Bayesian Methods In Private Funds Performance Evaluation

In recent years, with the rapid development of China's securities investment funds industry, how to correctly analyze and evaluate fund performance and how to establish funds performance evaluation methods and system applying to China's securities market is of great theoretical and practical significance for investors'making the right decisions and the healthy and steady development of the fund industry. But existing methods are mainly advanced for performance evaluation of public funds and new methods have less been put forward. These existing methods are generally based on traditional methods of risk adjusted using historical datas. Although traditional methods conducive to the objectivity of the evaluation, they did not attach importance of investor's intuition and experience about the fund performance situation to his actual decision when he evaluate fund performance. On the other hand, as a rising power in the securities market, private fund is still in its infancy in terms of performance evaluation. It's development is lagging behind and there is a lot to be desired. For these reasons, this article which bases on the traditional method and fund performance evaluation system takes private securities investment funds performance evaluation as the core of research, combined with the present situation of China's private fund and performance evaluation. This article studied Bayesian statistical inference methods in the application of private equity fund performance evaluation. Different from the traditional method, this method proceeds from an investor perspective and highlights investor'influence of prior information on fund performance evaluation.This article is about private fund performance evaluation in accordance with the logic thought as below: reviewing of fund performance evaluation theory--Bayesian method--establishing Bayesian performance evaluation model--empirical analysis. First, we draw prior information from intuition problem and thus establish a set of prior information about managerial skills. Then the model combine a prior information and Jensen measure, which is derived intercept of model- the posterior expection alpha. Finally, this article will use Bayesian performance evaluation index in empirical analysis of 87 Chinese private funds that are setted up before July 1, 2008 (sample period is from July 1, 2008 to June 30, 2011) and comprehensively sum up the results of the empirical analysis. Empirical studies have shown that using Bayesian methods ,we can not only see how prior information affect the final decision of the investors, but find a better solution to the problem that classical evaluation results largely influenced by the factors of data accuracy and significance level.This article is made up of five parts as follows: Chapter one: In-depth analysis of background research and status at home and abroad, give topic selection background and significance. Chapter two: Begins with a brief discussion on development history and current status of private fund, and then review the principle of fund performance evaluation and summarize the existing performance evaluation methods. Chapter three: A brief introduction to Bayesian statistical inference and it's essence, and illustrate the difference between Bayesian theory and classical performance evaluation method. And then we detail Bayesian performance evaluation methods, including models and mathematical deductions. Chapter four: make an empirical evaluation and analysis on private investment fund in China's securities market with Bayesian methods and draw empirical findings and conclusions. Chapter five: summarize the thesis and point out the shortages.Some research methods are adopted in this article, such as theoretical and empirical methods, combination of qualitative and quantitative analysis methods and methods of comparative analysis. Qualitative analysis methods include discussion on prior information and the economic environment of performance evaluation. Quantitative analysis methods contains establishing of the likelihood rate and the prior distribution, leading to prior distribution, establishing of posterior distribution model , and run a one-factor regression analysis and significant test based on setting excess benchmark earnings ratio as variable and excess fund return ratio as function. Comparative analysis is used when we illustrate the difference between the result of Bayesian methods and that of traditional performance evaluation methods to explain the relative superiority of Bayesian methods over the traditional theory. The innovations of this article mainly contain several following components: (1) In recent years there are few domestic private fund performance evaluation methods, this paper is an attempt of this respect; (2) The Bayesian model is used to evaluate private fund performance for the first time and it makes full use of the role of the prior information. (3) The prior distribution model in the original Bayesian methods of fund evaluation model is improved to make it more truly reflect the reality.Facing the rapid development of China's private fund market and actual needs of market participants on correctly and objectively evaluating the performance level of private fund, the research of this article not only shows the difference on private fund performance evaluation between Bayesian approach and classical methods and how the investors'prior information will affect their investment decisions, it also provides a new perspective and method for private securities investment fund performance evaluation. So it has very important theoretical value and it can enrich the current private fund performance evaluation theory so that China's private fund participants will be inspired and it will conduce to guide more social resources to investment fund managers who have matured and rational concept .The empirical conclusion has positive and practical significance to guide investors'investment decisions as well.