Unveiling the Nexus Between Crises,Investor Sentiment,and Volatility of Tourism-Related Stocks: Empirical Findings From Pakistan

Abstract

This study investigates the predictive power of sentiment metrics concerning the volatility of tourism-related stocks listed on the Pakistan Stock Exchange (PSX) during recent crises. We employed univariate and multivariate GARCH-family econometric models to analyze daily data collected from March 2020 to April 2023. The data is categorized into two periods: A (COVID-19 until February 2022) and B (the vote of no-confidence motion until April 2023). The findings indicate that Market Sentiment (MS) plays a more significant role than individual sentiment in predicting stock price volatility and negatively impacts return volatility. During period A, the Fear Index (FI) and in B, economic policy uncertainty (EPU) were identified as factors contributing to increased fluctuation in tourism-related stocks. Additionally, MS exhibited a heightened level of bearishness compared to the COVID-19 era. The multivariate BEKK-GARCH analysis revealed that MS past volatility negatively spillover onto the current conditional volatility of PSX firms. This highlights the paramount importance of MS in understanding stock price fluctuations in the PSX. Interestingly, the technology sector exhibited lower sensitivity across the analysis. This study provides valuable insights for decision-makers by elucidating the principles of behavioral finance theory within the context of the PSX. They should prioritize long-term strategies to institutionalize investors, enhance their emotional intelligence, and mitigate psychological biases in the market.

Keywords

tourism industry investor sentiment volatility behavioral finance GARCH family models

Introduction

The value of information has notably surged as a crucial asset amid recent global health crises in the financial market. Fama’s (1970) efficient market hypothesis (EMH) postulates that available market information is reflected in asset prices. On the contrary, the noteworthy work of De Long et al. (1990) contradicts (EMH), and affirms the behavioral finance theory, suggesting that noise players create bubbles that deviate from the stock price’s intrinsic value. Baker et al. (2012) examined the multifaceted determinants of stock prices, noting an investor’s interest in a bullish market as opposed to bearish periods. This contrast underscores the critical influence of investor sentiment in the dynamic of stock trading. Investor sentiment is typically used to assess investor cognitive bias, which is considered an erroneous perception of asset value (Zhou, 2018). However, a growing body of literature examined the impact of investor sentiment on the cross-section of stock prices, see for example (Beer & Zouaoui, 2013; Gong et al., 2022; ur Rehman, 2021). Several studies have utilized Principal components analysis (PCA) to formulate proxies reflecting investor sentiment in financial markets (Baker & Wurgler, 2007; D. Lei, 2019; Song et al., 2023). Abdelhedi and Boujelbène-Abbes (2019) used PCA to measure investor sentiment and the BEKK-GARCH model to capture volatility, demonstrating significant spillover to stock price fluctuation. On the other hand, several studies have focused on the exploration of news-based investor sentiment (Narayan & Bannigidadmath, 2017; Puzanova & Eratalay, 2021; Srinivas et al., 2023). This phenomenon has attracted significant interest across various disciplines. However, the tourism industry has received limited emphasis on psychological variables, encompassing investor sentiment, market fear, and even their potential impact on economic characteristics, such as the stock market return volatility and trading behavior (Khalid et al., 2021; Su et al., 2017). Few studies have specifically focused on examining the impact of investor sentiment on the volatility of tourism stock returns (Hu et al., 2021; K.-L. Peng et al., 2023). Zargar and Kumar (2023) explored the interplay between investor sentiment and the UK, US, EU, and Italy tourism sectors, revealing that amid COVID-19, the U.S. tourism returns were more sensitive than other to investor sentiment (Sorić, 2021; B. Wu et al., 2021).

Many studies focused on aggregate stock market to address the investor sentiment effect on emerging market Pakistan (W. Ahmed, 2018; M. A. Khan & Ahmad, 2018). ur Rehman (2021) conducted firm base analysis in Pakistan and developed both PCA and news sentiment; the findings affirm behavioral finance theory, showing irrational investor sentiment impacts stock market activities. Additionally, Parveen et al. (2023) examined in their study that the COVID-19 pandemic negatively affected the PSX investors’ behavior. To the best of the author’s knowledge and peer review, insufficient literature has been observed on the nexus between investor sentiment and tourism stock return volatility. This study pivots on this gap, specifically delving into Pakistan’s emerging market. Globally, the tourism industry contributed US$ 10 billion to the world GDP in 2019; it declined by −34.3% in 2020 due to the pandemic; compared to 2021, it rebounded with a 111% increase, equal to US$ 7.5 in 2022. The industry is expected to reach $9.5 billion in 2023 and is projected to contribute US$15.5 billion to global GDP by 2033 (Statista 14-12-2023). Similarly, in Pakistan, as per (TTCR, accessed 08-07-2023) Travel and Tourism Competitive Index 2017, tourism contributed US$ 328.3 million in 2015, representing 2.3% of the nation’s GDP. Furthermore, Hasni et al. (2021) reported in their study in 2019 that the tourism industry contributed 7.8% to the GDP, generated 1.5 million jobs, constituting 2.4% of total employment, and improved its global rank to 121out of 124 Pakistan, a three-point rise from 2017. In their recent study N. U. Khan et al. (2022) the tourism revenue plummeted by 22.9% amid the pandemic. The projection indicates a U-shape recovery, with an expected US$3,008.00 million in 2023, within the anticipated 6.14% annual growth, surpassing US$ 3817.00 million by 2027.

The tourism sector has the potential to considerably enhance financial growth and catalyze the overall economic development in Pakistan. This study seeks to investigate the precise research question,” To what extent do investor sentiment and economic policy uncertainty impact the tourism-related stocks return volatility? It makes noteworthy contributions toward accomplishing specific objectives, offering novel perspectives and insight to enrich the existing scholarly discourse. Firstly, prior studies mainly focused on the tourism industry from a qualitative perspective (N. Ali, 2013; Y. Ali et al., 2018; Nazir et al., 2023). This study enhances the quantitative perspective to engage investors in PSX strategies. Secondly, the time spans from March 2020 to April 2023, and encapsulates the pivotal events of this century, including the global health crises, the COVID-19 pandemic, and internal turmoil, like the vote of no-confidence filed by the (PDM) Pakistan democratic movement against former prime minister Imran khan. This event marked the inception of a subsequent crisis era in Pakistan’s current landscape. Thirdly, S. K. Khan and Saleem (2022) incorporated market proxies and employed PCA, while Naeem et al. (2021) utilized a Machine learning approach for data scraping. Therefore, we employed both methods to present a clear picture and adopted the (ML) approach within the Python environment, specifically the (Jupiter notebook), to facilitate data scraping from URL’s and construct individual investor sentiment indexes. Additionally, the well-known PCA (Principal Component Analysis) technique for market-based proxies to extract the market sentiment index. Fourthly, the COVID-19 pandemic and the vote of no confidence motion in 2022 drastically affected the Pakistani economy (Iqbal et al., 2023). The GDP fell fromUS$314 billion in 2018 to US 290 in 2022, amid COVID-19 and the vote-of-no-confidence motion, the external debt increased from $60 to $130 billion in 2022, and the currency lost 300% of its value against the US dollar, the central bank interest rate an all-time high 21% in April 2023 (S. U. Rahman & Jan, 2023). Therefore, this study follows Salisu and Akanni (2020) and develops a fear index for the COVID-19 era, and subsequently uses EPU as a proxy for internal crises to investigate their impact on tourism-related stock return volatility. Lastly, the STATA was employed for MS (market sentiment) and univariate analysis, and then the R programming language was used for further analysis, like the urca package for data stationarity. Then, the renowned BEKK-GARCH approach has been employed, demonstrating superior efficacy over the MTS package in capturing variance-covariance in complex settings (CRAN.R, Accessed July 15, 2023).

The subsequent section unfolds as follows. Section “Literature Review” consists of a literature review Section “Methodology” pertains to data collection and selection of statistical approaches. Section “Empirical Results” presents empirical results and discussion. Lastly, Section “Conclusion” reports the study conclusion and underscores policy implications.

Literature Review

Global connectivity is a double-edged sword it is advantageous, but it has recently unleashed unprecedented severe sparks, like COVID-19, a once-in-a-century calamity, swiftly ravaging global financial markets; it eclipsed the 2008 GFC global financial crisis. This event led to a $2 trillion loss in the global economy by early 2020, with several countries grappling with economic growth below 2.5% (Mishra, 2020). According to Bagchi et al. (2020), during COVID-19, the top ten companies listed on the S&P experienced a significant $1.4 trillion loss attributed to investor fear. Further, amid the COVID-19 pandemic, the International Air Transportation Association (IATA) reported a staggering $113 billion loss, heavily affected by stringent travel restrictions (IATA, Accessed December 15, 2023) Airlines (2020). A country like Pakistan has faced prolonged continuous, financial, natural, political, and terror turmoil for decades (Parveen et al., 2020). Arshad et al. (2018) forecasted Pakistan’s tourism industry to grow by $38.9 billion in a decade and ranked Pakistan 124th out of 136 in 2018. Mystery pneumonia COVID-19 impacted (PSX) causing a 10% index below 35,000 points, losses amounting to $610 million, tempting investor uncertainty (Aslam et al., 2020). Political crises induce uncertainty, posing a threat to the financial system. One grand issue practised for centuries, named the no-confidence vote, was filed for the first time against the British prime minister in 1782 (Mason, 2022). In the subcontinent, Bangladesh prohibited the no-confidence motion; conversely, in India, Prime Minister Vajpayee faced defeat in April 1999 by a narrow margin of one vote (269–270) (Hindustan Times. (2024) Accessed 2023-11-10). Since the 1980s, Pakistan has experienced three key votes of no-confidence: Benazir Bhutto in 1989, Shaukat Aziz in 2006, and recent vote of the no-confidence motion in 2022. Each of these instances has impacted its economy significantly (Ghani et al., 2022; Iqbal et al., 2023). In their recent study, Hussain Mirza and Kumar (2023) reported that the post-April 10 no-confidence motion contracted the GDP from 5% to 1%, with the currency depreciating by over 40% in a few weeks. This reflected and severely impacted emotional intelligence.

Investor Sentiment in the Tourism Market

An array of scholarly inquiries have analyzed investor sentiment impact on the tourism industry (Corbet et al., 2022; Dragouni et al., 2016; Huang et al., 2014). In their recent empirical study, Zargar and Kumar (2023), the US tourism sector’s returns amid COVID-19, exhibit that market fear, investor sentiment, and economic policy uncertainty are net volatility transmitters and the tourism sector predominantly is the net receiver. During the COVID-19 outbreak ML base analysis was piloted, uncovering a heterogeneous impact of tweeter-base economic policy uncertainty on the tourism sector (Shiljas et al., 2022). In the same vein, Rudež (2022) investigated twelve tourism-related stocks and observed that amid the COVID-19 pandemic, investor sentiment responded rationally, leading to the swift rebound of several stocks to their intrinsic value.

Investor Sentiment Across Industries

In behavioral finance, investor sentiment is identified as a key determinant that significantly affects industry-specific return volatility (Chen et al., 2013; Dash & Mahakud, 2013; Niu et al., 2021). According to Hong et al. (2020), amid COVID-19, a stronger asymmetric and herding effect has been observed in the manufacturing and IT sectors. Few researchers focus on the COVID-19 pandemic, it has been proven that behavioral finance theory has a diverse impact on tourism-related sectors (Olugbode et al., 2014; K.-L. Peng et al., 2023; Sorić, 2021). Yuan (2021) found that Intensified herding behavior exerts a strong influence on transportation, leasing, business service, and culture products, however, observed varied both short and long-term effects of investor sentiment excluding real estate. In the same line, Cui and Zhang (2020) conducted a firm-based analysis and concluded that managers’ overconfidence is more crash-prone. Recently Rohilla et al. (2023) examined the effect of investor sentiment on four S&P industries, finding a robust correlation with these industries in both short-term and long-term contexts. Furthermore, using the VAR-GARCH model on the aviation industry, the immense negative impact of investor mood was observed (Kaplanski & Levy, 2010). Zheng (2021) it is imperative to contemplate tourists’ psychological attitudes when implementing remedial actions or resilience policies.

