Cryptocurrency markets differ substantially from traditional financial markets due to their decentralized structures, high levels of volatility, and speculative price movements. These characteristics necessitate questioning the applicability of classical financial theories—particularly the Efficient Market Hypothesis—to crypto-assets. The primary aim of this study is to measure the level of market efficiency in cryptocurrency markets and to investigate the existence of calendar anomalies. Within this framework, the day-of-the-week and January effects are examined in order to determine whether investor behavior in cryptocurrency markets exhibits systematic patterns.

The study analyzes daily closing price data for four major cryptocurrencies—Binance Coin (BNB), Bitcoin (BTC), Ethereum (ETH), and Ripple (XRP)—covering the period from January 1, 2018, to December 31, 2023. Price series are converted into logarithmic returns, and the stationarity properties of the series are examined using the Augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) tests. According to the results of both tests, the series are found to be stationary in levels. Regression analyses are initially conducted using the Ordinary Least Squares (OLS) method; however, due to the presence of heteroskedasticity and autocorrelation in the models, the GARCH(1,1) model—which explicitly accounts for volatility—is subsequently employed.

The empirical findings indicate that, with respect to the day-of-the-week anomaly, Thursday yields are negative and statistically significant for BNB. For BTC, Monday yields are positive and Thursday yields are negative, both of which are statistically significant. In the case of ETH, a statistically significant negative return is observed on Thursdays, while for XRP, positive returns on Wednesdays and negative returns on Thursdays are identified as statistically significant. These results suggest that cryptocurrency markets are not fully efficient and that investors display differing expectations and behavioral patterns on specific days of the week.

Regarding the January anomaly, a statistically significant and positive return is identified for ETH at the 10% significance level. No significant January effect is observed for BTC. For BNB, positive returns are found in February (10% significance) and March (5% significance), while a negative return is observed in June at the 1% significance level. In the case of XRP, negative returns are identified in May and June at the 5% significance level, while a positive return is observed in July at the 1% significance level. These findings indicate that calendar anomalies may intermittently emerge in cryptocurrency markets and that investor behavior can be influenced by seasonal effects.

Overall, the study concludes that cryptocurrency markets are not weak-form efficient, as information is not instantaneously incorporated into prices and investor psychology and market sentiment play a significant role in price formation. The findings provide valuable insights for policymakers, investors, and researchers in terms of developing investment strategies and evaluating market efficiency.