatility test for the hypothesis of stock market rationality. Even though Mankiw-Romer-Shapiro (1985) correct some severe criticisms of Shiller's volatility test (1981), there are still some problems unresolved by their test the reduction in power of the test due to the arbitrary variable P$\sp0($t) and the use of the inappropriate assumption of a constant rate of return on the stock market which makes their test a joint test for stock market rationality and a constant rate of return. This research proposes a new testing methodology which is immune to the problems in the MRS test. First, in order to increase the power of the test, we develop a model which does not rely on the arbitrary variable P$\sp0($t) in the MRS test. Then, using Lucas' equilibrium asset pricing model, the inappropriate assumption of a constant rate of return is relaxed. Thus, the new test is a test for stock market rationality only under the Lucas model. Monte Carlo simulation shows that the new test is more powerful than the MRS test. When we apply the existing volatility tests including Shiller's test and the MRS test to our data, they continue to reject the hypothesis of stock market rationality. However, the empirical results of the new test suggest that we cannot reject the hypothesis of stock market rationality which is consistent with the majority of empirical studies which support the hypothesis of stock market rationality.
Na, Dong-Min, "A new test for stock market rationality" (1991). Dissertations available from ProQuest. AAI9211978.