Keywords

Covid-19, Coronavirus, Text Mining, Financial Impact, Economic Impact.

Introduction

According to The Union Health Ministry, India, as of on 18 April, 2020 eleven thousand six hundred sixteen people in India got inflicted by coronavirus infection and four hundred fifty-two people succumbed to death. On 17 April, 2020, the WHO (2020) reports that there are 2,074,529 confirmed cases, and the death toll is estimated as 139,378. Amidst the other rumours such as Covid-19 is global bio war tool such as SARS that out broke in China during end of November, 2002 and finally had spread to 26 countries. As SARS spread from the epicentre there were news of new economic ruptures after the bank failures and currency devaluations five years ago. We can see a similar cycle in 2020 when the mass economic initiatives by the World Bank and individual government bodies are being assessed to counteract the economic massacres on the account of reduced manpower and working hours as well as collapse of trade in general. There are ongoing economic negotiations for financial moratoriums to be distributed among various states in India. According to a report by The Times of India Covid-19 will make 195 million full-time jobs or 6.7 % working hours wiped out. There are debates going on about borrowing, lending, credit deferments, economic stimulus packages to combat the upcoming recession etc.

The report by International Labour Organization ILO (2020) classified issues such as solvency, income loss, layoffs, unprotected workers, changes in working hours, annual job losses. workforce displacement, limited access to health services and social protection, looming risk of escalated poverty PTI (2020), policy responses, scarcity public resources as major indexes for the economic deterioration in the present financial quarter. The International Labour Organization ILO (2020) also did a thematic analysis to create a documentation on “Interventions to support enterprises during the COVID-19 pandemic and recovery”.

This article analyses thirty important economics and financial market related newspaper articles by eminent analysts and did a content analysis to identify the main themes and subthemes associated. This research bring paper brings forth the main financial and economic themes at one vessel to be used for future researches by market analysts and researchers.

The value lies in the consolidation of the major setbacks during this pandemic, itemizing them in one list, and thereby essentially creating a vessel where all the peak economic and financial crisis had been accumulated.

Samples

Thirty-Three articles comprising of a total of thirty thousand fifty-nine words were collected from sources such as BBC, The Times, Aljazeera, The Economics Time, The New York Times, The Wall Street, Bloomberg, The Guardian, The Telegraph, The Washington Post, Lancet etc. Most of the Articles were composed by eminent scholars such as Nobel Laureates as well as other famous economists, and financial analysts. The articles talk about the global economic situations due to the outbreak of Covid-19 as well as emphasise the geographical markets such as Vietnam, China, Korea, India, The USA, Europe, The Middle East etc. The articles were collected between first and second week of April, 2020.

Method

A simple analytic consolidation of main themes and associated subthemes were conducted using Nvivo 12 plus software. NVivo 12 is software provided by QSR International, which is effective in analysing the thematic analysis as well as hierarchical clustering using word similarity and Pearson's coefficient (Azeem, 2012). The thirty-three articles comprising of 67 pages and 30059 words were put into a Microsoft word document and analysed for thematic extraction by machine algorithm supported automatic coding.

Results

Initially we “auto-coded” (software algorithm generated coding) the document using NVivo 12. The NVivo software uses a semi-supervised “machine-learning” algorithm to automatically code (Hai-Jew, 2014). Considering the length (67 pages) of the whole document we found the relying on the machine algorithm was more realistic and scientific approach rather than coding it through manual open/axial coding for identification of themes-subthemes. Simultaneously different levels of sentiment analysis were performed using the software as discussed in the following section.

The Sentiment count of identified “auto-coded” words are shown in Table 1. The total number of relevant key words with associated sentiments were 718, out of which only 64 words had positive sentiments associated with them; and 237 words were neutral. 286 of the identified keywords had Negative sentiments, and 131 words reflected Mixed (both positive and negative) sentiments. Therefore, only 9% of total number of identified words had positive sentiments Figure 1. The Neutral and Mixed sentiments were 33% and 18% respectively (See Figure 1). The Negative sentiments occurred for 40% and had the maximum share (Figure 1). Figure 2 explicitly shows the data in Figure 1 in a 2 dimensional matrix for an alternative appeal to the visual pallet.

Table 1 Number of Coding References for Coded Sentiments
Codes	Number of coding references
data_192	718
data_192 - Mixed	131
data_192 - Negative	286
data_192 - Neutral	237
data_192 - Positive	64

The main themes with more than 2% occurrences are tabulated in Table 2. The main themes identified are: supply, products, chains, companies, market, payment, unemployment, shocks, consumers, email, systems, coronavirus, online, delivery, debt. Table 3 has given the full list of the themes and associated subthemes. The main themes that appeared for 1% or so are: advertising, growth, benefits, chains, challenge, health, risk, retailers, strategy, credit, services, unemployment, policy, and debt.

