摘要：We develop higher order accurate estimators of integrated volatility in a stochastic volatility models by using kernel smoothing method and using different weights to kernels. The weights have some relationship to moment problem. As the bandwidth of the kernel vanishes, an estimator of the instantaneous stochastic volatility is obtained. We also develop some new estimators based on smoothing splines.
关键词：Stochastic Volatility; Kernel Estimator; Realized Volatility; Moment Problem; Rate of Convergence; Higher Order Asymptotics; Smoothing Spline
