The role of text-extracted investor sentiment in Chinese stock price prediction with the enhancement of deep learning

Whether investor sentiment affects stock prices is an issue of long-standing interest for economists. We conduct a comprehensive study of the predictability of investor sentiment, which is measured directly by extracting expectations from online user-generated content (UGC) on the stock message board of Eastmoney.com in the Chinese stock market. We consider the influential factors in prediction, including the selections of different text classification algorithms, price forecasting models, time horizons, and information update schemes. Using comparisons of the long short-term memory (LSTM) model, logistic regression, support vector machine, and Naïve Bayes model, the results show that daily investor sentiment contains predictive information only for open prices, while the hourly sentiment has two hours of leading predictability for closing prices. Investors do update their expectations during trading hours. Moreover, our results reveal that advanced models, such as LSTM, can provide more predictive power with investor sentiment only if the inputs of a model contain predictive information.     

Time-varying impact of monetary policy shocks on US stock returns: The role of investor sentiment

This paper investigates how monetary policy shock affects the stock market of the United States (US) conditional on states of investor sentiment. In this regard, we use a recently developed estimator that uses high-frequency surprises as a proxy for the structural monetary policy shocks, which in turn is achieved by integrating the current short-term rate surprises, which are least affected by an information effect, into a vector autoregressive (VAR) model as an exogenous variable. When allowing for time-varying model parameters, we find that, compared to the low investor sentiment regime, the negative reaction of stock returns to contractionary monetary policy shocks is stronger in the state associated with relatively higher investor sentiment. Our results are robust to alternative sample period (which excludes the zero lower bound) and model specification and also have important implications for academicians, investors, and policymakers.     

The impacts of investor sentiment on returns and conditional volatility of international stock markets

One of the main arguments of behavioral finance is that some properties of asset prices are most probably regarded as deviations from fundamental value and they are generated by the participation of traders who are not fully rational, thus called noise traders. Noise trader theory postulates that sentiment traders have greater impact during high-sentiment periods than during low-sentiment periods, and sentiment traders miscalculate the variance of returns undermining the mean-variance relation. The main objective of this research is to construct a model to evaluate the returns and conditional volatility of various stock market indexes considering the changes in the investor sentiment by measuring the effects of noise trader demand shocks on returns and volatility. EGARCH model is used to determine whether earning shocks have more influence on the conditional volatility in high sentiment periods weakening the mean–variance relation. This paper takes an international approach using weekly market index returns of U.S., Japan, Hong Kong, U.K., France, Germany, and Turkey. Weekly trading volumes of these indexes are regressed against a group of macroeconomic variables and the residuals are used as proxies for investor sentiment and significant evidence is found that there is asymmetric volatility in these market indexes and earning shocks have more influence on conditional volatility when the sentiment is high.     

Overnight stock returns,intraday returns,and firm-specific investor sentiment

In this study, we propose a new index for measuring firm-specific investor sentiment using overnight and intraday stock returns. We use actual equity data to construct the firm-level investor sentiment index and find that the new index has characteristics expected of a sentiment measure. In addition, we propose a novel sentiment-weighted trading strategy and apply it to momentum and short-term reversal strategies. We find that the sentiment-weighted trading strategy generates better performance in momentum and short-term reversal strategies. The sentiment-weighted trading strategy’s superior performance is evidence that our firm-level investor sentiment index possesses predictive powers with regard to future returns.     

Individual stock crowded trades,individual stock investor sentiment and excess returns

Recent behavioral asset pricing models and the popular press suggest that investors may follow similar strategies resulting in crowded equity positions to push prices further away from fundamentals. This paper develops a new approach to measure individual stock crowded trades, and further investigates the joint effects of individual stock crowded trades and individual stock investor sentiment on excess returns. Specifically, our results show that the combined effect of individual stock crowded trades and individual stock investor sentiment on excess returns is positive and significant, which reveals the importance of “anomaly factors” in asset pricing. Furthermore, our results suggest that increasing individual stock buyer-initiated crowded trades will increase excess returns simultaneously; however, increasing individual stock seller-initiated crowded trades will decrease excess returns simultaneously. Collectively, our results highlight the importance of individual stock crowded trades and individual stock investor sentiment on the formation of stock prices.     

