Validity of the adaptive market hypothesis in the foreign exchange markets of newly industrialized countries and an examination of relationship between return predictability and market conditions in Türkiye

The Adaptive Markets Hypothesis is a theory that argues market efficiency can change over time. This hypothesis attributes changes in the level of efficiency to many factors, primarily market conditions. The EMH argues that markets are always efficient and, therefore, investors cannot achieve excessive returns by relying on historical data. Recent studies have shown that markets are not always efficient and that investors sometimes exhibit irrational behavior. The Adaptive Markets Hypothesis was developed in parallel with these criticisms and argues that markets are dynamic and can exhibit varying levels of efficiency over time depending on different factors. This thesis examines the validity of the Adaptive Market Hypothesis in the foreign exchange markets of Newly Industrializing Countries. Furthermore, the relationship between return predictability and market conditions is investigated specifically in the Turkish foreign exchange market. In this regard, the foreign exchange markets of the countries in question against the US Dollar, monthly closing data for the period January 2000 – December 2023, were analyzed using AutomaticPortmanteau and Wild-Bootstrap Automatic Variance Ratio tests. These tests were applied to subsamples obtained using the rolling window method to examine the change over time in the degree of predictability of market returns. Then, Johansen cointegration and Granger causality analyses were used to examine the relationship between return predictability and market conditions. According to the findings obtained as a result of the analyses, the foreign exchange markets of the Newly Industrialized Countries have an efficient structure in some periods and an inefficient structure in other periods. This situation demonstrates that the Adaptive Markets Hypothesis is valid. Furthermore, according to the analysis results examined specifically for the Turkish foreign exchange market, the results of the Johansen cointegration test show that the series move together in the long term. The results of the Granger causality test indicate that some variables are related to the degree of efficiency in the Turkish foreign exchange market, while others are not. In this context, the variables related to the activity level of the Turkish foreign exchange market are the foreign trade balance, the dollar index, Brent crude oil price, inflation, and the policy interest rate. The consumer confidence index, gold price, and BIST100 have been identified as unrelated variables. These results indicate that changes in the Turkish foreign exchange market activity are shaped primarily by fundamental macroeconomic dynamics such as price stability, external balance, and global cost factors, rather than consumer perceptions and capital market returns.