“…With the notable exception of the rough Heston model [1,26,27,25,24] and its affine extensions [2,35], the absence of Markovianity of the fractional Brownian motion prevents any pricing tools other than Monte Carlo simulations; the simulation of continuous Gaussian processes, including fractional Brownian motion, is traditionally slow as soon as one steps away from the standard Brownian motion. However, the clear superiority-for estimation and calibration-of these rough volatility models has encouraged deep and fast innovations in numerical methods for pricing, in particular the now standard Hybrid scheme [10,43] as well as Donsker-type theorems [45,64], numerical approximations [6,42,36] and machine learning-based techniques [47,71].…”