Examine the effect of investor sentiment 
on JSE ETFs under changing market 

conditions

PA Shenjere
orcid.org 

0000-0001-6051-7763

Dissertation accepted in fulfilment of the 
requirements for the degree Master of  
Commerce in Risk Management at the 

North-West University 

Supervisor: Prof SJ Ferreira-Schenk
Co-supervisor:  Mr F Moodley

Graduation: May 2025 
Student number: 34775692



 
Declaration 

 

I declare that: “Examine the effect of investor sentiment on JSE ETFs under changing market 

conditions”  is my own work, and that I have provided full citations for all the sources that I 

have used and/or quoted, and that I have never before submitted this dissertation for any 

credit toward a degree at any other universities. 

 

 

 

Signature: __________________________________ 

 

 

Date: ______________21 November 2024________________________ 

 

 

 

  



ii 
 

 
Declaration by Language Editor 

 

  



iii 
 

List of Publications 

Title: Does Investor Sentiment Influence South African ETF Flow During Different Market    
Conditions? 

Journal: Economies 

Journal Accreditation: (SCimago Journal Rank: Quartile 2) 

Citation: 

Shenjere, P.A., Ferreira-Schenk, S. & Moodley, F. 2025. Does Investor Sentiment Influence 
South African ETF Flows During Different Market Conditions?. Economies, 13(1): 10. 
https://doi.org/10.3390/economies13010010  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



iv 
 

Acknowledgements 

 

“For I know the plans I have for you, declares the Lord, plans to prosper you and not to harm 

you, plans to give you hope and a future.” – Jeremiah 29:11 

 

I would like to take a moment and express my gratitude to the people who have been with me on 

this journey.  The success of this study would not have been possible without the support of the 

following people. 

- My parents, for encouraging and supporting me to take this road and achieve all my 

dreams and goals. I am grateful for their patience, motivation and love throughout this 

journey. I am grateful for their confidence and faith in me, as it encouraged me to 

persevere through the most challenging moments in this study.  

- My supervisors, Professor Sune Ferreira-Schenk and Mr Fabian Moodley, for the 

inspiration, guidance, support and advice you have given to me throughout this study. I 

am grateful for their involvement and feedback, and I sincerely appreciate the time and 

work they have put into helping me achieve this milestone. Their knowledge and guidance 

have helped form this study. 

- My siblings, I am truly grateful for your affection, understanding and unwavering support 

that provided me the willpower and commitment to finish my study. I am grateful for how 

you both had my back throughout this journey. 

- To all my friends, thank you for the support and encouragement, and for being my 

cheerleaders throughout this journey. I truly appreciate their support, encouragement, and 

all their words of affirmation. 

- My editor, Dolores Donovan, for her great feedback and suggestions, and her insightful 

comments as I was conducting this study. I appreciate her time and effort in reviewing and 

editing my work and ensuring that my work was up to par. Her insights have been a great 

contribution to refining my study. 

- Most importantly, I want to thank God Almighty, and all my heavenly angels, for blessing 

me with this opportunity. For guiding me, walking with me, and giving me all the strength 

I needed throughout this journey. I am grateful to God for giving me the courage, wisdom 

and drive I needed to finish my research. To God be the glory. 

 

  



v 
 

Abstract 

ETFs have become popular as an investment tool both globally and in South Africa. The 

exponential growth in popularity of ETFs and assets under management over the last three 

decades has solidified ETFs as an essential component of many investors' portfolios. Investors 

with low sentiment often exhibit greater caution and risk aversion, while investors with high 

sentiment typically have a more optimistic view and an eagerness to take on risk in response to 

shifting market conditions. When the economy is growing, high sentiment investors like to buy 

more ETFs to take advantage of the expansion, while low sentiment investors typically sell their 

ETFs on the exchange in order to reduce their exposure during recessions. Furthermore, investor 

behaviour during these times of market volatility may raise questions about the stability of the 

ETF market, which would detract from the attractiveness of ETFs and encourage investors to 

steer away from them.  

The primary objective of this research study is to determine the effect of investor sentiment on 

JSE-listed ETF returns under changing market conditions. The study followed a quantitative 

methodology using secondary data in answering the primary objective. The secondary data was 

collected from the McGregor BFA database and consisted of monthly closing prices of seven JSE 

ETFs and an investor sentiment index. A time series analysis was used for the converging period 

from October 2008 to December 2023. For a more complex understanding of how sentiment 

evolved and influenced market regimes, the Markov regime-switching model was integrated with 

Principal Component Analysis. The results of this research study found that investor sentiment 

had a significant impact on most of the ETFs in both the bull and bear market conditions, and 

both market conditions had a few insignificant impacts on the ETF returns. It was also found that 

the bull market condition was more dominant across the ETF returns. The findings showed that 

investor sentiment does affect the returns of ETFs on the JSE. Identifying the effect of investor 

sentiment on ETFs will result in ETF portfolios being less affected by changing market conditions 

by using risk management techniques and diversifying across asset classes and investing 

methods. 

 

The main contribution of the study lies in the understanding of how investor sentiment interacts 

with ETFs in times of fluctuating market sentiment under changing market conditions. As ETFs 

become more popular among investors because of their ease of use, affordability, and liquidity, 

this field of study becomes more important. It is important to understand how investor sentiment 

affects the volatility of ETFs since changes in sentiment can magnify market movement and 

increase ETF return volatility. As ETF products continue to offer diversification, and more 

investors engage in the market, these insights are important for the management of risk in the 

ETF market. The study underlines the factors behind the heightened sensitivity of ETFs to 



vi 
 

sentiment-driven behaviours, which are usually more apparent during times of volatile markets. 

The immediate fluctuation in prices of ETFs can be caused by investor sentiment, which can 

cause significant inflows and outflows of investments during periods of increased uncertainty or 

abrupt market changes. Therefore, it is essential to comprehend these sentiment-driven 

fluctuations in order to forecast possible risks related to ETFs in different market situations. 

To the author’s knowledge, no other research study has examined how investor sentiment affects 

ETF returns in the South African context. This study has added a significant component to the 

behavioural finance literature in South Africa. By examining the effect of investor sentiment on 

ETF performance, this study significantly adds to the knowledge of how behavioural factors affect 

investments in the South African market. The limitations of this study, as well as the introduction 

and background, open the door for further improvements of future research, and future studies 

can expand the dataset to include both international and domestic ETFs. Given that this research 

used an indirect measure, market-based metrics, future studies can incorporate different investor 

sentiment indexes or a different method to calculate investor sentiment, such as direct 

approaches like survey measures, or a more modern strategy like gathering information from 

online sources. 

Keywords: Bear market, bull market, EFTs, investor sentiment, JSE. 

JEL classification: G11; G14; G45 

  



vii 
 

 

Table of Contents 

Declaration ................................................................................................................................... i 

Declaration by Language Editor ............................................................................................... ii 

List of Publications ................................................................................................................... iii 

Acknowledgements ................................................................................................................... iv 

Abstract v 

Acronyms and Abbreviations ................................................................................................. xii 

List of Figures .......................................................................................................................... xiii 

List of Tables ........................................................................................................................... xiv 

Chapter 1: Introduction .............................................................................................................. 1 

1.1 Introduction and Background to the Study ................................................................. 1 

1.2 Problem Statement ......................................................................................................... 3 

1.3 Theoretical Perspective and Framework ...................................................................... 4 

1.3.1 Efficient market hypothesis ........................................................................................... 4 

1.3.2 Behavioural finance ....................................................................................................... 5 

1.3.3 Adaptive market hypothesis .......................................................................................... 6 

1.4 Research Objectives ...................................................................................................... 7 

1.4.1 Primary objective ............................................................................................................ 7 

1.4.2 Empirical and Theoretical objectives ........................................................................... 7 

1.5 Research Design and Methodology .............................................................................. 8 

1.6 Econometric Methods of Analysis .............................................................................. 10 

1.7 Ethical Considerations ................................................................................................. 11 

1.8 Contribution of the Study ............................................................................................ 11 

1.9 Chapter Classification .................................................................................................. 12 

Chapter 2:  Sentiment-Driven Trading and its Effects on Exchange-Traded Funds (ETFs)

 ................................................................................................................................... 13 

2.1 Introduction ................................................................................................................... 13 

2.2 Investor Sentiment ....................................................................................................... 13 

2.3 Financial Theories Relating to Investor Sentiment ................................................... 16 



viii 
 

2.3.1 Efficient market hypothesis ......................................................................................... 16 

2.3.2  Adaptive market hypothesis ........................................................................................ 17 

2.3.3 Behavioural finance ..................................................................................................... 18 

2.3.3.1  Herding behaviour ................................................................................................... 18 

2.3.3.2  Cognitive biases ....................................................................................................... 19 

2.4 Factors that Influence Investor Sentiment ................................................................. 21 

2.4.1 Economic factors ......................................................................................................... 22 

2.4.2 Political factors ............................................................................................................. 22 

2.4.3 Social factors ................................................................................................................ 22 

2.4.4 Market integration ........................................................................................................ 22 

2.4.5 Cultural aspects ............................................................................................................ 23 

2.5 How Investor Sentiment is Calculated ....................................................................... 23 

2.6 Investor Sentiment and the ETFs Market ................................................................... 25 

2.7 Exchange-Traded Funds (ETFs) ................................................................................. 29 

2.7.1 Definition of ETFs ......................................................................................................... 29 

2.7.2 Features of ETFs .......................................................................................................... 29 

2.7.3 Types of ETFs traded on the JSE ............................................................................... 30 

2.7.4 Popularity of ETFs ........................................................................................................ 31 

2.7.5 Other factors affecting the performance of ETFs ...................................................... 34 

2.7.5.1  Underlying assets of the ETFs ................................................................................ 34 

2.7.5.2  Market liquidity of the ETFs .................................................................................... 35 

2.7.5.3  Tracking error ........................................................................................................... 37 

2.7.5.4  Macroeconomic conditions that affect ETFs ......................................................... 38 

2.8 Rising Importance of Investor Sentiment in the Stock Markets............................... 39 

2.9 Conclusion .................................................................................................................... 39 

