On the universality of the volatility formation process: when machine learning and rough volatility agree
This work addresses volatility forecasting for financial markets, providing insights into universal mechanisms, but it is incremental as it builds on existing rough volatility models and machine learning approaches.
The authors tackled the problem of forecasting daily realized volatility across hundreds of liquid stocks, showing that a universal LSTM model consistently outperforms asset-specific parametric models, uncovering evidence of a universal volatility formation mechanism. They confirmed this universality by achieving the same performance level with a parsimonious parametric model combining rough fractional stochastic volatility and quadratic rough Heston models with fixed parameters.
We train an LSTM network based on a pooled dataset made of hundreds of liquid stocks aiming to forecast the next daily realized volatility for all stocks. Showing the consistent outperformance of this universal LSTM relative to other asset-specific parametric models, we uncover nonparametric evidences of a universal volatility formation mechanism across assets relating past market realizations, including daily returns and volatilities, to current volatilities. A parsimonious parametric forecasting device combining the rough fractional stochastic volatility and quadratic rough Heston models with fixed parameters results in the same level of performance as the universal LSTM, which confirms the universality of the volatility formation process from a parametric perspective.