Investor Sentiment and Stock Return

A widely recognized technique PCA (principal component analysis) employed on market proxies to establish an investor sentiment, aimed at forecasting stock returns (Baker & Wurgler, 2006; Baker et al., 2012; Y. Li & Ran, 2020; M. L. Rahman & Shamsuddin, 2019; Song et al., 2023). Recently K.-L. Peng et al. (2023) employed four market proxies: CCI (consumer confidence index), NOP (newly opened accounts), TR(turnover ratio), and P/E (price to earnings ratio); their findings unveiled high sensitivity of tourism-related stocks to investors mood in the Chinese stock market. Several researchers utilized a machine learning approach (Das et al., 2019; Herrera et al., 2022; Hu et al., 2021; Kim et al., 2020; Nti et al., 2020; H. Peng et al., 2018; Renault, 2020; Vijh et al., 2020). P. Wang et al. (2021) utilized Python to crawl data from Weibo; unveiling that amidst COVID-19, investor sentiment turned pessimistic toward the stock market. Scholarly research primarily presents a consensus on the existence of positive correlation (Ainin et al., 2020; Baker & Wurgler, 2006; Baker et al., 2012; Hsu & Tang, 2022; W. Y. Lee et al., 2002; Lux, 2011). Conversely, numerous studies have noted a slumping effect (Nicolaus et al., 2020; W. Wu et al., 2021; Shahzad & Caporin, 2020). various findings manifested range from mix to no existence association (Jansen & Nahuis, 2003; Kling & Gao, 2008; Rudež, 2022; Sigala, 2020). The multivariate GARCH base analysis demonstrates that investor sentiment significantly predicts volatility in the Mexican stock market returns (Liston-Perez et al., 2018). Narayan et al. (2021) investigated the G7 economies and concluded that government financial grant policies positively influence investor sentiment and stock market returns. Higher uncertainty exacerbates volatility, and stocks remain unpredictable (Nartea et al. (2020) concluded that EPU (Economic Policy Uncertainty) in conjunction with varying levels of investor emotions, indicates EPU trading strategy during low sentiment is more profitable. Further, the US-centric research (1995-2015) posited investor sentiment and EPU as driving factors for US sectoral returns (Rehman et al., 2021). López-Cabarcos et al. (2021) S&P 500, Kumari and Mahakud (2015) Indian stock market, applied univariate GARCH model, further Johnk and Soydemir (2015) US equity market, (Guerello, 2016) US stock market, Zeng (2020) for China applied multivariate GARCH model and concluded that investor sentiment is a robust predictor for stock return volatility. Ghorbel et al. (2014) and Trichilli et al. (2018) employed multivariate BEKK-GARCH technique, revealing that investor sentiment legged volatility significantly spillover to Islamic indices return volatility.

Hence, stock price fluctuations are contingent upon the level of investor confidence. During episodes of overconfidence, frequent trading introduces market noise, deviating stocks from their intrinsic path (Barber & Odean, 1999; Flynn, 2003; Rashid et al., 2019). Therefore, this article puts forth three hypotheses to test the behavioral finance theory of Blackledge and Lamphiere (2021) and K.-L. Peng et al. (2023) examine the sentiment index’s influence on the Pakistan tourism-related stocks market.

Hypothesis 1: Investor sentiment significantly influences the volatility of tourism-centric firms.

Hypothesis 2: COVID-19 has markedly escalated stock price fluctuation in tourism-related stocks.

Hypothesis 3: Economic policy uncertainty provokes a heterogeneous effect on stock price volatility across sectors.

Methodology

The sample is based on K.-L. Peng et al. (2023) and Geng et al. (2021) took thirty tourism-related stocks from five sectors listed on PSX (Pakistan stock exchange). Collected daily closing prices for firm stocks from Business Recorder and daily figures for fear index proxies; Table 1 presents data along with data sources. Meanwhile, market sentiment proxies were gathered on a monthly basis except for EPU and then converted to daily frequency using eviews-12 to facilitate more precise regression analysis. The textual sentiment detail is illustrated later in this section. The time period spanning from March 2020 to April 2023, bifurcated into two distinct phases; period A, marking the COVID-19 outbreak, initiated with the WHO pandemic declaration, and B, from March 2022 is, encapsulating the political turbulence of the vote of no-confidence motion in Pakistan, with a focus on its repercussions on the volatility in the tourism-related sectors. The goal is to examine how investors’ psychological responses to the stock market fluctuation across these two tumultuous periods. Further, the sentiment detail is delineated later in this section.

Table 1.

Variables and Data Source.

Food sector		Technology		A-hotel	Tobacco
QUICE	MFL	PTCL	MDTL	PHDL	PAKT
UPFL	MUREB	TRG	WTL	AKDHL	KHTC
SHEZ	NATF	PAKD	AVN	PSEL	Transport sector
NESTLE	TOMCL	NETSOL	TPL	NML	PIA
MFFL	FCEPL	HUMNL	TELE	HBL	PNSC
Firm closing prices		Business Recorder		Sarwar et al. (2019)
Economic policy uncertainty		www.policyuncertainty.com		Ghani et al. (2022)
News sentiment		https://www.centralbanking.com/organisations/state-bank-of-pakistan https://en.wikipedia.org/wiki/2022%E2%80%932023_Pakistani_economic_crisis		Chen et al. (2021); ur Rehman (2021)
Proxies for Market sentiment & COVID-19 index
Turnover		PSX(Pakistan)		Rashid et al. (2019)
Newly account open		CDC Pakistan		K.-L. Peng et al. (2023)
Consumer Confidence Index		SBP(Pakistan)		K.-L. Peng et al. (2023)
Price/earnings ratio		PSX(Pakistan)		K.-L. Peng et al. (2023)
COVID-19 new, total new, COVID-19 death, and total death cases		Our world in data		Salisu and Akanni (2020)

Note. PSX = Pakistan Stock Exchange; CDC = Central Depository Company; SBP = State Bank of Pakistan.

Data Normality Test

An outlier refers to an observation that deviates from the normal state, and necessitates appropriate treatment as it can influence the process of hypothesis testing (K. Lee et al., 2022). This study detected outliers in many stocks and applied normality tests to natural Log returns using statistical tools (Shapiro-Wilk, Konglomorov-Smirnov) and graphical tools (Histogram, Boxplot, QQ plot). Due to missing values, we set an average period-A, A, 550, and B, having 300 observations, if a variable exceeds the approximate limit, has trimmed; otherwise, it is winsorized. The results are not included in the study brief and are available upon request (Afzal et al., 2021; Bieri, 2020; Ehsan, 2021; Ejsmont et al., 2023).

PCA Index Development

This study formulated three indices using a variety of proxies. The market sentiment inspired by Baker et al. (2012) and K.-L. Peng et al. (2023), incorporates metrics such as TR (Turnover Ratio), CCI (Consumer Confidence Index), NOA (Newly Opened Accounts) and P/E (Price To Earnings Ratio). Secondly, the fear index Following Subramaniam and Chakraborty (2021) integrated data on COVID-19 new and total cases, as well as COVID-19 deaths and total death cases. Lastly the textual sentiment index, based on W. Zhang et al. (2021) and ur Rehman (2021), by scrapping data from two URL’s, as detailed in Table 1. Similarly, Tables A1 and A2 show the descriptive statistics and data stationarity, and the ARCH effect test for all variables.

The following are the raw sentiments of the first component after the rotation matrix

I S = 0.6932 P E C + 0.0061 S B P + 0.2466 L P E C - 0.6772 L S B P

(1)

FI = 0.5598 NC + 0.0273 TNC - 0.4668 DC + 0.6841 TDC

(2)

ME = 0.5395 CCI - 0.1816 \frac{T}{R} + 0.1982 NOA + 0.7979 \frac{P}{E}

(3)

Figure 1 demonstrates the correlation among proxy variables; the intensity of color and the legend on the right side explains the relationship. The dark blue presents a high correlation, and the light blue has a partial positive correlation among proxies. Furthermore, the red-shaded area spectacle negativity, but the dark red stands for high negativity, the light red otherwise. Table 2 specifies the KMO and barrelet sphericity test results, indicating that the variables are not orthogonal, providing concrete evidence for PCA fitness (Shrestha, 2021). The components above eigenvalue are considered for final sentiment index. Figure 2, exhibits cumulative data variation for Market Sentiment 81%, Individual or textual Sentiment 68%, and Fear Index or COVID-19 sentiment 75%, all these values are more significant than famous Baker and Wurgler (2007) and Kurbatskii and Mironenkov (2023) 53%.

Figure 1.

Correlation among proxy variables.

Table 2.

KMO Test and Bartlett’s Test for PCA.

Kaiser-Meyer-Olkine		Bartlett’s test of sphericity
Index	Measurement of sampling adequacy	K-Sq	DF	Significance
MS	0.72	9131.6	3	.000***
IS	0.63	7374.6	3	.000***
FI	0.69	364.08	7	.000***

Note. MS = Market Sentiment; IS = Individual Sentiment; FI = Covid-19, or fear index.

*, **, *** subscript 1%, 5%, 10% significance.

Figure 2.

The scree plot for ME (market emotions), IS (individual sentiment), FI (Fear index).

Individual Sentiment (IS) Construction

VADER (Valance Aware Dictionary and Sentiment Reasoner) lexicon does not need to train data; it takes human-labeled data (Hutto & Gilbert, 2014). This article followed Rehman (2021) who scraped Bloomberg News to develop textual sentiment for Pakistan, and we extended their technique and scraped data from URL’s mentioned in Table 1. The Python Jupiter notebook used and called the NLTK (Natural Language Tool Kit) library from NLP (Natural Language Toolkit) processing tool and practiced these steps (a) Scrape Data (b) Initial Screening Text (c) Sentence Tokenization (d) Remove Stop Words (e) Word Frequency (f) SIA(Sentiment Intensity Analyzer), the SIA can read the exclamation and word capitalization for the intensity (g) word-cloud visualization shown in Figure 3, (the supporting document available in Appendix Table A3). We extracted the top 50″ common words by employing the word frequency function and then took the relevant ten negatives and positives from each set. Match them with clean data for microblog selection. Applied the SIA function to each sentence and divided it into positive, negative, neutral, and polarity, then applied the following equation.

FC_n s_{t} = \frac{po s_{t} - ne g_{t}}{com p_{t}}

(4)

S {B_{ns}}_{t} = \frac{po s_{t} - ne g_{t}}{com p_{t}}

(5)

Where $po s_{t}, ne g_{t}$ , $com p_{t}$ exhibit the positive negative compound values. Moreover $FC_n s_{t}$ , $SB_n s_{t}$ for financial crises and State Bank of Pakistan news. Frequency observations were enhanced through EVIEWS-12 and taken their log applied PCA to construct news sentiment (Paramanik & Singhal, 2020; Puzanova & Eratalay, 2021).

Figure 3.

Show the word-cloud of SBP and financial crises 2022–23.

Econometric Specification

This study follows Murty et al. (2022), and employs both uni-multi-variate GARCH models to examine the influence of investor sentiment on returns volatility. The AR and MA models have the drawback of capturing time-varying clusters (Pilbeam & Langeland, 2015). The univariate GARCH family effectively captures the pattern in financial time-series where the small shock is followed by the small, and the large shock is followed by the large (Katsiampa, 2017). Furthermore, according to Yu et al. (2020) VAR can inspect dynamic relationships but falls short in capturing the non-linear complexities inherent in a dynamic time-series. Although copula models address these limitations, including asymmetries, they are subject to drawbacks concerning spillover direction. The BEKK-GARCH proposed by Engle and Kroner (1995) holds supreme importance in comprehensively exploring the volatility spillover relationship between time series analysis. This study employs uniquely for investor sentiment and tourism-related stocks.

GARCH Specification

The symmetric GARCH model proposed by (Bollerslev, 1986) is given as;

σ_{t} = ω + \sum_{i = 1}^{p} α_{i} ε_{t - i}^{2} + \sum_{j = 1}^{q} β_{j} σ_{t - j} + Ψ_{1} MS + Ψ_{2} IS + Ψ_{3} FI

(6)

ω, α, β > 0, α_{i} + β_{j} < 1, i = 1, \dots p, j = 1, \dots q)

(6)

$σ_{t}$ Is conditional variance at time $t$ $α_{i}$ is $squared residua l_{t - 1}$ , $β_{j}$ is the $conditional varianc e_{t - 1}$ , $Ψ_{1} MS + Ψ_{2} IS + Ψ_{3} FI$ Represents market sentiment, COVID-19(Fear) index and individual sentiment (Çelik, 2020).

EGARCH Model

GARCH has restriction of symmetry, Brooks (2019), to capture asymmetry(leverage effect in firm operation), E-GARCH proposed by (Nelson, 1991).

\log σ_{t} = ω + \sum_{j = 1}^{q} β_{j} \ln (\log σ_{t - 1}) + \sum_{p = 1}^{p} φ_{m} \frac{ε_{t - i}}{\sqrt{σ_{t - 1}}} + α_{i} \sum_{i = 1}^{p} [\frac{| ε_{t - 1} |}{\sqrt{σ_{t - 1}^{2}}} - \sqrt{\frac{2}{π}}] + Ψ_{1} MS + Ψ_{2} IS + Ψ_{3} FI

(7)

$ω, α, β > 0, α_{i} + β_{j} < 1, i = 1, \dots p, j = 1, \dots q)$ , further $Ψ_{1} MS + Ψ_{2} IS + Ψ_{3} FI$ are external shocks, $σ_{t}$ Is conditional variance at time $t$ , $α_{i}$ , and $β_{j}$ are lagged recent and old information parameters to current conditional volatility, $\log σ_{t}$ is non-negative parameter. Moreover if $φ_{m} < 0$ risky shock exhibit leverage effect, otherwise $φ_{m} > 0$ (D. H. Vo et al., 2022).

