Full length articleInvestor sentiment and stock market liquidity: Evidence from an emerging economy
Introduction
In this paper, I examine the relationship between aggregate stock market liquidity and investor sentiment in the non-linear conditional mean–variance framework. The extant literature focuses on how the liquidity affects the stock returns whereas a few studies examine the sources of the variations in the stock market liquidity (Chordia et al., 2001, Amihud, 2002). A couple of studies document the macroeconomic factors affecting the stock market liquidity and the intertemporal relationships between market liquidity, returns and volatility (e.g., Chowdhury et al., 2018). The recent evidence from the school of behavioral finance emphasizes the role of investor sentiment in the asset pricing (Baker and Wurgler, 2006, Baker et al., 2012), the behavior of returns and volatility. However, the literature has paid little attention to evaluate the relationship between investor sentiment and stock market liquidity.
Intuitively hypothesize that investor sentiment can have direct and indirect effects on the stock market liquidity. In the case of direct effects, bullish investor sentiment affects the liquidity through noise traders (Black, 1986) and irrational market makers (Shiller, 2000). The earlier theoretical framework of De Long et al., 1990a, De Long et al., 1990b, Kyle (1985), and Lee et al. (1991) incorporate the significant role of noise traders in asset pricing. Such function, in turn, expected to affect the market liquidity. In the indirect effects, Baker and Stein (2004) propose a model in which sentiment is higher when a large number of irrational market makers exists in the markets. As these irrational market makers are assumed to underreact or overreact to the information contained in the order flows, the price impact caused by such order flows is lower, and hence liquidity increases. The higher sentiment reflects a higher level of overconfidence in the markets, and such overconfidence increases stock market liquidity (Odean, 1998). Despite such theoretical linkage between liquidity and sentiment, empirical evidence on the relationship is scarce particularly in emerging markets (EMs) (e.g., Ogunmuyiwa, 2010, Debata et al., 2017). The relationship between microstructure variables such as liquidity and behavioral factors such as sentiment offers intriguing insights into working of real-world financial markets.
Against this backdrop, I examine the relationship between liquidity and sentiment and contribute to the nascent literature in many folds. First, the present study is first work on liquidity and sentiment in EMs. The work of Liu (2015) is the sole empirical study on the issue exclusively focusing on the US market. The work Liu (2015) ignored the time varying nature of liquidity and sentiment. Our study addresses this issue and thus departs from previous work. Further, the findings on the free market benchmark such as the US are inapplicable to the EMs such as India with the equal force due to peculiar characteristics of the latter. The friction such as non-synchronous trading, lack of liquidity, the dominance of institutional traders, higher volatility, poor disclosure norms, lack of regulation, etc. characterizes the EMs. At the same time, these markets offer higher returns and better diversification opportunities. The issue also assumes significance because of financial liberalization, international portfolio diversification, the exponential increase in the market capitalization, and faster economic growth of the EMs. Therefore, the sample from India, the second faster-growing economy is an ideal candidate for the empirical investigation. Thus, this study also extends the literature on sentiment in EMs.
Second, the present study develops an aggregate sentiment index (ASI) for India following the top-down approach of Baker and Wurgler (2007). Our sentiment index includes aggregate market sentiment indicators related to market performance, types of trading activity, derivative variables, and other sentiment proxies.
Third, I examine the role of institutional investor sentiment in determining the stock market liquidity. Previous research focuses on the impact of investor sentiment on the stock market liquidity in the developed markets in which the retail investors dominate and noise trading due to them has a substantial effect on the market (Brown and Cliff, 2004). The extant studies have not addressed the question of how institutional investor sentiment is priced in the EMs. Theoretically, the institutional investors are informed and rational arbitrageurs, they trade on fundamentals. Nevertheless, institutional investors in EMs such as India often trade against market fundamentals and act irrationally with the optimism and herding attitude to gain extra risk premium from the markets. Therefore, such traders drive the prices against intrinsic values. Hence, the issue of how institutional investor sentiment affects liquidity assumes significance in the EMs such as India.
