ABSTRACT

This paper analyses the effect of the COVID-19 pandemic spread the into Europe on the tourist stocks in Croatia by application of the event study methodology. The analysis starts with the descriptive overview of market wide performance of different sectors in the period before, during and after the initial pandemic outbreak and afterwards explicitly tests for the COVID-19 outbreak effects on the tourist sector. First, the 35 day long event window is specified so that important events related to the pandemic can be identified. Second, the first officially reported COVID-19 incidence in Italy and World Health Organization declaration of global pandemic are used as identified events in a shorter 10 day window event study estimation. The results point to the significant negative and robust effect of COVID-19 pandemic on the returns of tourist stocks in the Zagreb stock exchange. However, the overall results do not provide evidence of the relatively stronger COVID-19 effects on tourist sector but rather equal effects across different sectors.

INTRODUCTION

The coronavirus pandemic spread around the globe in just a few months after the first case was registered, leaving huge consequences in the form of threatening the health and human lives, great economic losses, and psychological fear. The presence of the virus has prevented potential tourists from feeling safe in the tourist destinations and produced a complete stall in the functioning of tourism business. According to the UNWTO (2021), international tourist arrivals (overnight visitors) dropped by 74% in 2020 compared to the 2019, due to the massive drop in tourist demand and travel restrictions. The pandemic produced a loss of USD 1.3 trillion in export revenues which is eleven times greater loss compared to the global financial crisis from 2008. UNWTO forecasted that it could pass between 2.5 or even 4 years for international tourism to return on the old track from 2019. The effectiveness of vaccines and the level of vaccination of the population will certainly play an important role and are expected to reduce the number of new cases, as well as to the mitigation of travel restriction measures and the restoration of consumer confidence. The consequences of the pandemic are especially pronounced in the tourism dependent countries like Croatia.

According to the UNWTO (2021), Croatia is among the 10 most vulnerable countries according to the criterion of the direct impact of tourism on the share of GDP. More precisely, in the first place is Macao (China) 48%, followed by: Fiji 13%, Jordan 12%, Spain 12% and Croatia with 11%. Many countries have introduced travel restrictions and knowing that the share of foreign tourists in Croatia is 89% (UNWTO, 2021), negative financial results are expected from companies that are directly or indirectly involved in tourism. Beside aforementioned facts, tourism is a significant contributor to national exports (35%) (UNWTO, 2021) and this contraction will have negative macroeconomic consequences. Although Croatia achieved a record high in 2019 in terms of the total number of arrivals and overnight stays of domestic and foreign tourists in 2020, due to the global corona pandemic, there was a steep decline compared to 2019. The pronounced drop in the tourism spilled over into the capital market and caused turmoil, so the purpose of this research is to investigate how tourism stock prices responded to the outbreak of the COVID-19 pandemic by using event study methodology. This methodology has proven to be especially useful in identifying the effect of a particular event on stock market returns.

This study contributes to the literature by analyzing the impact of global COVID-19 pandemic on tourism sector in Croatia. As a practical implication, this study will be particularly useful to current and potential investors as well as policy management during this and other future unexpected crises. The remainder of this study is structured as follows: the section 2 performs an overview of Croatian tourism before and during the Covid-19 pandemic, the section 3 outlines a detailed literature review, the section 4 presents data and methodology, the section 5 reports the results of empirical research and discussion and finally in the last section a conclusion, limitations and future research recommendations are summarized.

TOURISM IN CROATIA BEFORE AND DURING THE COVID-19 PANDEMIC

The Republic of Croatia systematically follows global tourist trends and is well positioned on the European tourist market. It is recognized as a stable, safe tourist destination with beautiful and rich natural and cultural-historical heritage. The Croatian tourism sector has been successful since Croatia’s independence and accession to the European Union, and has consistently recorded enviable results, until the outbreak of the coronavirus pandemic in 2020. According to the data published on the official website of the World Health Organization–WHO, by April 20, 2021, a total of 141,754,944 confirmed cases of COVID-19 were recorded, of which 3,025,835 deaths in the world. According to the same source, from 3 January 2020 to 20 April 2021, 310,306 confirmed COVID-19 cases with 6,643 deaths were recorded in Croatia on the total population of 4,058,165 therefore, it can be concluded that Croatia was severely affected by the pandemic. The SARS-CoV-2 virus has affected the whole world and caused historically unprecedented problems in all aspects of social and economic life. It has greatly affected the global economic activities. Measures taken to prevent virus spreading (social distancing, traffic restrictions, restrictions on commercial activity, borders closures, etc.) have strongest affected the service sector — tourism and hospitality. The analysis of selected indicators of tourism development in Croatia before and at the time of Covid-19 pandemic in Croatia, shows how much the pandemic affected the results in tourism.

