key: cord-1053126-w73kea4l
authors: Hadi, Dlawar Mahdi; Naeem, Muhammad Abubakr; Karim, Sitara
title: Impact of COVID-19 on the Connectedness across Global Hospitality Stocks
date: 2022-05-10
journal: Int J Hosp Manag
DOI: 10.1016/j.ijhm.2022.103243
sha: 154f31047bd40889f3ef913fe984625eafb2dc4e
doc_id: 1053126
cord_uid: w73kea4l

This study investigated the connectedness of top ten hospitality stocks in the world and the impact of the COVID-19 pandemic on this connectedness. For this purpose, we employ the time-varying parameter vector autoregressions (TVP-VAR) to examine the return connectedness among the world’s top ten hospitality stocks. We further utilize the wavelet coherence measure to test the impact of COVID-related indexes on the connectedness across the hospitality stocks from January 1, 2020 to July 16, 2021. Our findings explore a strong connectedness among the hospitality stocks, although the total connectedness index is considerably affected by the first wave, the second wave, and the approval of COVID-19 vaccines. France and UK hospitality stocks appeared to be dominant and were the highest net transmitters of spillover shocks to other sample stocks. We document that the COVID-19 pandemic is the prime driver of the hospitality stocks’ connectedness during the sample period. Aside from the contribution to hospitality and finance literature, our conclusions and implications can also benefit parties such as hospitality firm managers, investors, portfolio managers, and policymakers.

The COVID-19 pandemic has brought unprecedented challenges to the global economic world, leading to social, economic, and financial crises. Since the COVID-19 outbreak is reported to be a source of systematic risk (Shahzad et al., 2021; Balli et al., 2021; Naeem et al., 2021) , the higher uncertainty among the markets has resulted in the volatility of the stock markets, particularly when it comes to the share prices of hospitality, travel, and leisure stocks. Evidence suggests that whenever a particular pandemic hits the global world, the hospitality and tourism sector is the most affected, showing lower recovery to these pandemics (Zheng et al., 2021) . For example, about 3 million employees working for the tourism and hospitality sector lost their jobs due to the outbreak of SARS in 2002 and a remarkable economic loss of around USD 20 billion in Eastern Asia (World Travel and Tourism Council, 2003) . Similarly, a sharp decline of 22% of world tourists in the first quarter of 2020, after the COVID-19 outbreak, reflects the severity of the pandemic, specifically for the hospitality and travel sector (Sharma et al., 2021) . Reports documenting the job loss in the hospitality industry due to suggest that about one million employees lost their jobs (World Tourism Organization, 2020) . 1 Within the global hospitality and tourism industry, Europe (in particular, France and UK) is considered the chief contributor, accounting for approximately 50% of the market share of the international travel and leisure industry (Abbas et al., 2021) . 1 Please see: https://www.unwto.org/news/covid-19-international-tourist-numbers-could-fall-60-80-in-2020 J o u r n a l P r e -p r o o f

The tourism sector crucially contributes to the economic growth of developing and developed countries. Shahzad et al. (2017) argue that the tourism sector boosts the economy through various channels, such as employment, and that it generates tax revenues and catalyzes investment in human capital, infrastructure, and technology. It fosters the efficiency of domestic firms by raising competition among them and, most significantly, through the channel of service export, which reduces the deficit of the balance of payment. Moreover, several empirical studies have documented that the tourism sector drives the economy (Brida et al., 2016) . The pandemic has devasted the travel and leisure sector along with overall global economic activities.

Therefore, government intervention-applying different degrees of a myriad of fiscal, monetary, and supply-side measures-becomes necessary to save and support the survival of the tourism sector. In this context, Anguera-Torrell et al. (2020) examine the response of hotels to the COVID-19 outbreak as well as various public sector measures across various countries during the period between February 24 and April 24, 2020. The study shows that the hotel industry has been inversely impacted by the COVID-19 evolution and positively correlated to economic policies. On the same note, Yeh (2021) proposes that government-sponsored loans are crucial to the survival of the hospitality industry. Our study highlights vital implications for policymakers to consider in the current and future crises facing the hospitality sector.

In this way, the current study investigates the connectedness of the hospitality stocks of ten countries with the highest number of COVID-19 cases. The top ten major tourist destinations in terms of tourist arrivals, based on the 2020 UNWTO report, 2 are chosen in the sample. They Meanwhile, the selection of stocks is based on the destinations with the most coronavirus cases.

