Time-Varying Market Efficiency in the Turkish Stock Market: Evidence from an Entropy-Based Analysis

This study analyzes the time-varying informational efficiency of the Borsa Istanbul composite index (XU100) and its main sector indices from January 2017 to June 2023. The sample entropy method is used to measure market efficiency across multiple timescales ranging from 1 to 30 business days, and a rolling window approach is used to capture daily informational efficiency dynamics. Our findings indicate that market efficiency decreases as the timescale increases, which suggests that indices are more efficient in the short-term periods than in the long-term periods. Among the indices, BIST Financials (XUMAL) has the highest average efficiency, while BIST Technology (XUTEK) displays the lowest. The results also reveal that efficiency levels across all indices fluctuate widely on a daily basis, and a particularly sharp decline was noticed during the onset of the COVID-19 pandemic. This suggests that a major disruption in informational efficiency occurred during this time. The Russia-Ukraine war and the 2023 February earthquake further weakened the efficiency of all indices, though their impacts were comparatively less severe than COVID-19. An entropic correlation analysis reveals strong positive correlations between XU100 and BIST Industrials, which highlights that these indices are related in terms of efficiency dynamics. This study offers new insights into the dynamic nature of market efficiency in the Turkish stock market, emphasizing that informational efficiency is strongly influenced by sector-specific characteristics and external shocks. Our empirical findings strongly support the Adaptive Market Hypothesis (AMH), confirming that market efficiency evolves dynamically in response to changing market conditions and external events.