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

Abstract

The study investigates and confirms the spillover effects from investor fear, mood, sentiment and uncertainty to the US tourism sector returns. The findings indicate that market fear, investor mood and sentiment are net transmitter of shocks and economic uncertainty and the tourism sector is net receiver of shocks. We also provide evidence that media-hype, infodemic, media-coverage related to COVID-19 and infectious disease equity market volatility impacts the total and directional spillover of information from fear, mood, sentiment and uncertainty to the tourism sector.

Keywords

Investor fear investor mood investor sentiment economic policy uncertainty tourism sector COVID-19

1. Introduction

Tourism, being the most important sector of many countries around the world, is the worst affected by the COVID-19 pandemic (UNWTO, 2020). Lockdowns, restrictions on travel, social distancing, closure of borders, air travel ban, shutting down of accommodation and catering establishments, etc. have resulted in a total collapse of the sector. UNWTO (2020) claims that the impact of the COVID-19 pandemic on the US tourism sector will be worse than the Severe Acute Respiratory Syndrome (SARS) and the global financial crisis of 2008.

This has motivated a lot of tourism academic researchers to study the impact of the COVID-19 pandemic on the global tourism sector (Khalid et al., 2021). However, very little has been explored if we go away from the macroeconomic variables and move towards the psychological variables such as market fear, investor mood, and sentiment. It has been argued that the individuals’ mood and sentiment can influence many economic aspects such as the stock market returns (Baker and Wurgler, 2006), trading behaviour (Su et al., 2017) and consumer spending (Ludvigson, 2004). This study explores the spillover effects from market fear, investor mood, sentiment, and economic uncertainty to the tourism sector of the United States during the COVID-19 pandemic. The study contributes by outlining the significance of fear, mood, sentiment and economic uncertainty in the forecast error variance of tourism sector stock returns.

To investigate the spillover effects, we use Standard and Poor's 500 index (S&P 500) index as the proxy for investor mood (Yap and Allen, 2011). CBOE Volatility index (VIX) index is used as a proxy for market fear (Yap and Allen, 2011). The economic uncertainty variable is captured using Economic Policy Uncertainty index (Baker et al., 2016).

2. Data & methodology

2.1. Data

We consider daily data of Dow Jones US Travel & Leisure index (hereafter T&L), S&P500, CBOE VIX (hereafter VIX), economic policy uncertainty (hereafter EPU), and valid number of tweets related to COVID-19 (hereafter Tweet) for a period from 03–01-2020 to 22–03-2021. We also consider daily data of the Media-hype index, Infodemic index and Media-coverage index related to COVID-19 and infectious disease equity market volatility (hereafter EMV) for the same period. The data for T&L, S&P500, and VIX are obtained from Bloomberg. The data for EPU and EMV are obtained from https://www.policyuncertainty.com. The data for the Media-hype index, Infodemic index and Media-coverage index are obtained from https://www.ravenpack.com. The data for Tweet are obtained from COVID-19 Infodemics Observatory. We consider logarithmic returns of the S&P 500 and T&L series in the study.

2.2. Spillover index

To examine the information spillover effects on the tourism sector, we make use of the spillover index proposed by Diebold and Yilmaz (2012). Since this study aims to investigate the impact of the market fear, investor mood, sentiment, and the economic uncertainty on the tourism sector of the United States, we explore both net spillover as well as net-pairwise spillover. Net spillover will help us to identify whether the tourism sector of the United States is a net transmitter or net receiver of the shocks during the COVID-19 pandemic. Net-pairwise spillover will help us to identify the transmission of shocks between each variable under study and the tourism sector of the United States.