Chapter 3: Research Design and Methodology ..................................................................... 41 

3.1 Introduction ................................................................................................................... 41 

3.2 Research Design and Context ..................................................................................... 41 

3.3 Data Sample .................................................................................................................. 41 

3.4 JSE Listed-ETFs and Investor Sentiment Proxies Description ................................ 42 



ix 
 

3.4.1 JSE-listed ETFs ............................................................................................................ 42 

3.4.1.1  Satrix 40 ETF ............................................................................................................ 43 

3.4.1.2  Satrix SWIX Top 40 ETF ........................................................................................... 43 

3.4.1.3  Satrix FINI ETF .......................................................................................................... 44 

3.4.1.4  Satrix INDI ETF ......................................................................................................... 44 

3.4.1.5  Satrix dividend plus ETF ......................................................................................... 44 

3.4.1.6  Satrix RAFI 40 ETF ................................................................................................... 45 

3.4.1.7  FNB Top40 ETF ........................................................................................................ 45 

3.4.2 Investor sentiment proxies .......................................................................................... 45 

3.4.2.1  Share turnover .......................................................................................................... 46 

3.4.2.2  Equity issue ratio ..................................................................................................... 47 

3.4.2.3  Advance/Decline ratio .............................................................................................. 47 

3.4.2.4  Bid-ask spreads ....................................................................................................... 48 

3.4.2.5  Term structure of interest rates .............................................................................. 48 

3.5 Data Transformation on the EFT Returns .................................................................. 49 

3.6 Empirical Models .......................................................................................................... 50 

3.6.1 Principal Component Analysis .................................................................................... 50 

3.6.1.1  Advantages and disadvantages of PCA ................................................................ 50 

3.6.2 Markov Regime-Switching Model ............................................................................... 52 

3.6.2.1  Advantages and disadvantages of the Markov Regime-Switching model ......... 53 

3.6.3 Empirical model specification ..................................................................................... 53 

3.7 Preliminary and Diagnostic Tests ............................................................................... 55 

3.7.1 Preliminary tests ........................................................................................................... 56 

3.7.1.1  Multicollinearity ........................................................................................................ 56 

3.7.1.2  BDS test of nonlinearity .......................................................................................... 56 

3.7.1.3  Unit root and stationary tests ................................................................................. 57 

3.7.1.4  Correlation analysis ................................................................................................. 59 

3.7.1.5  Diagnostic tests ....................................................................................................... 59 

3.7.1.6  Normality test ........................................................................................................... 59 

3.8 Hypothesis of Preliminary and Diagnostic Tests ...................................................... 61 



x 
 

3.9 Conclusion .................................................................................................................... 61 

Chapter 4:  Results and Discussion ....................................................................................... 63 

4.1 Introduction ................................................................................................................... 63 

4.2 Preliminary Tests .......................................................................................................... 63 

4.2.1 Descriptive statistics and investor sentiment ........................................................... 63 

4.2.2 Multicollinearity test ..................................................................................................... 66 

4.2.3 BDS test ......................................................................................................................... 66 

4.2.4 Unit root and stationarity tests ................................................................................... 67 

4.2.5 Correlation analysis results ........................................................................................ 69 

4.3 Empirical Model Results .............................................................................................. 69 

4.3.1 Principal Component Analysis .................................................................................... 70 

4.3.2 Markov Regime-Switching Model ............................................................................... 72 

4.3.2.1  Markov Regime-Switching Model results .............................................................. 72 

4.3.2.2  Transition Probabilities and Constant Duration results ....................................... 80 

4.3.2.3  Smooth Regime Probabilities results .................................................................... 83 

4.4 Discussion of the Findings .......................................................................................... 91 

4.5 Conclusion .................................................................................................................... 95 

Chapter 5:  Conclusion, Limitations and Recommendations .............................................. 97 

5.1 Introduction ................................................................................................................... 97 

5.2 Summary of the Research Study ................................................................................ 97 

5.3 Main Findings of the Study .......................................................................................... 99 

5.3.1 Objective 1 .................................................................................................................... 99 

5.3.2 Objective 2 .................................................................................................................... 99 

5.3.3 Objective 3 .................................................................................................................. 100 

5.4 Conclusion .................................................................................................................. 100 

5.5 Main Contribution of the Study ................................................................................. 101 

5.6 Limitations and Recommendations .......................................................................... 102 

5.6.1 Limitations .................................................................................................................. 102 

5.6.2 Recommendations ..................................................................................................... 102 

LIST OF REFERENCES .......................................................................................................... 105 



xi 
 

APPENDICES .......................................................................................................................... 130 

Appendix A: EViews Markov Regime-Switching Output .................................................... 130 

Appendix 1: SATRIX_Top_40 ................................................................................................ 130 

Appendix 2: SATRIX SWIX Top 40 ........................................................................................ 130 

Appendix 3: SATRIX_FINI ...................................................................................................... 131 

Appendix 4: SATRIC_INDI ..................................................................................................... 132 

Appendix 5: FNB-TOP40 ........................................................................................................ 132 

Appendix 6: SATRIX RAFI ..................................................................................................... 133 

Appendix 7: SATRIX-DIVI ...................................................................................................... 134 

Appendix B: EViews Unit Root and Stationarity Test Results ........................................... 134 

Appendix 1: SATRIX_TOP40 ................................................................................................. 134 

Appendix 2: SWIX Top 40: ..................................................................................................... 136 

Appendix 3: Satrix FINI .......................................................................................................... 138 

Appendix 4: Satrix INDI .......................................................................................................... 140 

Appendix 5: Satrix RAFI ........................................................................................................ 142 

Appendix 6: Satrix DIVI .......................................................................................................... 144 

Appendix 7: FNB Top 40 ........................................................................................................ 146 

Appendix 8: Investor Sentiment ........................................................................................... 148 

Appendix C: Eviews Principal Component Analysis Results ............................................ 151 

Appendix D: Ethics Approval Form ...................................................................................... 152 

 

  



xii 
 

Acronyms and Abbreviations 

 

ALSI  All Share Index  

AMH  Adaptive Market Hypothesis  

APs  Authorised Participants 

BCI  Business Confidence Index  

BF  Behavioural Finance 

CCI  Consumer Confidence Index  

CEEF  Closed-end Equity Fund  

CEFD   Closed-end Fund Discount 

EFTs  Exchange-Traded Funds  

EMH  Efficient Market Hypothesis 

FMI  Financial Market Integration  

Forex  Foreign Exchange Market  

ICI  Investor Confidence Index  

JSE  Johannesburg Stock Exchange  

MCSI  Michigan Consumer Sentiment Index  

MICI  Momentum Investor Confidence Index  

MLA  Myopic Loss Aversion  

NAV  Net Asset Value 

NEWEUR Newwave Euro Exchange Traded Note  

NEWGBP Newwave Pound Sterling Exchange Traded Note  

NEWUSD  Newwave US Dollar Exchange Traded Note 

PCA  Principal Component Analysis  

PCR  Put-Call Ratio  

RAFI  Research Affiliates Fundamental Indexation 

STIR  Short-Term Interest Rate  

SWIX   Shareholder Weighted Index 

VIX  Volatility Index  

 

Keywords: Bear market, bull market, EFTs, investor sentiment JSE. 

 

  



xiii 
 

List of Figures 

Figure 1.1:  Methodological research framework ....................................................................... 11 

Figure 4.1:  Satrix Top 40 smooth regime probabilities .............................................................. 84 

Figure 4.2:  Satrix SWIX Top 40 smooth regime probabilities .................................................... 85 

Figure 4.3:  Satrix FINI smooth regime probabilities .................................................................. 86 

Figure 4.4:  Satrix INDI smooth regime probabilities .................................................................. 87 

Figure 4.5:  FNB Top 40 smooth regime probabilities ................................................................ 88 

Figure 4.6:  Satrix Rafi smooth regime probabilities ................................................................... 89 

Figure 4.7:  Satrix Divi smooth regime probabilities ................................................................... 90 

 

 

  



xiv 
 

List of Tables 

Table 1.1:  Summary of the research methods ............................................................................ 9 

Table 4.1:  Results of descriptive statistics of investor sentiment and ETFs .............................. 64 

Table 4.2:  Results of the Variance Inflation Test ....................................................................... 66 

Table 4.3:  Results of the BDS Test:  Dimension 2 .................................................................... 66 

Table 4.4:  Results for unit root test for the JSE ETFs returns ................................................... 67 

Table 4.5:  Results for unit root test for investor sentiment ........................................................ 68 

Table 4.6: Correlation analysis of the relationship between investor sentiment and JSE ETF 

returns ........................................................................................................................................ 69 

Table 4.7:  Results of the Principal Component Analysis ........................................................... 70 

Table 4.8:  Satrix Top 40 Markov Regime-Switching Model ...................................................... 72 

Table 4.9:  Satrix Swix 40 Markov Regime-Switching Model ..................................................... 73 

Table 4.10:  Satrix FINI Markov Regime-Switching Model ......................................................... 75 

Table 4.11:  Satrix INDI Markov Regime-Switching Model ........................................................ 76 

Table 4.12:  FNB Markov Regime-Switching Model ................................................................... 77 

Table 4.13:  Satrix Rafi Markov Regime-Switching Model ......................................................... 78 

Table 4.14:  Satrix Divi Markov Regime-Switching Model .......................................................... 79 

Table 4.15:  JSE ETF Transition Probabilities and Constant Duration ....................................... 80 

Table 4.16:  Summary of findings ............................................................................................... 91 

 

 



1 
 

Chapter 1: Introduction 
 

1.1 Introduction and Background to the Study 

Exchange-traded funds (ETFs) have gained popularity as an investment instrument since the mid-

1990s due to their overnight liquidity, low transaction costs, and low expense ratios (Ben-David et 

al., 2017). ETFs are an expanding asset class, which is essentially a basket of stocks exchanged 

on an exchange, and many investors now find indexing to be more affordable and straightforward 

(Da & Shive, 2018). ETFs allow trades to be made at any time during the trading day rather than 

only at the day's closing price (Tsalikis & Papadopoulos, 2019). The ability to access the market 

constantly and at a low cost of trading is what differentiates ETFs for investors, and as a result, 

ETFs might draw higher levels of high-frequency demand than other institutional portfolios, such 

as conventional index funds (Ben-David et al., 2018).The main advantages of ETFs over their 

closest competitors, such as conventional index funds, are low transaction costs, great liquidity, 

and the growing demand for passive investments (Liebi, 2020). The exponential growth in 

popularity and assets under management over the last three decades has solidified ETFs as an 

integral component of many investors' portfolios. The most significant factors for the explosive 

performance of ETFs in the last few years have been their transparency, low management costs, 

tax efficiency, and diversity (Tsalikis & Papadopoulos, 2019). These benefits are what make ETFs 

attractive to many investors, and the high demand for ETFs can result in high return volatility in 

the market, which can ultimately be influenced by investor sentiment. 