Fourth, noise trading drives the prices away from the fundamentals and often leads to higher volatility. Such higher volatility due to sentiment in the market eventually threatens the stability of financial markets (Shleifer and Summers, 1990). The higher volatility because of noise traders increases the cost of trading for the market participants and thus adversely affect the market liquidity. Therefore, the institutional investor sentiment and stock market liquidity relationship poses important research questions and the nexus is significant from the perspective of market reforms and microstructure changes in India. Fifth, the study employs the set of non-linear GARCH class of models to find out the positive and negative effects of institutional investor sentiment on the stock market liquidity volatility. The non-linear GARCH models are comprehensive tools and provide intriguing insights to understand the time variation of liquidity. I introduce sentiment in the mean–variance framework to capture the stylized facts such as volatility persistence, clustering and asymmetry characteristics for the stock market liquidity. I probe how time varying liquidity volatility is dependent on the time varying sentiment. The evidence from the previous work on the issue failed to capture the time variation and non-linearity.
The remainder of the paper organized as follows. In Section 2, I present the theoretical framework and the empirical literature and describe the methodology in Section 3. In Section 4, I explain the data and the construction of the liquidity and investor sentiment variables whereas I discuss the empirical results in Section 5. I conclude the paper in the last section.
Section snippets
Theoretical framework and empirical literature
The modern theory of finance postulates that asset prices tend to move closer to the fundamental values and self-correcting mechanism corrects any disruptions in the prices, and thus markets are efficient in the long-run. This proposition assumes market participants as rational risk-averse arbitragers and believes in the subjective expected utility maximization. Hence, the fundamental factors alone influence asset prices, cross-sectional returns, and liquidity in the financial markets. Of late,
Methodology
I employ the univariate non-linear conditional heteroskedastic models to estimate monthly time-varying conditional liquidity and the investor sentiment. I apply Bollerslev’s (1986) generalized autoregressive conditional heteroskedastic models (GARCH), Nelson’s (1991) exponential GARCH and threshold GARCH models of Zakoian (1994) and Glosten et al. (1993) models. The present study is motivated to analyze the non-linear data generating process (Campbell and Hentschel, 1992) as one where the
Data and variables
The study period spans from the April-2000 to March-2018 (the 204 monthly time series observations). The sample is subject to the availability of data on variables to construct the ASI and liquidity variables. The data sources are NSE India, AMFI and the CMIE Prowess and Bloomberg terminal.
Empirical findings and discussions
Conclusion
I empirically investigated the theoretical interlinkages between the stock market liquidity and the investor sentiment in India. Unlike previous work on EMs, I constructed the aggregate investor sentiment index (ASI) for India. Departing from the existing literature, this study probed the time varying liquidity relationship with time varying investor sentiment in the non-linear GARCH conditional framework. Further, I employed the nonlinear causality tests to understand the causal relationship
Acknowledgment
I thank anonymous reviewer for the impactful comments and insightful suggestions. Usual disclaimer applies.
References (79)
- et al.
Linear and non-linear granger causality between oil spot and futures prices: A wavelet-based test
J. Int. Money Financ.
(2014) Illiquidity and stock returns: cross-section and time-series effects
J. Financ. Mark
(2002)- et al.
Market liquidity as a sentiment indicator
J. Financ. Mark
(2004) - et al.
Global, local, and contagious investor sentiment
J. Financ. Econ.
(2012) Generalized autoregressive conditional heteroskedasticity
J. Econom.
(1986)- et al.
Alternative factor specifications, security characteristics, and the cross-section of expected stock returns
J. Financ. Econ.
(1998) - et al.
Investor sentiment and the near-term stock market
J. Empir. Financ.
(2004) - et al.
No news is good news: An asymmetry model of changing volatility in stock returns
J. Financ. Econ.
(1992) - et al.
Liquidity and macroeconomic management in emerging markets
Emerg. Mark. Rev.
(2018) - et al.
A new statistic and practical guidelines for nonparametric granger causality testing
J. Econ. Dyn. Control
(2006)