The VAR model of Diebold and Yilmaz (2012) is estimated as

y_{t} = \sum_{i = 1}^{q} β_{i} y_{t - i} + ε_{t}

(1)

The above VAR model is represented in moving-average form as

y_{t} = \sum_{j = 1}^{p} A_{j} ε_{t - j}

(2)

Based on Pesaran and Shin (1998), we make use of the generalized H-step ahead forecast error variance decomposition to estimate the directional, total, net, and net-pairwise spillover as follows

θ_{i j} (H) = \frac{σ_{j j}^{- 1} \sum_{h = 0}^{H - 1} {(e_{i}^{'} A_{h} \sum e_{j})}^{2}}{\sum_{h = 0}^{H - 1} (e_{i}^{'} A_{h} \sum A_{h}^{'} e_{j})}

(3)

where

σ_{j j}

represents the standard deviation of the error term, ∑ represents the variance matrix, and

ε_{t}

is a vector with each i-th element to be one and all other elements to be zero. The main diagonal elements represent the contribution by the own-variable and the off-diagonal elements represent the contribution by other-variable.

Here, the sum of the other-variable and own-variable variance contribution is not equal to one in the case of the generalized decomposition. Therefore, we normalize each entry of the matrix by the row sum as follows

{\tilde{θ}}_{i j} (H) = \frac{θ_{i j} (H)}{\sum_{j = 1}^{N} θ_{i j} (H)}

(4)

Using equations (3) and (4), we estimate the total spillover index (TSI) index as

T S I (H) = \frac{\sum_{i j = 1, i \neq j}^{N} {\tilde{θ}}_{i j} (H)}{\sum_{i j = 1}^{N} {\tilde{θ}}_{i j} (H)} \times 100 = \frac{\sum_{i j = 1, i \neq j}^{N} {\tilde{θ}}_{i j} (H)}{N} \times 100

(5)

Next, the directional spillover index (originating from variable i to all j variables) is estimated as

D S I_{i \to} (H) = \frac{\sum_{j = 1, i \neq j}^{N} {\tilde{θ}}_{j i} (H)}{\sum_{i j = 1}^{N} {\tilde{θ}}_{j i} (H)} \times 100 = \frac{\sum_{j = 1, i \neq j}^{N} {\tilde{θ}}_{j i} (H)}{N} \times 100

(6)

Similarly, the directional spillover index (originating from all j variables to variable i) is estimated as

D S I_{i \leftarrow} (H) = \frac{\sum_{j = 1, i \neq j}^{N} {\tilde{θ}}_{i j} (H)}{\sum_{i j = 1}^{N} {\tilde{θ}}_{i j} (H)} \times 100 = \frac{\sum_{j = 1, i \neq j}^{N} {\tilde{θ}}_{i j} (H)}{N} \times 100

(7)

Equations (6) and (7) help us estimate the net spillover index (NSI) (from variable i to all j variables) as follows

N S I_{i} (H) = D S I_{i \to} - D S I_{i \leftarrow}

(8)

Finally, the net-pairwise spillover index (NPWSI) is estimated as

N P W S I_{i j} (H) = (\frac{{\tilde{θ}}_{i j} (H) - {\tilde{θ}}_{j i} (H)}{N}) \times 100

(9)

3. Empirical results

3.1. Whole-sample spillover table

Table 1 presents the whole-sample results of the spillover analysis. The xy^th term represents the contribution from the shocks to series y to the forecast error variance of series x indicating the intra-series connectedness (diagonal elements) and inter-series connectedness (non-diagonal elements) of shocks. The higher diagonal values indicate stronger intra-series connectedness of shocks than the inter-series connectedness of shocks. The ‘Contribution To others’ for the given y series indicates the directional spillover of shocks to the other series from y. The findings indicate that approximately 51.637% of directional spillover of shocks is sourced from the tourism sector to other series. The ‘From’ column indicates the gross directional spillover from other series to the given x series. The findings indicate that approximately 54.053% of the shocks associated with other series (investor mood, fear, policy uncertainty and sentiment) spills over to the tourism sector. The total spillover index of 33.082% supports the moderate interconnectedness among the given series. The results based on the net spillover indicates S&P 500, VIX and Tweet are the net transmitters of shocks to other series and the T&L and EPU series are the net receiver of the shocks from others indicating that the shocks from the other series spillover to the tourism sector during the COVID-19 period.

Table 1.