A significant factor influencing global financial markets is investor sentiment, which is often 

characterised by the attitude or mood of market participants as a whole. Investor sentiment can 

be defined as the perception of risk and return that is not supported by reality (Wang et al., 2021). 

Corredor et al. (2015), and Baker and Wurgler (2006, 2007) characterised it as optimism (high 

sentiment) or pessimism (low sentiment) toward equities generally, yet they also associate it with 

an affinity for speculation. Investors with low sentiment often exhibit greater caution and risk 

aversion, while investors with high sentiment typically have a more optimistic view and an 

eagerness to take on risk in response to the shifting market conditions. The most optimistic 

opinions about many stocks tend to be excessively exaggerated during times of high market 

sentiment, resulting in many stocks being overpriced (Shen et al., 2017). However, Shen et al. 

(2017) further stated that the most optimistic opinions about many stocks tend to come from 

rational investors during low-sentiment times, therefore, mispricing is less likely during those times. 

There are internal factors that influence investor sentiment levels, including expected returns, risk 

and valuation, overconfidence, and the influence of "noise" traders (Adamson, 2017). An irrational 

person who makes trading decisions based on inaccurate or incomplete information is known as 

a noise trader (Collins, 2020).  



2 
 

Moreover, investor sentiment in financial markets can be impacted by concerns about the political, 

economic, and financial landscape (Kunjal, 2022). Sentiment changes will result in more noise 

trading, more mispricing, and excess volatility if ignorant noise traders base their trading choices 

on sentiment and risk-averse arbitrageurs will be restricted in their ability to engage in arbitrage 

(Da et al., 2015). According to the liquidity trading hypothesis, nonfundamental disturbances in the 

ETF sector could cause the underlying assets to become more volatile (Liebi, 2020). Although it 

could also be the result of other processes, a causal relationship between ETF flows and the price-

to-fundamentals relationship is consistent with nonfundamental demand shocks driving prices 

away from fundamental values (Zou, 2019). Investor sentiment can be considered a 

nonfundamental shock, as this sentiment may be caused by market information or social trends 

rather than fundamental changes. Investors who make irrational ETF investment decisions may 

allow their emotions to take dominance over reason, which could lead to excessive trading 

aggression, poor market timing, or inaccurate estimates (Kunjal & Peerbhai, 2021). Kunjal and 

Peerbhai (2021) further stated that the efficiency of ETF markets may suffer as a result of an 

increase in trading volumes and ETF return volatility. Depending on the current status of the 

market, investor sentiment may have different effects on it, such as asset price fluctuations and 

market volatility. Two additional factors that contribute to the mispricing are arbitrage restrictions 

and uninformed demand shocks, that persist due to investor attitude (Baker & Wurgler, 2006; 

Wang et al., 2022).  

The news effect, whether positive or negative, can intensify or dull the general attitude of the stock 

market, depending on whether it is in a bullish or bearish mood (Hanna et al., 2020). Bull and bear 

conditions refer to extended periods of time during which prices have either extensively climbed 

or declined, respectively, when applied to the entire market, and these intervals naturally form 

alternating phases because they are bordered by peaks and troughs (Hanna, 2018).  ETFs have 

high illiquidity during periods of financial distress, as during this time investor sentiment typically 

becomes negative. When investor mood is negative, the market can still increase during normal 

economic conditions, however, during recessions and moments of economic expansion, the 

markets reflect the prevailing sentiment (Adamson, 2017). During recessions, low-stimulation 

investors tend to sell their ETFs on the exchange, reducing their exposure, while high-stimulation 

investors increase their exposure to Johannesburg Stock Exchange (JSE) ETFs during economic 

expansions to benefit from the growth. Furthermore, investor behaviour during these periods of 

market instability may cast doubt on the ETF market's stability, making ETFs less attractive and 

prompting investors to move away from them. Therefore, this research seeks to delve into 

understanding the influence of investor sentiment on JSE ETFs under changing market conditions.  

Previously, financial markets were believed to be rational, meaning that there was perfect 

information and share prices available. In 1936, Keynes argued that investors' "animal spirits" may 



3 
 

cause the market to vary drastically and affect prices in ways that had nothing to do with 

fundamentals (Da et al., 2015). In recent academic research, the focus has changed from looking 

for entirely rational explanations for the behaviour of share prices to looking at the roles that 

psychological biases and investor behaviour contribute (Adamson, 2017). As ETFs become more 

popular, it is important to investigate the reasons behind their volatility in returns to manage risk in 

the ETF market. Although an increasing amount of empirical research has focused on the return 

volatility of ETFs, little has been done on the effect of investor sentiment on ETFs listed on an 

exchange under changing market conditions. Similar academic research has been explored in 

regard to investor sentiment in the South African context. For example, Rupande et al. (2019) 

examined investor sentiment and stock return volatility, Muguto et al. (2022) investigated the 

impact of investor sentiment on sectoral returns and volatility from the Johannesburg Stock 

Exchange, Dalika and Seetharam (2014) performed an analysis on investor sentiment in the South 

African market, and Muguto et al. (2019) examined investor sentiment and foreign financial flows 

in South Africa. Other studies looked into investor behaviour in the South African market; for 

example, Kunjal and Peerbhai (2021) examined investor overconfidence in the South African ETF 

market and investor herding during COVID-19 in the South African ETF market, and Charteris and 

Musadziruma (2017) examined feedback trading in stock index futures in South Africa. 

 

This dissertation seeks to determine how investor sentiment played a significant role in shaping 

market conditions and the performance of ETFs on the JSE that also influenced trading volumes, 

asset prices, and market trends in the South African context. The study aims to examine the effect 

that investor sentiment has on ETF returns listed on the JSE under bull and bear market 

conditions. As ETFs become more popular in South Africa, these insights are important for the 

management of risk in the ETF market and understanding the heightened sensitivity of ETFs to 

sentiment-driven behaviours, which are usually more apparent during times of volatile markets. 

 

1.2 Problem Statement 

When market conditions change, investors with low and high sentiments respond differently, which 

leads to a contradiction in the market, leading to unpredictable markets and increased risk. High 

investor sentiment can fuel bullish behaviour during economic expansions, increasing demand for 

assets and causing prices to rise.  On the other hand, low investor sentiment can fuel bearish 

behaviour during economic recessions, decreasing demand for assets and causing prices to fall.  

These changes in investor sentiment can contribute to changes in trading volumes, market 

liquidity, and price volatility (Da et al., 2015), which affects the efficiency and stability of financial 

markets. A broad range of stocks can be affected by the market-wide component of investor 

sentiment, including ETFs. Markets can quickly absorb changes in investor psychology and 



4 
 

sentiment, which in turn affects investors' risk aversion and portfolio selection without regard to 

cash flow projections or fundamental value measures (Ahmed, 2020).  

There is still not enough reliable empirical data regarding investor sentiment's predictability over 

time on the overall stock market (Mbanga et al., 2019). Little research has been carried out on 

how investor sentiment affects ETF returns in the South African context. However, similar 

academic research has been explored in regard to investor sentiment in South Africa. This study 

highlights the gap in the literature by examining the role sentiment plays in ETF volatility and 

providing a more comprehensive understanding of how sentiment interacts with market conditions 

to affect ETF pricing in the South African context. Investor sentiment is one of the factors that 

influence financial markets. The immediate fluctuation in prices of ETFs can be caused by investor 

sentiment, which can cause significant inflows and outflows of investments during periods of 

increased uncertainty or abrupt market changes. 

 It is essential to comprehend these sentiment-driven fluctuations to forecast possible risk related 

to ETF portfolios of South African investors under changing market conditions. Identifying the 

effect of investor sentiment on ETFs could result in ETF portfolios being less affected by changing 

market conditions by using risk management techniques and diversifying across asset classes 

and investing methods. It is imperative that portfolio managers look out for changes in sentiments, 

as this affects the outcomes of the portfolios of South African investors. Understanding the effect 

of investor sentiment on JSE ETFs is deemed beneficial to investors, policymakers, and market 

participants. It was unclear, as far as the author was aware, exactly how investor sentiment 

interacted with ETFs in times of fluctuating market sentiment in the South African context. Return 

predictability fluctuates over time due to changing market conditions, including bubbles, crashes, 

crises, and institutional factors (Al-Khazali & Mirzaei, 2017). Thus, there was a need for more 

studies and deeper analysis to determine the effect. This helped to understand the relationship 

between investor sentiment and ETFs under changing market conditions in a South African 

context.  

Therefore, the research question guiding this study was: What is the effect of investor sentiment 

on ETF returns under changing market conditions? 

 

1.3 Theoretical Perspective and Framework 

1.3.1   Efficient market hypothesis 

One of the foundations of contemporary finance is the efficient market hypothesis (EMH), which 

is founded on the assumption that all publicly available information is accessible to financial 

investors, stock market participants, and other financial market participants; and therefore, asset 

prices always include and reflect all pertinent information (Spulbar et al., 2021). There are three 



5 
 

distinct forms of market efficiency: the weak form, the semi-strong form, and the strong form of the 

efficient market theory (Yildirim, 2017). Investor sentiment falls into the semi-strong form. 