Table 1. Tourist arrivals and overnight stays in Croatia in the period 1980–2020

According to the historical review of data on tourist arrivals and overnight stays in Croatia (table 1), a constant growth in tourist arrivals and tourist overnight stays can be observed until 2020. If we compare the nights and arrivals in 2019 with those from 2010, it could be stated that the results have almost doubled. Despite the favorable results, a sharp decline followed in 2020 with decrease in tourist arrivals by 64.2% and decrease in overnight stays by 55.3% in commercial accommodation in 2020 compared to 2019 year. The decline in arrivals and overnight stays followed among domestic and foreign tourists, and was more pronounced among foreign tourists, which can be seen from Charts 1 and 2.

Figure 1. Arrivals and overnight stays of foreign and domestic tourists in Croatia in the last five years (before and during the pandemic)

According to the data from Figure 1, it can be concluded that the decline of domestic tourist arrivals in 2020 (-34.24%) is much smaller compared to the decline of foreign tourist arrivals (-68.05%). The same effect is present in realized overnight stays, domestic overnight stays in 2020 decreased by -23.68%, while realized overnight stays of foreign tourists had a decrease of -57.95%. These results are a consequence of the fear of the COVID-19 pandemic and lockdown measures.

LITERATURE REVIEW

Event study methodology is used in a variety of application settings with the aim of measuring the impact of unforeseen or unanticipated events on prices or business performance. This methodology is has therefore found wide usage in the field of finance for measuring corporate announcements on financial performance, investor behavior or similar. The implicit assumption for the application of event study methodology is that markets are efficient, i.e. publicly available information are reflected in the prices of securities, so it was used in economics for testing the market efficiency on the capital market (Kothari and Warner (⊕2007Kothari, S., and Jerold B. Warner. 2007. “Econometrics of Event Studies.” In.)). However, Novak (⊕2019Novak, I. 2019. “EFFICIENT Market Hypothesis: CASE of the Croatian Capital Market.” In.) rejected the weak-form efficiency hypothesis on Croatian capital market. One of the first authors which showed the usefulness of the event study methodology were Ball and Brown (⊕1968Ball, Ray, and P Brown. 1968. “EMPIRICAL Evaluation of Accounting Income Numbers.” Journal of Accounting Research 6 (2): 159–78. https://EconPapers.repec.org/RePEc:bla:joares:v:6:y:1968:i:2:p:159-178.) who researched the impact of earnings surprises on stock prices. Brown and Warner (⊕1980Brown, Stephen, and Jerold B. Warner. 1980. “Measuring Security Price Performance.” Journal of Financial Economics 8 (3): 205–58. https://EconPapers.repec.org/RePEc:eee:jfinec:v:8:y:1980:i:3:p:205-258.) have compared different event study methodologies and concluded that complicated methodologies will not benefit with better results in comparison with a simple one factor market model. Asquith and W (⊕1983Asquith, Paul, and Mullins W. 1983. “The Impact of Initiating Dividend Payments on Shareholders’ Wealth.” The Journal of Business 56 (1): 77–96. https://EconPapers.repec.org/RePEc:ucp:jnlbus:v:56:y:1983:i:1:p:77-96.) concluded that initiating a dividend policy as an information has a strong and positive impact on the market reaction. Miletic (⊕2011Miletic, M. 2011. “Stock Price Reaction to Dividend Announcement in Croatia.” Economic Research-Ekonomska Istraivanja 24: 147–56.) analyzed impact of dividend announcement on Croatian capital market by event study methodology. Results confirmed that increase or decrease of dividend significantly affect stock price in the same direction while dividend retention had no affect on stock price. Abarbanell and Park (⊕2017Abarbanell, Jeffery, and Hyungshin Park. 2017. “Do Bright-Line Earnings Surprises Really Affect Stock Price Reactions?” Manag. Sci. 63 (4): 1063–84. https://doi.org/10.1287/mnsc.2015.2376.) found that companies with larger ex ante earnings response coefficients are linked to a greater propensity to positively bias earnings surprise and more negative intercepts in regressions of announcements returns on earnings surprises. Skrinjaric and Orlović (⊕2019Skrinjaric, Tihana, and Zrinka Orlović. 2019. “Effects of Economic and Political Events on Stock Returns: Event Study of the Agrokor Case in Croatia.” Croatian Economic Survey 21: 47–86.) applied event study methodology in order to test whether political events regarding the concern Agrokor affect stock prices. Their results indicate that stocks which belong to Agrokor concern suffered from lowering returns while other liquid stocks on the Zagreb stock exchange were not affected.