Moreover, we employed the unique methodology, the time-varying parameter vector autoregression (TVP-VAR) approach, to examine the connectedness of the top ten hospitality stocks. The TVP-VAR approach uniquely offers two benefits over other techniques. First, the TVP-VAR technique caters to the problem of arbitrary selection of rolling-window size; second, the model potentially avoids the loss of valuable observations to make the sample suitable (Adekoya & Oliyide, 2021) . Moreover, TVP-VAR model determines possible changes in the connectedness of hospitality stock markets to provide evidence of whether the linear structure is 2 Please see: https://www.e-unwto.org/doi/book/10.18111/9789284422456 J o u r n a l P r e -p r o o f derived from the probability of the shocks or from the extension of the change in responses. The model also offers peculiar characteristics to recognize potential structural breaks in time-varying dynamics and provides significant reasons to recognize the relationship among variables.

In addition, we employed wavelet coherence analysis to further investigate the impact of COVID-induced indexes on the connectedness of hospitality (TLS) stocks. Specifically, the COVID-19 indexes are: (1) panic index, which measures the level of news and rumors referring to panic or hysteria during coronavirus; (2) media coverage index, which calculates the percentage of all news sources covering the topic of the novel coronavirus; (3) sentiment index, which measures the level of sentiment across all entities mentioned in the news alongside the coronavirus; (4) fake news index, which measures the level of media gossip about the novel virus that makes reference to misinformation or fake news alongside COVID-19; (5) media hype index, which measures the percentage of news about the novel coronavirus; and (6) infodemic index, which calculates the percentage of all entities (e.g., places and companies) that are somehow linked to In addition, all these indexes are available from January 2020 to calculate the media sentiments and analysis during the pandemic to provide a comprehensive tool for examining the impact of media sentiment indexes on the connectedness of travel and leisure stocks during COVID-19.

In the light of the above arguments, the current study contributes to the extant literature in multiple ways. First, this is the pioneer study to examine the connectedness of the hospitality stocks during COVID-19 of the top ten countries with the highest number of coronavirus cases.

Second, the study employs the time-varying parameters vector autoregressions (TVP-VAR) approach, proposed by Primiceri (2005) and extended by Antonakakis and Gabauer (2017) , to advance in the current body of literature. Third, we employed wavelet coherence analysis to investigate the co-movements of the hospitality stocks and COVID-induced indexes. Fourth, this is the first study that cumulatively examines the impact of six COVID-related indexes that significantly influence the connectedness of travel and leisure stocks. Finally, we proposed useful implications for hospitality firms, managers, stakeholders, investors, portfolio managers, policymakers, and regulatory authorities to be considered in the future during episodes of uncertainty. The remainder of the study is structured as follows. Section 2 explains prior studies of the research field in a literature review. Section 3 describes sample countries' TLSs data and period, COVID-19 measures, and descriptive statistics. Section 4 demonstrates the methodological framework of TVP-VAR and wavelet coherence techniques. Finally, in Section 5, we conclude and offer significant implications for various constituents of the hospitality and travel and leisure sector.

Using various methodologies, prior research examined the dependence structures between the hospitality stocks for different periods. The application of network analysis in the hospitality and tourism industry identifies various interlinks of different travel and leisure stocks. The study of Stokowski (1994) is considered as one of the developer studies that employed network analysis to study the relationship between travel and leisure stocks and their groups. Numerous studies applied connectedness approaches (for instance, Baggio, 2007; Baggio et al., 2010; Scott et al., 2011) . The main exercise of a connectedness investigation is to obtain the spillover analysis and choose the most attractive (destructive) hospitality nations; while its other purpose is to offer useful policy implications for businesses, the hospitality industry, and the stakeholders. In J o u r n a l P r e -p r o o f addition, the connectedness analysis investigates the time-varying attributes, addressing the evolution of different periods and the connectedness of hospitality stocks over a particular period to draw some policy and practical ramifications (Carley & Columbus, 2013) .

