Rough Volatility European Call Calculator

Posted by Gary Pai

This calculator follows the Monte Carlo pseudo-code in rough_vol_with_pseudo_code.pdf. It prices a vanilla European call under the rough fractional stochastic volatility (RFSV) model by generating fractional Brownian motion paths, evolving log-volatility, and simulating the correlated stock path under the risk-neutral measure.

Model inputs

Initial asset price S0
Initial volatility \( \sigma_0 \)
Hurst parameter H0 < H < 0.5
Volatility of volatility nu
Mean reversion speed \( \alpha \)
Long-run log-volatility level \( m \)\( m = \log(\text{long-run volatility}) \)
Leverage correlation \( \rho \)\( -1 < \rho < 1 \)
Risk-free rate rannualized
Strike K
Maturity Tyears
Time steps2 to 128
Monte Carlo paths100 to 50000
Outputs
European call price
Monte Carlo standard error
95% confidence interval
Runtimems

Model equations

\( dY_t = \left(r - \frac{1}{2}\sigma_t^2\right)dt + \sigma_t\,dW_t^S \)

\( dX_t = \nu\,dW_t^H - \alpha(X_t - m)dt,\ \text{with}\ X_t = \log \sigma_t \)

\( d\langle W^S, W^H\rangle_t = \rho\,dt \)

The implementation mirrors the PDF structure: build the fractional Brownian covariance matrix, apply Cholesky decomposition, simulate rough-volatility paths, and discount the average terminal payoff from the pseudo-code simulation.

Leverage note: the PDF pseudo-code approximates the leverage effect by correlating the stock increment with the Gaussian innovations used to build the rough path. This page follows that same discrete-time approximation.

Notes: the page uses browser-side Monte Carlo simulation and the covariance-matrix construction is cubic in the number of time steps. Smaller values of N_steps and N_paths will return results faster. The long-run level \( m \) is the mean of log-volatility, not of volatility itself.

Tagged: Rough Volatility, RFSV, Monte Carlo

 •  First Version: Sep 21, 2026  • 

 •  Last Edited: Sep 21, 2026  • 

Copyright 2012-2026