Testing the Adaptive Market Hypothesis on the Swedish Stock Market - Empirical evidence between 1990-2019

Abstract: This study examines if the adaptive market hypothesis holds for the Swedish stock market between 1990 and 2019. We use Affärsvärldens Generalindex and test for time-varying return predictability by implementing a variance ratio test and an autocorrelation test. To track how market efficiency evolves over time we use a two-year moving subsample. We also use a logistic regression to determine which market conditions are related to an efficient or inefficient market. Finally, we implement three simple trading strategies to test the practical implications of the theory. We find that the market change between periods of efficiency and inefficiency, which supports the adaptive market hypothesis. From the regression analysis, we find a significant connection between different market conditions and inefficiency. From the implementation of our trading strategies, we find that one out of three is able to generate excess returns compared to the market return. In particular, our findings suggest that investors should not invest in the market when it is deemed inefficient, but only in an efficient state.