Studies examined the impact of the earlier SARS pandemic on the hospitality sector of selected Asian countries. Specifically, severe consequences of the SARS outbreak on Hong Kong hotels have been drawn by Pine and McKercher (2004) ; on the Korean hotel industry by Henderson and Ng (2004) and Kim, Chun, and Lee (2005) ; on Hong Kong restaurants by Alan, So, and Sin (2006) ; and on Taiwanese hotels by Chen, Jang, and Kim, (2007) . Hence, we argue that the response of hospitality sector segments to the SARS outbreak across different countries in the East Asia region is evident for the connectedness of TLS during health crises. The COVID-19 outbreak has a more pronounced and long-lasting inverse impact on the tourism industry and economy (Kaushal & Srivastava, 2021; Škare et al., 2021) . On the same note, Fotiadis, Polyzos, and Huan (2021) capture a huge decline (about 98%) in international tourism demand during 2020 and expect it to continue in 2021. It is noteworthy that international tourism demand has a strong direct impact on the performance of equity indexes (Balli et al., 2021) . Moreover, as a consequence of the COVID-19 outbreak, many countries' hospitality, travel, and leisure operations have been virtually shut down, and international and domestic travel has all but ceased.

Although scholars, hospitality firm managers, investors, and portfolio managers must know the deep potential impact of the COVID-19 pandemic outbreaks on the hospitality, travel, and leisure industry, thus far the literature lacks empirical studies on the pandemic's influences on the performance of hospitality, travel, and leisure firms. Additionally, the current studies have eminent limitations concerning the samples and periods considered. For instance: Song, Yeon, and Lee (2020) explore the vital role that firm characteristics play in the stock decline in response to the COVID-19 pandemic shock in the US from January 3 to May 15, 2020. Gil-Alana and Poza (2020) study the behavior of tourism stocks in Spain in the period of May 14, 2018 to May 14, 2020 , and conclude that the coronavirus crisis has increased the persistence in the data, moving in some of the series from a mean-reverting process to a non-mean reverting one. By reviewing the aftermath of the earlier crisis on the travel and leisure sector, Škare et al. Further, by using accounting data, Crespí-Cladera, Martín-Oliver, and Pascual-Fuster (2021) examined the financial distress of Spanish and Portuguese hospitality firms during the COVID-19 outbreak. The study observes that financial distress affects mainly small firms. On the same note, considering the period from January 6, 2020 to March 23, 2020, Kaczmarek et al. (2021) show that global travel and leisure firms with low valuations, limited leverage, and high investments have been more immune to the pandemic-induced crash between January 1, 2020

and July 16, 2021. Further, a recent study by Zargar and Kumar (2021) confirmed the spillover of shocks related to investor mood, fear, sentiment, and policy uncertainty to the tourism sector in the United States during the COVID-19 era. Lastly, Shahzad, Hoang, and Bouri (2021) explored that the results show that the bad contagion among tourism firms significantly increased in the United States with the outbreak of the COVID-19 pandemic, and spillover among firms is still high. To this end, unlike earlier studies, our study offers a comprehensive analysis of the contribution of the COVID-19 pandemic to the world's top TLS.

The above literature reveals that the hospitality industry substantially contributes to economic well-being. However, the COVID-19 pandemic has posed enormous challenges to the sector as it worked to confirm its survival during the outbreak with the restrictions imposed by the World Health Organization (WHO, 2020) . For all these reasons, the current study is novel in presenting its contribution to the extant literature for the given study period to position the hospitality countries with the most susceptibility to the outbreak risk and subsequently extract beneficial implications, particularly during COVID-19.

We utilize the Travel and Leisure sector indexes from Datastream to represent international tourism stock markets. The top ten major tourist destinations in terms of tourism arrivals, based on the 2020 UNWTO report, 4 are chosen in the sample. They include France (FRA), Spain is based on the destinations with the most coronavirus cases. We introduce travel and leisure indexes to represent the hospitality firms' stocks in the top tourist destinations in the world (Hadi et al., 2020a; Hadi, Irani, and Gökmenoğlu, 2020; Balli et al., 2021) . We retrieved the stock prices from Datastream. Further, use six COVID-19 proxies in our analysis: (1) panic index; (2) media coverage index; (3) sentiment index; (4) fake news index; (5) media hype index; and (6) infodemic index. 5 Lastly, since our focus is on the COVID-19 period, we utilize daily data from January 1, 2020 to July 16, 2021. 6 Additionally, for our analyses, the stock prices are converted to change rates, calculated as LN(P it -P it-1 ) * 100, where P i is the daily stock price for each country's hospitality index at day t. 7

Descriptive statistics are explored in Table 1 (Abbas et al., 2021; Rastegar et al., 2021; Kim et al., 2021; Song et al., 2021) also reported that the COVID-19 pandemic had brought unprecedented challenges to the hospitality, travel, and leisure industry.

