Machine learning sentiment analysis,COVID-19 news and stock market reactions

The recent COVID-19 pandemic represents an unprecedented worldwide event to study the influence of related news on the financial markets, especially during the early stage of the pandemic when information on the new threat came rapidly and was complex for investors to process. In this paper, we investigate whether the flow of news on COVID-19 had an impact on forming market expectations. We analyze 203,886 online articles dealing with COVID-19 and published on three news platforms (MarketWatch.com, NYTimes.com, and Reuters.com) in the period from January to June 2020. Using machine learning techniques, we extract the news sentiment through a financial market-adapted BERT model that enables recognizing the context of each word in a given item. Our results show that there is a statistically significant and positive relationship between sentiment scores and S&P 500 market. Furthermore, we provide evidence that sentiment components and news categories on NYTimes.com were differently related to market returns.     

Does investor sentiment predict bitcoin return and volatility? A quantile regression approach

The study investigates hypotheses relating to the effect of investor sentiment on predicting bitcoin returns and volatility. Using moments quantile regression, we present robust empirical evidence for the period 2017–2021. Our findings demonstrate that investor interest and emotions are significant predictors of bitcoin returns and volatility, while VIX and Bitcointalk.org forum are the most suitable predictors for representing investor emotions and interest, respectively. The findings also indicate a nonlinear relationship between investor sentiment and bitcoin returns and volatility, with predictable power changing based on the market conditions. Thus, the study enriches existing literature by providing empirical evidence to affirm the viability of behavioral finance theories in the bitcoin market and complements investors with more information to seek profits in different market conditions.     

Wall Street analysts as investor relations officers

This paper examines the practice of hiring financial analysts as investor relations officers (IRO). We posit that analysts-turned-IROs (AIROs) have a competitive advantage in communicating with investors, thereby lowering the effort expended by the investment community to process corporate disclosures. Using a unique manually-collected dataset on the employment history of IROs compiled from LinkedIn, Capital IQ, RelationshipScience.com, and appointment press releases, we show that disclosure readability in 8-K and 10-K filings improves and that companies are more likely to host analyst/investor days after hiring former analysts as IROs. Most importantly, we find increases in analyst following, institutional investors, and stock liquidity after hiring a former analyst as IRO. We conclude that both a disclosure and a network channel are at play in the relation between AIROs and increased interest from the investment community. Overall, our findings suggest that firms benefit from hiring Wall Street analysts as IROs.     

Herd Behaviour and Cascading in Capital Markets: a Review and Synthesis

We review theory and evidence relating to herd behaviour, payoff and reputational interactions, social learning, and informational cascades in capital markets. We offer a simple taxonomy of effects, and evaluate how alternative theories may help explain evidence on the behaviour of investors, firms, and analysts. We consider both incentives for parties to engage in herding or cascading, and the incentives for parties to protect against or take advantage of herding or cascading by others.     

THE BROKERAGE FIRM EFFECT IN HERDING: EVIDENCE FROM INDONESIA

We examine herding behavior of domestic and foreign investors in the Indonesian stock market. We document that both domestic and foreign investors from a particular brokerage firm tend to herd. The foreign investors exhibit a greater propensity to herd than domestic investors. However, when examining investor trading across brokerage firms, we find only weak evidence of herding by domestic investors and no herding by foreign investors. Our overall findings suggest a strong brokerage firm effect on herding but a weak marketwide effect. Moreover, we find evidence that the strong brokerage effect on herding is likely driven by acting on common information.     

Herd mentality in the stock market: On the role of idiosyncratic participants with heterogeneous information

This paper examines herd behaviour using aggregate market data for stocks, with a focus on the role of idiosyncratic participants with heterogeneous information. We look at herding asymmetry between up and down markets, taking into consideration the daily price limits and the impact of the recent financial crisis. We also improve upon existing tests for fundamental and non-fundamental herding, as well as proposing a method for investigating herd behaviour of different groups of investors. Empirical evidence based on the Ho Chi Minh Stock Exchange in Vietnam reveals a greater level of herding on up compared to down market days, and a significant reduction in the magnitude of herding following the crisis. We document robust intentional herding even when unintentional (fundamental) herding is factored out. Our empirical results also uncover potential within-group herding and between-group interactions among arbitrageurs and noise traders in the market.     

Market sentiment: a key factor of investors’ imitative behaviour

The aim of this paper is to explore herding behaviour among investors to determine its rational and emotional component factors and identify relationships among them. We apply causality tests to evaluate the impact of return and market sentiment on herding intensity. The herding intensity is quantified using the measure developed by Patterson and Sharma (2006) . The research was conducted during the period 1997–2003 in the Spanish stock market, where the presence of herding has been confirmed. The results reveal that the herding intensity depends on past returns and sentiment or subjective assessments and confirm the presence of both a rational and an emotional factor.     

