New Econometric Methods and Their Applications to Japanese Financial Markets

本研究项目的主要目的是开发新的计量经济学方法，并将其应用于分析日本金融市场的数据。至于项目的理论方面，我们主要从标准抽样方法和贝叶斯方法两方面研究了时间序列分析中的统计问题，如单位根问题和不可逆转性问题。同时，我们还提出了一种新的连续时间随机模型中波动参数的估计方法。我们还利用日本金融市场的数据进行了一些实证研究。我们发现股票收益的分布通常与正态分布不同，并且具有像美国那样的大峰度。本文还研究了时变系统风险，并建立了日本股票市场的动态因子模型。项目成员在本项目的资助下撰写了大量论文。其中一些已经在学术期刊上发表，其余的作为工作论文提供。我们的正式报告中已经列出了我们论文的全部清单。

The main purpose of this research project was to develop new econometric methods and apply them to analyze the data in Japanese financial markets.As for the theoretical side of the project, we have examined mainly statistical problems such as the unit root problem and the noninvertibility problem in the time series analysis from both the standard sampling approach and the Bayesian approach. Also we have proposed a new estimation method of the volatility parameter in a continuous time stochastic model.We also have conducted a number of empirical studies using the data in Japanese financial markets. We found that the distributions of stock returns are often different from the normal distributions and have large kurtosis as in the United States. We also have investigated the time varying systematicrisks and developed a dynamic factor model for the Japanese stock markets.The members of the project have written a large number of papers with the financial assistance of this project. Some of them have been already published in academic journals and the rest of them are available as working papers. The complete list of our papers have been given in our formal report.