GJR-GARCH Model

to capture unknown shock effects to the current conditional volatility GJR-GARCH proposed by (Glosten et al., 1993)

h_{t} = ω + \sum_{i = 1}^{p} α_{i} ε_{t - l}^{2} + \sum_{i = 1}^{p} δ_{n} ε_{t - i}^{2} ψ_{t - 1} + \sum_{j = 1}^{q} β_{j} h_{t - j} + Ψ_{1} MS + Ψ_{2} IS + Ψ_{3} FI

(8)

$ψ_{t - 1} = 1 if ε_{t - i} < 0$ Bad news, otherwise $0, ε_{t - i} \geq 0$ , $ω, α, β > 0, α_{i} + β_{j} < 1, i = 1, \dots p, j = 1, \dots q)$ , further $Ψ_{1} MS + Ψ_{2} IS + Ψ_{3} FI$ aredefined above, $σ_{t}$ Is conditional variance at time $t$ $α_{i}$ is shock and, $β_{j}$ is volatility parameter. Further, if $δ_{n} > 0$ , and significant the negative shock will have more impact on current conditional variance then positive, Otherwise $δ_{n} < 0$ (Brooks, 2019; D. H. Vo et al., 2022). Note the $Ψ_{3 b}$ will be used for EPU in period B.

Bivariate BEKK-GARCH Model

The BEKK-GARCH proposed by R. F. Engle and Kroner (1995), to inspect the volatility spillover effect between investor sentiment and tourism-related stock returns.

R_{t} = π_{t} χ_{t} + ε_{t}, ε_{t} | τ_{t - 1} ~ N (0, h_{t})

(9)

h_{t} = α_{0} + α_{1} ε_{t - 1}^{2} + β_{t} h_{t - 1}

(10)

$π_{t}$ Stands for explanatory parameter, $χ_{t}$ exhibits the vector coefficient. $ε_{t}$ is residuals, $h_{t}$ is the conditional variance of equation of univariate model. Further, the conditional co-variance matrix has denoted by $H_{t}$ , equations are as follows

H_{t} = C_{0}^{'} C_{0} + \sum_{i = 1}^{q} A_{1}^{'} ε_{t - 1} ε_{t - 1}^{'} A_{i} + \sum_{i = 1}^{q} B_{1}^{'} H_{t - 1} B_{i}

(11)

C_{0} = (\begin{matrix} ω_{1} & 0 \\ ω_{2} & ω_{3} \end{matrix}), A = (\begin{matrix} α_{11} & α_{12} \\ α_{21} & α_{22} \end{matrix}), B = (\begin{matrix} β_{11} & β_{12} \\ β_{21} & β_{22} \end{matrix})

(12)

Where $C_{0}$ is intercept coefficient matrix, A is the shocks coefficients matrix, and B is the volatility matrix. The conditional co-variance of the BEKK-GARCH (1, 1) model is write as:

h_{t} = (\begin{matrix} h_{11, t} & h_{12, t} \\ h_{21, t} & h_{22, t} \end{matrix}) = {(\begin{matrix} ω_{1} & 0 \\ ω_{2} & ω_{3} \end{matrix})}^{'} (\begin{matrix} ω_{1} & 0 \\ ω_{2} & ω_{3} \end{matrix}) + {(\begin{matrix} α_{11} & α_{12} \\ α_{21} & α_{22} \end{matrix})}^{'} (\begin{matrix} ε_{1, t - 1} \\ ε_{2, t - 1} \end{matrix}) {(\begin{matrix} ε_{1, t - 1} \\ ε_{2, t - 1} \end{matrix})}^{'} (\begin{matrix} α_{11} & α_{12} \\ α_{21} & α_{22} \end{matrix}) + {(\begin{matrix} β_{11} & β_{12} \\ β_{21} & β_{22} \end{matrix})}^{'} (\begin{matrix} h_{11, t - 1} & h_{12, t - 1} \\ h_{21, t - 1} & h_{22, t - 1} \end{matrix}) (\begin{matrix} β_{11} & β_{12} \\ β_{21} & β_{22} \end{matrix}) .

(13)

The above equation can be write like:

h_{11, t} = ω_{1}^{2} + α_{11}^{2} ε_{1, t - 1}^{2} + 2 α_{11} α_{21} ε_{1, t - 1} ε_{2, t - 1} + α_{21}^{2} ε_{2, t - 1}^{2} + β_{11}^{2} h_{11, t - 1} + 2 β_{11} β_{21} h_{12, t - 1} + β_{22}^{2} h_{22, t - 1}

(14)

\begin{matrix} h_{22, t} = (ω_{2}^{2} + ω_{3}^{2}) + α_{12}^{2} ε_{1, t - 1}^{2} + 2 α_{22} α_{12} ε_{1, t - 1} ε_{2, t - 1} \\ + α_{22}^{2} ε_{2, t - 1}^{2} + β_{11}^{2} h_{11, t - 1} + 2 β_{12} β_{22} h_{12, t - 1} \\ + β_{22}^{2} h_{22, t - 1} \end{matrix}

(15)

\begin{matrix} h_{12, t} = ω_{1} ω_{2} + α_{11} α_{12} ε_{1, t - 1}^{2} + (α_{11} α_{12} + α_{21} α_{12)} ε_{1, t - 1} ε_{2, t - 1} \\ + α_{22} α_{21} ε_{2, t - 1}^{2} + β_{11} β_{12} h_{11, t - 1} \\ + (β_{11} β_{22} + β_{21} β_{12}) h_{12, t - 1} + β_{21} β_{22} h_{22, t - 1} \end{matrix}

(16)

$h_{11, t}$ Denotes the conditional variance of market sentiment $h_{22, t}$ represents the conditional variance, of $ith$ firm. $h_{12, t}$ is the conditional co-variance coefficient of both(market sentiment and PSX firms).the main diagonal elements $α_{11}$ and $β_{11}$ . $α_{22}$ and $β_{22}$ denotes the own lagged shock and volatility of market sentiment and PSX tourism-related firms. However the off-diagonal elements $α_{12}$ and $β_{12}$ are the short and long-term transmission parameters of market sentiment to $ith$ firm. $α_{21}$ and $β_{21}$ are the residual and variance spillover effect of $ith$ firm on market sentiment.

Empirical Results

Univariate Analysis

Table 3 exhibits the results of GARCH for period A, and the intercept demonstrates the statistical significance in nearly all sectors across period 1, implying that the current mean return relies on previous mean returns. The $α$ parameter in PF5, PT6, PT9, and PTT2, as well as the $β$ in PT3, PA3, firms shows insignificance. The remaining reveals that the tourism-related firms’ current conditional volatility highly relies on past-volatility rather than past-shock. Additionally, the α + β coefficient except in PF8, PF10, the remaining are less than 1, indicating a tendency for mean reversion in the upcoming days. Regarding external shocks $Ψ_{1}$ is nearly significant in all sectors but predominantly exhibits negative coefficients, reflecting a bearish sentiment observed in all, except PT sector. These findings align with Kou et al. (2021), who observed that amid the COVID-19 pandemic, technological firms exhibit less sensitivity to investor sentiment shocks in relevant other sectors. $Ψ_{2}$ is statistically significant in eighteen firms, the PTB, sector relies at a 1% level. However, $Ψ_{3}$ exhibits no discernable relationship in five firms, while PT, and PTB, sectors are nearly witnessed positively compared to others. In comparison to $Ψ_{2}$ (individual sentiment), $Ψ_{1}$ (market sentiment) and $Ψ_{3}$ (fear index) are crucial predictors for PSX stock return volatility. The outcomes are in path with; rationality depends upon streaming information’s during the pandemic. Further this we replicate with D. H. Vo et al. (2022) that COVID-19significantly impact the current conditional volatility. Moreover, both bullish, bearish sentiments observed but particularly the bearish MS prevails which induce pressure on PSX tourism-related stock price fluctuation.

Table 3.

GARCH(1, 1) Segment A.

s.no	$ω$	α	β	α + β	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{3}$
PF1	.005***	.111**	.856***	.967	.014 ***	−.001*	−.703***
PF2	.004**	.054*	.937***	.991	−.009***	−.005***	−.177**
PF3	.033***	.720***	.251***	.971	−.002***	.002**	−.246**
PF4	.046***	.363**	.596***	.959	.052**	−.006	.678***
PF5	.012**	−.007	.983***	.976	−.001***	−.061	.016
PF6	.001***	.196***	.788*	.984	.003***	−.068**	−.072
PF7	.002***	.358*	.573***	.931	−.051***	−.006	−.305***
PF8	.004***	.199***	.830***	1.029	−.025***	−.004	−.352***
PF9	−.006**	.228**	.743*	.971	.027***	.010*	.420***
PF10	.052***	.121**	.853***	.974	−.014***	−.013*	−.741***
PT1	.006**	.220***	.803**	1.023	.790***	−.115***	−.197***
PT2	.001**	.039*	.907***	.946	.213***	.009	−.328**
PT3	.000*	.910***	.071	.981	.066***	.002**	−.003***
PT4	.002**	.166*	.808***	.974	.136**	−.004	.017
PT5	.007*	.661***	.283***	.944	.287***	.038***	.081***
PT6	.009***	.024	.907***	.931	.138***	.005**	.024
PT7	.007***	.090**	.867***	.957	.219***	.002	.049***
PT8	.004**	.251***	.721***	.972	.326***	−.001	.136***
PT9	.026***	.091	.773***	.864	.056**	.013*	.023**
PT10	.006**	.548***	.371**	.919	.051**	−.105***	.045***
PA1	.005**	.048**	.932***	.98	−.009***	−.058*	−.774***
PA2	.016***	.155***	.801***	.956	−.018**	−.006	−.022
PA3	.000***	.115**	.871	.986	−.003***	−.015	.558***
PA4	.004*	.437***	.486***	.923	.003***	.011**	.154**
PA5	.023***	.281***	.576***	.857	−.011***	−.065***	−.248***
PA6	.032***	.167*	.460**	.627	−.131***	−.016	−.426***
PTT1	.003***	.561***	.431***	.992	−.017***	−.002*	−.397***
PTT2	.006***	.026	.933***	.959	−.027***	−.006	−.637***
PTB1	.020	.701***	.214**	.915	.096***	−.009***	2.117***
PTB2	−.001**	.131**	.727***	.858	−.005**	−.005***	.149***

Note.*,**,***, 1%, 5%, 10% significance, P stand for Pakistan, PF(food), PT(technology), PA(average tourism), PTT(Travel& transport), PTB(Tobacco) sectors. $Ψ_{1}$ (Market sentiment) $Ψ_{1}$ (Individual sentiment) $Ψ_{1}$ (Fear Index).

Table 4 demonstrates that the impact of lagged-variance is almost overweight on current conditional volatility as compared to previous squared residuals. The ( $α + β$ ) exhibits mean reversion. However, the PT2, 5, PA3 coefficient value exceeds 1, which indicates the abnormal condition. Notably $φ_{m} \neq 0$ in eighteen firms show asymmetric behavior, but $φ_{m} < 0$ in PF2, 6, 8, PT6, and PTT1, signifying the presence of leverage effect in respective firms’ operations. This indicates that most tourism-related firms in PSX have less leverage inclusion in their operations. Similar outcomes Çelik (2020) reported for Turkish tourism firms amid the pandemic E-GARCH indicated a minor leverage effect in firm operations. $Ψ_{1}$ , denotes market sentiment is nearly significant in all sectors, within PT sector having five and PF having six positive coefficients. While most other sectors exhibit negativity, the PTT and PTB exhibit 1% significance. The outcomes are consistent with K.-L. Peng et al. (2023), that amid the pandemic, tourism-related stocks in China are significantly affected by investor sentiment. $Ψ_{2}$ , is momentous in seventeen firms but only PF1, PT2, PA4, 5, and PTT1, having positive coefficients. $Ψ_{3}$ , is significant for twenty-four firms, except PF1,7,9 PT1,5,9 PA3, and PTB 2, the remaining sixteen carry negative coefficients. This signifies that global disease infection negatively impacts PSX tourism-related firms’ volatility (Shaikh, 2022). Further, these findings persisted with Fakhfekh and Jeribi ( 2020) and Umar et al. (2021)that both positive and negative asymmetry were observed, further Amidst the outbreak, the influx of rumors substantially affects noise players’ behavior, driven by emotions rather than rationality, they overreact in hunt of elevated expected returns, ultimately their frequent trading leading market to instability and fluctuation.

Table 4.