The Response of Dynamic Herd Behavior to Domestic and U.S. Market Factors: Evidence from the Greater China Stock Markets

ABSTRACT

We investigate the dynamic reaction of stock market herding in China, Hong Kong, and Taiwan to unexpected shocks from domestic and U.S. market factors. In China and Taiwan, herding is more pronounced, and the investors tend to herd with the rising stock market returns. Overconfident investors will herd on the subsequent trading days under market stress. Compared with the response to the domestic market factors, the responses of herding in the Greater China stock market to the U.S. market factors are weaker. After the 2007–8 financial crisis, the U.S. market factors highly explain the forecast error variance of herding in the Shanghai A-share and Taiwan markets.     

Herding and Feedback Trading by Institutional and Individual Investors

We document strong positive correlation between changes in institutional ownership and returns measured over the same period. The result suggests that either institutional investors positive-feedback trade more than individual investors or institutional herding impacts prices more than herding by individual investors. We find evidence that both factors play a role in explaining the relation. We find no evidence, however, of return mean-reversion in the year following large changes in institutional ownership—stocks institutional investors purchase subsequently outperform those they sell. Moreover, institutional herding is positively correlated with lag returns and appears to be related to stock return momentum.     

Short-selling bans and institutional investors' herding behaviour: Evidence from the global financial crisis

The literature on short-selling restrictions focusses mainly on a ban's impact on market efficiency, liquidity and overpricing. Surprisingly, little is known about the effects of short-sale constraints on herd behaviour. Since institutional investors have come to dominate mature stock markets and rely extensively on short sales, constraining these traders may influence the asset pricing process. We investigate six stock markets that faced bans during the recent global financial crisis. Our empirical evidence shows that short-selling restrictions exhibit either no influence on herding formation or induce adverse herding. This implies a higher dispersion of returns around the market compared to rational asset pricing, which can be interpreted as an increase in uncertainty among stock market investors.     

Institutional herding in international markets

This paper studies herding behavior of institutional investors in international markets. First, we document the existence of wide-spread herding in 41 countries (referred to as “target countries” hereafter) in the sample. We then examine the relation between contemporaneous institutional demand and future returns and find that institutional herding stabilizes prices. Next, we examine the relation between institutional investors’ herding behavior and the level of information asymmetry in the target countries. We measure the degree of information asymmetry in each target country along five dimensions: (1) stock market development, (2) ease of access to information, (3) corporate transparency, (4) investor rights, and (5) macroeconomic factors that relate to the information environment. We find evidence that institutional investors herd more in markets characterized by low levels of information asymmetry (high level of information transparency). This result suggests that institutional investors’ herding behavior is likely driven by correlated signals from fundamental information. Lastly, we show that price adjustment is faster in informationally transparent markets.     

Credit rationing in P2P lending to SMEs: Do lender-borrower relationships matter?

This paper studies the role of individual P2P investors that are acquainted with the borrower in mitigating credit rationing in P2P lending to SMEs. I use proprietary data provided by one of the biggest Dutch P2P lending platforms, on which personal acquaintances of the borrower are able to invest before other P2P investors do. I find that P2P investors invest more in loans of borrowers to whom they are personally acquainted. More initial investment by investors acquainted with the borrower is subsequently associated with a higher likelihood of obtaining a second loan from the P2P lender, larger investments by other P2P investors and lower ex post defaults. These results are consistent with informal lenders having superior information or monitoring skills and rational herding following informal investors' investment decisions.     

Dynamic herding analysis in a frontier market

We use constant coefficient and time-varying parameter approaches to examine herding in the context of a frontier market. Our sample comprises of all companies listed on the Trinidad and Tobago Stock Exchange from January 2001 to December 2014. We find significant evidence of herding across the market, which is more prominent for smaller stocks. Microstructures, including liquidity and volatility, intensify herd behavior, except for larger firms. Additional analyses show that herding is present in both up and down markets, but is stronger during rising markets. The time-varying analysis, based on a state-space Kalman filter, further establishes that herding, though quite prevalent, is not a static feature of the market but evolves throughout the sample period. Specifically, it oscillates between greater herding to anti-herd behavior, as investors identify themselves with crises and better information access respectively.     

Individual and institutional herding and the impact on stock returns: Evidence from Taiwan stock market

Using high frequency intraday data, this paper investigates the herding behavior of institutional and individual investors in the Taiwan stock market. The study finds evidence of herding by both investors but a stronger herding tendency among institutional than among individual investors. Institutional investors herd more on firms with small capitalizations and lower turnovers and they follow positive feedback strategies. The portfolios that institutional investors herd buy outperform those they sell by an average of 1.009% during the 20 days after intense trading episodes. By contrast, individual investors herd more on firms with small sizes and higher turnovers, and they crowd to buy (sell) stocks with negative (positive) past returns. The portfolios that individual investors herd buy underperform those they sell by an average of − 0.829% during the following 20 days. Moreover, these return differences of both investors are more pronounced under a market with higher pressure and among small stocks. These findings suggest that the herding of institutional investors speeds up the price-adjustment process and is more likely to be driven by correlated private information, while individual herding is most likely to be driven by behavior and emotions.     

