Cryptocurrency volatility forecasts depend more on model than loss choice
Loss Choice or Model Choice? The Role of Forecast Level in Cryptocurrency Volatility Forecasting
Artificial IntelligenceComputational Engineering, Finance, and Science
Summary
Forecasting cryptocurrency price swings helps manage financial risks, but different methods can give different results. The paper shows that the way forecast errors are measured (loss choice) changes the average predicted volatility level, which can be misleading. After adjusting forecasts to the same level, the choice of forecasting model matters more than the loss function used. This finding helps clarify how to compare and improve future volatility predictions for cryptocurrencies.
What this means in practice
- •For financial risk teams: Improve cryptocurrency risk assessments by choosing forecasting models after aligning forecast levels rather than relying solely on error metrics.
- •For quantitative trading teams: Enhance trading strategies by focusing on model selection for volatility forecasts once the average forecast level is calibrated.
Authors
Andrzej Tokajuk, Jarosław A. Chudziak
Abstract
Volatility forecasts play a central role in financial risk management because their overall level and day-to-day movements affect downstream decisions. Most studies compare forecasting models while keeping the training loss fixed. Yet losses emphasise different errors and can target different properties of future volatility, so raw comparisons may combine persistent forecast-level differences with differences in daily forecast movements. This leaves unresolved whether the importance of loss choice comes mainly from the forecast level it targets or from differences that remain after level adjustment. We address this gap through a comparison of seven losses and five models across major cryptocurrencies. Validation-based alignment adjusts the forecast level before the raw and aligned forecasts are evaluated using statistical scores and one-day Value-at-Risk. Before alignment, marginal score variation is greater across losses. After alignment, model choice becomes the larger source of variation in the full five-model comparison, while cross-loss differences in VaR breach rates narrow substantially. Our contribution is a comprehensive evaluation of loss and model choice that shows why losses can appear so influential in raw comparisons and how this interpretation changes when forecast level and downstream risk are considered explicitly.