E-GARCH (1,1).

s.no	$ω$	α	β	α + β	$φ_{m}$	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{3}$
PF1	.083***	.124***	.383***	.507	.013***	.015***	.011***	.059*
PF2	.091***	.042*	.843***	.885	−.067***	.071***	−.050	−.635***
PF3	.044***	.236***	.333***	.579	.098***	.082*	.008	−.740***
PF4	.027***	.615***	.268**	.833	.975***	.064*	−.016*	−.344***
PF5	.016**	−.053**	.987***	.934	.017**	−.021***	−.069	.017
PF6	−.005*	−.028*	.994***	.966	−.010*	−.086**	−.013*	.077
PF7	.081***	.319***	.596***	.915	.991***	.011***	−.052*	.080***
PF8	.220**	.094	.884***	.978	−.088***	−.031	−.032	−.232***
PF9	−.005**	.517***	.422***	.939	.002	.015***	−.033**	.324***
PF10	.084**	.159***	.386***	.545	.011***	−.010***	−.014***	.029
PT1	.119**	.087*	.838***	.925	−.057	.076***	−.001***	7.676***
PT2	.028***	.223**	.816***	1.039	−.135	.139***	.004***	−12.231***
PT3	.071***	.735***	.144**	.879	.094***	.058***	−.016**	−1.722***
PT4	.017**	.351***	.609***	.96	−.039	.055***	.009	.017
PT5	.018***	.193**	.832***	1.025	−.066	.085***	−.009*	8.772***
PT6	.005***	−.027*	.993***	.966	−.034***	−.003***	−.004	−.509***
PT7	.011***	−.037*	.991***	.954	.002	−.078***	−.003	−5.328***
PT8	.022***	.536***	.369**	.905	.027**	−.007**	−.041	−.230***
PT9	.010*	.499***	.447***	.946	−.100	−.088**	−.023**	.274***
PT10	−.090**	.053*	.882***	.935	.008*	−.001***	−.004	.0535
PA1	.013***	.274***	.691***	.965	.024***	−.046*	.046	−.138**
PA2	.016***	.145*	.585**	.737	−.0136	−.0135**	−.004*	−.091
PA3	.002	.189***	.841***	1.03	.148***	.019***	−.039**	.074**
PA4	.039	.616***	.364**	.98	−.268	−.028***	.006*	−3.303***
PA5	.052***	.775***	.071	.846	.794***	−.052***	.010***	−.354***
PA6	.037***	−.033*	.991***	.958	−.002	−.014***	−.006	−.438***
PTT1	.088***	.333***	.415***	.748	−.897***	−.003***	.030**	−1.189***
PTT2	.025***	−.040*	.940***	.90	.066	−.013***	−.042	−.257***
PTB1	.019***	.520***	.332**	.852	1.052***	.025***	−.093***	−.174***
PTB2	.035***	.556***	.387***	.943	.079	−.067***	−.006	8.358***

Note.*,**,*** subscript 1,5,10% further $Ψ_{1}$ (Market sentiment) $Ψ_{2}$ (individual Sentiment) $Ψ_{3}$ (Fear Index).

Table 5. Presents GJR-GARCH (A). The lagged volatility to the current conditional volatility consistently leads over the lagged shock. Additionally, with the exception of PF1, 9 PT2, PA4, the absolute (α + β) exhibit mean reverting behavior. The $δ_{n} \neq 0$ for twenty-two firms, while $δ_{n} > 0$ in nineteen firms; in terms of sectors PF has eight out of ten, and PTT has both firms exhibit asymmetry. This suggests that previous bad news has more impact than previous good news on the current conditional volatility of PSX tourism-related stocks, particularly the firms in these two sectors. Similarly, Puzanova and Eratalay (2021) examined the impact of investor sentiment on two Swedish banks using GJR and observed that unknown shocks elevate current volatility. $Ψ_{1}$ Except in PTB1, is significant in all sectors, PT sector exhibits positivity signifies well performance during the outbreak. Among all sectors $Ψ_{2}$ among nineteen firms, PF sector witnessed the most relevant. $Ψ_{3}$ Except in PA2, and 6, significance is observed in all sectors; these outcomes align with previous GARCH and E-GARCH that rationale investors stopped their trading activities due to the fear of the pandemic. This implies that PT sector performed well during the outbreak, with similar results reported by Liew (2022) for the gold sector, which is a safe opportunity for investment during the outbreak. The summary demonstrates that $Ψ_{3}$ surge the volatility, in the same vein, bad asymmetry observed. $Ψ_{1}$ Consistently prevails $Ψ_{2}$ exhibits that herd mentality has a strong bearish impact on the current conditional volatility of PSX tourism related stocks, which accepts our first two hypotheses. Outcomes are consistent with Shaikh (2022) amid COVID-19; the current volatility is highly dependent on past volatility. Further, the asymmetry and investor sentiment heightened the return volatility of the S&P 500 energy sector.

Table 5.

GJR-GARCH (1, 1).

s.no	$ω$	α	β	α + β	$δ_{n}$	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{3}$
PF1	.071***	.270**	.792***	1.062	−.223*	−.010***	−.011**	.0598
PF2	.006***	.127***	.857***	.984	.0551**	−.044***	−.004**	−.451***
PF3	.005***	.636***	.279***	.915	−.235	−.002***	−.003	.367***
PF4	.047***	.225*	.614***	.839	.618*	.091***	−.012*	−.687***
PF5	.081***	.126*	.711***	.837	.717*	−.018***	−.017	−.594***
PF6	.039***	.312**	.597***	.909	.647**	−.083***	.021	−.405***
PF7	.002***	.236*	.561***	.797	1.022*	−.086***	.064*	−.312***
PF8	.003***	.124**	.637***	.761	.086**	−.012***	−.040**	−.486***
PF9	−.014*	.233***	.794***	1.027	.205**	−.023***	−.004**	.400***
PF10	.060***	.166**	.752***	.918	.233*	.018***	−.002**	−.835***
PT1	.015***	.752***	.238*	.99	.065	.075***	−.006	3.039***
PT2	.002***	.835***	.191**	1.026	−.016	.134***	.022**	−14.025***
PT3	.010***	.607***	.357***	.946	.056**	.077***	.001***	17.219***
PT4	.017***	.505***	.230***	.735	.112*	.054***	−.017	.244**
PT5	.017***	.312*	.659***	.971	.073	.085***	−.005	8.922***
PT6	.006**	.759***	.227**	.986	−.007	.128***	−.004***	9.185***
PT7	.024**	.184*	.563***	.747	.061**	−.008***	−.006**	−.313***
PT8	.003***	.218*	.620***	.836	.080*	−.013**	−.037	−.497***
PT9	.002*	.216***	.782***	.998	.286**	−.010***	−.002**	.332***
PT10	.004***	.086*	.857***	.943	.158*	−.013***	−.015**	−.617***
PA1	.047***	.245*	.642***	.887	.461*	−.001***	−.006	−.664***
PA2	.002***	.114***	.875***	.989	.025	−.019**	.055**	−.121
PA3	.058*	.094**	.880***	.974	.348**	−.002***	.041*	.528***
PA4	.026*	.521***	.489***	1.01	.111**	.035***	.057**	.231***
PA5	.016***	.195**	.765***	.96	.058***	−.001***	−.007	−.11*
PA6	.054*	.155**	.809***	.964	−.021	−.50**	.076*	−.127
PTT1	.013**	.181*	.801***	.982	.039**	−.008***	−.005	−1.170*
PTT2	.069***	.011	.946***	.957	.037*	−.018***	−.008	−.995***
PTB1	.004**	.365***	.624***	.989	2.154***	−.038	−.006***	.456***
PTB2	.019***	.606***	.211**	.811	.124	.079***	.002***	2.833***

Note. *,**,*** subscript 1%, 5%, 10% further $Ψ_{1}$ (Market sentiment) $Ψ_{2}$ (individual Sentiment) $Ψ_{3}$ (Fear Index).

Table 6 displays the finding of GARCH period 2 (internal crises); intercept is nearly significant in all. $α$ is insignificant in PF4, PT2,4 PA5 and PTT2, $β$ is irrelevant in PF3. Moreover, the previous lag variance substantially affects the current conditional volatility. The absolute value of $(α + β)$ in five firms displays abnormal fluctuation, while the remaining firms demonstrate gradual mean reversion in the next few days. Higher absolute values indicate a slower reversion speed. Similarly reported by D. H. Vo et al. (2022) that current conditional volatility is the cause of past volatility rather than past shock . $Ψ_{1}$ Is statically significant in entire sectors except in PF10; the coefficients are almost negative. $Ψ_{2}$ Is significant in sixteen firms and mostly carries negative coefficients except in the PTT sector; these findings correspond with Parveen et al. (2023), that crises affected the investors’ emotional intelligence, and their overreaction deviated from the intrinsic value of stock prices in PSX. $Ψ_{3 b}$ is momentous in all sectors except in PA 3 and PTTB1, it exhibits a downtrend in the market. This we replicate with Hussain Mirza and Kumar (2023) findings; the vote of the no-confidence motion, fuel uncertainty and the per capita income declined by 20%; however, the currency depreciated by approximately 40%. Further $Ψ_{1}$ Succeeds over $Ψ_{2}$ in both periods’ exhibits, herding behavior strongly impacts current conditional volatility in PSX-listed tourism-related stocks. These outcomes are consistent with Ryu et al. (2023) and ur Rehman (2021); elevated uncertainty reduces available information, as herd mentality engage in noise trading by instigating market volatility.

Table 6.

GARCH(1,1) period 2 (Internal crises).

s.no	$ω$	α	β	α + β	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{3 b}$
PF1	.007***	.402***	.309*	.711	−.075***	.006	.040***
PF2	.044***	.604***	.412***	1.021	−.028***	−.010***	.087***
PF3	.013**	.763***	.225	.988	−.007**	.023***	.044***
PF4	.066**	.291	.769***	1.06	.002*	.022	−.005*
PF5	.009***	.524***	.387***	.911	−.007***	.021	.015***
PF6	.005*	.230***	.587***	.817	.088**	−.003	.088**
PF7	.050***	.072*	.910***	.982	−.037***	−.003**	.076***
PF8	−.001**	.118***	.884***	1.002	.008***	−.004	.015**
PF9	.016***	.275***	.707***	.982	−.013***	−.036	.023***
PF10	.003*	.041*	.865***	.906	−.019	.001	.012*
PT1	.060***	.097**	.893***	.99	−.039***	−.003	.085***
PT2	.010**	.056	.945***	1.001	−.085***	.046*	.015***
PT3	.007**	.612***	.186*	.802	−.092***	−.016**	.013**
PT4	.012***	.107	.874***	.981	−.092***	−.006*	.018***
PT5	.001***	.050*	.922***	.972	−.094***	−.031	.0182***
PT6	.010**	.130*	.839***	.969	−.734***	.0106**	.014**
PT7	.011***	.074*	.841***	.915	−.091***	−.008*	.016***
PT8	.023***	.340***	.679***	1.019	−.008***	.059	.009***
PT9	−.089**	.085**	.907***	.992	.089***	−.072	.008***
PT10	.048**	.601***	.217**	.824	.029**	−.038**	.092***
PA1	.016**	.087*	.914***	1.001	−.017***	−.002*	.034***
PA2	.033**	.085*	.884***	.969	−.026***	.004*	.053***
PA3	.006*	.150**	.704***	.854	−.017*	.001	.017
PA4	.019**	.437***	.527***	.964	.011***	.024***	−.027***
PA5	−.006*	.022	.909***	.931	−.014***	−.001	.022**
PA6	.006**	.259**	.693***	.952	−.010***	−.027*	.094**
PTT1	.007*	.204*	.658***	.862	−.011**	−.004*	.021*
PTT2	.018**	.036	.887***	.923	.018***	.002	.036***
PTB1	.001**	.112**	.878***	.99	.015*	.003**	.018
PTB2	.004**	.283**	.641***	.924	−.011**	.002*	.078*

Table 7. Presents the E-GARCH estimation amid internal crises; the comparison of α and β signifies that the current conditional volatility of tourism-related stocks significantly relies on past-volatility. (α + β) Abnormal behavior was detected in five firms, particularly in the PF sector, leading to prolonged market volatility where even minor shocks could create turbulence and will have a long-lasting effect on the current conditional volatility. Further, the asymmetric parameter $φ_{m} \neq 0$ in 23 firms, but $φ_{m} < 0$ in PF 6, PT4, PA3, 5, 6 PTT2, and PTB2, exhibit the presence of leverage effect in PSX firms operations, similar to Liew (2022), that the current conditional volatility is also the cause of leverage in firm operations. The coefficient of $Ψ_{1}$ , is significant consistently in all sectors, except PTB1, featuring principally bearish behavior in all sectors. is relevant in sixteen firms, PF, PTB sectors are more sensitive. $Ψ_{3 b}$ Except for PT6, PTB1 is significant in all sectors, meanwhile PF sector demonstrates significance at 1%. Recently Feng et al. (2022) concluded that in the hoteling sector, EPU significantly affects volatility amid crises. The summary exhibits that harmful asymmetry (the leverage effect) increases as compared to segment A. The significance of $Ψ_{3 b}$ exhibits that EPU creates bubbles and stops investors from trading due to unknown market situations, consequently reducing expected returns. Consistently $Ψ_{1}$ and $Ψ_{3 b}$ are both strong predictors indicates, during high uncertainty, herding behavior has bearish influence on the current conditional volatility of PSX tourism-related firms. The outcome are also consisted with (Jiang et al., 2021; Paramanik & Singhal, 2020).

Table 7.