A finite-time singularity in the dynamics of the US equity market: Will the US equity market eventually collapse?

Fitting Dow Jones 30 index data for the 1790–1999 period into a log-periodic power-law singularity (LPPLS) model, the seminal paper by Johansen and Sornette (2001) was the first to show that the US equity growth rate is accelerating such that the market is growing as a power law toward a spontaneous singularity. Their model suggests that the US equity market will reach this critical point in the year 2052 ± 10 years, signaling an abrupt transition to a new regime. This study re-examines this important issue using (i) a novel approach to calibrate the LPPLS model and (ii) a different data set including >20 years of additional data. The extended data account for the dot.com bubble burst (2000), the Global Financial Crisis period (2008–2009), the COVID−19 crisis (2020−2022), and the ongoing Russian–Ukrainian war (starting in 2022), which are all events with severe consequences for the global economy. The calibrated LPPLS model suggests that the US equity market will reach a singularity condition by June 2050.     

Herding in the German Mutual Fund Industry

This paper analyses the trading activity of German mutual funds in the 1998–2002 period to investigate whether German mutual fund managers are engaged in herding behaviour. Another objective of the study is to determine the impact of this herd‐like trading on stock prices. Our results provide evidence of herding and positive feedback trading by German mutual fund managers. We show that a significant portion of herding detected in the German market is associated with spurious herding as a consequence of changes in benchmark index composition. Investigating the impact of mutual fund herding on stock prices, we find that herding seems to neither destabilise nor stabilise stock prices.     

Textual sentiment of comments and collapse of P2P platforms: Evidence from China's P2P market

Textual sentiment affects the investment activities of investors in traditional financial markets. Peer-to-Peer (P2P) lending market, as one of the emerging and active Internet financial markets, has recently received considerable attention from academia. However, few related studies are available. This work examines the relationship between the textual sentiment derived from investors’ comments on P2P platforms and probability of platform collapse. We collect comments from an authoritative Chinese third-party P2P lending consulting platform and use a weakly supervised convolutional neural network to calculate the textual sentiment of each comment. Empirical results show that the extracted textual sentiment has a significant influence on a P2P platform's collapse. Furthermore, the “agreement” and “disagreement” from other investors of each comment are pivotal in predicting a P2P platform's failure. We find that the textual sentiment of comments regarding P2P platforms from investor communities provide insights into predicting platforms’ collapse in the near future.     

Herd behaviour & investor sentiment: Evidence from UK mutual funds

The aims of this paper are to detect evidence of institutional investor herding behaviour and examine the role that investor sentiment plays in institutional investor herding behaviour. The herding behaviour is investigated by examining the dispersion of time varying beta of UK open-end and closed-end funds. The study finds evidence of fund managers' herding behaviour, which suggests they are likely to herd on market portfolio, size, and value factors. UK market-wide investor sentiment index is used for investigating the effects of investor sentiment on institutional herding behaviour. We find a unidirectional investor sentiment effect on the herding of UK mutual fund managers. We also reveal that the sentiment factors affecting UK open-end and closed-end fund managers herding behaviour are different due to the differences in fund structure.     

Gender and herding

This study uses a unique dataset from a large anonymous brokerage firm to examine the herding behavior of Chinese individual investors. The empirical evidence reveals that females are more inclined to follow the behavior of ‘same-sex’ investors. Market conditions and stock characteristics affect females and males similarly in that individual investors herd more intensively in the bull market, on stocks with better liquidity and larger market capitalization. We find female investors generally yield lower returns than males when they herd intensively, and this finding is more pronounced during a bull-market period. Outcomes from individual-level herding measurements suggest that portfolio turnover drives the difference in herding between genders.     

Herding and market volatility

In this paper, we explore the impact of investor herding behavior on stock market volatility. We adopt a direct herding measure based on the variation of cross-sectional stock betas. The measure can be readily separated into positive and adverse components, whereby investors herd towards and away from the market portfolio, respectively. Using A-shares listed in the Chinese equity market from August 2005 to March 2021, we show that the market volatility is Granger caused by the measure, and that there exists an asymmetric effect between positive and adverse herding on volatility. Furthermore, we provide robust evidence that the information contained in the herding measure helps generate significantly improved volatility forecasts and add economic value to investors. Our paper not only contributes to the volatility forecasting literature but also advances our understanding of herding in the equity market.