Is eSports a bubble?

Authors: Armando Martino, Financial Analyst, Carmine Apice, Data Scientist at EY

Summary

Investing in the eSports market appears to be very profitable, however a more careful analysis shows how, net of the excellent performance of the stocks that belong to this world, it's still necessary to make arrangements on one's investing strategy. This does not prevent us from considering this market very interesting, also in light of the recent pandemic period that has led many people to give attention to eSports (both as casual users of contents and from a financial investment point of view). The analysis shows how the US security market could offer profitable opportunities if properly exploited.

eSports Overview

An eSport, a.k.a. electronic sport, is a sport competition between professional players, played individually or in team, using video games. The most common declinations of this discipline are:

• Multiplayer online battle arena;
• First-person shooter;
• Fighting;
• Real-time strategy.

Compared to the other more classical sports (Soccer, NBA, NFL, etc.), the rapid growth of eSports in the last years could be associated to some specific advantages: on top of these, flexibility to change size or scale. Indeed, due to their reliance on digital and mobile platforms, they can avoid location and physical limits, reducing barriers to entry for players and viewers (i.e. to see an eSport match, it's not necessary to phisically move to a stadium/arena or paying some expensive ticket, but you only need an internet connection).

Last year electronic gamings have reached a total of over 211 million US dollars (with over 4000 tournaments) in prize money, increasing of almost 30% compared to the 2018.

eSports Industry

The turnover related to the world of eSports is estimated to reach 1.6 billion US dollars in 2023, growing at an average annual rate of 19% in the period 2020-2023, and having a global audience of approximately 646 million viewers (for additional info, see section Sources).

These data are in line with what has been observed in the last few years, when this market has seen a relevant growth in terms of audience.

In 2019 the majority of that was concentrated in the Asia-Pacific region, followed by Europe and Rest of the World.

Currently (2020) the market total revenue is estimated to be 0.95 billion dollars (increasing even faster in 2023), distributed on several streams: Sponsorships (the most relevant), Media Rights, Publisher Fees, Merchandise & Tickets, Digital events and Streaming.

North-America represents the largest slice of the market in terms of total revenue, accounting for 37% in 2019, making it very attractive from an investor point of view. The global emergency of COVID-19 has allowed the players operating in this context to grow further and to establish their already solid market positions.

How is it possble then to exploit market opportunities to extract value from companies operating in such a context?

The analysis proposed in the next paragraph shows the techniques used by the authors to build an equity portfolio based on shares that refer to US companies active in the financial world of eSports and gaming in general.

Portfolio Construction and Benchmark Comparison

Using the equity screener accessible on Yahoo Finance we have created our universe of stocks, considering the following parameters:

• Market: US
• Sector: Communication Services
• Industry: "Electronic gaming and multimedia" or "Internet content & information"

Then, taking into account some web searches and articles, we selected a total of 74 stocks.

In addition, a filter was applied to the initial set, excluding almost all OTC (Over The Counter) securities to minimize potential costs deriving from the counterparty and liquidity risk (we left only Tencent Holding due to its relevant role in the sector).

This process generated a total of 20 remaining stocks.

Having obtained the desired list, we dowloaded the time series of the daily security prices (USD) from 03/01/2017 to 30/10/2020, and we built a portfolio, evaluating its performance over time against specific benchmarks.

Below an excerpt of the R code

########################
### Upload packages ####
########################


tryCatch(library(tidyquant),error=function(e)install.packages('tidyquant'))

tryCatch(library(tibble),error=function(e)install.packages('tibble'))

tryCatch(library(PerformanceAnalytics),error=function(e)install.packages('PerformanceAnalytics'))

tryCatch(library(chron),error=function(e)install.packages('chron'))

tryCatch(library(gridExtra),error=function(e)install.packages('gridExtra'))



###########################
### Ticker to Download ###
###########################

ticker=c("SE","ATVI","EA","TTWO","BILI","ZNGA",'GLUU','GRVY','AVID','SCPL',
         'INSE','RNWK','GIGM','BHAT','SLGG','NCTY','YVR','DOYU','HUYA',
         'TCEHY')
ticker_benchmark=c("^NDX","^IXIC","^GSPC")
name=c('Sea Limited','Activision Blizzard, Inc.','Electronic Arts Inc.',
       'Take-Two Interactive Software, Inc.','Bilibili Inc.','Zynga Inc.',
       'Glu Mobile Inc.',"Gravity Co., Ltd.","Avid Technology, Inc.",
       "SciPlay Corporation","Inspired Entertainment, Inc.","RealNetworks, Inc.",
       "GigaMedia Limited","Fujian Blue Hat Interactive Entertainment Technology Ltd.",
       "Super League Gaming, Inc.","The9 Limited","Liquid Media Group Ltd.",
       "DouYu International Holdings Limited",
       "Huya Inc.","Tencent Holdings")
name_benchmark=c("Nasdaq 100","Nasdaq Composite","S&P 500")



