Efficient Quasi-Bayesian Estimation of Affine Option Pricing Models Using Risk-Neutral Cumulants
提出一种基于拉普拉斯型估计和序贯蒙特卡洛的通用快速方法，利用风险中性累积量的闭式表达式来估计仿射期权定价模型，在模拟和真实数据上均优于基准方法。
We propose a general, accurate and fast econometric approach for the estimation of affine option pricing models. The algorithm belongs to the class of Laplace-Type Estimation (LTE) techniques and exploits Sequential Monte Carlo (SMC) methods. We employ functions of the risk-neutral cumulants given in closed form to marginalize latent states, and we address parameter estimation by designing a density tempered SMC sampler. We test our algorithm on simulated data by tackling the challenging inference problem of estimating an option pricing model which displays two stochastic volatility factors, allows for co-jumps between price and volatility, and stochastic jump intensity. Furthermore, we consider real data and estimate the model on a large panel of option prices. Numerical studies confirm the accuracy of our estimates and the superiority of the proposed approach compared to its natural benchmark.