Paper: arXiv 2604.10402

Authors: Tenghan Zhong

Abstract

Volatility forecasting becomes challenging when market conditions shift and model performance varies across market states. Motivated by this instability, we develop a risk-sensitive specialist routing framework for ETF volatility forecasting. The framework uses online risk-sensitive evaluation and state-dependent gating to combine different forecasting specialists across calm and stressed market states. Using a daily panel of six ETFs under a rolling walk-forward design, we find that the strongest forecaster is regime-dependent rather than stable across all states. Relative to the rolling-best baseline, the proposed routing framework reduces high-volatility forecast loss by about 24% and underprediction loss by about 22%. These results suggest that specialist routing provides a practical forecasting architecture that adapts to changing market conditions.

Complexity vs Empirical Score

  • Math Complexity: 5.5/10
  • Empirical Rigor: 7.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper introduces a practical and novel risk-sensitive routing mechanism for combining volatility specialists across market regimes. Empirical evaluation on ETF data is rigorous, demonstrating significant improvements in high-volatility and underprediction losses over standard baselines.

Research Flowchart

  flowchart TD
    A["Research Goal:<br>Adapt ETF Volatility Forecasting to Market Regimes"] --> B["Data Inputs:<br>Daily Panel of 6 ETFs"]
    B --> C["Rolling Walk-Forward Evaluation<br>& Risk-Sensitive Scoring"]
    C --> D["State-Dependent Gating<br>& Specialist Routing Framework"]
    D --> E["Key Outcome 1:<br>Optimal Forecaster is Regime-Dependent"]
    D --> F["Key Outcome 2:<br>High-Volatility Loss Down 24%"]
    D --> G["Key Outcome 3:<br>Underprediction Loss Down 22%"]

Cited References (27)

  1. Autoregressive Conditional Heteroscedasticity with Estimates of the Variance of United Kingdom Inflation (1982) — link
  2. Modeling and Forecasting Realized Volatility (2003) — link
  3. A Simple Approximate Long-Memory Model of Realized Volatility (2009) — link
  4. Prediction of Realized Volatility and Implied Volatility Indices Using AI and Machine Learning: A Review (2024) — link
  5. Forecasting Stock Market Volatility with Regime-Switching GARCH Models (2005) — link
  6. Evaluating Forecast Performance with State Dependence (2023) — link
  7. Predicting the Volatility of the iShares China Large-Cap ETF : What Is the Role of the SSE 50 ETF ? (2019) — link
  8. A False Discovery Rate Approach to Optimal Volatility Forecasting Model Selection (2024) — link
and 19 more
  1. Forecasting Downside and Upside Realized Volatility: The Role of Asymmetric Information (2024) — link
  2. Forecasting Realised Volatility Using Regime-Switching Models (2025) — link
  3. Fractionally Integrated Generalized Autoregressive Conditional Heteroskedasticity (1996) — link
  4. A Machine Learning Approach to Volatility Forecasting (2023) — link
  5. Forecasting Realized Volatility with Machine Learning: Panel Data Perspective (2023) — link
  6. DeepVol : Volatility Forecasting from High-Frequency Data (2024) — link
  7. Forecasting Realized Volatility: Does Anything Beat Linear Models? (2024) — link
  8. Forecasting Realized Volatility: A Bayesian Model-Averaging Approach (2009) — link
  9. Online Prediction Under Model Uncertainty via Dynamic Model Averaging: Application to a Cold Rolling Mill (2010) — link
  10. Forecasting Realised Volatility: A Markov Switching Approach with Time-Varying Transition Probabilities (2019) — link
  11. Adaptive Mixtures of Local Experts (1991) — link
  12. Hierarchical Mixtures of Experts and the EM Algorithm (1994) — link
  13. Forecast Combinations: An Over 50-Year Review (2023) — link
  14. Forecasting Exchange Rate Volatility: The Superior Performance of Conditional Combinations of Time Series and Option Implied Forecasts (2012) — link
  15. On the Estimation of Security Price Volatilities from Historical Data (1980) — link
  16. Volatility Forecast Comparison Using Imperfect Volatility Proxies (2011) — link
  17. A Practical Guide to Harnessing the HAR Volatility Model (2021) — link
  18. Comparing Predictive Accuracy (1995) — link
  19. A Simple, Positive Semi-Definite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix (1987) — link

0 of 27 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.