Conditions under which stochastic correlation matters more than stochastic volatility

Determine whether there exist configurations of multi-asset basket quanto call options—such as specific numbers of correlated underlying assets or other setup parameters—for which modeling inter-asset correlation as stochastic has a larger effect on Monte Carlo pricing accuracy than modeling asset-return volatility as stochastic; and, if so, ascertain and characterize these conditions, including any threshold number of underlying assets at which the relative importance reverses.

Background

The authors argue that correlation risk should grow with the number of correlated underlyings in basket quanto options, potentially increasing the value of stochastic correlation modeling. They suggest there might be a point where stochastic correlation contributes more to pricing accuracy than stochastic volatility.

They explicitly state uncertainty about whether such conditions exist and encourage further paper to characterize when stochastic correlation modeling has the dominant impact.