E-GARCH (1,1) Internal Crises Estimations.

s.no	$ω$	α	β	α + β	$φ_{m}$	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{3 b}$
PF1	−.024***	−.083**	.907***	.824	.104**	−.013***	.003**	.042***
PF2	.006***	.206***	.809***	1.015	.414***	−.002**	−.006**	−.71***
PF3	.007**	.066**	.883***	.949	.043***	−.008***	.001**	−.007**
PF4	.039***	.290**	.689***	.979	.024	.011**	.005**	−.030**
PF5	−.010**	.170**	.785***	.955	.785***	−.016**	−.061	.031**
PF6	.190***	.327**	.663**	.989	−.024**	−.184***	.046**	−.127***
PF7	.014**	.078	.962	1.04	.063**	−.015***	.001	.029***
PF8	.002**	.191**	.774***	.964	.075**	−.098***	.069	.018***
PF9	.020**	.178**	.809***	.987	.090**	−.022***	−.028**	.046***
PF10	.005**	.200***	.863***	1.063	.077**	−.047***	−.001	.085***
PT1	−.028**	.353**	.494**	.847	.025**	.021***	.002	.044***
PT2	.099**	.173**	.807***	.98	.045	−.032***	−.012	.057***
PT3	.012***	.234***	.547***	.781	.743***	−.120***	.011***	.028***
PT4	.001***	−.027	.998***	.971	−.197***	−.001***	−.018***	.027***
PT5	.016***	.186**	.803***	.989	.020	−.012**	−.001	.020**
PT6	.004***	.136***	.802***	.932	.073***	.005**	.013**	.006
PT7	.017**	.091**	.886***	.977	.030	−.018***	.001	.038***
PT8	.006	.339**	.657***	.996	.088**	−.010**	.002	.024**
PT9	.011**	.216***	.762***	.978	.131***	−.020***	−.004**	.038***
PT10	.009**	.116**	.799***	.915	0129**	−.009**	−.001	.029**
PA1	.011**	.101**	.885***	.986	.051**	.053***	−.016	.028***
PA2	.033**	.126**	.968***	1.094	.035	−.027***	.045**	.55***
PA3	.009**	.322**	.667***	.989	−.017**	−.016**	−.011	.028**
PA4	.033***	.250***	.716***	.966	.025	.012**	.037***	−.036***
PA5	−.009**	−.173***	.931***	.758	−.064***	−.054***	−.004	.259***
PA6	.089**	.243**	.650**	.893	−.159**	−.024***	−.033**	.022***
PTT1	.004**	.128**	.899***	1.027	.023	−.014***	−.005**	.028**
PTT2	.020**	.123**	.862***	.895	−.159**	.018***	.021	.038***
PTB1	.010**	.216**	.764***	.98	.265**	−.035	.040**	.010
PTB2	.06***	.206***	.729***	.935	−.414***	.002**	−.006**	−.071***

Note.*,**,*** subscript 1%, 5%, 10% further $Ψ_{1}$ (Market sentiment) $Ψ_{2}$ (individual Sentiment) $Ψ_{3 b}$ (EPU).

Table 8. Presents the finding of GJR-GARCH (B). The past-volatility prevails over past-shock. The Absolut (α + β) value in PF1,7, PT3, PA2, 4 and PBT2 exhibits abnormal volatility; the remaining are mostly close to unity but expected to mean-revert in near future. Further, $δ_{n} > 0$ in eleven PF, PT sectors exhibit 4 firms with harmful asymmetric shock. The shock is equal to $α + δ_{n}$ Signify the previous unknown negative shock prevails over the same magnitude of positive shock on current conditional volatility. Similar results presented by Çelik (2020) that unknown shock heightened the fluctuation in the tourism sector. $Ψ_{1}$ is momentous in all sectors, and negativity is observed in nearly all sectors. $Ψ_{2}$ is observed momentously in eighteen firms, with PF having five and PT having four significant coefficients. $Ψ_{3 b}$ is significant in all except in PTB2, partially align with Ashraf and Shen (2019), that any internal political unrest in a country significantly affects its stock market. It also proved our third hypothesis and connected with GARCH, E-GARCH period 2. Similarly, Ryu et al. (2023) and H. Vo et al. (2021), stated that herding behavior creates bubbles; in the same vein, uncertainty also strongly influences the current conditional volatility. Further, the bad news impact counted less compared to segment (A), and the PT sector was robust in PSX interims of volatility. Herd mentality remains consistently bearish, a strong predictor in both periods, reflecting surge volatility in PSX.

Table 8.

GJR-GARCH (1,1).

s.no	$ω$	α	β	α + β	$δ_{n}$	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{3 b}$
PF1	−.023***	.082**	.938***	1.02	−.135**	.0134***	.002*	.042**
PF2	.003**	.628***	.313***	.941	.316**	−.008***	−.003**	.023**
PF3	−.012***	.504***	.302**	.806	−.013	−.006**	.017***	.043***
PF4	.040***	.108**	.818***	.926	.047	.012**	.005*	−.031*
PF5	.010*	.105**	.791***	.896	.024	−.017***	−.002	.033***
PF6	.017**	.153**	.835***	.988	.101*	−.018***	−.001	.038**
PF7	.004***	.114***	.911***	1.025	.115**	−.401***	.006*	.007***
PF8	.008***	.062*	.906***	.968	.241***	−.007***	.008**	.012**
PF9	.017**	.151**	.805***	.956	−.146**	.021***	−.002	.043**
PF10	.003*	.146***	.817***	.963	.137	−.003**	−0001	.006**
PT1	−.006***	.111*	.816***	.927	.468**	−.035***	.037***	.084***
PT2	.032**	.098**	.899***	.997	−.068	−.29***	.001	.059***
PT3	.035***	.927***	.125***	1.052	−.029*	−.028**	.007*	.055***
PT4	.002***	.116***	.790***	.906	.566**	−.015***	−.023***	.033**
PT5	.015***	.022	.844***	.866	.678**	−.001***	.005**	.022***
PT6	.002***	.053*	.940***	.993	.331***	.11***	.017***	.002***
PT7	.018**	.114*	.883***	.997	−.030	−.019**	.001	.039***
PT8	.006*	.251***	.728***	.979	−.162*	.010**	.018	.024**
PT9	.018***	.163	.812***	.975	−.128***	−.025***	−.004**	.047**
PT10	.005**	.642***	.178**	.82	.056	−.003*	−.003	.008**
PA1	.013**	.086**	.906***	.992	−.052*	−.060***	−.018*	.030***
PA2	.033**	.116**	.887***	1.003	−.094	.028**	.004*	.054***
PA3	.004	.163*	.679***	.842	.084**	−.913**	.016	.020*
PA4	.0322***	.443***	.569***	1.012	.063	.012**	.014***	−.032***
PA5	.008	.051	.946***	.997	.051*	.046***	−.007	.024**
PA6	.013**	.095*	.698***	.793	.048	−.049***	−.003**	.027***
PTT1	.007	.741***	.042	.783	.0742**	−.016**	−.006**	.033**
PTT2	.020**	.113*	.876***	.989	.045	−.018***	.002	.038***
PTB1	.004	.123**	.864***	.987	−.046	−.014**	.001	.025*
PTB2	.005	.140**	.883***	1.023	−.058**	−.010*	.004*	.014

Note. *,**,*** subscript 1%, 5%, 10% further $Ψ_{1}$ (Market sentiment) $Ψ_{2}$ (individual Sentiment) $Ψ_{3 b}$ (EPU).

Univariate Summary Tables

Table 9 presents that the current return volatility highly relies on past-volatility. The overall surge was perceived in leverage shocks, while notably, the harmful news impact diminished from P1 to P2. Further $Ψ_{1}$ revealed paramount significance across the periods, besides enough up-down trends detected in $Ψ_{1} and Ψ_{2}$ from P1 to P2. The results presented in Table 10, $Ψ_{3}$ in P1, and $Ψ_{3 b}$ in P2, demonstrates statistical significance in GJR-GARCH estimation, in addition, prominent negativity observed in the E-GARCH estimation for both $Ψ_{3}$ in P1 and $Ψ_{3 b}$ in P2.

Table 9.

Summary of Univariate GARCH Family Models.

Sectors	firms	P1	P2	P1	P2		P1		P2	P1	P2
		$α$	$α$	$β$	$β$		$Ψ_{1}$		$Ψ_{1}$	$Ψ_{2}$	$Ψ_{2}$
Food	10	9	9	10	9		10		9	6	3
Technology	10	8	8	9	10		10		10	6	6
Average tourism	6	6	5	5	6		6		6	3	4
Travel & Transport	2	1	1	2	2		2		2	1	1
Pakistan tobacco	2	2	2	2	2		2		2	2	2
E-GARCH(1,1)
		$α$	$α$	$β$	$β$	$φ_{m}$ < 0	$φ_{m}$ < 0	$Ψ_{1}$	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{2}$
Food	10	9	9	10	9	3	1	9	10	6	6
Technology	10	10	9	10	10	1	1	10	10	5	4
Average tourism	6	6	6	5	6	0	3	6	6	4	3
Travel and Transport	2	2	2	2	2	1	1	2	2	1	1
Pakistan tobacco	2	2	2	2	2	0	1	2	1	1	2
GJR-GARCH(1,1)
		$α$	$α$	$β$	$β$	$δ_{n}$ > 0	$δ_{n}$ > 0	$Ψ_{1}$	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{2}$
Food	10	10	10	10	10	8	4	10	10	7	6
Technology	10	10	8	10	10	6	4	10	10	6	6
Average tourism	6	6	5	6	6	4	2	6	6	4	4
Travel and transport	2	1	2	2	1	2	1	2	2	2	1
Pakistan tobacco	2	2	2	2	2	1	0	1	2	1	1

Note. P1, P2 stand for global and internal crises, the $φ_{m}$ < 0, $δ_{n}$ > 0, only added harmful asymmetry.

Table 10.

Negative Shocks Summary of Univariate GARCH Family Models.

sectors	firms	P1	P2	P1	P2	P1	P2	P1	P2
		$α$	$α$	$β$	$β$	$Ψ_{1}$	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{2}$
Food	10	0	0	0	0	6	6	4	2
Technology	10	0	0	0	0	0	8	2	5
Average tourism	6	0	0	0	0	5	5	2	2
Travel and Transport	2	0	0	0	0	0	1	1	1
Pakistan tobacco	2	0	0	0	0	0	1	2	0
EGARCH(1,1)
		$α$	$α$	$β$	$β$	$Ψ_{1}$	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{2}$
Food	10	2	1	0	0	4	9	5	2
Technology	10	2	1	0	0	5	8	4	2
Average tourism	6	0	1	0	0	5	3	2	1
Travel and Transport	2	0	0	0	0	2	1	0	1
Pakistan tobacco	2	0	0	0	0	1	1	1	1
GJR-GARCH(1,1)
		$α$	$α$	$β$	$β$	$Ψ_{1}$	$Ψ_{1}$	$Ψ_{2}$	$Ψ_{2}$
Food	10	0	0	0	0	8	7	6	1
Technology	10	0	0	0	0	4	8	4	2
Average tourism	6	0	0	0	0	5	3	0	2
Travel and transport	2	0	0	0	0	2	2	2	1
Pakistan tobacco	2	0	0	0	0	1	2	2	0

Note. P1, P2 stand for global and internal crises, 0 stance for non-negative coefficients.

Table 11 summarize that own lagged-shock and volatility except in EGARCH, the rest of all observed positive The overall negativity of $Ψ_{1}$ (market sentiment) increased and $Ψ_{2}$ (individual sentiment) decreased from P1 to P2.The GARCH (1, 1) depicts an increase of $Ψ_{1}$ in PT, PTT and PTB, while $Ψ_{2}$ diminished in PF and PTB sectors. EGARCH consistent with GARCH $Ψ_{1}$ negativity increased particularly in PF and PT sectors, in addition $Ψ_{2}$ taper in PF, PT, PA sectors from P1 to P2. The GJR-GARCH is in the same vein as GARCH and EGARCH, where the coefficient negativity of market sentiment surges and individual sentiment diminishes from P1 to P2.

Table 11.

The Fear Index and EPU Summary of GARCH Family Models.

Model	firms	GARCH	E-GARCH	GJR-GARCH	GARCH	E-GARCH	GJR-GARCH
Periods		P1	P1	P1	P2	P2	P2
coefficients		$Ψ_{3}$	$Ψ_{3}$	$Ψ_{3}$	$Ψ_{3 b}$	$Ψ_{3 b}$	$Ψ_{3 b}$
Food	10	8	7	9	10	10	10
Technology	10	8	8	10	10	9	10
Average tourism	6	5	5	4	5	6	6
Travel and transport	2	2	2	2	2	2	2
Pakistan tobacco	2	2	2	2	1	1	1
Negative shocks
Food	10	6	4	7	1	4	1
Technology	10	3	5	4	0	0	0
Average tourism	6	4	4	2	1	1	1
Travel and transport	2	2	2	2	0	0	0
Pakistan tobacco	2	0	1	0	0	1	0

Note. P1, P2 stand for period 1&2, $Ψ_{3}$ stance fear index (global crises) in P1, $Ψ_{3 b}$ outlook EPU (internal crises) in P2.

Table 12 exhibits the findings of the BEKK-GARCH model, the main diagonal elements $α_{11} α_{22}$ denotes the MS (market sentiment) and PSX firm shock, similarly $β_{11}$ and $β_{22}$ denotes the volatility persistence of MS and all firms, witnessed positive and statistically significant, confirming that both market sentiment and PSX firms are highly relying on lagged-shock and volatility; in their most recent study, L. Lei et al. (2023) reported that the current conditional volatility of PSX is the cause of previous lagged shock and volatility; however, the lagged volatility prevails.

Table 12.