According to Yildrim (2017), the semi-strong form states that information that is available to the 

public and past prices have an impact on how prices are formed. This suggests that investor 

sentiment would not independently impact stock prices under EMH unless it is based on new, 

relevant data that affects decisions on whether to buy or sell. Additionally, this implies that stock 

prices will rapidly reflect investor optimism and pessimism, making it difficult for investors to 

regularly outperform the market using only sentiment-driven methods. 

According to the EMH, return rates are memoryless, which implies that traders cannot use 

modified negotiating tactics for arbitration to get exceptional returns in the financial markets (Dias 

et al., 2020). Investment decisions are made by investors mostly based on all of the fundamental 

information about the assets, rather than on instincts and other irrational emotions.  This makes 

financial markets more unstable and uncertain. Traditional stock market theories did not view 

investor sentiment as a significant factor in their understanding of market dynamics, which was 

framed by the EMH and random walk theory (PH & Rishad, 2020). Prices fluctuate randomly and 

unexpectedly because they follow the EMH's "random walk" theory; as a result, investors are 

unable to generate greater long-term returns (Mandaci et al., 2019). French mathematician 

Bachelier (1900) established the random walk hypothesis after he provided strong evidence that 

commodities speculation in France was a "fair game" meaning that neither buyers nor sellers could 

anticipate making money (Agwu et al., 2020). EMH suggests that investor sentiment has no effect 

on ETF returns because rational investors did not appear to have been the driving force behind 

increasing the value of assets in relation to the then-current level of projected cash flows (Baker 

& Wurgler, 2007; PH & Rishad, 2020). This occurs because irrational investors' mispricing is 

corrected by rational investors, which helps the mispricing align with its fundamental value. Thus, 

markets are efficient. 

1.3.2 Behavioural finance 

According to the theory of behavioural finance, biases and cognitive errors which affect investors 

and market participants are not entirely rational (Bansal, 2020). Bansal (2020) furthermore stated 

that the primary parts of behavioural finance are psychology, which explains human fallibility, and 

limits to arbitrage, which contends that irrationality may have a long-lasting and substantial effect 

in an economy with both rational and irrational traders. It is an alternative model to the EMH, the 

EMH assumes that all investors receive all news and information at the same time. However, 

behavioural finance asserts that investors make their own judgments about investments based on 

their analysis of news and information (Yildirim, 2017). The reliance of investors on ad hoc 

heuristics renders them irrational at times and leads to cognitive biases that in turn cause markets 

to be inefficient (Moodley, 2020).  



6 
 

Research on behavioural finance has reached a consensus regarding the many abnormalities that 

investors display, which ultimately result in poor judgment and mispricing in financial markets 

(Baker et al., 2021). The behaviour of investors is mirrored in stock prices, and investor psychology 

ultimately shapes market swings, which in turn influences the market (PH & Rishad, 2020). The 

debate over how investor attitude affects asset returns in the integrated stock market was sparked 

by the emergence of behavioural finance theories (PH & Rishad, 2020). Stock market activity is 

influenced by investor attitude, for instance, through market activity asset valuation (Brown & Cliff, 

2005; Angeles Lopez-Cabarcos et al., 2020). Investor sentiment impacts asset values as overly 

optimistic or pessimistic investors push prices above or below fundamental values (Brown & Cliff, 

2005). Therefore, showing that behavioural finance illustrates that ETF returns can be influenced 

by investor sentiment, and that the effect would be linear. Asset prices differ from their underlying 

(fundamental) values due to systematic mispricing in the capital markets, which is caused by 

sentimental indicators and the irrational behaviour of investors (PH & Rishad, 2020). Since 

investors who are influenced by sensation often outweigh rational ones, mispricing in the market 

remains uncorrected, and price alignment with their fundamentals becomes difficult in a 

competitive financial market by investors' persistent use of non-fundamental information to 

outperform the market and achieve excess returns. Thus, financial markets are not efficient dur to 

irrational behaviour. 

As ETFs gain popularity the market for them has expanded, and their prices are subject to 

fluctuations based on investor sentiment. The topic of behavioural finance is a developing field 

that is expanding and encompassing various domains, and empirical evidence supports the 

importance of investor mood within it (Angeles Lopez-Cabarcos et al., 2020). According to the 

EMH, every investor is rational, and securities are valued rationally, but in actuality, however, 

investors' emotions, biases, and incentives influence how they make decisions (Mandaci et al, 

2019). Recent studies show that investor sentiment affects market returns, including ETF returns 

(Kadiyala, 2020; Lee & Chen, 2020; Zi-Long, 2021; Kunjal & Peerbhai, 2021; Lee et al., 2021; 

Phan et al., 2023). Nevertheless, there is still some debate where some scholars show that 

investor sentiment has no effect on market returns (

Andrianto & Mirza, 2016). This shows that the latter supports EMH and the former supports 

behavioural finance. 

1.3.3 Adaptive market hypothesis 

According to Lo (2004) and Al-Khazali and Mirzaei (2017), the adaptive market hypothesis (AMH) 

applies the concepts of natural selection, adaptation, and competition from evolution to financial 

interactions to balance behavioural options with market efficiency. According to the AMH, markets 

respond to events and structural changes and change over time, with varying degrees of market 

efficiency (Khuntia & Pattanayak, 2018). The AMH attempts to reconcile the EMH and behavioural 



7 
 

finance by stating that the alternating efficiency of financial markets are due to shifts in market 

conditions. Return predictability, which can fluctuate over time due to changing market conditions, 

such as bubbles, crashes, crises, and institutional factors, is a crucial implication of AMH (Al-

Khazali & Mirzaei, 2017). According to the AMH, which was seen to be contrary to the EMH, 

arbitrage possibilities occasionally occur because of market trends, panics, bubbles, and crashes; 

however, market timing is crucial in order to seize these profit chances (Mandaci et al., 2019). 

Hence, investor sentiment should have a varying effect on ETF returns in a bull market condition 

and in a bear market condition, and that effect will not be the same under each condition. 

 

Lo asserts that market efficiency fluctuates with time and market conditions in the adaptive market 

hypothesis theory (Li et al., 2021). Investor sentiment changes in different market conditions, some 

investors are either optimistic or pessimistic. According to the AMH, despite acting in their own 

self-interest, investors frequently make mistakes and then learn from those mistakes and modify 

their behaviour (Mandici et al., 2019). It also stated that investing opportunities change with time, 

and different business situations have an impact on investment techniques (Khursheed et al., 

2020). From a revolutionary point of view, there were prospects for profit as long as there were 

liquid markets (Moodley, 2020). Mandaci et al. (2019) furthermore stated that the AMH contends 

that the risk/reward relationship is shaped by the stock market environment and investor 

demographics and that the risk premium fluctuates over time. AMH stated that the size and 

preferences of market participants in a particular capital market dictated the relationship between 

risk and reward and that if these characteristics changed over time, the relationship between risk 

and reward was impacted (Lo, 2004). Nassirtoussi et al. (2014) stated that, consequently, there is 

a high nonlinear dependence throughout, although the linear dependence of stock returns 

fluctuates over time. The shift from linear to non-linear has not yet reached finality, and hence, this 

study is needed.  

 

1.4  Research Objectives 

1.4.1 Primary objective 

The primary objective of this research study was to determine the effect of investor sentiment on 

JSE ETFs returns under changing market conditions. 

1.4.2 Empirical and Theoretical objectives 

To facilitate the achievement of the primary objective, the following empirical and theoretical 

objectives were formulated: 

A.) Empirical objectives 

 Compare the effect investor sentiment has on JSE-listed ETF returns in a bull regime; 

 Determine the effect of investor sentiment on JSE-listed ETF returns in a bear regime; 



8 
 

 Compare the levels of bull and bear market conditions across JSE-listed ETFs 

B.) Theoretical objectives 

 To determine the relative strengths of bull and bear markets; 

 Investigate investor sentiment, under financial theories such as EMH, BF, and AMH; 

   To determine the relationship between ETFs and the effect of investor sentiment. 

 

1.5 Research Design and Methodology 

This study followed a quantitative method using secondary data. A quantitative method is a method 

that deals with numbers and measurable forms; it uses a systematic way of investigating events 

or data. The reason for using this method was to achieve the objective of determining the effect of 

investor sentiment on JSE ETFs under changing market conditions, and this method answered 

those questions to justify relationships with measurable variables to explain, predict, or control a 

phenomenon. The approach was appropriate for the study as it used quantitative data (JSE ETFs’ 

returns) and an empirical model (Markov regime-switching model) to examine the effect of investor 

sentiment on JSE ETFs under shifting market conditions. The approach was also beneficial as 

quantitative methods use a larger sample and do not take as long to collect data (Rahman, 2016). 

It was necessary since the bear and bull markets occurred over a long period and at various times 

(Moodley, 2020). Sentiments were measured using the Principal Component Analysis (PCA), 

which is a statistical technique that uses an orthogonal transformation to transform correlated 

observations into linearly uncorrelated values (Karamizadeh et al., 2020). Furthermore, for a more 

complex understanding of how sentiment evolved and influenced market regimes, the Markov 

regime model was integrated with the PCA. The monthly prices were transformed into monthly 

returns. Analysing the returns of ETFs assisted in providing valuable insight into the effect of 

investor sentiment. The price returns were used as this study is analysing the ETF price movement 

influenced by investor sentiment.



