Bitcoin Price Forecasting with Global Liquidity using TimeXer
Research Paper#Financial Forecasting, Time Series Analysis, Deep Learning🔬 Research|Analyzed: Jan 3, 2026 16:33•
Published: Dec 26, 2025 15:36
•1 min read
•ArXivAnalysis
This paper addresses the challenge of Bitcoin price volatility by incorporating global liquidity as an exogenous variable in a TimeXer model. The integration of macroeconomic factors, specifically aggregated M2 liquidity, is a novel approach that significantly improves long-horizon forecasting accuracy compared to traditional models and univariate TimeXer. The 89% improvement in MSE at a 70-day horizon is a strong indicator of the model's effectiveness.
Key Takeaways
- •Bitcoin price forecasting benefits from incorporating global liquidity data.
- •The TimeXer architecture, when conditioned on global liquidity, outperforms other models.
- •Long-horizon forecasts are significantly improved by considering macroeconomic factors.
Reference / Citation
View Original"At a 70-day forecast horizon, the proposed TimeXer-Exog model achieves a mean squared error (MSE) 1.08e8, outperforming the univariate TimeXer baseline by over 89 percent."