BEKK-GARCH Model.

s.no	$α_{11}$	$α_{12}$	$α_{21}$	$α_{22}$	$β_{11}$	$β_{12}$	$β_{21}$	$β_{22}$
PF1	.235**	−.031**	.133	.321***	.943***	−.019*	.016	.866***
PF2	.289***	.145***	−.006*	.952***	.936***	−.171***	.014	.145*
PF3	.107***	.018	.013	.127*	.898***	−.005*	.018*	.912***
PF4	.117**	.083	.037**	.161***	.910***	.007	−.054*	.925***
PF5	.345***	.024*	−.006**	.553***	.811***	−.004*	.009*	.730***
PF6	.149***	.020**	.007	.261***	.902***	.015	−.016***	.873***
PF7	.165***	.016*	−.004**	.282**	.887***	.002	−.006*	.873***
PF8	.110***	.016	.050**	.148**	.886***	−.009**	−.024**	.904***
PF9	.190***	.019**	.002**	.293***	.807***	−.023**	−.019*	.894***
PF10	.134*	.017	.006*	.198***	.915***	.004*	−.010	.835***
PT1	.149***	.020**	−.001**	.098	.935***	.008**	.001*	.903***
PT2	.314***	.014*	−.028**	.106**	.92***	−.009**	.023	.915***
PT3	.337***	.035	−.021**	.045**	.918***	−.013**	.012	.924***
PT4	.272***	.023	−.004*	.105*	.935***	−.018**	.017*	.907**
PT5	.263***	.033*	−.016**	.122***	.938***	−.025*	.009**	.944***
PT6	.295***	.064**	.011	.101**	0854***	−.011**	.002	.917***
PT7	.359***	.009**	.017*	.181**	.910***	−.005*	.001**	.896***
PT8	.295***	.021	−.072	.122*	.928***	−.015**	.018*	.881***
PT9	.340***	.020	.036**	.154*	.913***	−.007*	−.003**	.954***
PT10	.367***	.023*	−.038**	.703***	.907***	−.009**	.024*	.658***
PA1	.118***	.009*	.018	.112*	.905***	−.018**	.048	.902***
PA2	.259***	.007**	−.005**	.413***	.929***	.017	.006	.846***
PA3	.321***	.012	−.063*	.530***	.917***	−.003**	.010*	.755**
PA4	.327***	.039**	−.026**	.648***	.920***	−.015	.018**	.711**
PA5	.217***	.003	.001*	.333***	.942***	.016**	−.005*	.890***
PA6	.126***	.020**	.004**	.172*	.917***	.012	−.004**	.923***
PTT1	.195***	.017	.008*	.177**	.899***	−.004**	−.003*	.934***
PTT2	.278***	−.009***	.007	.425***	.887***	.013*	.008**	.802***
PTB1	.121***	.015*	.021**	.175**	.919***	−.021**	−.025**	.905***
PTB2	.292***	.050***	−.010**	.706***	.932***	−.044**	.026***	.561***

Note. $α_{11}$ , $β_{11}$ for MS $α_{22}, β_{22}$ are ARCH, GARCH parameters for $ith$ firm, $α_{12,} β_{12}$ and $α_{21}, β_{21}$ are cluster and persistent volatility transmission from MS to $ith$ firm and reciprocal.

The non-diagonal element $α_{12}$ lagged shock spillover effect among nineteen firms, except PF1, PTT2, the remaining has positive shock transmission from MS to firm. Further $α_{21}$ is momentous in twenty-three firms in all sectors; fourteen negative coefficients were detected. PT, PA, sectors transmitting negativity to MS. This ratifies the volatility clusters spillover effect is mostly positive from MS to firms and from firms to MS experienced mostly negative, among sectors PT transmitting negativity to MS. Further, the lagged volatility parameter $β_{12}$ is significantly spillover to twenty-four firms except PF10, PT1, PA5, PTT2 all others exhibit negative coefficients. $β_{21}$ is discerned significant in twenty-two firms, while ten firms carry negative coefficients; moreover PF sector hold five out of ten negative. It confirms that the lagged volatility spillover and negative spillover are high, ranging from investor sentiment to firm stocks, and the transmission varies across the firms. These outcomes align with (A. D. Ahmed & Huo, 2021; Trichilli et al., 2020). The short-term spillover leads from firms to sentiment with high negativity, while persistency prevails from sentiment to firms with high negativity. Persistence and positivity were observed from PT sector to MS.

Multivariate Summary Table

Table 13 presents the BEKK-GARCH summary, positive short-term spillover observed from MS to firms; the PT and PTB sectors are the most positive transmitters. The Lagged volatility spillover is highly negative among PT and PTB sectors, from firm to MS short term harmful transmission is high. Further, both short- and long-term bi-directional spillover is significantly witnessed in the PT and PTB sectors. This implies that any noise activities of herding behavior can strongly influence the conditional volatility of Pakistan’s tourism-related stocks.

Table 13.

Summary of BEKK-GARCH.

Periods	MS to firm		Firm to MS		Bi-directional spillover
coefficients	$α_{12}$	$β_{12}$	$α_{21}$	$β_{21}$	$α_{12}$ , $α_{21},$	$β_{12}, β_{21}$
Food and beverages
N-Spillover	1	6	3	6	4	4
P-Spillover	4	1	4	2	4	4
Technology
N-Spillover	0	9	6	1	5	7
P-Spillover	6	1	1	6	5	7
Average tourism
N-Spillover	0	2	3	2	3	2
P-Spillover	4	1	2	2	3	2
Travel and transport
N-Spillover	1	1	0	1	0	2
P-Spillover	0	1	1	1	0	2
Pakistan tobacco
N-Spillover	0	2	1	1	2	2
P-Spillover	2	0	0	1	2	2

Note. $α_{12}$ , $α_{21},$ and $β_{12}, β_{21}$ are only exhibits the bidirectional effect. N, P stand for negative, positive.

Conclusion

The present study concentrated on recent crises to investigate the impact of investor sentiment on PSX (Pakistan Stock Exchange) tourism-related stocks price volatility. The time spans from March 2020 to April 2023 divided into two portions to cover the most significant events of this century, such as the global pandemic and the internal political crises in Pakistan. Collected daily data for 30 firms and employed both univariate and multivariate GARCH family models. The results vary across the sectors, similar to the firm base analysis of Sarwar et al. (2019) on the Pakistan Stock Exchange (PSX). The following are the results.

The univariate-multi-variate analysis exhibits that the current conditional volatility of tourism-related firms significantly relies on past-volatility rather than lagged-shocks. Amidst the global health crisis (COVID-19), leverage and bad news were detected in the univariate analysis. The bearish sentiment observed, market-sentiment overshadowing individual sentiment, and fear-index significantly impacted the current conditional volatility.

In period B, both MS and economic policy uncertainty EPU (internal political unrest) exerted significant influence, and EPU led investor sentiment to ambiguity. In relevance to period A, the MS bearish sentiment amplified. It indicates that internal political crises are more prone to crash investors’ behavior, resulting in heightened volatility in tourism-related stocks. Conversely, the “smart lockdown” policy shielded Pakistan from economic crises, earning global commendation amid the outbreak (C. Wang et al., 2021). Furthermore, compared to period A, the asymmetry shows an increase in leverage effect from five to eight firms, accompanied by a decrease in bad news from nineteen to eleven. The PF, sector demonstrates heightened susceptibility to bad news. Despite being affected compared to period A, the PT sector demonstrated less susceptibility to these crises across the analysis. The multivariate analysis indicates a uni-bi-directional volatility transmission, with heightened short-term transmission from firm to sentiment. While the long-term spillover along with negativity observed from market sentiment to firms. Consequently, market sentiment emerges as the superior influential predictor for tourism-related sectors in the Pakistan stock exchange.

Implications

The empirical findings of this study align with prior literature; the bearish sentiment affects stock return volatility; aside from investor sentiment, the COVID-19(Fear index) and EPU have a prominent influence on current conditional volatility. The finding offers policy recommendations in the following.

Policy-makers ought to implement stringent measures to foster enduring political stability. Furthermore, it is advisable for them to educate both individual and institutional investors to augment their emotional intelligence, alleviate biases, and curtail noise trading through sensible decision-making.

Precautionary measures, such as text rate adjustment, supplementary financial grants, and lower interest rates, will encourage investment quality. Recently, in April 2023, SBP announced an all-time high interest rate of 21% (S. U. Rahman & Jan, 2023). These imprecise regulations and policies deter investment and generate uncertainty in the market.

Technology, a primarily demanded industry for tourism development, has demonstrated remarkable stability throughout the analysis. This stability positions it as a recommended opportunity for a secure investment in the Pakistan stock (PSX).

Limitation and Future Research

Tourism isn’t confined but comprises inter-related sectors, with findings extending partly to other PSX sectors. However, these limitations are emotional factors, limited time, and the unavailability of multiple data modes. Other global factors affect the return volatility; therefore, future research may involve exploring Pakistan’s reliance on oil imports. Utilizing multivariate DCC-GARCH and BEKK-GARCH can examine the joint impact of investor sentiment and crude oil spot prices on PSX return volatility, considering the Russia-Ukraine conflict.

Footnotes

Appendices

Table A3.

Clean Data Sentences.

	https://en.wikipedia.org/wiki/2022%E2%80%932023_Pakistani_economic_crisis	P	N	Nt	pt
Negative sentence
1	Pakistan’s highest inflation badly affect the economy	0.0	0.256	0.744	−0.4764
2	Covid-19 had a severe negative impact on Pakistan’s economy in 2020 and 2021	0.0	0.387	613	−0.743
3	textile industry distressed due to political unrest	0.0	0.318	0.682	−0.4221
4	For the past 25 years, Pakistan’s debt has increased about every five years, having a dreadful impact on all stockholders	0.089	0.23	0.681	−0.5106
5	The dissolution of the national assembly by the president, Dr. ArifAlvi, has seriously hampered Pakistan’s political stability	0.0	0.091	0.909	−0.1779
6	Pakistan’s foreign exchange reserves are just enough to cover five weeks’ worth of imports, which is a poor omen for the country’s financial stability	0.073	0.119	0.808	−0.296
7	If the stakeholders did not act fast and wisely, the CEO of Habib Bank warned that it might be a “devastating shock to the economy	0.088	0.283	0.629	−0.7351
8	The rupee remain worthless 20% versus the dollar in a few days when Pakistan loosened the artificial ceiling on its currency	0.0	0.123	0.868	−0.4404
9	Covid-19 BADLY hit the Pakistan stock market in the financial year 2020-2021!	0.0	0.292	0.708	−0.628
10	A default for Pakistan is a serious possibility, according to the rating agency in New York and Moody’s!!.	0.0	0.112	0.888	−0.2225
Positive sentence
1	The decision to outlaw imported luxury items is excellent for the economy of Pakistan.	0.222	0.0	0.778	0.5719
2	The China-Pakistan Economic Corridor (CPEC) will strengthen Pakistan’s economy, which is like a galaxy on earth	0.27	0.0	0.73	0.5859
3	With a 6% GDP increase, Imran Khan a former Pakistani prime minister kept the country’s economy growing.	0.235	0.0	0.756	0.235
4	From 2018 until 2022, the Khan administration granted Pakistan high political stability!!.	0.19	0.0	0.81	0.3786
5	High profits were made in the first quarter of fiscal year 2023 by the textile sector.	0.162	0.0	0.838	0.4404
6	Agriculture experienced consistent growth in FY 22, exceeding the 3.5% growth objective.	0.342	0.0	0.658	0.6369
7	Remittances increased significantly, rising by a remarkable USD 26.1 billion (7.6%) from the preceding period	0.322	0.0	0.678	0.6908
8	With an increase of 1.65% from the prior period, FDI totaled USD 1.46 billion in financial year 10 months of 2022!!.	0.126	0.0	0.874	0.4374
9	KSE-100 index climbed to 47,356 points in 2021, which is the highest growth After COVID-19	0.157	0.0	0.843	0.3818
10	Pakistan reveals that the KSE-100 index will CLIMB swiftly due to the appeal of overseas investors!!.	0.148	0.0	0.852	0.4184
https://www.centralbanking.com/organisations/state-bank-of-pakistan Negative sentence
1	Due to political turmoil, foreign investors stopped investing in Pakistan’s stock market	0.0	0.306	−0.694	−0.5267
2	The country has been affected by high inflation rates.	0.0	0.186	0.814	−0.1531
3	The oil prices have TERRIBLE implication on each corner of the country.	0.0	0.228	0.772	−0.5904
4	The IMF has put pressure on Pakistan to tighten monetary policy, and end subsidies.	0.0	0.145	0.855	−0.296
5	Honda Atlas is still closing its plant because of import problems in Pakistan!!.	0.0	0.215	0.785	−0.508
6	During the most recent quarter, the State Bank of Pakistan’s (SBP) foreign exchange reserves VERY BADLY affected and it decreased by $354 million.	0.0	0.261	0.739	−0.8143
7	The sudden economic crises in Pakistan create fear for foreign investors.	0.0	0.224	0.629	−0.2732
8	The Pakistani rupee is in a serious phase against the US dollar!!.	0.0	0.159	0.841	−0.2225
9	As a result of the SBP’s fearful taxation, the companies stopped operating.	0.0	0.362	0.638	−0.6249
10	The tsunami in Pakistan’s northern regions in 2022 had a devastating impact on the country’s already fragile economy.	0.0	0.212	0.788	−0.6486
Positive sentence
1	The CPEC flagship project would boost the economy of both nations and the positive connection.	0.391	0.0	0.609	0.7717
2	To prevent Pakistan from going into default, the IMF loan money is a positive gesture for the already fragile economy	0.217	0.0	0.783	0.5719
3	The Tribune Pakistan offers better advice on safe investment avenues in the Pakistani stock market.	0.309	0.0	0.691	0.7003
4	In comparison to the same period last year, the electricity and food & beverage sectors make 2.5% higher profit in the second and third quarters of the fiscal year 2022.	0.094	0.0	0.906	0.4404
5	China contributed US$ 249.6 million to the CPEC flagship project as a helpful gesture for Pakistan’s economy.	0.231	0.0	0.769	0.4939
6	For both domestic and foreign investors, real estate investment in Pakistan offers the best opportunities.	0.343	0.0	0.657	0.7783
7	The Technology, healthcare, and tourism industries are the most lucrative investment prospects for American-Pakistani investors to move their capital to Pakistan!!.	0.132	0.0	0.868	0.4674
8	Contributing to Roshan Digital, which has just recently seen its greatest daily deposits—$57 Million in a 24-hour period—on June 30, 2022.	0.181	0.0	0.819	0.6369
9	The financial action task force’s affirmative recommendation to remove Pakistan from the FATF’s gray list.!!	0.113	0.0	0.887	0.1984
10	Former prime minister Imran khan insisted that for better economic stability needs political stability.	0.182	0.0	0.818	0.4404

Note. N (negative), P(positive), Nt(neutral), Pt(polarity).