"Place Table 2 about here"

This study first uses the time-varying parameter vector autoregressions (TVP-VAR), which measured the connectedness of top hospitality stock markets. Second, we test the time-frequency impact of COVID-related factors on the connectedness of hospitality stocks by employing wavelet coherence.

We use the time-varying parameter vector autoregressions (TVP-VAR) model of Primiceri (2005) , which was later extended by Antonakakis and Gabauer (2017) . This technique discloses two merits: (1) it assists in overcoming the challenge of arbitrarily choosing the optimal rollingwindow size, and (2) it circumvents the issue of loss of valuable observations, thus making it suitable for a short sample as well (Adekoya & Oliyide, 2021) . The TVP-VAR model determines the possible changes in the connectedness of hospitality stock markets to provide evidence of whether the linear structure is derived from the probability of the shocks or from the extension of the change in responses. The model also offers peculiar characteristics to recognize potential structural breaks and provides significant reasons to recognize the relationship among variables.

In contrast to this, the dynamic copula technique proposed by is based on a bivariate model and measures tail-dependence between two set of variables. Since the current study is based on a multivariate model, the dynamic copula approach cannot be employed on the given set of variables. Therefore, the current study measures the time-varying connectedness of J o u r n a l P r e -p r o o f hospitality stocks using the TVP-VAR approach. Thus, a detailed methodology is given in the following sections.

The model is stated as:

(1)

Here, y t is an (n × 1) vector for the dependent variable and denotes (n × n) time-varying coefficients, which are rewritten as t  matrix. X t represents an (n × k) matrix comprising intercepts and lags of the time-dependent variables. u t denotes structural shocks with (n × 1) heteroskedastic distribution term, with zero mean and time-varying variance-covariance matrix Ω t . Given the log-differenced returns of green, Islamic, and conventional financial markets, the variance-covariance matrix is segregated as:

( 3) Here,

shows simultaneous relationships of variables and H t presents stochastic connectedness. Moreover, the transitions in the time-varying parameters are observed as follows,

Eqs. (4) and (5) estimate the time-varying parameters following a random walk process, and Eq.

(6) examines the stochastic connectedness following the independent random walk. Primiceri (2005) proposed that coefficients among variables change independently for simplifying the inference and increasing the efficiency of the estimates. It denotes that the main equation error term is independent of the transition equation.

To examine the time-frequency impact of COVID-related factors on the connectedness of hospitality stock markets, we employ a wavelet coherence that provides a widespread analysis of time-series, irrespective of the sample period. Wavelet coherence is divided into two sub-J o u r n a l P r e -p r o o f categories: (1) first cross-wavelet power; and (2) cross-wavelet transform. Torrence and Compo (1998) explained that cross wavelet transform is clarified by two-time sequences, that is, a(t) and b(t), which is stated as:

where and illustrate two continuous transforms of a(t) and b(t) independently;

p is the location index; q is the measure; and composite index is shown by (*). In addition, crosswavelet transform is used to calculate wavelet power by . The spectra of cross wavelet 

M reveals a smoothing mechanism whereas indicates a range of squared wavelets, assuming closeness to zero indicates no correlation, and closeness to unity indicates high correlation.

Before we examine the effect of the COVID-19 pandemic on the connectedness of the hospitality stocks, it is essential to estimate the spillover among the sample hospitality stocks. In Table 3 , the total directional spillovers from and to specific stocks, net directional spillovers, and total connectedness index (TCI) for country travel and leisure stocks are reported. The rows in Table 3 represent the contribution of each TLS to the forecast error variance of one individual TLS in the system, while the columns represent the effect of one specific TLS on all other TLSs separately. Table 3 about here"

The results in Table 3 To extend our analysis, Figure 1 was restricted by February 2, 2020. 14 Notably, the dynamic connectedness of TLS is at the highest [above 80%] level at the pandemic's peak in March 2020 and it starts to decline when governments implement various fiscal and monetary policies to relieve their economies (Akhtaruzzaman et al., 2021) . These actions by states changed investors' sentiments to become optimistic about the private sector's future performance, including the hospitality firms.

We further find another peak of spillovers at the beginning of the second pandemic wave in November 2020, and then a considerable decrease due to further fiscal and monetary policies and disaster.

