如何把控金融市场的运行规律，更好的规避场外因素对金融价格波动产生的影响

How to control the operating rules of the financial market and better avoid the impact of off-market factors on financial price fluctuations; how to reasonably invest in assets to obtain maximum economic returns has always been a hot issue of concern. The study of diffusion process volatility as the core of financial market risk measurement has become an important work for scholars from all over the world. <br><br>In the context of high-frequency data, Andersen (1997) proved that the ARCH model has a good estimate of volatility. Since then, more and more scholars have closely combined the study of volatility with high-frequency data; Barndorff-Nielsen and Shephard (2002) It is proposed that when the asset price process satisfies the diffusion model, the consistent estimator of the integral volatility is the realized volatility; <br>Kristensen (2010) links the estimated volatility with the instantaneous volatility and proves that the second power variation The asymptotic normality and uniform consistency of the kernel estimator. Tang Yong and Liu Fengtao (2005) compared the currently used three models to predict volatility and found that the realized volatility is better than the SV (Stochastic Volatility) model and the GARCH model; we are also in the context of high-frequency data, Use the realized volatility model for research. <br><br>This paper is mainly influenced by Kristensen (2016) on the study of volatility. Based on the two-step estimation method of the stochastic volatility model, the proof of the uniform consistency of the estimator is optimized, and a new convergence rate is given. At the end of this article, we test the estimator by numerical simulation: the shorter the time interval we choose, the better the fit between the true value line and the simulated line, and the smaller the error.

How to control the operation law of financial market, better avoid the influence of over-the-market factors on financial price fluctuations, how to reasonably invest in assets in order to maximize economic returns, has always been a hot issue of concern to people. It is an important work for scholars in various countries to study the volatility of diffusion process as the core of risk measurement in financial markets.<br><br>After Andersen (1997) proved that the ARCH model had a good estimate of volatility in the context of high-frequency data, a growing number of scholars combined volatility research with high-frequency data;<br>Kristensen (2010) linked realized volatility to transient volatility estimates and demonstrated the progressive normality and consistent fit of the secondary power variability nucleation estimates. Tang Yong and Liu Fengtao (2005) compared the prediction of volatility using three models and found that the realized volatility was better than the SV and GARCH models, and we also studied the realized volatility model against the background of high-frequency data.<br><br>This paper is mainly influenced by Kristensen (2016) on the study of volatility, on the basis of the two-step estimation of random volatility model, the consistent convergence of estimates is optimized, and a new convergence velocity is given. At the end of this article, we test the estimate by numerical simulation: the shorter the interval we select, the better the fitting of the true and analog lines, and the smaller the error.

如何把控金融市场的运行规律，更好的规避场外因素对金融价格波动产生的影响；怎样合理的对资产进行投资，以获得最大化的经济收益，一直以来都是人们关心的热点问题。将扩散过程波动率作为金融市场风险测量的核心进行研究，成为各国学者的重要工作。在高频数据背景下Andersen(1997)证明了ARCH模型对波动性的良好估计，此后，越来越多的学者将波动率的研究与高频数据紧密结合起来；Barndorff-Nielsen和Shephard(2002)提出，当资产价格过程满足扩散模型时，积分波动率的一致估计量就是已实现波动率；Kristensen(2010)将已实现波动率与瞬时波动率的估计联系起来，并证明了二次幂变差核估计量的渐进正态性和一致相合性。唐勇和刘峰涛(2005)将目前使用较多的三种模型对波动率的预测进行比较，发现已实现波动率优于SV(Stochastic Volatility)模型和GARCH模型；我们也是在高频数据背景下，利用已实现波动率模型进行研究。本文主要是受Kristensen(2016)对波动率研究的影响，在随机波动率模型的两步估计法的基础上，对估计量的一致相合性的证明进行优化，并给出了新的收敛速度。在本文的最后，我们通过数值模拟的的方法对该估计量进行检验：我们选取的时间间隔越短，真值线和模拟线的拟合程度越好，误差也越小。<br>