9
 

 T
ab

le
 1

.1
: 

 S
u

m
m

a
ry

 o
f 

th
e

 r
e

s
e

a
rc

h
 m

e
th

o
d

s

R
e

s
e

a
rc

h
 M

e
th

o
d

 
D

e
s

c
ri

p
ti

o
n

 
A

d
v

a
n

ta
g

e
s

 
D

is
a

d
v

a
n

ta
g

e
s 

E
x

a
m

p
le

 

P
ri

n
ci

p
a

l c
o

m
p

o
n

e
n

t 
a

n
a

ly
si

s 
(P

C
A

)
P

C
A

 is
 u

se
d

 t
o 

co
n

ve
rt

 
co

m
p

le
x 

sp
e

ct
ra

l d
a

ta
se

ts
 

in
to

 u
n

d
e

rs
ta

n
d

a
b

le
 

in
fo

rm
a

tio
n

 b
y 

re
co

g
n

is
in

g
 

re
cu

rr
e

n
t 

p
a

tt
e

rn
s 

in
 t

h
e

 
d

a
ta

 w
h

ile
 m

in
im

is
in

g
in

fo
rm

a
tio

n
 lo

ss
 (

B
e

a
tt

ie
 &

 
W

h
ite

, 
2

0
2

1
).

 

It
 h

a
s 

th
e

 a
dv

a
n

ta
g

e
 o

f 
b

e
in

g
 s

im
p

le
 a

n
d

 c
o

n
ci

se
 

in
 it

s 
p

re
se

n
ta

tio
n

 o
f 

a
 

d
a

ta
se

t,
 w

h
ic

h
 m

a
ke

s 
it 

p
o

w
e

rf
u

l a
n

d
 v

e
rs

a
til

e
 

(B
ea

tt
ie

 &
 W

h
ite

, 
2

0
2

1
).

 

A
 d

ra
w

b
a

ck
 w

ith
 P

C
A

 is
 

th
a

t 
us

in
g

 h
ig

h
-

d
im

e
n

si
o

n
a

l d
a

ta
 a

s 
p

ro
je

ct
io

n
 m

e
th

o
d

 in
p

u
t 

ca
n

 r
e

su
lt 

in
 s

ta
tis

tic
a

lly
 

si
gn

ifi
ca

n
t 

is
su

e
s 

(G
e

w
e

rs
 

e
t 

a
l.,

 2
0

2
1

).
 

P
C

A
 c

an
 b

e
 u

se
d

 t
o

 
co

m
p

a
re

 t
h

e
 r

e
tu

rn
s 

o
f 

n
u

m
e

ro
u

s 
e

q
u

iti
e

s 
a

n
d

 
u

n
co

ve
r 

u
n

de
rl

yi
n

g
 t

re
n

d
s 

o
r 

g
ro

u
p

in
g

s 
o

f 
st

o
ck

s 
th

a
t 

m
o

ve
 t

o
g

e
th

e
r.

 

M
a

rk
o

v 
st

a
te

 s
w

itc
h

in
g

 
m

o
d

e
ls

 
T

h
e

 M
a

rk
o

v 
R

e
g

im
e

-
S

w
itc

h
in

g
 m

o
d

e
l i

s 
us

e
d

 t
o

 
id

e
n

tif
y 

d
iff

er
e

nt
 s

ta
te

s 
o

r 
re

g
im

e
s 

in
 t

im
e

 s
e

ri
e

s 
d

a
ta

.  

It
 b

e
ca

m
e

 p
o

ss
ib

le
 to

 
o

b
se

rv
e

 c
om

p
le

x 
d

yn
a

m
ic

 
p

a
tt

e
rn

s 
b

e
ca

u
se

 t
h

e
 

m
o

d
e

l h
a

d
 m

a
n

y 
e

q
u

a
tio

n
s 

th
a

t 
so

rt
e

d
 t

h
e

 
tim

e
 s

er
ie

s 
b

e
h

a
vi

o
u

r 
in

to
 

d
iff

er
e

n
t 

re
g

im
e

s 
(M

o
o

d
le

y,
 2

0
2

0
).

 

T
h

e
 v

a
ri

a
b

le
s 

o
r 

fu
n

ct
io

n
a

l 
sp

e
ci

fic
a

tio
ns

 t
h

a
t 

ca
n

 b
e

 
u

se
d

 t
o

 d
e

sc
ri

b
e

 t
h

e
 

d
yn

a
m

ic
s 

in
 t

h
e

 t
ra

n
si

tio
n

 
p

ro
b

a
b

ili
tie

s 
a

re
 n

o
t 

a
lw

a
ys

 s
tr

a
ig

h
tf

o
rw

a
rd

 
(B

a
zz

i e
t 

a
l.,

 2
0

1
7

).
 

M
o

d
e

lli
n

g
 t

h
e

 le
n

g
th

 o
f 

u
n

e
m

p
lo

ym
e

n
t 

d
u

ri
n

g
 

e
co

n
o

m
ic

 b
o

o
m

s 
a

n
d

 
b

u
st

s 
u

si
n

g
 d

iff
e

re
n

t 
ri

sk
 

ra
te

s.
 

C
o

n
st

a
n

t 
tr

a
n

si
tio

n
 

p
ro

b
a

b
ili

tie
s 

C
o

n
st

a
n

t 
tr

a
n

si
tio

n
 

p
ro

b
a

b
ili

tie
s 

ch
a

ra
ct

e
ri

se
 

a
 s

itu
a

tio
n

 in
 w

h
ic

h
 t

h
e

 
p

ro
b

a
b

ili
ty

 o
f 

sh
ift

in
g

 f
ro

m
 

o
n

e
 r

e
g

im
e

 (
st

a
te

) 
to

 t
he

 
n

e
xt

 r
e

m
a

in
s 

co
n

st
a

n
t 

a
cr

o
ss

 t
im

e
.

C
o

n
st

a
n

t 
p

ro
b

a
b

ili
tie

s 
a

re
 

e
a

sy
 t

o
 c

o
m

p
re

h
e

n
d

, 
e

va
lu

a
te

, 
a

nd
 a

p
p

ly
 in

 
m

o
d

e
ls

. 

C
o

n
st

a
n

t 
tr

a
n

si
tio

n
 

p
ro

b
a

b
ili

tie
s 

d
o

 n
o

t 
a

cc
o

u
n

t 
fo

r 
th

e
 im

p
a

ct
 o

f 
ch

a
n

g
in

g
 e

xt
e

rn
a

l 
va

ri
a

b
le

s 
o

r 
in

te
rn

a
l 

d
yn

a
m

ic
s 

on
 t

h
e

 
p

ro
b

a
b

ili
ty

 o
f 

sw
itc

h
in

g
 

re
g

im
e

s
(G

o
m

e
s 

e
t 

a
l.,

2
0

2
4

).
 

S
w

itc
h

e
s 

in
 le

ve
l o

f 
tim

e
 o

r 
ve

ct
o

r 
tim

e
 s

e
ri

e
s,

 
vo

la
til

iti
e

s,
 o

r 
co

rr
e

la
tio

n
 

(B
a

zz
i e

t 
a

l.,
 2

0
1

7
) 

S
ou

rc
e:

 A
u

th
o

r’s
 o

w
n 

co
m

p
ila

tio
n

 (
20

24
).

 



10 
 

1.6 Econometric Methods of Analysis  

This study used preliminary and diagnostic tests such as stationarity tests without structural 

breaks and unit roots tests, correlation tests, and normality tests, as discussed below. 

Correlation test analysis 

The correlation test was used to look at the strength and direction of the association between 

the independent and dependent variables. The correlation coefficient has a range of -1 to +1. 

A positive correlation between two variables suggested that as one variable increased, so did 

the other variable; this implied that the variables moved in the same direction. On the other 

hand, a negative correlation suggested that as one variable increased, the other one 

decreased; this implied that the variables moved in opposite directions. 

Stationarity tests 

Unit root and stationarity tests prevented the formation of autocorrelation and spurious 

regressions. If the statistical value> test critical levels and the probability value <0.05 were 

both shown by the root test results, then the data were considered stationary in the average 

(Djakasaputra et al., 2020). These tests were essential for verifying the stationarity of data 

series, which was a prerequisite for numerous econometric models to produce precise and 

significant results.  

Normality tests 

Normality tests were used to determine the type of the distribution in the data. The primary 

goal was to analyse whether the goodness of fit of a statistical model accurately captured how 

well it fitted a collection of data - a measure of goodness of fit of a model usually provides an 

overview of the difference between its observed and expected values (Khatun, 2021).  

 



11 
 

Step 1: Price Returns of ETFs 

Step 2: Econometric Methods of Analysis

Step 3: Principal Component Analysis 

Step 4: Markov Switching Regime

 

Step 5: Constant Transition Probabilities

Figure 1.1:  Methodological research framework 

                                     Source: Author’s own compilation (2024).  

 

1.7 Ethical Considerations 

This research study was in accordance with the ethical guidelines and principles of the North-

West University (NWU, 2016:15). The significance of the ethical concerns surrounding the 

conduct of the research is paramount, even beyond the significance of choosing a suitable 

research technique and procedures (Fleming & Zegwaard, 2018). Ethical considerations are 

crucial since they affect not just primary research in particular, but also the use of secondary 

data sets because they have to do with impartial and fair source and analysis choices 

(Rahman, 2016). The researcher will ensure that the strategies, techniques, and assumptions 

employed in the study are ethical and that they are obtained and discussed ethically. By 

following ethical rules and principles, researchers can make it morally right with regard to their 

research study. 

 
1.8 Contribution of the Study

As ETFs become more popular, it is important to investigate the reasons behind their volatility 

in return to manage risk in the ETF market. It is evident from the review of the empirical 

 

 



12 
 

literature that the majority of international and domestic studies focus on the risk and 

macroeconomic determinants of ETFs. This study introduces the aspect of investor sentiment, 

which, as far as the author is aware, has never been done in South Africa. Studies focus on 

the linear relationship between the determinants of ETFs. However, this study introduces the 

aspect of changing market conditions, which has been supported by AMH and is yet to be 

done in South Africa. 

This study aimed to contribute significantly to academia and the financial market by examining 

the effect of investor sentiment on JSE-listed ETFs under bull and bear market conditions. 

1.9 Chapter Classification 

Chapter 1: Introduction and background to the study: This chapter introduced the topic of 

this research paper, as well as background information to provide the idea and direction of the 

study, which was the effect of investor sentiment on JSE ETFs under changing market 

conditions. 