Being highly vulnerable to political, environmental, and pandemic outbreak shocks, firms in the travel, hospitality, and leisure sectors are the most affected by the COVID-19 pandemic. Using the TVP-VAR approach, we examine the spillover among the world's top ten travel and leisure stocks. Unlike earlier studies, our sample covers the period of first and second waves of the pandemic as well as the approval of the vaccine and its availability. TCI revealed a strong volatility spillover among the stocks, and the pronounced connectedness between the stocks J o u r n a l P r e -p r o o f managers are required to make a rational trading decision and establish the best hedging and diversification strategies, as the pandemic has proven itself to be a drastic factor for investment avenues. Thus, investors and portfolio managers must prudently assess the pros and cons of their investment streams. Correspondingly, policymakers need to reformulate their existing travel and hospitality sector policies to ensure the health and safety of their affiliates and foreigners, and to provide them a real-time experience of safe travel. Finally, regulatory bodies and their constituents must introduce effective plans to anticipate the pandemic's drastic impacts, continue their inter-state affairs at a smooth pace, and maintain foreign alliances to increase the number of travelers in the hospitality and leisure sector. Note: This figure shows the time-varying total return connectedness between the country tourism stocks using TVP-VAR model from January 1, 2020 to July 16, 2021. Note: Wavelet coherence between the connectedness and COVID related indexes. The 5% significance level against the noise is shown as a thick contour. The relative phase relationship is shown as arrows (with in-phase pointing right which implies positive co-movement, anti-phase pointing left which implies negative co-movement, and COVID related index leading connectedness by 90• pointing straight down, which implies a strong positive comovement between the connectedness and the COVID-19 indices).

Exploring the impact of COVID-19 on tourism: Transformational potential and implications for a sustainable recovery of the travel and leisure industry

How COVID-19 drives connectedness among commodity and financial markets: Evidence from TVP-VAR and causality-in-quantiles techniques

COVID-19 media coverage and ESG leader indices

Crisis management and recovery: How restaurants in Hong Kong responded to SARS

COVID-19: Hotel industry response to the pandemic evolution and to the public sector economic measures

The web graph of a tourism system

Spillovers from tourism demand to tourism equity indices

Has the tourism-led growth hypothesis been validated? A literature review

Basic lessons in Ora and Automap

The stock price reaction of the COVID-19 pandemic on the airline, hotel, and tourism industries

The impact of the SARS outbreak on Taiwanese hotel stock performance: An event-study approach

Financial distress in the hospitality industry during the Covid-19 disaster

The adjustment of stock prices to new information

The good, the bad and the ugly on COVID-19 tourism recovery

The impact of COVID-19 on the Spanish tourism sector

External determinants of the stock price performance of tourism, travel, and leisure firms: Evidence from the United States

The vulnerability of tourism firms' stocks to the terrorist incidents

Responding to crisis: Severe acute respiratory syndrome (SARS) and hotels in Singapore

How to survive a pandemic: The corporate resiliency of travel and leisure companies to the COVID-19 outbreak

COVID-19 pandemic waves and global financial markets: Evidence from wavelet coherence analysis

Hospitality and tourism industry amid COVID-19 pandemic: Perspectives on challenges and learnings from India

The effects of SARS on the Korean hotel industry and measures to overcome the crisis: A case study of six Korean five-star hotels

Tourism's vulnerability and resilience to terrorism

A new time-varying optimal copula model identifying the dependence across markets

Tourism and sustainability in times of COVID-19: The case of Spain

Religion vs ethics: Hedge and safe haven properties of Sukuk and green bonds for stock markets pre-and during COVID-19

COVID-19 and the stock market: Impacts on tourism-related companies

The impact of SARS on Hong Kong's tourism industry

Time varying structural vector autoregressions and monetary policy

Network analysis methods for modeling tourism inter-organizational systems

From pandemic to systemic risk: Contagion in the US tourism sector

Tourism-led growth hypothesis in the top ten tourist destinations: New evidence using the quantile-on-quantile approach

Tourism and COVID-19: Impacts and implications for advancing and resetting industry and research

Impact of COVID-19 on the travel and tourism industry

Impact of the COVID-19 pandemic: Evidence from the US restaurant industry

Leisure in society: A network structural perspective

A practical guide to wavelet analysis

Tourism recovery strategy against COVID-19 pandemic

Market fear, investor mood, sentiment, economic uncertainty and tourism sector in the United States amid COVID-19 pandemic: A spillover analysis

Afraid to travel after COVID-19? Self-protection, coping and resilience against pandemic ‗travel fear

Note: This figure shows the time-varying NET connectedness between the country tourism stocks using TVP-VAR model from January 1, 2020 to July 16, 2021. The negative territories imply that the stock is a net receiver of shock in that period, and the positive territories imply that the stock is a net transmitter.