(???) researched new COVID-19 official cases announcements and death ratio on the financial markets volatility index (VIX). Author has concluded that new cases reported outside China have positive impact on VIX, death ratio has significant positive impact on VIX, and the spread of the COVID-19 increase financial volatility. Chia, Liew, and Rowland (⊕2020Chia, Ricky Chee-Jiun, Venus Liew, and Racquel Rowland. 2020. “Daily New Covid-19 Cases, the Movement Control Order, and Malaysian Stock Market Returns.” MPRA Paper. University Library of Munich, Germany. https://EconPapers.repec.org/RePEc:pra:mprapa:107988.) research the relation between Malaysian stock market and variables related to COVID-19 their results suggest that daily new cases had negative but insignificant impact on the indices returns. On the other hand authors also found that movement control order had significant and positive impact on all indices’ returns which is quite surprising. There is a significant body of recent reseach that explores the impact of COVID-19 on the capital markets (Mazur, Dang, and Vega (⊕2020Mazur, M., M. Dang, and M. Vega. 2020. “COVID-19 and the March 2020 Stock Market Crash. Evidence from S&P1500.” Finance Research Letters 38: 101690–0.); Wagner (⊕2020Wagner, A. 2020. “What the Stock Market Tells Us About the Post-Covid-19 World.” Nature Human Behaviour, 1–1.); Phan and Narayan (⊕2020Phan, D. H. B., and P. Narayan. 2020. “Country Responses and the Reaction of the Stock Market to Covid-19—a Preliminary Exposition.” Emerging Markets Finance and Trade 56: 2138–50.); Narayan, Devpura, and Wang (⊕2020Narayan, P., Neluka Devpura, and Hua Wang. 2020. “Japanese Currency and Stock Market—What Happened During the Covid-19 Pandemic?☆.” Economic Analysis and Policy 68: 191–98.); Contessi and De Pace (⊕2020Contessi, Silvio, and Pierangelo De Pace. 2020. “The International Spread of Covid-19 Stock Market Collapses.” Economics Department, Working Paper Series. Economics Department, Pomona College. https://EconPapers.repec.org/RePEc:clm:pomwps:1013.); Erdem (⊕2020Erdem, Orhan. 2020. “Freedom and Stock Market Performance During Covid-19 Outbreak.” Finance Research Letters 36 (C): S1544612320306735. https://EconPapers.repec.org/RePEc:eee:finlet:v:36:y:2020:i:c:s1544612320306735.); Onali (⊕2020Onali, Enrico. 2020. “COVID-19 and Stock Market Volatility.” Capital Markets: Asset Pricing & Valuation eJournal.); Rahman, Amin, and Mamun (⊕2020Rahman, Lutfur, Abu S. Amin, and M. Mamun. 2020. “The Covid – 19 Outbreak and Stock Market Reactions: Evidence from Australia.” Finance Research Letters, 101832.); Harjoto, Rossi, and Paglia (⊕2021Harjoto, Maretno Agus, Fabrizio Rossi, and John K. Paglia. 2021. “COVID-19: Stock Market Reactions to the Shock and the Stimulus.” Applied Economics Letters 28 (10): 795–801. https://doi.org/10.1080/13504851.2020.1781767.); Okorie and Lin (⊕2020Okorie, David Iheke, and Boqiang Lin. 2020. “Stock Markets and the Covid-19 Fractal Contagion Effects.” Finance Research Letters 38: 101640–0.) 2021) by using the event study methodology. Panyagometh (⊕2020Panyagometh, Kamphol. 