Forecasting abnormal stock returns and trading volume using investor sentiment: Evidence from online search

We examine the ability of online ticker searches (e.g. XOM for Exxon Mobil) to forecast abnormal stock returns and trading volumes. Specifically, we argue that online ticker searches serve as a valid proxy for investor sentiment — a set of beliefs about cash flows and investment risks that are not necessarily justified by the facts at hand — which is generally associated with less sophisticated, retail investors. Based on prior research on investor sentiment, we expect online search intensity to forecast stock returns and trading volume, and also expect that highly volatile stocks, which are more difficult to arbitrage, will be more sensitive to search intensity than less volatile stocks. In a sample of S&P 500 firms over the period 2005-2008, we find that, over a weekly horizon, online search intensity reliably predicts abnormal stock returns and trading volumes, and that the sensitivity of returns to search intensity is positively related to the difficulty of a stock being arbitraged. More broadly, our study highlights the potential of employing online search data for other forecasting applications.     

Hard to arbitrage,hard for analysts to forecast

This paper examines whether limits to arbitrage (LA) affect analysts' earnings forecast accuracy. Using the LA index, which is constructed from unique trading constraints in the Chinese stock market and other commonly used measures, we find that forecast accuracy is much lower for stocks with high LA. Moreover, our results are more suited to explanations of cognitive bias that turn to investor sentiment or limited attention and cannot be fully explained by more objective factors, including analyst ability, broker size, broker experience, and commission pressure. We also find that LA amplifies analyst forecast dispersion. Such results indicate that LA distorts analysts’ earnings expectations and provides new insight into how LA affects anomaly returns.     

The momentum and reversal effects of investor sentiment on stock prices

In this paper, we illustrate the real function relationship between the stock returns and change of investor sentiment based on the nonparametric regression model. The empirical results show that when the change of investor sentiment is moderate, the stock return is positively correlated with the change of investor sentiment, presenting an obvious momentum effect. However, the stock return is negatively correlated with the change of investor sentiment if the change of investor sentiment is dramatic, presenting significant reversal effects. Moreover, the degree of reversal effect caused by extremely optimistic sentiment is greater than that driven by extremely pessimistic sentiment, which shows a significant asymmetry. Our findings offer a partial explanation for financial anomalies such as the mean reversion of stock returns, the characteristic of slow rise and steep fall in China's stock market and so on.     

New insights into the role of global factors in BRICS stock markets: A quantile cointegration approach

Because of the acceleration in marketization and globalization, stock markets in the BRICS (Brazil, Russia, India, China, and South Africa) countries are affected by various global factors, for example, oil prices, gold prices, global stock market volatility, global economic policy uncertainty, financial stress, and investor sentiment. This paper offers new insights into the short- and long-run linkages between global factors and BRICS stock markets by applying the quantile autoregressive distributed lags (QARDL) approach. This novel methodology enables us to test short- and long-run linkages accounting for distributional asymmetry. That is, the nonlinear dynamic relationship between the global factors and BRICS stock prices depends on market conditions. Our empirical results show that the effects of gold prices and global stock market volatility on BRICS stock prices are more significant in the long run than in the short run. A decrease in global stock market volatility is associated with higher stock prices, while gold prices demonstrate upward co-movement in dynamic correlations with stock markets. Irrational factors, such as economic policy uncertainty, financial stress, and investor sentiment, play a critical role in the short term, and negative interdependence is dominant. Finally, the rolling-window estimation technique is used to examine time-varying patterns between major global factors and BRICS stock markets.     

Stock Market’s responses to intraday investor sentiment

We investigate the effect of intraday sentiment on subsequent stock returns. Mispricing caused by intraday sentiment is not corrected immediately; rather, it lasts for about 30 min. After 30 min, however, investor sentiment negatively affects stock returns, suggesting that mispriced stocks are at least partially but not entirely adjusted back to their fundamental values. We also show that the effect of intraday sentiment depends on the degree of arbitrage. Intraday sentiment has little effect on firms that are easy to arbitrage. For these firms, the difference in the one-minute returns of firms with high and low sentiment is nearly zero, implying that any mispricing caused by intraday sentiment is immediately corrected for this group of firms. In contrast, among firms that are hard to arbitrage, the difference in the returns of firms with high and low sentiment lasts for about half an hour. This difference in the effect of intraday sentiment is not caused by the firms’ liquidities.     