Whole-sample spillover analysis.

x↓/y→	T&L	S&P500	VIX	EPU	Tweet	FROM
T&L	45.947	32.724	20.102	0.816	0.410	54.053
S&P 500	28.011	44.734	25.721	0.842	0.691	55.266
VIX	21.463	23.018	54.188	0.035	1.296	45.812
EPU	2.062	2.050	1.786	92.584	1.518	7.416
Tweet	0.101	0.955	0.780	1.031	97.134	2.866
Contribution TO others	51.637	58.747	48.389	2.724	3.915	Total spillover
Contribution including own	97.584	103.482	102.578	95.308	101.049	33.082
Net spillovers	−2.416	3.482	2.578	−4.692	1.049	—

Note: T&L: Travel & Leisure index; EPU: economic policy uncertainty; S&P 500: Standard and Poor's 500 index; VIX: Volatility index

3.2. Time-varying spillover analysis

To analyze the evolution of the spillover effect, we undertake a time-varying analysis using a 30-day moving window with 10-days ahead forecasts. Figure 1(a) presents the estimates of the time-varying total spillover index. Even though the total spillover estimate based on a whole sample is 33.082%, the time-varying estimates mainly lie between 27.3% and 67% and fluctuate across the study period. We observe the highest peak in the spillover index on 26 Feb 2020 when due to a sudden rise in COVID-19 cases the global financial market started experiencing an environment of uncertainty and recession. This was followed by travel restrictions resulting in economic and social harm to the tourism sector in the United States. The second, third and fourth peaks were observed on 22-06-2020, 29-10-2020, and 28-01-2021, respectively, can be related to recovery in markets observed due to actions taken by regulatory bodies to boost the economy. Next, we analyze how investor mood, investor fear, policy uncertainty and sentiment impacted the tourism sector based on the directional spillover to the tourism sector and net pair-wise spillover analysis.

Figure 1.

(a) and (b): Total spillover and directional spillover.

3.3. Directional spillover analysis

Figure 1(b) reports the time-varying directional spillover from the given variables to the tourism sector. The estimates of directional spillover from these series indicate a moderate influence of investor mood, investor fear, policy uncertainty and sentiment on the tourism sector. Here also, we notice a similar pattern in the evolution of directional spillover as we observed in the total spillover index indicating the shocks in these variables impacted the performance of the tourism sector in the United States during the COVID-19 period.

3.4. Net pair-wise spillover effect

The static analysis indicates that investor mood, fear and sentiment are the net transmitters of shocks. Figure 2 presents the time-varying net pair-wise spillover effect between the tourism sector and other given variables. The results indicate that the shocks to the investor mood, fear, policy uncertainty and sentiment frequently switch their role between net-transmitter and net-receiver across the study period regarding shock spillover with respect to the tourism sector. We notice the difference in magnitude of net spillover estimates across different pairs with wider fluctuations for T&L-VIX and T&L-Tweet pairs and comparable narrower fluctuations for T&L-S&P500 and T&L-EPU pairs indicating the impact of the given variables on the tourism sector is heterogeneous. During the peak periods (as observed from total spillover index estimates and directional spillover estimates), all indicator variables are mainly acting as a net transmitter of shocks.

Figure 2.

Net pairwise spillover.

3.5. Impact of media-hype, infodemic, media-coverage and infectious disease EMV related to COVID-19 on total and directional spillover

Table 2 reports the results indicating the significant impact of media-hype, Infodemic index, media coverage and infectious disease EMV (related to COVID-19) on total and directional spillover effects. The impact of media coverage, infodemic and infectious disease EMV is stronger than media-hype. The findings indicate that excess information including deceptive information related to COVID-19 significantly impacted the tourism sector by directing information spillover from investor’s mood, fear and sentiment and policy uncertainty.

Table 2.

Impact of media-hype, Infodemic index, media-coverage and infectious disease EMV related to COVID-19 on total and directional spillover.