2020. “The Effects of Pandemic Event on the Stock Exchange of Thailand.” Economies 8 (4): 1–22. https://ideas.repec.org/a/gam/jecomi/v8y2020i4p90-d433708.html.) used a sample of 46 stocks listed in the Stock Exchange of Thailand in order to analyze stock price reactions during pandemic. After applying event study methodology in order to empirically measure abnormal returns and volatility, research results suggest that the majority of stocks in the Stock Exchange of Thailand have been negatively affected by Covid-19 pandemic. On the other side, the author found that some stocks had positive returns, in particular the stocks from commerce sector which are included in distribution of pharmaceutical products and services. Irfan, Kassim and Dhimmar (2021) analyzed the impact of COVID-19 on the performance of Indian stock exchange and Indonesian Stock Exchange. Authors included three different event windows to check results in different time periods. The analysis of both countries showed opposite results since Indian stock market showed downward sloping trend after WHO declaration, and Indonesian upward. He et al. (⊕2020He, Pinglin, Yulong Sun, Ying Zhang, and Tao Li. 2020. “COVID–19’s Impact on Stock Prices Across Different Sectors—An Event Study Based on the Chinese Stock Market.” Emerging Markets Finance and Trade 56 (10): 2198–2212. https://doi.org/10.1080/1540496X.2020.178.) investigated impact of COVID-19 on stock prices of different Chinese industries. Authors concluded that pandemic greatly affected transportation, mining, electric and heating, and environmental industries, while manufacturing, information technology, education and health-care industries have resisted the impact of the COVID-19 pandemic. Liu et al. (⊕2020Liu, Haiyue, Aqsa Manzoor, C. Wang, L. Zhang, and Zaira Manzoor. 2020. “The Covid-19 Outbreak and Affected Countries Stock Markets Response.” International Journal of Environmental Research and Public Health 17.) evaluated short-term impact of COVID-19 on stock indices of various countries (Japan, Korea, Singapore, USA, UK etc.) Their results suggested that COVID 19 had significant negative effect on all affected countries. Authors found that Asian stock markets had more negative abnormal returns compared to other countries. Their regression results also support the evidence of adverse effect of confirmed cases on stock indices abnormal returns. Kandil Göker, Eren, and Karaca (⊕2020Kandil Göker, İlkut Elif, Binali Selman Eren, and Süleyman Serdar Karaca. 2020. “The Impact of the Covid-19 (Coronavirus) on the Borsa Istanbul Sector Index Returns: An Event Study.” Gaziantep University Journal of Social Sciences. Gaziantep Üniversitesi Yerleşkesi Şehitkamil-Gaziantep: Gaziantep University. https://doi.org/10.21547/jss.731980.) researched impact of COVID-19 on the Istanbul sector indices returns by applying the event study methodology. Authors confirmed that most of the sectors have negative abnormal returns. Sectors which were hit the most are sport, tourism and transportation.