The role of an illiquidity risk factor in asset pricing: Empirical evidence from the Spanish stock market

The main purpose of this study is to construct an illiquidity risk factor for the Spanish stock market over the 1994–2002 period. Because of the absence of consensus in empirical research about the most appropriate liquidity measure, we applied the Amihud [Amihud, Y. (2002). Illiquidity and stock returns: Cross-section and time-series effects. Journal of Financial Markets 5, 31–56] illiquidity ratio that shows the price response associated with one euro of trading volume. Moreover, we generated an illiquidity factor using the Fama and French [Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds. Journal of Financial Economics 33, 3–56] orthogonal approach and analyzed whether it enters the stochastic discount factor as an additional state variable. We conclude that systematic illiquidity should be a key ingredient of asset pricing.     

Investor sentiment and country exchange traded funds: Does economic freedom matter?

Using unbalanced panel data of 27 iShares MSCI country-specific exchange traded funds (ETFs) over the period 1996–2014, this paper applies quantile regression to examine the impacts of global, foreign, and U.S. investor sentiments on the returns of the ETFs traded in the U.S. markets. We further investigate whether a country’s economic freedom affects the relationship between investor sentiments and ETF returns. We find that ETF returns are strongly determined by investor sentiments and the ETF expense ratio. The quantile regression approach reveals that high-return ETFs are positively sensitive to changes in global sentiment (measured by market turnover, VIX, U.S. federal funds rate), foreign sentiment (measured by current account balance, inflation, market turnover, public debt), U.S. sentiment, currency exchange ratio, and expense ratio, while negatively influenced by economic freedom and Asian proxy. The effects of VIX and foreign inflation are a reversal; that is, returns from lower (higher) quantiles have a negative (positive) relation with VIX and foreign inflation. Not all components of economic freedom affect returns equally.     

基于投资者情绪的IPO首日收益研究

本文借鉴国内外学者关于投资者情绪度量指标选取的方法,遴选出适合中国证券市场特征的四类投资者情绪指标,通过主成分分析方法提取反映投资者情绪的主要因素,并在此基础上,利用VAR模型对投资者情绪与IPO首日收益之间的领先滞后关系进行刻画,主要考察投资者情绪的新息冲击对IPO首日收益动态脉冲影响.结果表明,投资者情绪显著影响IPO首日收益,投资者对未来的预期同样受到IPO首日表现的影响,投资者情绪指标之间存在显著的扩散和弥漫效应.     

An Investigation of Stock Price Declines Following Corporate Financing Events: Further Evidence

Prior research documents a large downward drift in stock prices following issuances of debt and equity by US firms. We conduct tests based on both stock price and trading volume to provide evidence on the reasons for this apparent market anomaly. We document evidence of earnings management through accruals prior to external financing and lower operating performance afterward that is associated with the amount of capital raised. The earnings management that precedes external financing and the amount of capital raised are associated with both the post-financing decline in stock price and trading volume around earnings announcements that follow for a period of three years. This evidence is consistent with the proposition that firms raise external capital prior to predictable declines in their operating performance and they release upward biased earnings before these events to manage investor expectations. The failure of many investors to incorporate this information into their trading decisions in a timely manner consistent with limited attention and over-confidence appears to drive stock mispricing. Our evidence does not support the conjecture that the financing anomaly is primarily a statistical artifact or that it is a manifestation of the accrual anomaly.     

Impact of network investor sentiment and news arrival on jumps

This paper studied the influence of news announcements and network investor sentiment on Chinese stock index and index futures market jumps. A machine learning text analysis algorithm was employed to measure investor forum sentiment. It was found that news arrivals were an important reason for jump occurrences, jumps were significantly associated with network investor sentiment, and while occasionally the news and network investor sentiment resulted in simultaneous market jumps, they appeared to be relatively independent. The network investor sentiment time-lag and asymmetric effects were also tested, from which it was found that network investor sentiment had a significant asymmetric effect on the jumps, but time-lag effects had little influence. News announcements and the top 25% of the extreme network sentiments were found to explain more than 50% of the jumps, with extreme sentiments tending to increase the volatility of the news-related jumps and persistently influencing returns after the news-related jumps.     

