.png){kind=icon}
.png){kind=icon}
Return
Regime-Aware Adaptive Forecasting Framework for Bitcoin Prices Using Probabilistic Generative Models
.png){kind=icon}
PRE
AI
O
DOI：10.1007/s10614-026-11338-3.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
This research presents a regime-aware hybrid forecasting framework for the Bitcoin market's nonlinear, nonstationary and regime-switching behavior. The architecture integrates econometric models, neural forecasting and meta-learning, unified under a regime-detection mechanism using probabilistic inference. Central to the approach is a Hidden Markov Model (HMM) trained on log returns, which infers latent market regimes, bull, bear and sideways, based on statistical characteristics rather than arbitrary thresholds. Each detected regime triggers a specialized forecasting model: ARIMAX for volatile bear markets, SARIMAX for cyclical sideways periods and NeuralProphet for nonlinear bullish dynamics. These models leverage historical returns (Jan. 2012-Jun. 2025) and external signals, including technical indicators (RSI, MACD, Bollinger bands) and volatility metrics. A meta-learning layer, implemented via XGBoost, dynamically selects the optimal model at each time step based on the regime. This enables real-time adaptation to evolving market conditions. Predictions are made on log returns and translated into price forecasts through exponentiation. The framework's performance is evaluated using R-2, Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE). The regime-aware model outperforms the no-regime model significantly across all metrics, especially in error reduction (MAE cut by similar to 56%) and higher explanatory power (R-2 increased from 0.82 to 0.91). Ablation results confirm the structural validity of the proposed framework, with the regime-model assignment (ARIMAX for bear, SARIMAX for sideways, NeuralProphet for bull) achieving the lowest forecasting error (MAE = 736, R-2 = 0.93) at the yearly level and outperforming alternative configurations. The inferred regimes exhibit economically meaningful persistence (average durations 14.8-22.4 days) and transition stability (diagonal probabilities 0.91-0.94). The meta-learning component shows coherent and interpretable behavior, with regime labels and recent model errors explaining nearly 70% of decision weight and regime-consistent model selection exceeding 80%. These forecasting gains translate into tangible economic benefits: in a six-month backtest, the proposed strategy delivers the highest return (19%), lowest drawdown (19%) and highest Sharpe ratio (1.01), outperforming all benchmarks.
Keywords:
Adaptive forecasting
Bitcoin market
Meta-learner
Technical indicators
Bull, bear, sideways regime
Journal
C
IF:
2.2
Papers:
255
Citations:
2.3K
Organization
Cited Papers