Chapter 2: Literature review: This chapter provided the theoretical background of the effect 

of investor sentiment on JSE ETFs under changing market conditions, and also provided 

insight into previous studies that have been conducted on this topic.  

Chapter 3: Research design and methodology: This chapter included the quantitative 

method used in this research paper using secondary data collected from the McGregor BFA 

database. 

Chapter 4: Results and findings: The research and findings were discussed once the 

statistical analyses had been employed, and the results from the tests and models were 

presented and discussed.  

Chapter 5: Conclusions, limitations, and recommendations: This chapter included the 

research conclusions, together with the limitations and recommendations. 

 

 



13 
 

Chapter 2:  Sentiment-Driven Trading and its Effects on Exchange-Traded 

Funds (ETFs)

2.1 Introduction 

Exchange Traded Funds (ETFs) were among the few investments that evaded the 2007/2008 

financial crisis, contributing to their popularity among investors (Goltz & Schröder, 2011), and 

have become a significant part of many investors' portfolios due to their exponential rise in 

popularity and assets under management over the past three decades. The popularity of ETFs 

has raised questions about how they can affect market efficiency in general because of their 

simplicity and low cost of trading, which appeal to individual investors who are more likely to 

follow trends (Chau et al., 2011). Investor sentiment has a substantial impact on global 

financial markets (Wang et al., 2021). Studies have indicated that the attitude of investors 

affects asset values, which in turn affects market patterns; and this thesis delves into 

examining the influence of sentiment in the ETF market. By analysing the part sentiment plays 

in ETF volatility and offering a more thorough comprehension of how sentiment interacts with 

market conditions to impact ETF pricing, this study fills a gap in the research in the South 

African context. 

The main aim of this Chapter focuses on the comprehensive research on the effect of investor 

sentiment on ETFs returns under changing market conditions and provides findings from other 

research studies. To achieve the primary goal and address the gap of the study, the following 

theoretical objectives were formulated: (i) To determine the relative strengths of bull and bear 

markets; (ii) To investigate investor sentiment, under financial theories such as EMH, BF, and 

AMH; and (iii) To determine the relationship between ETFs and the effect of investor 

sentiment. The Chapter is as follows, Section 2.2 provides an overview of investor sentiment, 

followed by Section 2.3 that provides theories relating to investor sentiment. Section 2.4 

highlights the factors that influence investor sentiment, and Section 2.5 shows how investor 

sentiment is calculated. This is followed by Section 2.6 which explains investor sentiment and 

the ETFs market. Subsequently, Section 2.7 provides an overview of ETFs, followed by 

Section 2.8 explaining the importance of investor sentiment in stock markets, and lastly, 

Section 2.9 with the conclusion of the Chapter 

 

2.2 Investor Sentiment 

An investor's general perspective of the financial markets or asset classes is normally 

expressed in their sentiment (Shen et al., 2017). Investor sentiment refers to the level of 

optimism or pessimism that an investor has about the future performance of the stock market 

(Angeles Lopez-Cabarcos et al., 2020). The bias in the discrepancy between an asset's 



14 
 

present price and its sustained price based on fundamentals is also included in the definition

of sentiment (Reis & Pinho, 2021). Chau et al. (2011) stated that it is conceivable that the 

relationship between trading behaviour and investor sentiment could differ depending on the 

market regime. Investors display different behaviours in bullish and bearish regimes.  When 

optimism is dominant, investors typically exhibit bullish behaviour, and when pessimism is 

dominant, investors typically exhibit bearish behaviour. When referring to entire markets, bull 

and bear markets are typically understood to represent long periods of time during which 

prices have generally risen or fallen, respectively (Hanna, 2018). Bull and bear market cycles 

have a substantial effect on the general economy and social welfare in addition to reflecting 

investor confidence and economic growth (Chi et al., 2016). 

The bear market, in contrast to the bull market, is a sequence of intermediate highs and lows 

that are broken up by a sequence of intermediate lows and highs in a long-term downward (as 

opposed to upward) movement (Bejaoui & Karaa, 2016). There is a possibility that crashes 

during extremely bearish market periods will result in asset underpricing and, eventually, a 

quick recovery with positive returns (Andleeb, 2024). Conversely, a bullish investor anticipates 

returns that will exceed a certain average value, and they may be impacted by other investors 

as well as have an impact on them, and bull markets inherently occur when these investors 

control the majority of trading activity (Hanna, 2018). When the market is extremely bullish, 

bubbles may form, which could cause asset overpricing and a rapid correlation with profit 

losses (Andleeb, 2024). During bull markets, a positive shift in investor sentiment increases 

equity returns, but during bear regimes, a negative shift in sentiment has the reverse effect; 

on the other hand, during bull markets, a negative shift in investor sentiment decreases equity 

returns, but during bear regimes, a positive shift in sentiment has the reverse effect (Wang et 

al., 2022).  

Extreme optimism can lead to overinvestment in financial securities, leading to market bubbles 

and exaggerated asset values, whereas extreme pessimism causes investors to sell riskier 

assets and invest in risk-free ones, leading to market crashes and undervaluation of asset 

values (Andleeb, 2024). Reis and Pinho (2021) further highlighted how rising market co-

movements are usually linked to times of crisis and instability marked by sharp swings in the 

financial markets. Furthermore, Shen et al. (2017) stated that in times of high market 

sentiment, stocks often become overvalued due to excessive optimism, while in low-sentiment 

periods, rational investors’ views prevail, reducing the likelihood of mispricing. Kuhnen and 

Knutson (2011) observed that whereas negative sentiment states discourage investors from 

taking on riskier investments, positive sentiment states encourage them to do so. The way 

investors behave has a significant impact on the stock market and the overall market 

performance. When there is too much positive news, investors tend to overreact, which results 



15 
 

in negative returns during the correction; conversely, when there is insufficient good news, 

they underreact (Griffith et al., 2020).  

Investors prefer to monitor and pay more attention to the information provided by preliminary 

positive news, but steer away from subsequent information derived from unfavourable 

preceding news (Karlsson et al., 2009; Mbanga et al., 2019). This is because different types 

of investors exhibit different decision-making behaviours (Phan et al., 2023). Passion 

promotes optimistic risk assessments and risk-seeking decisions, while fear encourages 

pessimistic risk estimates and risk-averse decisions (Griffith et al., 2020). Su et al. (2020) 

highlighted that investors focus more on intuition and other irrational sentiments than on 

fundamental information about assets when making decisions about their investments, which 

makes markets more unpredictable and uncertain. Liu (2015) further highlighted that 

sentiment includes both mood and emotions and can influence decisions unconsciously.  De 

Long et al. (1990) stated that even in the absence of fundamental risk, noisy trader risk can 

cause asset prices to deviate dramatically from their basic values.  

The rational asset pricing theory states that when sentiment is low, there is a broad trade-off 

between mean and variance, implying a positive relationship between variance and the 

anticipated market return over time, and vice versa (Uygur & Tas, 2014). Due to their lack of 

expertise, noise traders are prone to overestimate return variance, which would cause the 

mean-variance relationship to be distorted by their involvement (Yu & Yuan, 2011). During 

times of high and low sentiment, noise traders respond differently to protect their positions. In 

high-sentiment periods, they trade more aggressively and participate more than in low-

sentiment periods (PH & Rishad, 2020). Furthermore, the market is more likely to be rational 

during times of low sentiment because marginal investors are more likely to be rational, and 

pessimistic investors tend to stay out of the market because of short-sale constraints (Shen et 

al., 2017). Wang et al. (2022) stated that high levels of optimism produce high 

contemporaneous returns, and that overpricing would eventually be corrected by the mean-

reverting feature with low later returns. On the other hand, high levels of pessimism produce 

low contemporaneous returns, and under-pricing would eventually be corrected by the mean-

reverting feature with high later returns (Wang et al., 2022). 

Brown and Cliff (2005) and Wang et al. (2021) furthermore highlighted that the mean reversion 

property implies that investor sentiment and future stock returns are negatively correlated, 

indicating that future stock returns might be low or high depending on whether investor 

sentiment drives firm prices above or below basic values. The behaviour of investors is 

mirrored in stock prices, and investor psychology ultimately shapes market fluctuations, which 

in turn influence the market (PH & Rishad, 2020). As shown in the research of Shiller (2000), 

who correctly predicted the 2008 financial crisis, emphasised that investor sentiment can also 



16 
 

influence bull and bear markets (Phan et al., 2023). Financial markets have seen numerous 

manias, panics, and collapses throughout history, none of which can be fully comprehended 

by fundamental research alone (Zhou, 2018). Additionally, more noise in the trading due to 

high sentiment could result in more market liquidity. Liu (2015) stated that higher sentiment 

results in more irrational market makers and a lower price impact, resulting in increased 

market liquidity. Similarly, Ben-David et al. (2012) further show that during times of market 

difficulties, arbitrageurs drastically cut back on their trading activity, which might worsen the 

liquidity and market conditions.  

 

2.3 Financial Theories Relating to Investor Sentiment 

Reis and Pinho (2021) stated that stock prices may become irrational if the share price 

deviates from its theoretical price and sentiment may be a major factor in supporting that 

difference. The efficient market hypothesis (EMH), which holds that security prices represent 

all available information, has long dominated the financial markets. Rational investors have 

largely adhered to this theory, but as time passed and there were several significant global 

financial crises, the idea of behavioural finance began to take shape (Rehman, 2013). An 

essential theoretical foundation for balancing behavioural models and the EMH, and 

explaining anomalies is offered by the adaptive market hypothesis (AMH) (Tripathi & Dixit, 

2020). The three theories that provide an insightful view of how financial markets operate and 

how investor sentiment is involved are the Efficient Market Hypothesis EMH, Adaptive Market 

Hypothesis and, Behavioural Finance. 