The COVID-19 pandemic uncertainty,investor sentiment,and global equity markets: Evidence from the time-frequency co-movements

We use daily data of the Google search engine volume index (GSVI) to capture the pandemic uncertainty and examine its effect on stock market activity (return, volatility, and illiquidity) of major world economies while controlling the effect of the Financial and Economic Attitudes Revealed by Search (FEARS) sentiment index. We use a time–frequency based wavelet approach comprising wavelet coherence and phase difference for our empirical assessment. During the early spread of the COVID-19, our results suggest that pandemic uncertainty, and FEARS sentiment strongly co-move, and increased pandemic uncertainty leads to pessimistic investor sentiment. Furthermore, our partial wavelet analysis results indicate a synchronization relationship between pandemic uncertainty and stock market activities across G7 countries and the world market. Our results are robust to the inclusion of alternative pandemic fear measure in the form of equity market volatility infectious disease tracker. The pandemic uncertainty and associated sentiment implications could be one plausible reason for increased volatility and illiquidity in the market, and hence, policymakers should look upon this issue for the financial market stability perspective.     

Industries' heterogeneous reactions during the COVID-19 outbreak: Evidence from Chinese stock markets

This study examines the heterogeneous effects of the COVID-19 outbreak on stock prices in China. We confirm what is already known, that the pandemic has had a significant negative impact on stock market returns. Additionally, we find, this effect is heterogeneous across industries. Second, fear sentiment can directly cause stock prices to fall and panic exacerbates the negative impact of the pandemic on stock returns. Third, and most importantly, we demonstrate the underlying mechanisms of four firm characteristics and find that those with high asset intensity, low labor intensity, high inventory-to-revenue ratio, and small market value are more negatively affected than others. For labor-intensive state-owned firms, in particular, stock performance worsened because of higher idle labor costs. Finally, we created an index to measure the relative position of an industry in the supply chain, which shows that downstream companies were more vulnerable to the effects of the pandemic.     

America's decoupling from China: A perspective from stock markets

America's decoupling-from-China debate started after July 2018, reached its peak in August 2020, and is likely to continue even if it may not be a high priority for the Biden administration. Many studies have examined various aspects of this issue, especially the potential economic impacts on the US economy. Unlike previous research, this study looks at the response of stock markets. Using Google Trends data, this study created a weekly dataset from January 2020 to June 2021 to measure investor sentiment towards the US decoupling from China. Employing the generalised autoregressive conditional heteroskedasticity (GARCH) models, the study finds that concern over decoupling is associated with significant variations in stock market prices. From this we can infer that the overall effects of decoupling on the US economy are likely to be considerable.     

Evolution of price effects after one-day abnormal returns in the US stock market

This paper provides a comprehensive analysis of price effects after one-day abnormal returns and their evolution in the US stock market, using Dow Jones Index over the period 1890–2018. We utilise several statistical tests and econometric methods (the modified cumulative abnormal return approach, regression analysis with dummy variables, R/S analysis (Hurst, 1951), and the trading simulation approach). The results suggest that a strong momentum effect between 1940 and 1980 after a day of positive abnormal returns was present in the US stock market, and it was exploitable for profit. However, after the 1980s this has since disappeared. Overall, price effects after one-day abnormal returns during the analysed period tend to be unstable in terms of their strength and direction (momentum or contrarian effect). Nowadays, the evidence for the price effects after one-day abnormal returns in the US stock market is weak. Our results, therefore, are consistent with the Adaptive Market Hypothesis (Lo, 2004).     

Testing for linear and nonlinear granger causality in the stock price-volume relation: Korean evidence

In this paper, linear and nonlinear Granger causality tests are used to examine the dynamic relationship between daily Korean stock returns and trading volume. We find evidence of significant bidirectional linear and nonlinear causality between these two series. ARCH-ype models are used to examine whether the nonlinear causal relations can be explained by stock returns and volume serving as proxies for information flow in the stochastic process generating volume and stock returns respectively. After controlling for volatility persistent in both series and filtering for linear dependence, we find evidence of nonlinear bidirectional causality between stock returns and volume series. The finding of strong bidirectional stock price-volume causal relationships implies that knowledge of current trading volume improves the ability to forecast stock prices. This evidence is not supportive of the efficient market hypothesis. Another finding is that the nonlinear relationship is sensitive to institutional, organizational, and structural factors. The results of this study should be useful to regulators, practitioners and derivative market participants whose success precariously depends on the ability to forecast stock price movements.