Testing the fractal market hypothesis in international markets with the MFDFA method

The aim of this study is to examine the fractal market hypothesis with the MFDFA method for cryptocurrency, commodity, foreign exchange and stock markets. In this context, Bitcoin, Ethereum, crude oil, gold, EUR/USD, USD/JPY, S&P 500 and BIST 100 were selected to represent each market, and the multifractal structure of daily returns between 01.01.2018-19.12.2022 was analysed. The findings suggest that financial assets exhibit a multifractal structure. Hurst exponents reveal that Bitcoin, Ethereum, oil, and BIST 100 have long memory, while gold, EUR/USD, USD/JPY, and S&P 500 have short memory. The results show that time series that do not exhibit random walk are inefficient and the fractal market hypothesis is valid. The multifractality degree is low for the foreign exchange market and high for the cryptocurrency and commodity markets. Considering the stock market, S&P 500 has a low multifractality degree, while BIST 100 has the highest multifractality degree among the variables. The time series graphs of these financial assets, which exhibit a multifractal structure, show similar patterns that repeat at different scales. By using the past returns of these financial assets, it is possible to predict the future and beat the market by earning abnormal returns. In addition to benefiting investors, this study is expected to contribute to the literature, which focuses on a single market and has limited studies where the hypothesis is tested with the MFDFA method.