2.3.1 Efficient market hypothesis 

Traditional financial theory has advanced rapidly since Fama first used the efficient markets 

hypothesis in 1970.  When all available information about securities is reflected in the price, a 

market is considered efficient in terms of information (Naseer & bin Tariq, 2015). According to 

the efficient market theory, rational investors will optimise their portfolios in a way that 

encourages competition, which will lead to asset prices reaching a rational equilibrium (Su et 

al., 2020). Based on rational and representative agents, EMH stated that the stock price 

should reflect intrinsic value because arbitrage makes it easy to rectify any discrepancy 

(Naeem et al., 2021). The efficient market theory states that actual prices represent basic 

values, and that contrary to popular belief, news circulates quickly and, upon disclosure of 

new information, is immediately reflected in stock prices (Uygur & Tas, 2014). Diffusion of 

information leads to equilibrium and "fairly" priced assets that simply represent basic value, 

where mispricing is transient and swiftly fixed by arbitrageurs at almost little risk or expense 

(Friedman, 1953). 



17 
 

The random walk theory and the efficient market hypothesis were unable to account for the 

various ways in which investors behaved in the capital market because they did not view 

investor emotions as a significant factor (PH & Rishad, 2020). Conventional asset pricing 

theory makes the explicit assumption that investor sentiment is irrelevant in order to define 

equilibrium relations and determine the relationships between changes in stock price and 

intertemporal variation in key state parameters in a manageable manner (Chau et al., 2016). 

EMH stated that investors cannot make above-average risk adjusted profits or predict the 

future based on analysis of historical and current data or new information that is easily 

absorbed into securities prices and market activities (Naseer & bin Tariq, 2015). Therefore, 

EMH views investor sentiment as noise that should not consistently impact pricing. 

 

2.3.2  Adaptive market hypothesis 

The adaptive market hypothesis applies evolutionary ideas to financial markets, providing an 

integration between EMH and behavioural finance. AMH proposes far more intricate market 

dynamics, taking into account the fact that, depending on the circumstances, arbitrage 

possibilities can occasionally arise (Chu et al., 2019). According to the AMH, market efficiency 

varies in degree at different times and evolves as a result of events and structural changes 

(Khuntia & Pattanayak, 2018). Specific market conditions may have an impact on how 

predictable stock returns are, since these conditions can have a significant psychological 

impact on market players and how they interpret information, which can influence how they 

make decisions (Urquhart & McGroarty, 2016). AMH takes into account market frictions and 

maintains that markets change over time, in contrast to EMH, which posits a frictionless market 

(Hiremath & Kumari, 2014). According to the AMH, times of turbulence or change may require 

investors to adjust their well-developed heuristics to reflect the new reality because both the 

investment environment and investor behaviour are subject to change (Hall et al., 2017).  

According to Lo (2004), the AMH allows the EMH and calendar anomalies to coexist in an 

intellectually coherent way; this means that people trade financial assets for their own 

interests, learn from their mistakes, and adapt accordingly, which can spur competition and 

innovation in the market. A structural change from the outside, like the financial crisis, forces 

change in the environment in which investors work, pushing associated institutions and 

investor behaviour to adjust or perish (Hall et al., 2017). In contrast to the weak form of 

efficiency that ignores historical prices, AMH suggests that the risk-reward relationship 

changes over time as a result of shifting market participant preferences, that previous price 

movements impact current preferences through natural selection, and that arbitrage 

opportunities periodically emerge and disappear, reflecting the continuous creation and 

disappearance of profit opportunities from an evolutionary perspective (Hiremath & Kumari, 



18 
 

2014). Furthermore, environmental factors can also cause an investment strategy's 

profitability to momentarily rise or fall (Lo, 2004; Mandaci et al., 2019). 

2.3.3 Behavioural finance 

In contrast to conventional economic theory, the behavioural finance theory could explain the 

irrationality and illogicality in behaviours, hence the focus of the discussion switched from the 

efficient market model to the behavioural and psychological characteristics of market players 

(Naseer & bin Tariq, 2015). Rupande et al. (2019) highlighted how in behavioural finance, the 

notion that investors are prone to a variety of behavioural biases means that they do not 

always behave rationally when making investment decisions, which contradicts the 

assumption that investors always act rationally. According to the behavioural finance 

hypothesis, investor sentiment, which is a nonfundamental component, can potentially affect 

future returns (Naeem et al., 2021). Developed in the 1970s and 1980s, behavioural finance 

uses investor behaviour as its research subject and applies sociological theory to explain the 

abnormal phenomena of the financial market (Hu et al., 2021). The theory of behavioural 

finance contends that arbitrage has its bounds, permitting significant and persistent price-

influencing investor irrationality (Ogunlusi & Obademi, 2021). The behavioural finance theory 

suggests that irrational investors' cognitive and selection biases can impact financial asset 

prices, this is due to their incomplete rationality in belief formation, preference, and decision-

making, and the ineffectiveness of rational investors' arbitrage  

Behavioural finance suggests that human behaviour is irrational when it comes to making 

decisions and that this behaviour influences choices about investments, portfolio construction, 

and when to buy and sell stocks (Parveen et al., 2020). Kanojia et al. (2018) stated that there 

are numerous behavioural aspects that influence investors while making an investment 

decision, which include overconfidence, cognitive dissonance, herd behaviour, and mood, 

among others.  Investor sentiment furthermore includes traits such as herding behaviour, and 

cognitive biases such as anchoring, loss aversion, overconfidence, and representative bias, 

as discussed below. 

2.3.3.1 Herding behaviour 

Herding bias occurs when an investor follows the decisions made by other investors without 

first performing independent research using tactics or fundamental analysis (Armansyah, 

2022). Imitation is a common human habit in social situations and the financial markets as 

well, and ignorant investors copy the actions of others because they believe they are more 

knowledgeable and have greater access to information (Tlili et al., 2023). There are two types 

of herding behaviour; irrational herding which is driven by investors' inclination to ignore their 

own analysis and conform to market consensus, and rational herding, motivated by a need to 

protect one's image, informational cascades, and reward systems (Choi & Yoon, 2020).  



19 
 

Herd formation has been shown to exacerbate financial system instability and lead to financial 

market booms and crashes (Sheikh et al., 2023). Speculative bubbles are mostly caused by 

investor herding behaviour, which suggests that traders make similar decisions that might lead 

stock prices to move away from their underlying values (Hirdinis, 2021). There may be 

inaccuracies in stock pricing due to a bias in assessing risk and projected returns when 

herding behaviour takes place, since the market price of stocks does not accurately reflect the 

condition of the economy (Salasiah et al., 2021). 

2.3.3.2 Cognitive biases 

Investor psychological variables, that is, personality traits, have an impact on the probability 

that they may display cognitive biases when making investing decisions (Ahmad, 2020). 

Cognitive bias refers to an investor's divergence in understanding, processing, and making 

decisions based on information or facts, and information processing bias is included in this 

category (Armansyah, 2022). These psychological biases are intrinsic to human behaviour 

and can result in irrational and poor decision making (Wang, 2023). These biases include 

anchoring, loss aversion, overconfidence, and representative bias. 

(a) Anchoring  

 

People take a long time to embrace changes since they are uncomfortable and come with 

Anchoring bias arises when individuals make decisions based on preexisting information or 

the first available information (Kartini & Nahda, 2021). Kartini and Nahda (2021) furthermore 

stated that because of this bias, investors often base their investment decisions on a single 

piece of information, whether it was the first piece of information obtained or was the only 

information that was accessible, with a high degree of dependence on it; and anchoring bias 

can be identified and reduced to improve ways of making decision processes and improving 

financial outcomes. Anchoring bias has a significant impact on financial decision-making 

because people tend to base their judgments and choices on the initial information they 

receive, rather than considering all relevant factors (Wang, 2023). Wang (2023) highlighted 

how price and valuation choices can also be impacted by anchoring bias. For example, the 

first asking price might act as an anchor during a price negotiation and have an impact on the 

ultimate agreed-upon price. 

(b) Loss Aversion 

 

Loss aversion refers to the idea that people have different perspectives on gains and losses, 

A propensity known as loss aversion 

occurs when investors prioritise preventing losses over achieving gains because they are so 



20 
 

afraid of losing money, and the more losses an individual endures, the more likely it is that 

they may develop loss aversion (Kartini & Nahda, 2021). Loss aversion makes people risk 

averse and inclined to choose less-than-ideal options to minimise losses (Wang, 2023). 

Investors prioritise preventing losses over making profits, which explains the characteristics of 

individual investors, who prefer to take on risk during periods of loss when the potential for 

loss is twice that of profit (Bihari et al., 2023). 

Myopic loss aversion (MLA) refers to the inclination of investors who are risk cautious to 

assess their portfolios excessively frequently, leading them to allocate insufficient funds to 

high-risk assets (Lee & Veld-Merkoulova, 2016). Investors with myopic risk aversion are more 

prepared to accept risk and examine their investments less frequently for higher returns. 

Conversely, investors with little financial understanding or a fear of loss may not accept 

unknown risks and instead accept lower returns with known risks (Khan, 2017). Investors 

typically exhibit a lower level of optimism regarding potential returns and a greater fear of 

losing their principal investment when weighing the potential gains and losses from an 

investment (Gupta & Shrivastava, 2022). 

(c) Overconfidence 

 

Overconfident investors disregard non-confirming facts and place too much weight on their 

own abilities, knowledge, experience and analytical capabilities (Ahmad, 2020). In such cases, 

investors frequently overestimate their own worth, opinions, beliefs and abilities (Kanojia et 

al., 2018). This excessive belief in something is known as overconfidence in an investor, and 

it causes them to overestimate their knowledge and underestimate their projections because 

investors believe that they are better than others (Armansyah, 2022). Excessive or unrealistic 

optimism is the state in which people forecast future events with excessive optimism without 

make investors often underestimate their predictions and ignore risks. 