Table 2 % Occurrence of Coded Themes
THEMES	%  OCCURANCE
supply	4.53%
products	3.79%
chains	3.41%
companies	3.2%
market	3.07%
payment	2.61%
unemployment	2.51%
shocks	2.43%
consumers	2.36%
email	2.36%
systems	2.3%
coronavirus	2.05%
online	2.05%
delivery	2.04%
debt	2.01%

Table 3 Themes and Subthemes from the Thirty Three Articles
Auto-coded Themes
advertising	growth	products
advertising arrangements	economic growth rate	dairy products
advertising commitments	economic growth targets	factory production
advertising partners	forecast growth	finished product
digital advertising	initial growth path	health-relevant products
asset	target growth rate	hygiene products
asset valuations	health	medicine production
equity asset class	global health	product groups
risk assets	health conditions	productive activity
safe assets	health crisis	scheduling products
wide asset classification forbearance	health infrastructure	protections
benefits	mental health support	anti-virus protection
awarding benefits	public health policies	coronavirus protections
core benefit	hit	protective gear
unemployment benefits	direct hit	rate
capacity	financial hit	economic growth rate
adding capacity	hitting manufacturing	household savings rates
excess capacity	indirect hit	key interest rate
logistical capacities	household	occupancy rates
chains	household exposure	policy rates
2016 value chain model	household savings rates	target growth rate
fair share	household wealth contracts	recession
global value chains	industry	global recession
share values	banking industry	last recession
supermarket chain	extractive industries	policy recessions
supply chains	particular industry	recession risk
challenge	theatre industry	recovery
challenging categories	infrastructure	recovery cycle
challenging proposition	city infrastructures	recovery dynamics
current coronavirus challenge	health infrastructure	recovery patterns
legal challenge	road infrastructure	recovery strategy
major challenge	instructions	response
companies	changed payment instructions	bungled response
aviation companies	urgent instructions	crisis response
big companies	wiring instructions	monetary policy response
chinese companies	internet	retailers
company guidelines	internet router	bricks-and-mortar retail
german company	internet use	electronic retailer
gig economy companies	password-protected internet connections	large retailers
large companies	legacy	risk
lingerie company	macroeconomic legacy	additional transmission risk
luxury goods company	microeconomic legacy	recession risk
several pharma companies	loan	risk assets
tourism companies	bad loans	risk profiles
consumers	bank loan repayment	services
chinese consumers	loan covenants	backup child care service
consumer comfort	past loans	cleaning services
consumer confidence correlate	market	delivery services
keeping consumers	financial market performance	remote meeting services
coronavirus	financial market sell-offs	shocks
coronavirus crisis	key markets	demand shocks
coronavirus governments	key stock market	economic shock
coronavirus protections	market intelligence services    portal	exogenous shocks
coronavirus strain	market signals	prior shocks
coronavirus struggle	market worldwide	supply shocks
current coronavirus challenge	wholesale money market	supply-side shock
credit	media	shopping
corporate credit	media giant	bricks-and-mortar shopping
credit intermediation	social media accounts	offline shopping worlds
credit spreads	social media content	online shopping
crisis	models	store
coronavirus crisis	2016 value chain model	physical stores
crisis planning	door-to-door models	store cleaning
crisis response	economic model	store closing
health crisis	mechanical models	strategy
cut	national	post-recovery strategy
cut backs	foreign nationals	recovery strategy
surprise cut	national concerns	supply
temporary pay cuts	national developments	critical supplies
damage	national governments	essential supplies
actual damage	oil	requisition supplies
economic damage	crude oil futures	sufficient supplies
structural damage	global oil demand	supply chains
debt	mentha oil futures	supply disruption
bank holiday	oil importer	supply shocks
debt holiday provision	online	supply shortage
generous provision	online purchases	three-year supply pact
government debt	online sales channels	support
delivery	online shopping	fiscal support
delivery services	online spaces	mental health support
efficient delivery system	transacting online	policy support
enabled delivery systems	packaging	taking support
home delivery	economic rescue package	systems
demand	economic stimulus packages	company systems
demand shocks	payment	digital inventory system
exogenous demand	changed payment instructions	efficient delivery system
global oil demand	paid time-off	enabled delivery systems
economy	payment leeway	intuitive reasoning system
advanced economies	post-holiday payments	political systems
gig economy companies	resumed payments	transmission
major economies	temporary pay cuts	additional transmission risk
real economy shock	planning	classic transmission
email	crisis planning	plausible transmission channels
365 email accounts	planned broadcast	transmission mechanism
business email interruption scams	post-recovery planning	unemployment
internal company email	policy	unemployment benefits
phishing email	monetary policy response	unemployment claims
goods	policy action	unemployment insurance appeal board
ancillary goods chemicals	policy error	unemployment insurance fund
luxury goods company	policy interventions
material goods	policy rates
government	policy recessions
government debt	policy support
individual governments	public health policies
national governments

A deep look into Figure 3 will divulge that Mixed sentiments are actually dominating most of the thematic words. For a few words (companies, systems, unemployment, challenge, products, market, benefits, loan, industry, economy, protect, support, delivery, internet, infrastructure, media, advertising, consumers, email, capacity etc.) the Neutral sentiments governed the second biggest chunk after Mixed sentiment. So even when the death toll is rising and it seems that the world is coming to a standstill, the economic measures and media news seem to have kept a neutral and mixed sentiment ambience for the most deeply economically hit sectors. The Negative sentiments are observed for words such as supply, products, coronavirus, payment, rate, stocks, growth, debt, risk, recession, demand, damage, good, retailers, credit, cut, crisis, transmission, national, models, planning, response. A focused review of these words shows that the direct economic and financial components, and the retail industry had been shaken badly and the fear is emanating from and resonating in these domains.

