Investor gambling preference and asset pricing

Taking investors’ gambling preferences as a starting point, this thesis focuses on gambling preferences for lottery-like and policy-related stocks. The related trading behavior, corresponding market consequences, and internal economic mechanism are deeply discussed in this thesis. The specific conclusions of this thesis are as follows.

In the first essay of this thesis, idiosyncratic skewness, which is a concept from “skewness preference theory,” is used to measure investors’ gambling preferences for lottery-like stocks. This part deeply explores the gambling behavior and trading characteristics of investors in the Chinese stock market. In addition, the underlying reasons for the strong gambling atmosphere of the Chinese stock market are also discussed. This section finds that there is an obvious gambling phenomenon in the Chinese market. Furthermore, using social media data (i.e., Baidu search data, Sina News data, and Hexun News data), this part constructs an individual investor’s attention index. In addition, the “Dragon and Tiger list” is used as the exogenous attentiongrabbing event of individual investors and uses high frequency trading data to construct the trading aggressiveness of individual investors. With the above measures, this portion of the paper finds that investor attention is an important driving factor in individual investor’s gambling behavior and the subsequent individual investor’s aggressive trading is the main internal channel. Finally, using the relevant factors of the trading mechanism and stock characteristics, this section constructs comprehensive limits of the arbitrage index. The results confirm that higher arbitrage restrictions in the Chinese market may further stimulate the gambling preferences of investors, thus intensifying the gambling atmosphere of the market.

In the second essay of this thesis, this study uses the maximum stock daily return (MAX) as a more direct measure of investors’ gambling preferences for lottery-like stocks. This section analyzes the underestimation of the gambling effect caused by the price limit rule and the inside economic mechanism. This portion finds that due to the existence of price limits, after stock returns reach the upper price limit, the MAX factor does not fully reflect the investors’ sentiments and relevant trading information, so the real gambling effect may be underestimated. To determine this, this section calculates the modified MAX (RMAX) factor that can fully reflect the relevant information. The results indicate that using RMAX as a new measure of investors’ gambling preferences can capture stronger gambling behavior. Subsequently, this part also explores the economic mechanism behind the MAX factor’s underestimation of the gambling effect. For the above purpose, the Baidu search volume is used to capture individual investor’s attention. Investors’ social forum data, reflected by the Eastmoney stock forum, are used to capture investment sentiment. High frequency trading data are used to construct the trading indicators of individual investors. Using the above indicators, this section finds that when stock returns reach the daily upper price limit, investors’ attention and the corresponding gambling sentiment will be aroused, and more individual investors will trade aggressively. However, the original MAX factor cannot capture the corresponding change in sentiment and trading resulting in the underestimation of the gambling effect. 

In the third essay, this study uses individual stocks’ exposure to economic policy uncertainty to capture investors’ gambling preferences for policy-related stocks and explores the related speculation behavior of individual investors and the market consequences in detail. This section finds that individual investors demonstrate high gambling preferences for economic policy uncertainty and the corresponding speculation impedes price discovery efficiency. The trading aggressiveness index of individual investors measured by high frequency trading data indicates that the speculative trading of individual investors is the primary economic mechanism that impedes price discovery efficiency. In addition, this section also conducts many heterogeneity tests on stock characteristics (i.e., high liquidity, high attention, and high positive information shock) and market states (i.e., UP market, high sentiment, and lunar new year) that more easily stimulate investors to gamble. The results indicate that investors do speculate more aggressively. Finally, combined with the impact of exogenous laws and changes in the development of the capital market (i.e., cancellation of short-selling constraints and the development of the analyst market), this study finds that relevant policy speculation is weakened in a more developed and transparent market environment. 

Overall, this thesis effectively helps us to achieve a more comprehensive understanding of gambling behavior in the Chinese stock market. More importantly, this thesis innovatively uses high frequency trading data, social media data, and online search data, deeply discusses the internal mechanism of investors’ gambling preferences, and explores the trading characteristics of investors’ gambling behavior. The above results significantly help us to achieve a deeper understanding of investors’ gambling behavior and internal mechanisms. In addition, this thesis also analyzes gambling behavior in the Chinese stock market in terms of several institutional factors. The exploration of relevant institutional factors provides a reference for a more comprehensive understanding of the heavy gambling atmosphere of the Chinese market. Moreover, analysis of the related institutional factors and the reform of the trading mechanism can also provide important theoretical support to reduce gambling behavior and guide the reasonable development of the stock market.