METHODOLOGY

RESULTS AND DISCUSSION

In this analysis we broadly define the event in terms of COVID-19 pandemic transitioning from global shock to taking measurable effect in Europe. To proxy the event effect, the first reported case in Italy on Feb, 21th is chosen and extended one week prior to the event to allow for the information to get incorporated in the market valuations. The reason for extending the event period is because investors were already aware of the ongoing pandemic and the virus was already spreading in Italy before the official announcement. In a similar fashion, the end of the event period is extended for three trading days after the official lockdown was introduced in Croatia on March 19th to allow investors to process the information and adjust the market valuation accordingly. For the given period, the results of the several parametric event study tests are presented in the Table 2. The table shows clusters of high significance across different tests on the first trading day after the lockdown was introduced in Italy (Feb 24th) as well as the following day (Feb 25th) but with somewhat less robustness. Clusters of significance across different tests are also observed around March 11th when World Health Organization declared COVID-19 a global pandemic pointing to the significant effect on the listed tourist firms on ZSE. It is also noticeable that robustly significant COVID-19 effects can be observed on March 9th, 12th, 16th, and 17th pointing to the investor`s anticipation of official declarations in the local market.

Table 2: Table 2: The event study reuslts for the parametric tests

Date	BW80	sig	BW85	sig	t-stat	sig	pt	sig	BH	sig	LMB	sig
2020-02-17	0.155		0.154		0.500		0.938		0.899		0.154
2020-02-18	0.552		0.548		1.100		1.799		1.642		0.547
2020-02-19	-0.730		-0.725		-1.806		-2.519	**	-1.536		-0.724
2020-02-20	0.259		0.257		1.312		1.052		1.345		0.256
2020-02-21	0.869		0.863		0.868		1.248		1.166		0.862
2020-02-24	-4.217	***	-4.187	***	-2.027		-11.658	***	-2.235	**	-3.407	***
2020-02-25	-2.819	***	-2.799	***	-1.751		-4.203	***	-1.632		-2.739	***
2020-02-26	-0.482		-0.479		-0.527		-0.120		-0.072		-0.471
2020-02-27	-0.187		-0.186		-0.259		0.408		0.191		-0.184
2020-02-28	-0.659		-0.654		-0.759		-1.527		-0.875		-0.628
2020-03-02	1.810		1.797		1.335		3.738	***	1.835		1.792
2020-03-03	0.258		0.256		0.267		0.089		0.062		0.255
2020-03-04	-0.376		-0.374		-0.268		0.230		0.117		-0.373
2020-03-05	-0.554		-0.550		-0.646		-2.726	***	-1.337		-0.547
2020-03-06	-1.390		-1.380		-1.053		-5.260	***	-1.565		-1.314
2020-03-09	-5.554	***	-5.515	***	-1.900		-8.950	***	-2.210	**	-3.254	***
2020-03-10	-0.117		-0.116		-0.076		-2.800	***	-1.078		-0.115
2020-03-11	-4.943	***	-4.908	***	-2.259	**	-7.922	***	-2.397	**	-4.122	***
2020-03-12	-2.011	**	-1.997	**	-0.633		-5.292	***	-1.338		-1.103
2020-03-13	0.607		0.603		0.335		2.030	**	0.691		0.475
2020-03-16	-2.549	**	-2.531	**	-0.945		-7.119	***	-1.648		-1.761
2020-03-17	-2.212	**	-2.196	**	-1.693		-2.267	**	-0.638		-2.118	**
2020-03-18	-1.314		-1.304		-0.734		-4.098	***	-0.984		-1.183
2020-03-19	-0.440		-0.437		-0.241		0.248		0.060		-0.431
2020-03-20	-1.086		-1.078		-1.051		-2.758	***	-1.096		-1.022
2020-03-23	-0.901		-0.895		-0.559		-3.662	***	-1.064		-0.819

To check the robustness of the results, several nonparametric tests are additionally run and results are shown in the Table 3. Nonparametric tests show overall less significance relative to the parametric tests but point in the same direction. First significance cluster can be observed on Feb 20th, a one day before the first officially reported case in Italy and another cluster is related to March 11th, a day when WHO declared global pandemic. Also, cluster of significance across tests is observed on the March 17th and could be related to the investor`s anticipation of the introduction of lockdown in the Croatia on March 19th.