Investors may not always be more aware of the dangers, and when they ignore the first 

warning indications of possible damage, they end up doing more harm than good (Kartini & 

Nahda, 2021).  Overconfidence bias is common among inexperienced investors who seek to 

act quickly and confidently in order to generate large returns; this conduct can result in larger 

trading volumes (Hirdinis, 2021). Overconfidence can be used to explain why active portfolio 

managers are hired by investment and pension funds, why portfolio managers trade so much 

and so often, and why even financial economists frequently own actively managed portfolios; 

these individuals all have the same level of confidence in their ability to identify winners and 

 

(d) Representative bias 



21 
 

Representative bias states that when people respond to a question about whether a particular 

object belongs to a certain specific category based on the information provided, they may 

disregard the objective reference probability and instead create a belief by determining the 

degree of affiliation to this specific category based on the information provided (Ye et al., 

2020). Investors utilise the representative heuristic when they want to purchase stocks from 

the stock market since it encourages them to purchase previous winners (Parveen et al., 

2020). Investors overvalue experiences that are recent or that immediately come to mind, and 

 stated that investors frequently choose stocks based only on 

recent price movements, completely ignoring the fundamental value of the stocks, and assume 

that stocks that have historically produced returns above average will continue to do so in the 

future, ignoring the possibility of a well-known regression towards the mean, or average value. 

The negative effect of representativeness bias is that it tends to rely on estimations based on 

small samples and uses simple categories rather than complex data to support their 

assumptions (Adiputra, 2021). 

In the field of behavioural finance, it is acknowledged that investors are not always logical 

players and that their decisions about investments are frequently impacted by social, 

emotional, and cognitive biases. Yu and Yuan (2011) demonstrate that sentiment has a direct 

impact on the mean-variance trade-off on the market portfolio. Rupande et al. (2019) found 

that the impact of investor behavioural biases on asset pricing is substantial, despite the fact 

that typical asset pricing models ignore them, and found that there is a strong correlation that 

exists between the stock returns volatility in the South African market and investor sentiment 

across the board. Ogunlusi and Obademi (2021) revealed how behavioural finance and 

investment decisions are positively correlated, confirming earlier findings and advancing 

generality. Brown and Cliff (2004), Baker and Wurgler, 2007, and Phan et al. (2023) 

highlighted that stock price variations that conventional financial theory based on strong 

assumptions about investor rationality is unable to explain can be explained by behavioural 

finance. Behavioural finance has emerged as a fascinating area of study since behavioural 

functions are used to analyse financial markets and ethics and emotions have an impact on 

financial performance (Khan et al., 2017; Cuomo et al., 2018; Angeles Lopez-Cabarcos et al., 

2020).  

 

2.4 Factors that Influence Investor Sentiment

The influence of investor sentiment on future share rate returns lies in country-specific factors, 

including the degree of market integration and distinctive cultural aspects (Al-Jabouri & 



22 
 

Oleiwe, 2020). Country-specific characteristics also include political, social, and economic 

aspects among the particular factors that influence investor sentiment. 

2.4.1 Economic factors 

Since individual investors are not as adept at gathering knowledge as institutional investors 

are, changes in economic policy can significantly alter their expectations for risk and projected 

returns (Sun et al., 2021). This implies that changes in macroeconomic factors can influence 

investors' views and predictions. During recessions in the economy, investors become more 

sensitive to news, which results in a strong prediction power (Griffith et al., 2020). This is 

because market volatility and equity markets are generally anticipated to be impacted by 

changes in monetary policy. However, certain market segments and industries may have 

different impacts (Boring, 2023). Therefore, economic changes and policies, such as inflation 

or interest rates, can affect the sentiment of investors and their decision-making. 

2.4.2 Political factors 

Events related to the political landscape of nations that lack political stability should be taken 

into consideration, as it could have a significant effect on the stock market (Maqsood et al., 

2020). Maqsood et al. (2020) furthermore stated that the country's economic and political 

conditions, as well as the sentiment of the community, influence the volatility of stock markets 

in addition to the non-linear nature of data and economic regulations. Dai and Ngo (2021) 

stated that previous research has shown that elections are linked to increased political 

unpredictability. This unpredictability can influence investor sentiment on future stock 

conditions and affect their decision-making, and volatility in the market and investor anxiety 

can be caused by political turmoil, corruption, or abrupt changes in policy. 

2.4.3 Social factors 

Social networks are seen as important information sources by markets and investors for stock 

markets (Piñeiro-Chousa et al., 2021). Sun et al. (2021) highlighted that, since most investors 

these days get their news and announcements online, it stands to reason that how Internet 

users respond to emergencies can have an impact on investor sentiment. Griffith et al. (2020) 

pointed out how the media plays a crucial role in spreading information and influencing market 

movements. Social factors such as public opinion of business and innovation, education 

levels, and demographic trends also play a role in the influence of investor sentiment. 

2.4.4 Market integration 

Financial market integration (FMI) is the state in which financial markets from several nations 

operate together and display the same projected risk-adjusted returns (Patel et al., 2022). 

Organisations with a worldwide network are thought to have more access to finance than those 



23 
 

that rely solely on local financial sources (Mmolainyane & Ahmed, 2015). Furthermore, 

Mmolainyane and Ahmed (2015) stated that when there is a lack of regulation, financial 

integration frequently results in overpopulation in the banking industry and unfair competition 

between domestic and international enterprises, making it harder for locals to benefit. Patel et 

al. (2022) furthermore stated that regarding markets that are not yet fully integrated, investors 

might allocate money to the most productive market or region, or diversify by distributing 

across other locations to obtain a better risk-return trade-off (Patel et al., 2022). Financial 

integration might limit investors' capacity to diversify their portfolios as a risk hedge by reducing 

the profitability disparities of financial assets between nations (Kablan & Guesmi, 2016). This 

may affect investor sentiment and decision making. 

 

2.4.5 Cultural aspects 

Investor sentiment is significantly influenced by cultural factors, such as societal conventions, 

values, and behaviours. Khan (2017) stated that cultural values also have an impact on 

investors' cognitive biases. As a component of culture and social capital, trust is not created 

in a vacuum; rather, it is shaped by a variety of factors, including history, politics, religion, 

ethnicity, upbringing, education, and formal institutions (Pevzner et al., 2015). Pevzner et al. 

(2015) further demonstrated how the capital market's response to organisation financial 

disclosure is influenced by both official and informal national institutions, particularly national 

culture and social trust. This could influence investor sentiment, as investment perspectives, 

perceptions of risk, and confidence in financial institutions are all influenced by culture. 

2.5 How Investor Sentiment is Calculated 

Investor sentiment cannot be observed but must be calculated, and there are several ways in 

which investor sentiment can be measured. Several sentiment measures have been 

proposed, including measures derived from market factors (also known as indirect measures) 

or investor surveys (often known as direct measures) (Smales, 2017). By using surveys or 

other data sources, direct indicators are created to gather investors' subjective sentimental 

attitudes; whereas, when determining investor sentiment, indirect indicators use data and 

matching objective indicators that are already available in the financial markets as a stand-in 

(Li et al., 2021). Investor sentiment can be measured in different ways, such as surveys, 

market-based indicators, and media sentiment analysis.  

Survey measures are a direct measure of investor sentiment, and Zi-long et al. (2021) stated 

that direct measurements use investor surveys to look at investor opinions. Survey measures 

offer direct insight into investors' views, expectations, and impressions of the market and 

individual securities. Survey-based metrics are intended to measure the general attitude of 

the market among both active investors and those waiting for a good opportunity to buy (Chau 



24 
 

et al., 2016). Zhou (2018) stated that survey measures are significant because the data enable

analysis of whether certain sentiment metrics derived from market data align with investor 

beliefs. The information content obtained from survey-based indicators is highly valuable due 

to its compilation of prospective investors' expectations and market perspective (Ahmad, 

2020). The Consumer Confidence Index (CCI), Investor Confidence Index (ICI), and Business 

Confidence Index (BCI) are some examples of survey measures. There are other surveys-

based metrics like the Investors' Intelligence or the Michigan Consumer Sentiment Index 

(MCSI) (Angeles Lopez-Cabarcos et al., 2020).  

Measurements based on the market are dependent on quantifiable market data, with the 

assumption that the underlying factors include latent investor sentiment (Ung et al., 2024). 

Market-based metrics represent an indirect approach in which market data on investor 

sentiment are chosen to create an investor sentiment index (Hu et al., 2021). Indirect 

measurements measure investor sentiment by using financial market factors derived from 

financial theories and are predicated on financial and economic variables that represent 

investors' perspectives (Khan & Ahmad, 2018). Market-based measures include the Volatility 

Index (VIX) and the Advance-Decline line. Closed-end equity fund (CEEF) discount and equity 

put-call ratio (PCR) and closed-end fund discount (CEFD) are some of the sentiment proxies 

(Bathia & Bredin, 2016) (Angeles Lopez-Cabarcos et al., 2020). Muguto et al. (2019) 

measured investor sentiment by creating an investor sentiment index which included seven 

proxies (share turnover; R/$ bid-ask spread; R/P bid-ask spread; R/Euro bid-ask spread; 

Advance decline ratio; term structure of interest rates; and equity issue ratio). Additionally, 

investor sentiment can be measured by indices constructed by other market variables such 

as the Baker and Wurgler index (Baker & Wurgler, 2006, 2007).  

A more modern strategy is to gather sentiment from sources like newspapers and Internet 

forums using textual (media- or search-based) approaches (Ung et al., 2024). Sentiment 

analysis, broadly defined as the study of textual data using techniques that process natural 

language to capture people's attitudes toward a topic, is applied in the evaluation of investor 

sentiment utilising media-based data sources (Johnman et al., 2018). Kräussl and 

Mirgorodskaya (2017) contend that the coverage in the media influences investors' emotions 

in addition to reflecting economic events. Media-based investor sentiment measures have 

become more and more common, and these metrics are based on textual analysis of media 

items, including blogs, message boards, newspaper columns, and Google search results (Sun 

et al., 2016). Zhou (2018) stated that, based on text and media, perspectives are taken from 

media outlets, texts, recordings, events, and a variety of online activities. Tetlock (2007) finds 

that abnormally high or low pessimism in the media forecasts large market trading activity, 

and high pessimism in the media predicts a decline in market prices foll