	Intercept	Media-hype	Infodemic	Media-coverage	Infectious disease EMV
Total spillover	82.240^#	−0.151^*	0.249^#	−0.697^#	0.250^#
Directional spillover	18.170^#	−0.042	0.176^#	−0.223^#	0.101^#

# and * mean significant at 1% and 5% levels of significance, respectively.

4. Conclusion

The study confirms the presence of spillover of shocks related to investor mood, fear, sentiment, and policy uncertainty to the tourism sector in the United States. The findings also provide evidence of the significant role of media-hype, infodemic, media coverage and infectious disease equity market volatility in impacting the total spillover and directional spillover from investor mood, fear, sentiment and policy uncertainty to tourism sector returns. The study has implications for investors, portfolio managers and tourism industry stakeholders with significance in forecasting tourism stocks returns.

Moreover, the findings have various policy implications. During periods of higher psychological and socio-economic uncertainties, tourism policymakers and authorities should implement strategic plans to empower tourism activities. The economic support during such uncertain periods can add value to the tourism firm’s performance. Different partners of the tourism sector including legislatures of host nations, tourism firms and affiliations need to team up in creating and implementing contingency plans to subside the impact of uncertainties.

Footnotes

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iDs

Faisal Nazir Zargar

Dilip Kumar

Appendix

Table A1:

Details of the variables.

Variables	Definition
T&L	It represents the returns of the Dow Jones U.S. Travel & Leisure index. It is considered to represent the performance of the travel, tourism, and leisure companies in the United States (Hadi et al., 2020). It measures the stock price performance of the US travel, tourism, and leisure companies.
S&P 500	This represents the returns of the S&P 500 index. It is taken as a proxy of investor mood (Yap and Allen, 2011).
VIX	It tracks the expected volatility of the U.S. stock market based on the real-time data derived from the S&P 500 index options and is used as a proxy for market fear (Yap and Allen, 2011).
EPU	It measures the policy-related economic uncertainty in the United States (Baker et al., 2016).
Tweet	It represents the valid number of tweets related to COVID-19.
Media-hype	It measures the percentage of distinct stories mentioning coronavirus.
Infodemic	It measures the percentage of all entities (places, companies, organizations, etc.) that are reported in the media alongside COVID-19.
Media-coverage	It is estimated as the daily count of distinct sources of news mentioning coronavirus divided by the total daily count of distinct sources.
Infectious disease EMV	It measures the economic uncertainty induced due to infectious diseases (Baker et al., 2020).

Note: T&L: Travel & Leisure index; EPU: economic policy uncertainty; S&P 500: Standard and Poor's 500 index; VIX: Volatility index

Author biographies

Faisal Nazir Zargar is an Assistant Professor at BITS Pilani Dubai Campus, UAE. His research interest includes volatility modelling, crypto economics and tourism economics.

Dilip Kumar is an Associate Professor at the Indian Institute of Management Kashipur, India. His research interest includes risk analysis including interdisciplinary research with different domains including tourism economics.

References

Baker

Wurgler

(2006) Investor sentiment and the cross‐section of stock returns. The journal of Finance 61: 1645–1680.

Baker

Bloom

Davis

(2016) Measuring economic policy uncertainty. The Quarterly Journal of Economics 131: 1593–1636.

Baker

Bloom

Davis

, et al. (2020) Covid-induced Economic Uncertainty. National Bureau of Economic Research.

Diebold

Yilmaz

(2012) Better to give than to receive: predictive directional measurement of volatility spillovers. International Journal of Forecasting 28: 57–66.

Hadi

Katircioglu

Adaoglu

(2020) The vulnerability of tourism firms’ stocks to the terrorist incidents. Current Issues in Tourism 23: 1138–1152.

Khalid

Okafor

Burzynska

(2021) Does the size of the tourism sector influence the economic policy response to the COVID-19 pandemic? Current Issues in Tourism: 1–20.

Ludvigson

(2004) Consumer confidence and consumer spending. Journal of Economic perspectives 18: 29–50.

Pesaran

Shin

(1998) Generalized impulse response analysis in linear multivariate models. Economics Letters 58: 17–29.

Fang

Yin

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

10.

Yap

Allen

(2011) Investigating other leading indicators influencing Australian domestic tourism demand. Mathematics and Computers in Simulation 81: 1365–1374.

11.

World Tourism Organization. Market Intelligence, & Promotion Section (2007) UNWTO world tourism barometer (Vol. 5). World Tourism Organization.