<![CDATA[The high volatility and fat-tailed return distribution of cryptocurrency markets require a more adaptive risk management approach than conventional static models. This study aims to implement and evaluate the effectiveness of a bet-sizing model adapted from Game Theory using the Kelly Criterion as an adaptive risk management framework. A quantitative approach was applied to four major cryptocurrencies—Bitcoin (BTC), Ethereum (ETH), Solana (SOL), and XRP—using 2020 to 2022 as the in-sample period and 2023 to 2025 as the out-of-sample evaluation period. The out-of-sample results demonstrate that the Kelly Criterion-based bet-sizing model outperformed the traditional risk management approach, generating a total return of one hundred thirty-eight point sixty percent and a Sharpe Ratio of 1.34, compared with twelve point eleven percent and 1.22, respectively, for the conventional model. However, this superior performance was accompanied by a substantially larger Maximum Drawdown (MDD) of thirty-three point thirty-three percent, compared with four point twenty-three percent under the traditional approach. These findings indicate that an adaptive transaction-level risk allocation strategy is better able to respond to the dynamic characteristics of cryptocurrency markets. Nevertheless, the relatively limited number of trades and the restricted evaluation period make the bet-sizing model sensitive to changes in sample characteristics, particularly under more extreme market conditions. Despite these limitations, the study incorporates in-sample and out-of-sample validation, as well as transaction costs and slippage, into the performance evaluation, providing a more realistic assessment of the proposed adaptive risk management strategy.]]>
