The Predictive Power of Bitcoin Halving: Assessing Price Implications

Abstract

This study investigates the effect of Bitcoin mining reward halvings on price fluctuations and enhances prediction models by incorporating a novel factor termed "halving impact weight". This weight quantifies the delayed influence of halving events on Bitcoin prices with an exponentially decaying model. Incorporating this factor significantly enhances forecast accuracy: Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) decrease by 12.05% and 12.49%, respectively. Our findings demonstrate the critical importance of accounting for mining reward halvings in predictive models of Bitcoin prices.