A Research On The Window Dressing Of Open-End Funds In China

Early in the 19th century, the world's first closed-end fund has been produced in the UK. In the early twentieth century, it flourished in Europe and America. In 1924, the United States produced the world's first corporate type open-end fund. The securities investment fund market in China started late compared to Britain and other developed countries. Until the 90s of last century, the Southern Fund Management Company established the first closed-end fund in China. Since China's first open-end fund-Hua'an Innovation Fund was set up in September 2001, China's open-end Fund began to develop rapidly. Not only the types of fund products increased quickly, but also more innovations emerged in fund market. Open-end fund quickly became one important part of China's securities market and attracted more scholars'attention. Since the fund's own mode constraints, securities investment fund industry inevitably faces principal-agent problem. Investors invest in funds and entrust their money to fund managers, but without real-time monitoring, asymmetric information leads to managers'sufficient motive to sacrifice the interests of investors in order to satisfy personal gains. In the current investment environment, investors increasingly value fund performance, so both fund managers and the fund company's entire future and fund performance are closely related. As a result, fund managers tend to do whatever it takes to maintain a good trading record. A common way to deceive investors before the end of the quarter is to buy good performance stocks and sell poor performance shares to disclose the higher (lower) than the real proportion of good performance (poor performance) ratio of the stock positions, and ultimately achieve the cosmetic results. This phenomenon is known as the Window Dressing.Window dressing was firstly proposed when the foreign scholars tried to explain the "January effect" of the stock market, which means there is often a positive return exception in the stock market in January. Some scholars believed that the January effect was produced in the United States' Tax-Losing Effect, and some scholars proposed the January effect was related to window dressing in the fund market in US. Other scholars claimed January effect produced because that many festivals gathered at the end of year, and companies would give employees year-end award. The foreign literature has tested on the window dressing's existence covering different investment groups, different fund types, different models, different frequency data, and in different markets. But the motivations of the fund managers' window dressing were rarely studied. Compared to foreign research, domestic literatures pay less attention to window dressing. Limited literature mentions the window dressing, mainly as an explanation of calendar effects. Many domestic literatures roughly inspect window dressing by testing the fluctuations of funds' yield. Most of the existing literatures have common problems, such as lacking of theoretical support and deep-seated reasons for the cause of window dressing, modeling roughly, lacking index to accurately measure the existence and extent of the window dressing. In terms of the use of the data, since the semi-annual fund data for all holdings of securities cases were not available until 2004, and it's difficult to deal with the increasing number of all stock holdings data, so almost all the domestic literatures use the disclosed top ten quarterly fund data, which may leave out important information contained in the whole holding positions.The presence of window dressing, not only lead investors to accept false information and misjudge the ability of fund managers and then suffer losses which could have been avoided, but also increase the administrative expenses and other unnecessary costs of the fund. As the development of Chinese fund market, institutional investors become increasingly influential in the financial markets. The existence of window dressing hampers the efficient allocation of resources in the financial markets, distorts the efficient allocation of resources, and weakens the market efficiency. Therefore, it is necessary to research on and analyze the domestic window dressing. The research on the existence of window dressing could examine the irrational investment behavior of Chinese open-end securities investment funds, and to identify and analyze the factors influencing window dressing could provide a more accurate method to evaluate fund managers' ability. Studying window dressing not only help cultivate rational investors, but also allows fund managers and fund companies recognize the dangers of window dressing to promote the fund industry self-regulation. What's more, it can help supervisors to identify and constraint the irrational behavior of the fund industry, and to stabilize the market and increase the market efficiency.Based on the previous literature, this paper constructs a relative indicator and a absolute indicator to measure window dressing in China's open-end fund market. On this basis, the paper further analyze the window dressing motives of fund managers using data from June 2006 to June 2015.After that, this paper use the sort grouping method, the fixed effects model and Logit model to test three effective factors of the window dressing's-the performance of the fund in the past, the fund manager's ability, as well as fund style. Eventually, through the analysis of the momentum factor's significant in the Carhart's four-factor model, this paper separates window dressing from the momentum effect, and then prove the constructed measure index capture the window dressing rather than the momentum effect, strengthening this article's conclusion robustness.Follow the ideas above, this article is divided into five chapters, the article is organized as follows:The first chapter is the introduction. The introduction part introduces the research background, the methods used, the whole article content framework and the innovation of this paper. The second chapter introduces three main directions of the study of window dressing. Finally, the existing literatures were reviewed and summarized.The third chapter is the design of key variables and the introduction of data used. This chapter describes how to construct two indicators to measure window dressing. In addition, this chapter describes the construction method of other key indicators. Finally, the data used are introduced. The fourth chapter is empirical study part. This paper chooses open-end funds in China as the research object from 2006 to 2015, and selects the stock holdings data in the fund semi-annual and annual reports. This paper uses packet sorting method, fixed effects model, Logit model and T-test methods. On the basis of inspection the motives of fund manager window dressing, this article studies the presence and influence factors of window dressing from multiple perspectives. Finally, the results obtained are taken robust test using Carhart's four factor model, excluding the impact of momentum trading on the experimental results.The fifth chapter summarizes and analyzes the empirical results of this paper. Then this chapter proposes policy suggestions according to the empirical results. According to results of this study, investors are concerned about the fund's past performance, but also increasingly concerned about the stock holdings of the Fund disclosed in the half year end and year-end. Those funds with worse past performance are more motivated to window dressing. In the case of fund managers lacking the ability, the paper also draws a conclusion that the window dressing will be more pronounced. In addition, actively managed funds compared to passive managed funds are more prone to window dress. For policy suggestions, this paper puts forward four proposals. Specifically, to improve the information disclosure mechanism, to establish a scientific incentive mechanism, to improve the legal regulation and self-management, and to enhance investor education and foster rational investors can effectively reduce the generation of the fund market window dressing.The main innovation of this paper is reflected in the following three aspects: First, in the empirical research, since the window dressing is difficult to be directly observed and measured, there is no widely accepted indicator in our country. Learning from Aarwal et al (2014), this article starts from the most detailed open-end fund stock holding data, and constructs two indicators which can concrete measure window dressing. Second, this paper studies the direct motivation of fond managers to window dress. Finally, since the window dressing and momentum strategies are similar in buying success stocks and selling failure stocks, this paper uses the Carhart's four-factor model to test the robustness of the empirical results.Although this article tied best to analyze window dressing, there are still shortcomings. Specifically, there are three limitations:Firstly, the data from this study are from the semi-annual and annual reports. While the semi-annual data can capture overall information that quarterly data can't, but due to the low frequency of time, the information during the six months may have been overlooked, as a result the window dressing in the market may be underestimated and investors can't make decisions quickly based on the latest information. Thus, if combined with the semi-annual and quarterly disclosed information to build a more reasonable indicator to test the window dressing, we may get more instructive conclusions. Secondly, the stock positions adjustment in 2014 causes lack of the research data, which makes it difficult to separately study the open-ended equity funds and open-ended hybrid funds. If there is a way to solve this problem, more accurate results may be available. Finally, the article simply uses the significant factor of the Carhart's four-factor model as the robustness test, which is relatively rough. If there is more quantifiable method, it will be more conducive to the consolidation of the results of this paper.