Table 3: Table 2: The event study reuslts for the nonparametric tests

Date	SIGN	sig	GSIGN	sig	CSIGN	sig	RANK	sig	MRANK	sig	WLCX	sig
2020-02-17	0.000		0.823		1.103		1.017		1.017		40
2020-02-18	-0.577		0.229		-0.276		0.696		0.696		43
2020-02-19	-0.577		0.229		0.827		-0.226		-0.226		20
2020-02-20	1.155		2.009	**	1.378		1.755		1.755		55
2020-02-21	-1.155		-0.364		0.689		0.557		0.557		26
2020-02-24	-0.577		0.229		-0.276		-0.711		-0.711		19
2020-02-25	0.000		0.823		0.000		-0.191		-0.191		27
2020-02-26	-0.577		0.229		-0.276		-0.497		-0.497		33
2020-02-27	0.000		0.823		0.551		0.832		0.832		41
2020-02-28	0.577		1.416		0.551		0.892		0.892		38
2020-03-02	0.577		1.416		1.103		1.654		1.654		55
2020-03-03	-0.577		0.229		-0.276		-0.148		-0.148		34
2020-03-04	0.000		0.823		0.827		0.269		0.269		38
2020-03-05	-0.577		0.229		0.276		0.314		0.314		29
2020-03-06	0.000		0.823		0.276		0.358		0.358		31
2020-03-09	-0.577		0.229		-0.276		-0.653		-0.653		15
2020-03-10	-1.155		-0.364		-0.551		-1.050		-1.050		27
2020-03-11	-2.309	**	-1.551		-1.103		-2.225	**	-2.225	**	14	**
2020-03-12	0.000		0.823		0.276		0.284		0.284		37
2020-03-13	-0.577		0.229		-0.276		-0.528		-0.528		35
2020-03-16	0.000		0.823		0.000		0.033		0.033		34
2020-03-17	-2.309	**	-1.551		-1.103		-2.036	**	-2.036	**	13	**
2020-03-18	1.155		2.009	**	0.551		0.959		0.959		43
2020-03-19	0.577		1.416		0.827		1.288		1.288		50
2020-03-20	-1.732		-0.958		-0.827		-1.526		-1.526		22
2020-03-23	0.577		1.416		0.276		0.396		0.396		34

Cumulative changes in the return series are additionally examined for two events, the first COVID-19 incidence in Italy and WHO declaration of COVID-19 a global pandemic, that have been characterized by clustering of significance across multiple parametric and nonparametric tests. In this case the event window is set to be shorter (10 days). The Figure 1 shows the results when no control variables are included in the model and Figure 2 is related to the extended model (i.e. market model) with Crobex index returns as a control variable. The left panel shows the results for the first COVID-19 incidence in Italy and the right panel is related to the WHO declaration of global pandemic. In both cases the null hypothesis of the abnormality of returns can’t be rejected since the full line, representing the stock returns, is inside 95% confidence intervals denoted by the dotted lines.

Overall, the results point to the significant negative effect of COVID-19 pandemic on the returns of tourist stocks listed on the ZSE. When the event window is defined in broader terms we observe clusters of significance across variety of tests around two events, the first official incidence of COVID-19 in Italy and WHO declaration of global pandemic. This result implies the importance of international and global events for trends on the Croatian capital market. This result is robustly confirmed with the rage of parametric and nonparametric tests. Furthermore, this finding is corroborated with shorter event window specification and shows robustness with respect to inclusion of external variable controlling for the overall market return. The descriptive statistics point to the significant effect of COVID-19 pandemic on the tourist stocks in the comparable size to the general market and also very similar to the other stock market sectors. Therefore, we conclude that COVID-19 pandemic had an equal market wide effect and no particular sector effect is can be identified. It is also interesting to note that results imply that international events potentially carry a higher importance for the local stock market trends than local epidemiological policy.

CONCLUSION

Coronavirus pandemic outbreak has largely affected economic activities all over the world and governments around the globe implemented policies that reduce movement of people and goods. These measures are expected to have a negative impact on all economic activities and especially pronounced in the service sector like tourism. This analysis provides empirical evidence on the direct reactions of ZSE listed tourist firms to the outbreak of COVID-19 global pandemic by application of the event study methodology. The results confirm that COVID-19 pandemic has affected tourist stocks by triggering the negative above average cumulative return during the event period. Two events related to the first official COVID-19 incidence and WHO declaration of global pandemic could have been identified as most important for the returns. The findings are robust to different length specification of the event window and controlling for the general market returns. Interestingly, the pandemic effects seem to be similar in size across different sectors and tourist sector is no exception to this rule. This result might be due to the low liquidity of tourist stocks on the ZSE or structure of Croatian economy and financial market but further inquiry would open an interesting area for future research as well as further examining the reasons for the unequal sectoral distribution of pandemic effects in the local and global markets.