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by Project of Hainan Academician Innovation Platform Scientific Research (YSPTZX202210).

ORCID iD

Aziz Ullah

Data Availability Statement

Data is Available upon reasonable request.

References

Abdelhedi

Boujelbène-Abbes

(2019). Transmission of shocks between Chinese financial market and oil market. International Journal of Emerging Markets, 15(2), 262–286.

Afzal

Haiying

Afzal

Mahmood

Ikram

(2021). Value-at-risk analysis for measuring stochastic volatility of stock returns: Using GARCH-based dynamic conditional correlation model. Sage Open, 11(1), 21582440211005760.

Ahmed

A. D.

Huo

(2021). Volatility transmissions across international oil market, commodity futures and stock markets: Empirical evidence from China. Energy Economics, 93(C), 104–141.

Ahmed

(2018). Stock market return, volatility and the role of investor sentiments: A case study on Pakistan stock exchange. International Journal of Research and Scientific Innovation, 5(10), 7–16.

Ainin

Feizollah

Anuar

N. B.

Abdullah

N. A.

(2020). Sentiment analyses of multilingual tweets on halal tourism. Tourism Management Perspectives, 34, 100658.

Airlines. (2020). Potential for revenue losses of $113bn due to COVID-19 “crisis”. Retrieved December 15, 2023, from https://airlines.iata.org/2020/03/09/potential-revenue-losses-113bn-due-covid-19-crisis

Ali

(2013). Researcher reflexivity in tourism studies research: Dynamical dances with emotions. In Ateljevic

Morgan

Pritchard

(Eds.), The critical turn in tourism studies (pp. 13–26). Routledge.

Ali

Shah

Z. A.

Khan

A. U.

(2018). Post-terrorism image recovery of tourist destination: A qualitative approach using Fuzzy-VIKOR. Journal of Tourism Analysis: Revista de Análisis Turístico, 25(2), 129–153.

Arshad

M. I.

Iqbal

M. A.

Shahbaz

(2018). Pakistan tourism industry and challenges: A review. Asia Pacific Journal of Tourism Research, 23(2), 121–132.

10.

Ashraf

B. N.

Shen

(2019). Economic policy uncertainty and banks’ loan pricing. Journal of Financial Stability, 44, 100695.

11.

Aslam

Mohmand

Y. T.

Ferreira

Memon

B. A.

Khan

(2020). Network analysis of global stock markets at the beginning of the coronavirus disease (Covid-19) outbreak. Borsa Istanbul Review, 20(Suppl. 1), S49–S61.

12.

Bagchi

Chatterjee

Ghosh

Dandapat

Bagchi

Chatterjee

Ghosh

Dandapat

(2020). Impact of COVID-19 on global economy [Conference session]. Coronavirus Outbreak and the Great Lockdown: Impact on Oil Prices and Major Stock Markets Across the Globe, 15–26. Springer.

13.

Baker

Wurgler

(2006). Investor sentiment and the cross-section of stock returns. The Journal of Finance, 61(4), 1645–1680.

14.

Baker

Wurgler

(2007). Investor sentiment in the stock market. Journal of Economic Perspectives, 21(2), 129–151.

15.

Baker

Wurgler

Yuan

(2012). Global, local, and contagious investor sentiment. Journal of Financial Economics, 104(2), 272–287.

16.

Barber

Odean

(1999). Do investors trade too much. American Economic Review, 89(5), 262.

17.

Beer

Zouaoui

(2013). Measuring stock market investor sentiment. Journal of Applied Business Research (JABR), 29(1), 51–68.

18.

Bieri

(2020). An Application of Extreme Value Theory to Finance-An Empirical Study of Whether Stock Shape Parameters and the Differences Between Corresponding Actual and BSM Put Option Values Are Positively Correlated. Available at SSRN 3555755.

19.

Blackledge

Lamphiere

(2021). A review of the fractal market hypothesis for trading and market price prediction. Mathematics, 10(1), 117.

20.

Bollerslev

(1986). Generalized autoregressive conditional heteroskedasticity. Journal of Econometrics, 31(3), 307–327.

21.

Brooks

(2019). Introductory econometrics for finance. Cambridge university press.

22.

Çelik

G. G.

(2020). Volatility modelling for tourism sector stocks in Borsa Istanbul. International Journal of Economics and Financial Issues, 10(3), 158.

23.

Chen

M.-P.

Chen

P.-F.

Lee

C.-C.

(2013). Asymmetric effects of investor sentiment on industry stock returns: Panel data evidence. Emerging Markets Review, 14, 35–54. https://doi.org/10.1016/j.ememar.2012.11.001

24.

Chen

Liang

Umar

(2021). Is investor sentiment stronger than VIX and uncertainty indices in predicting energy volatility?. Resources Policy, 74, 102–123.

25.

Corbet

Hou

Oxley

(2022). Did COVID-19 tourism sector supports alleviate investor fear? Annals of Tourism Research, 95, 103434.

26.

Cui

Zhang

(2020). Does investor sentiment affect stock price crash risk? Applied Economics Letters, 27(7), 564–568.

27.

Das

Wiener

H. J. D.

Kareklas

(2019). To emoji or not to emoji? Examining the influence of emoji on consumer reactions to advertising. Journal of Business Research, 96, 147–156.

28.

Dash

S. R.

Mahakud

(2013). Investor sentiment and stock return: Do industries matter? Margin: The Journal of Applied Economic Research, 7(3), 315–349.

29.

De Long

J. B.

Shleifer

Summers

L. H.

Waldmann

R. J

. (1990). Noise trader risk in financial markets. Journal of Political Economy, 98(4), 703–738.

30.

Dragouni

Filis

Gavriilidis

Santamaria

(2016). Sentiment, mood and outbound tourism demand. Annals of Tourism Research, 60, 80–96.

31.

Ehsan

Z.-A.

(2021). An empirical analysis on the weak form market efficiency in the Bangladeshi pharmaceutical industry-A case study of Renata Ltd. East West University.

32.

Ejsmont

Milošević

Obradović

(2023). A test for normality and independence based on characteristic function. Statistical Papers, 64, 1861–1889.

33.

Engle

R. F.

Kroner

K. F.

(1995). Multivariate simultaneous generalized ARCH. Econometric Theory, 11(1), 122–150.

34.

Fakhfekh

Jeribi

(2020). Volatility dynamics of crypto-currencies’ returns: Evidence from asymmetric and long memory GARCH models. Research in International Business and Finance, 51, 101075.

35.

Fama

E. F.

(1970). Efficient capital markets: A review of theory and empirical work. The Journal of Finance, 25(2), 383–417.

36.

Feng

Sun

(2022). Oil price, economic policy uncertainty and tourism development: Evidence from China. Procedia Computer Science, 214, 560–564.

37.

Flynn

S. M.

(2003). The portfolio allocation effects of investor sentiment about the ability of managers to beat the market. IDEAS/RePEc Search.

38.

Geng

Liu

Liao

(2021). Operating performance of tourism listed companies in China: The perspective of economic value added. SAGE Open, 11(1), 2158244020981064.

39.

Ghani

Guo

(2022). Forecasting Pakistan stock market volatility: Evidence from economic variables and the uncertainty index. International Review of Economics & Finance, 80, 1180–1189.

40.

Ghorbel

Abdelhedi

Boujelbene

(2014). Assessing the impact of crude oil price and investor sentiment on Islamic indices: Subprime crisis. Journal of African Business, 15(1), 13–24.

41.

Glosten

L. R.

Jagannathan

Runkle

D. E.

(1993). On the relation between the expected value and the volatility of the nominal excess return on stocks. The Journal of Finance, 48(5), 1779–1801.

42.

Gong

Zhang

Wang

(2022). Investor sentiment and stock volatility: New evidence. International Review of Financial Analysis, 80, 102028.

43.

Guerello

(2016). The effect of investors’ confidence on monetary policy transmission mechanism: A multivariate GARCH approach. The North American Journal of Economics and Finance, 37(C), 248–266.

44.

Hasni

M. J. S.

Farah

M. F.

Adeel

(2021). The technology acceptance model revisited: Empirical evidence from the tourism industry in Pakistan. Journal of Tourism Futures. Advance online publication. https://doi.org/10.1108/JTF-09-2021-0220

45.

Herrera

G. P.

Constantino

J.-J.

Naranpanawa

(2022). Renewable energy stocks forecast using Twitter investor sentiment and deep learning. Energy Economics, 114, 106285.

46.

Hindustan Times. (2024). When NDA’s first PM Atal Bihari Vajpayee lost no-confidence motion by 1 vote. Author. Retrieved October 11, 2023, from https://www.hindustantimes.com/india-news/

47.

Hong

Lee

C.-C.

(2020). Investor herding in the China stock market: An examination of CHINEXT. Rom J Econ Forecast, 23(4), 47–61.

48.

Hsu

Y.-L.

Tang

(2022). Effects of investor sentiment and country governance on unexpected conditional volatility during the COVID-19 pandemic: Evidence from global stock markets. International Review of Financial Analysis, 82, 102186.

49.

Shao

Hua

(2021). Using investor and news sentiment in tourism stock price prediction based on XGBoost model [Conference session]. 2021 IEEE/ACIS 6th International Conference on Big Data, Cloud Computing, and Data Science (BCD), 20–24. Piscataway, New Jersey, United States.

50.

Huang

Yang

Sheng

(2014). An empirical study of the effect of investor sentiment on returns of different industries. Mathematical Problems in Engineering, 2014, 545723.

51.

Hussain Mirza

S. H.

Kumar

. (2023). Regime change in Pakistan 2022: Implications for political and economic stability. Pakistan Journal of International Affairs, 6(1), 160–174.

52.

Hutto

Gilbert

(2014). Vader: A parsimonious rule-based model for sentiment analysis of social media text. Proceedings of the International AAAI Conference on Web and Social Media, 8(1), 216–225.

53.

Iqbal

Shahzad

M. N.

Ali

Aslam

Asif

(2023). No confidence politics in Pakistan: A historical analysis. Journal of Positive School Psychology, 7(1), 869–881.

54.

Jansen

W. J.

Nahuis

N. J.

(2003). The stock market and consumer confidence: European evidence. Economics Letters, 79(1), 89–98.

55.

Jiang

Zhu

Zhang

Yan

Shen

(2021). Investor sentiment and stock returns during the COVID-19 pandemic. Frontiers in Psychology, 12, 708537.

56.

Johnk

D. W.

Soydemir

(2015). Time-varying market price of risk and investor sentiment: Evidence from a multivariate GARCH model. Journal of Behavioral Finance, 16(2), 105–119.

57.

Kaplanski

Levy

(2010). Sentiment and stock prices: The case of aviation disasters. Journal of Financial Economics, 95(2), 174–201.

58.

Katsiampa

(2017). Volatility estimation for Bitcoin: A comparison of GARCH models. Economics Letters, 158, 3–6.

59.

Khalid

Okafor

L. E.

Burzynska

(2021). Does the size of the tourism sector influence the economic policy response to the COVID-19 pandemic? Current Issues in Tourism, 24(19), 2801–2820.

60.

Khan

M. A.

Ahmad

(2018). Measurement of investor sentiment and its bi-directional contemporaneous and lead–lag relationship with returns: Evidence from Pakistan. Sustainability, 11(1), 94.

61.

Khan

N. U.

Alim

Begum

Han

Mohamed

(2022). Examining factors that influence the international tourism in Pakistan and its nexus with economic growth: Evidence from ARDL approach. Sustainability, 14(15), 9763.

62.

Khan

S. K.

Saleem

(2022). Nexus between investor sentiment and equity returns in Pakistan stock exchange. Audit and Accounting Review, 2(2), 25–47.

63.

Kim

Chang

Song

J. W.

(2020). Predicting the direction of US stock prices using effective transfer entropy and machine learning techniques. IEEE Access, 8, 111660–111682.

64.

Kling

Gao

(2008). Chinese institutional investors’ sentiment. Journal of International Financial Markets, Institutions and Money, 18(4), 374–387.

65.

