Are stock prices in the US nonstationary? Evidence from contemporary unit root tests

Abstract

This article extends the empirical literature on the efficiency of stock markets in the US by applying a battery of unit root tests to empirically ascertain whether stock prices are mean reverting. This article, unlike previous studies, employs a disaggregated approach using the daily closing values of the Dow Jones industrial average, NASDAQ composite and S&P 500 index covering the period 5 February 1971 to 31 December 2009 to investigate the integration properties of the US stock market. The empirical findings reveal that the three major stock price series are nonstationary, indicating that they do not follow a trend path. The primary implication is that trading strategies that simply rely on mean reversion of stock prices are valueless.

Keywords:

• random walk
• efficient market
• unit root

JEL Classification::

• G14
• G12