Volatility modeling of cryptocurrencies according to different investment horizons

The Cryptocurrency market has emerged as a revolutionary innovation in the financial ecosystem in recent years and become a focus of great interest because of its rapid development, structure that challenges traditional financial systems and high return potential. This market is based on blockchain technology, which provides security and transparency. With these features, it has attracted the attention of both a broad investor base and portfolio managers and researchers. This study investigates cryptocurrencies according to different investment horizons; The fractal structure, efficiency and long memory feature of four different cryptocurrencies with the highest market value (Bitcoin, Ethereum, Ripple and Binance Coin) were evaluated using Maximum Overlap Wavelet Transform (MODWT), Rescaled Range (R/S) analysis and volatility models. was researched. While daily return data between 01/01/2017 and 22/11/2023 was used for Bitcoin, Ethereum and Ripple, daily closing prices between 06/11/2017 and 21/11/2023 were used for Binance Coin. As a result of the MODWT analysis, it was observed that the returns of BTC, and ETH, According to the calculated Hurst exponent coefficients, it was determined that the returns of these cryptocurrencies in the short and medium-term investment horizons are discontinuous, deviate from the average, have negative autocorrelation, is, they exhibit non-permanent behaviour in the series. According to this finding, it was observed that Efficient Market Hypothesis would be completely rejected and they exhibited a chaotic structure. It was observed that in the returns of these cryptocurrencies and their returns in long and term investment horizons, Fractal Market Hypothesis is valid because the information shocks entering the market are eliminated at a hyperbolic rate, future returns can be predicted by using past returns, therefore EPH can be violated and it creates repeating trends. Finally, the most appropriate volatility models for the selected cryptocurrencies were determined (see Tables 19, 21, 23, 25). According to the findings, it was observed that BTC and ETH, In short, and medium-term investment horizons, it was determined that both volatility and information shocks are temporary but change. It can be inferred that the findings obtained from both Hurst coefficients and the best-fit volatility models are compatible and the results will cover the entire Cryptocurrency market. It may benefit both individual and institutional investors and portfolio managers who want to invest in the Cryptocurrency market to determine investment strategies according to different investment horizons in terms of risk levels and portfolio diversification.