Kou

Olgu Akdeniz

Ö.

Dinçer

Yüksel

(2021). Fintech investments in European banks: A hybrid IT2 fuzzy multidimensional decision-making approach. Financial Innovation, 7(1), 39.

66.

Kumari

Mahakud

(2015). Does investor sentiment predict the asset volatility? Evidence from emerging stock market India. Journal of Behavioral and Experimental Finance, 8, 25–39.

67.

Kurbatskii

Mironenkov

(2023). Estimating the quality of life using weighted principal components method. ELIT–Economic Laboratory for Transition Research Dz. Washingtona 4/5, 19(1), 7.

68.

Lee

Jeong

Joo

Yoon

Y. S.

Han

Baik

(2022). Outliers in financial time series data: Outliers, margin debt, and economic recession. Machine Learning with Applications, 10, 100420.

69.

Lee

W. Y.

Jiang

C. X.

Indro

D. C.

(2002). Stock market volatility, excess returns, and the role of investor sentiment. Journal of Banking & Finance, 26(12), 2277–2299.

70.

Lei

(2019). Black–Litterman asset allocation model based on principal component analysis (PCA) under uncertainty. Cluster Computing, 22, 4299–4306.

71.

Lei

Aziz

Sarwar

Waheed

Tiwari

A. K.

(2023). Spillover and portfolio analysis for oil and stock market: A new insight across financial crisis, COVID-19 and Russian-Ukraine war. Resources Policy, 85, 103–645.

72.

Ran

(2020). Investor sentiment and stock price premium validation with Siamese twins from China. Journal of Multinational Financial Management, 57, 100655.

73.

Liew

V. K.-S.

(2022). The effect of novel coronavirus pandemic on tourism share prices. Journal of Tourism Futures, 8(1), 109–124.

74.

Liston-Perez

Torres-Palacio

Bayram

S. G.

(2018). Does investor sentiment predict Mexican equity returns? International Journal of Managerial Finance, 14(4), 484–502.

75.

López-Cabarcos

M. Á.

Pérez-Pico

A. M.

Piñeiro-Chousa

Šević

(2021). Bitcoin volatility, stock market and investor sentiment. Are they connected? Finance Research Letters, 38, 101399.

76.

Lux

(2011). Sentiment dynamics and stock returns: The case of the German stock market. Empirical Economics, 41, 663–679.

77.

Mason

(2022). Vote of no confidence: A brief history. History Extras.

78.

Mishra

M. K.

(2020). The world after COVID-19 and its impact on global economy. ZBW – Leibniz Information Centre for Economics.

79.

Murty

Victor

Fekete-Farkas

(2022). Is Bitcoin a safe haven for Indian investors? A GARCH volatility analysis. Journal of Risk and Financial Management, 15(7), 317.

80.

Naeem

Mashwani

W. K.

Ali

Uddin

M. I.

Mahmoud

Jamal

Chesneau

(2021). Machine learning-based USD/PKR exchange rate forecasting using sentiment analysis of Twitter data. CMC-Computers Materials & Continua, 67(3), 3451–3461.

81.

Narayan

P. K.

Bannigidadmath

(2017). Does financial news predict stock returns? New evidence from Islamic and non-Islamic stocks. Pacific-Basin Finance Journal, 42, 24–45.

82.

Narayan

P. K.

Phan

D. H. B.

Liu

(2021). COVID-19 lockdowns, stimulus packages, travel bans, and stock returns. Finance Research Letters, 38, 101732.

83.

Nartea

G. V

Bai

(2020). Investor sentiment and the economic policy uncertainty premium. Pacific-Basin Finance Journal, 64, 101438.

84.

Nazir

Santos

Silveira

(2023). Heritage tourism and peace interdependency: News content analysis on the destruction and the rebuilding drives of Jahanabad Buddha Statue, Swat, Pakistan. International Journal of Tourism Cities. Advance online publication. https://doi.org/10.1108/IJTC-09-2022-0222

85.

Nelson

D. B.

(1991). Conditional heteroskedasticity in asset returns: A new approach. Econometrica: Journal of the Econometric Society, 59(2), 347–370.

86.

Nicolau

J. L.

Sharma

Shin

(2020). The tourism effect of President Trump’s participation on Twitter. Tourism Management, 81, 104133.

87.

Niu

Wang

(2021). Does investor sentiment differently affect stocks in different sectors? Evidence from China. International Journal of Emerging Markets. Advance online publication. https://doi.org/10.1108/IJOEM-11-2020-1298

88.

Nti

I. K.

Adekoya

A. F.

Weyori

B. A.

(2020). Predicting stock market price movement using sentiment analysis: Evidence from Ghana. Applied Computer Systems, 25(1), 33–42.

89.

Olugbode

El-Masry

Pointon

(2014). Exchange rate and interest rate exposure of UK industries using first-order autoregressive exponential GARCH-in-mean (EGARCH-M) Approach. The Manchester School, 82(4), 409–464.

90.

Paramanik

R. N.

Singhal

(2020). Sentiment analysis of Indian stock market volatility. Procedia Computer Science, 176, 330–338.

91.

Parveen

Satti

Z. W.

Subhan

Q. A.

Jamil

(2020). Exploring market overreaction, investors’ sentiments and investment decisions in an emerging stock market. Borsa Istanbul Review, 20(3), 224–235.

92.

Parveen

Satti

Z. W.

Subhan

Q. A.

Riaz

Baber

S. F.

Bashir

(2023). Examining investors’ sentiments, behavioral biases and investment decisions during COVID-19 in the emerging stock market: A case of Pakistan stock market. Journal of Economic and Administrative Sciences, 39(3), 549–570.

93.

Peng

Cambria

(2018). Learning multi-grained aspect target sequence for Chinese sentiment analysis. Knowledge-Based Systems, 148, 167–176.

94.

Peng

K.-L.

C.-H.

Lin

P. M. C.

Kou

I. E.

(2023). Investor sentiment in the tourism stock market. Journal of Behavioral and Experimental Finance, 37, 100732.

95.

Pilbeam

Langeland

K. N.

(2015). Forecasting exchange rate volatility: GARCH models versus implied volatility forecasts. International Economics and Economic Policy, 12(1), 127–142.

96.

Puzanova

Eratalay

M. H.

(2021). Effect of real estate news sentiments on the stock returns of Swedbank and SEB Bank. Econometric Research in Finance, 6(2), 77–117.

97.

Rahman

M. L.

Shamsuddin

(2019). Investor sentiment and the price-earnings ratio in the G7 stock markets. Pacific-Basin Finance Journal, 55, 46–62.

98.

Rahman

S. U.

Jan

(2023). Pakistan in 2023: On the brink of a full-blown crisis. India Quarterly, 79(3), 293–303.

99.

Rashid

Fayyaz

Karim

(2019). Investor sentiment, momentum, and stock returns: An examination for direct and indirect effects. Economic Research-Ekonomska Istraživanja, 32(1), 2638–2656.

100.

Rehman

M. U.

Sensoy

Eraslan

Shahzad

S. J. H.

X. V.

(2021). Sensitivity of US equity returns to economic policy uncertainty and investor sentiments. The North American Journal of Economics and Finance, 57, 101392.

101.

Renault

(2020). Sentiment analysis and machine learning in finance: A comparison of methods and models on one million messages. Digital Finance, 2(1–2), 1–13.

102.

Rohilla

Tripathi

Bhandari

(2023). Does investor’s sentiment affect industries’ return? – A case of selected Indian industries. Business Analyst Journal, 44(2), 106–127.

103.

Rudež

H. N.

(2022). The impact of the Covid-19 pandemic on investor sentiments: Evidence from 12 selected major tourism-related companies. Managing Global Transitions, 20(1), 43–58.

104.

Ryu

Yang

(2023). Whose sentiment explains implied volatility change and smile? Finance Research Letters, 55, 103838.

105.

Salisu

A. A.

Akanni

L. O.

(2020). Constructing a global fear index for the COVID-19 pandemic. Emerging Markets Finance and Trade, 56(10), 2310–2331.

106.

Sarwar

Shahbaz

Anwar

Tiwari

A. K.

(2019). The importance of oil assets for portfolio optimization: The analysis of firm level stocks. Energy Economics, 78(C), 217–234.

107.

Shahzad

S. J. H.

Caporin

(2020). On the volatilities of tourism stocks and oil. Annals of Tourism Research, 81(C), 102705.

108.

Shaikh

(2022). Impact of COVID-19 pandemic on the energy markets. Economic Change and Restructuring, 55(1), 433–484.

109.

Shiljas

Kumar

Bashir

H. A.

(2022). Nexus between Twitter-based sentiment and tourism sector performance amid COVID-19 pandemic. Tourism Economics, 29(8), 2200–2205. https://doi.org/10.1177/13548166221123102.

110.

Shrestha

(2021). Factor analysis as a tool for survey analysis. American Journal of Applied Mathematics and Statistics, 9(1), 4–11.

111.

Sigala

(2020). Tourism and COVID-19: Impacts and implications for advancing and resetting industry and research. Journal of Business Research, 117, 312–321.

112.

Song

Gong

Zhang

(2023). Investor sentiment based on scaled PCA method: A powerful predictor of realized volatility in the Chinese stock market. International Review of Economics & Finance, 83, 528–545.

113.

Sorić

(2021). COVID-19 and the stock markets: A glance at the initial effects of the pandemic. EkonomickýČasopis, 69(5), 472–493.

114.

Srinivas

Gadela

Sabu

Das

Nath

Datla

(2023). Effects of daily news sentiment on stock price forecasting. ArXiv Preprint ArXiv:2308.08549.

115.

Fang

Yin

(2017). The role of news-based implied volatility among US financial markets. Economics Letters, 157(C), 24–27.

116.

Subramaniam

Chakraborty

(2021). COVID-19 fear index: Does it matter for stock market returns? Review of Behavioral Finance, 13(1), 40–50.

117.

Trichilli

Abdelhédi

Boujelbène Abbes

(2018). Googling investor’s sentiment: Powerful measure in conventional and Islamic MENA financial markets. International Economic Journal, 32(3), 454–469.

118.

Trichilli

Boujelbène Abbes

Masmoudi

(2020). Predicting the effect of Googling investor sentiment on Islamic stock market returns: A five-state hidden Markov model. International Journal of Islamic and Middle Eastern Finance and Management, 13(2), 165–193.

119.

Umar

Mirza

Rizvi

S. K. A.

Furqan

(2021). Asymmetric volatility structure of equity returns: Evidence from an emerging market. The Quarterly Review of Economics and Finance, 87(4), 330–336.

120.

ur Rehman

M. A

. (2021). The impact of investor sentiment on returns, cash flows, discount rates, and performance. Borsa Istanbul Review, 22(1), 352–362.

121.

Vijh

Chandola

Tikkiwal

V. A.

Kumar

(2020). Stock closing price prediction using machine learning techniques. Procedia Computer Science, 167, 599–606.

122.

D. H.

C. M.

Dang

T. H.-N.

(2022). Stock market volatility from the COVID-19 pandemic: New evidence from the Asia-Pacific region. Heliyon, 8(9), e10763.

123.

Trinh

Q.-D.

Nguyen

T.-N.

(2021). Does economic policy uncertainty affect investment sensitivity to peer stock prices? Economic Analysis and Policy, 72(C), 685–699.

124.

Wang

Abbas

Duan

Mubeen

(2021). Global financial crisis, smart lockdown strategies, and the COVID-19 spillover impacts: A global perspective implications from Southeast Asia. Frontiers in Psychiatry, 12, 643783.

125.

Wang

Shi

Jiao

(2021). Sentiment analysis of rumor spread amid COVID-19: Based on weibo text. Healthcare, 9(10), 1275.

126.

Cai

Zhang

(2021). Investor sentiment and stock price crash risk in the Chinese stock market. Journal of Mathematics, 2021, 6806304.

127.

Lee

C.-C.

Xing

S.-J.

(2021). The impact of the COVID-19 outbreak on Chinese-listed tourism stocks. Financial Innovation, 7, 22.

128.

Zha

Stafylas

Liu

(2020). Dependences and volatility spillovers between the oil and stock markets: New evidence from the copula and VAR-BEKK-GARCH models. International Review of Financial Analysis, 68, 101280.

129.

Yuan

(2021). Herding effect in China’s a share stock market under COVID-19 pandemic [Conference session]. 2021 7th International Conference on E-Business and Applications, 70–76.

130.

Zargar

F. N.

Kumar

(2023). Market fear, investor mood, sentiment, economic uncertainty and tourism sector in the United States amid COVID-19 pandemic: A spillover analysis. Tourism Economics, 29(2), 551–558.

131.

Zeng

(2020). Volatility modelling of Chinese stock market monthly return and investor sentiment using multivariate GARCH models. International Journal of Accounting & Finance Review, 5(1), 123–133.

132.

Zhang

Gong

Wang

(2021). Predicting stock market volatility based on textual sentiment: A nonlinear analysis. Journal of Forecasting, 40(8), 1479–1500.

133.

Zheng

(2021). The herding behavior in Chinese stock markets: Evidence from a-share markets during COVID-19. Emerging Markets Finance and Trade, 56(15), 3578–3587.

134.

Zhou

(2018). Measuring investor sentiment. Annual Review of Financial Economics, 10, 239–259.