The only holistically positive and mixed sentiments are ruling the following themes: infrastructure, advertising, apps, internet, capacity, loan, and protection. Figure 4 provides the most dominating hundred words in a machine generated basic word cloud.

The reason advertising was seen to be in a mixed and positive zone due to digital applications Table 3. Infrastructure seems to be more of an established component with three sub-components as represented in Table 3: city infrastructures, health infrastructure, and road infrastructure. The loan market shows a better mood due to bank loan repayment, loan covenants, and past loans. The protection key word Table 3 has a better impact as it talks about anti-virus protection, coronavirus protections, and protective gear. However, economic protection is totally neglected. The logistical capacities are the main reason why we had a primacy dominance of mixed feeling and secondary dominance of positive sentiment in the keyword “capacity”.

Discussions

The scope of this article, as stated before, to identify the opportunities during the mightiest pandemic crisis in the last century. The article simply finds the right scope of in-depth research problems for better analytically presented managerial solutions. In this line this research found certain scopes for researchers to target certain issues.

The auto-coded themes and subthemes in Table 3 highlighted a few important points as follows:

1. Growth: The growth is related to forecasting as well as the growth cycle, growth target and rate. There are abundant scopes for research in term of the whole economic cycle at this phase.

2. Assets: The asset (financials) need to be explored more crucially from the point of risk, safe, equity etc. How these have long term individual and collective impact should reach the research community for better management system for future crisis onsets.

3. Chains: The values chain models, global and individual retail chains ask for magnified and micro analysis.

4. Industry: Beyond banking, how the theatre industry is taking the hit is also of concern.

5. Recession: Are the current researchers put adequate weightage of importance for policy recession as the onset of any global war, bio-warfare, and pandemics?

6. Health: How the global health, health crisis, health infrastructure, and public health policies are adequate? Is there gap in the mental health support system?

7. Advertising: How are the advertisement arrangements, commitments are proceeding? How the advertising partners are getting affected due to this pandemic? Is there more scope for digital advertising as the physical advertisements is facing an impediment?

8. Products: The research expresses to specifically pour more focus on the industries such as dairy products, factory production system, health-relevant, medicine, hygiene products. The production activity and scheduling of production system also stood out.

9. Recovery: the text mining specifically pointed towards the recovery cycle, recovery dynamics, recovery patterns, and recovery strategy.

Table 3 also points to the specific possibilities of explorations of various companies affected by the covid-19 pandemic, consumers, retailers, shopping, services that are crucial during the pandemic, and economics facets such as debt, loan, shocks, risk, credit, crisis, cut, supply, demand, unemployment, policy, and payment.

The research itself does not provide any solution, nor it does explore any gap in the current economic analysis. It simply explores the possibilities of certain themes and related sub-themes and point at the major research needs under the spell of the pandemic. This crisis can be a learning phase in industrial 4.0 era in order to control, plan, and find feasible solutions at the quickest time. And the research simply put the major issue, that are observed at this point of time, upfront for concise and precise outlooks at one go for immediate concentration.

References

1. Azeem, M., Salfi, N.A., & Dogar, A.H. (2012). Usage of NVivo software for qualitative data analysis. Academic Research International, 2(1), 262-266.
2. Hai-Jew, S. (2014). Using NVivo: An unofficial and unauthorized primer. Retrieve from: http://scalar. usc. edu/works/using-nvivo-an-unofficial-and-unauthorized-primer/index.
3. ILO (2020). Interventions to support enterprises during the COVID-19 pandemic and recovery. Retrieved from https://www.ilo.org/wcmsp5/groups/public/---ed_emp/---emp_ent/documents/publication/wcms_741870.pdf
4. ILO (2020). LO Monitor: COVID-19 and the world of work. Second edition. Retrieved from https://www.ilo.org/wcmsp5/groups/public/@dgreports/@dcomm/documents/briefingnote/wcms_740877.pdf
5. Ministry of Health and Family Welfare. Government of India. Retrieved from https://www.mohfw.gov.in/
6. PTI (2020). About 400 million workers in India may sink into poverty: UN report. The Economic Times. Retrieved from  https://economictimes.indiatimes.com/news/economy/indicators/about-400-million-workers-in-india-may-sink-into-poverty-un-report/articleshow/75041922.cms
7. WHO (2020). Coronavirus disease 2019 (COVID-19) Situation Report –88. Retrieved from https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200417-sitrep-88-covid 191b6cccd94f8b4f219377bff55719a6ed.pdf?sfvrsn=ebe78315_6