#####################################################################
### Rebalancing Freq., start-date, end-date and transaction-costs ###
#####################################################################

freq_reb='3 months' #One of the possible options
start_date=as.Date('2017-01-01')
end_date=as.Date('2020-10-31')
costs=0.0050 # 50 basis point


###################################
### Download time series prices ###
###################################

lista=list()
for(j in 1:length(ticker)){
hist_data=tq_get(ticker[j],get='stock.prices',from=start_date,to=end_date)
hist_data2=as.xts(read.zoo(hist_data[,c('date','adjusted')]))
colnames(hist_data2)=ticker[j]
lista[[j]]=hist_data2
}



for(j in 1:length(ticker_benchmark)){

hist_data_benchmark=tq_get(ticker_benchmark[j],get='stock.prices',from=start_date,to=end_date)
hist_data_benchmark2=as.xts(read.zoo(hist_data_benchmark[,c('date','adjusted')]))
colnames(hist_data_benchmark2)=ticker_benchmark[j]
lista[[length(lista)+1]]=hist_data_benchmark2
}



prices=Reduce(merge,lista)
head(prices)


###################################
### Calculate the Gross Portfolio #
###################################

# N.b. we leave to the reader the construction of matrix of weights: as suggestion, for each rebalance date we assign an equal weight considering only the number of securities available at that date based on our price history (i.e. if a given security's price is not present on a given date, it does not participate in the calculation of the weights).


  ptf=Return.portfolio(rend,weights = weights,verbose = T)
  beginweights=ptf$BOP.Weight
  endweights=ptf$EOP.Weight
  transactions=beginweights-lag(endweights)
  transactions[is.na(transactions)]=0
  turnover=xts(rowSums(abs(transactions)),order.by = index(transactions))
  turnover[1,1]=rowSums(weights[1,])
  TO=turnover[turnover!=0,]
  averageTurnover=colMeans(TO[2:dim(TO)[2],])*100
  


###################################
### Calculate the Net Portfolio ###
###################################


  transactions_costs=turnover*(costs)
  ptf_with_costs=ptf$returns-transactions_costs



###################################
### Plot the Portfololio Analysis #
###################################

  compare=na.omit(cbind(ptf$returns,ptf_with_costs,rend_benchmark))
  colnames(compare)=c('Portfolio',paste0('Portfolio with ',costs*100,' % 
  costs'),name_benchmark)                
  charts.PerformanceSummary(compare)
  table.AnnualizedReturns(compare)
 

The benchmark indexes used are: Nasdaq Composite, Nasdaq100, S&P500. The first two were used as benchmarks for the eSports market, while the third was used as a reference for the US market as a whole.

Several rebalancing frequencies have been tested, net of transaction costs (estimated at a maximum of 50 basis points, following a very conservative approach), assuming equal weights.

The choice of the best rebalancing frequency depends on the investor's level of risk aversion. In particular, if the goal is to maintain a low volatility (but always higher than the benchmarks considered) and keep the returns high, the investor will prefer a 5-days rebalancing frequency, otherwise he will prefer the strategy without any rebalancing to maximize the return.

Conclusions

PROS

• The return per unit of total risk (i.e. sharp-ratio) calculated on the eSports portfolio is larger with respect to all berchmark considered.
• Such over-performance is consistent over-time.
• N.B. Among the different ways in which we can invest in eSports market, the public one is the more "liquid", with lower costs and risks, so it's in general preferable.

CONS

• Net to the above statistics, the constructed portfolio is not very diversified as it is composed of a small quantity of securities belonging to the same industry (the average number of securities cross re-balancing strategy lies between 13 and 16).
• Some securities contribute to the total return in a preponderant way, driving its performance (e.g. Activision Blizzard, Inc., Gravity Co., Ltd., Liquid Media Group Ltd.).
• To mitigate the drawdown's recovery time (which, as can be seen in the overall graph presented above, is slower than the one of the reference market) and the overall risk of the portfolio, less risky assets should be associated with the selected stocks.

To answer our initial question, that is whether or not the eSports market is a bubble destined to burst, we can say that the indicators analyzed describe the profile of a fast growing market, with interesting numbers and prospects still guided by a series of trends that have been confirmed over the years.

Sources