We show that, for a class of mean-reverting models, the correlation function of stochastic variance (squared volatility) contains only one -- relaxation -- parameter. We generalize and simplify the expression for leverage for this class of models. We apply our results to specific examples of such models -- multiplicative, Heston, and combined multiplicative-Heston -- and use historic stock market data to obtain parameters of their steady-state distributions and cross-correlations between Weiner processes in the models for stock returns and stochastic variance.
↧
