Predicting In-Play Match Decisions

We investigate the effect of the crowd on the decisions of match officials within a professional sports environment. We do this using data from 9,835 football (soccer) matches, comparing matches played behind closed doors because of the COVID-19 pandemic with those played before a crowd. We find that home advantage in terms of in-play decisions by match officials is significantly reduced in the absence of crowds. Examining the decisions of football referees, we find that away teams receive fewer yellow and red cards when playing in an empty stadium compared to matches with a crowd. This suggests the decisions of officials are influenced by the social pressure on match officials of a crowd, and that forecasts of match-related events and outcomes should be adjusted accordingly.

Who are the most important players in team sports

We present a statistical model using box scores from games as data to develop a theory to determine which players are the most important to success in team sports. We apply this to NBA basketball, NFL football and NHL hockey. The results show that the most important players for team success are not necessarily the most outstanding players. Moreover, its generally much more successful to have several good players rather than a single outstanding player. In NBA basketball, NFL football, and NHL hockey defensive players often stand out. The results are useful for good team construction, game strategy and recruiting of players.

Holiday Effects in the US Equity Futures Markets

We investigate the holiday effect in US equity futures markets during three sub-periods 1993-2011, 1993-2020, and during the 2020 covid-19 year for small cap stocks measured by the Russell2000 and large cap stocks measured by the S&P500. All the days from -3 before the holiday to -1 had gains and for the large caps there were gains on +1 and +2. The effect is stronger for the small caps. The year 2020 had results similar to the longer series with positive gains. We show the various holidays by holiday day and observe that the -3 day had gains on all the holidays whereas the other days did not. The effect has diminished in the 1990s and 2000s and only the -3 day is statistically significant. The -3 day in the futures anticipates the cash move on -1 day.

Revisiting the Balanced Book Hypothesis – Bettor Preferences and Results in the NFL

Data on betting percentages, both in terms of number of bets and actual money wagered, is still difficult to find. Sports Action Network, in their premium access service, does provide this data in terms of both number of bets and money bet. For the2020 NFL season, it was found that the balanced book could be rejected as bettors were shown to prefer road favorites, big favorites, and the over at the highest totals. Allowing this imbalance in the sides market appeared profitable for the book as the underdog won more often than implied by efficiency, while totals were evenly split.

Does the impact of the COVID-19 pandemic influence the FX? A note.

This note shows that the effective response of a country in its battle against COVID-19 influences the exchange rate of its currency. Particularly, we examine the GBPUSD, AUDUSD and AUDGBP pairs of currency during the COVID-19 outbreak and the results show that the domestic currency of the country which documents more COVID-19 cases in each pair is depreciated against the foreign one. Therefore, a country which cannot effectively mitigate the impact of COVID-19 and whose currency is depreciated may present further economic consequences in the future. Such consequences extend beyond economic recession and may include sovereign and interest rate risk. These findings may be useful for policy makers in order to estimate the cost of the pandemic.

COVID-19 and Two different Stock Markets: An Event Study Analysis

In this study, we examine two different stock markets’ response to the COVID-19 pandemic using event study methodology and a novel linear regression model. We use LSE (UK) as a proxy for the developed countries stock market and DSE (Bangladesh) as a proxy for the developing countries stock market. Using the daily COVID-19 confirmed cases and deaths and stock market returns data from these two countries (UK and Bangladesh) over the period November 01, 2020 to August 07, 2020. Our main research question was, which stock market suffered more during the COVID-19 pandemic, whether developed countries stock market or developing countries stock market. We find that developed countries stock markets (LSE as proxy) responded negatively to the growth in COVID-19 confirmed cases and deaths in COVID-19. We further find that developing countries stock markets (DSE as proxy) did not responded to the growth in COVID-19 confirmed cases and deaths in COVID-19.

Is There a Method to Madness? Predicting Success of Bollywood Movies

The objective of the study is to develop a parsimonious model to predict the box office success of a Bollywood movie before its release. A movie is considered successful if the revenue generated is greater than its budget, in other words, a Revenue to Budget Ratio (RBR) greater than 1. An original data set of 1698 Hindi movies released across a period of 13 years is used to identify the success factors of a movie in the Indian context. Predictive models are developed using traditional methodologies like multiple regression and logistic regression, as well as, contemporary approaches like regression trees and classification trees. The results highlight a unique mix of elements that a producer should consider to ensure the success of a movie in the highly competitive Indian movie market.

Dividend Changes as Predictors of Future Profitability

This study investigates “the information content of dividends hypothesis” using data on UK firms from 1990-2015. Dividends act as an important conveyor of information. Dividend changes may trigger changes in stock prices because they may convey new information about the firm’s future earnings and profitability. Why do companies pay dividends (or analogously why are stockholders interested in receiving dividends), given that it is well known that dividends are often taxed heavily? This question is of special interest in the UK, where the dividend tax is higher than the capital gain tax. Previous research has used a number of dividend policy theories to explain the dividend policy puzzle. We carry out several estimations and find out that contrary to some other studies, there is no evidence that dividend increases (decreases) provide information about the future profitability or earnings of UK firms.

Tracking Errors and Their Determinants

The purposes of this study are to compare the tracking error between 53 sampled physical and 15 over-the-counter (OTC) swap-type exchange-traded funds (ETFs) on the Tokyo Stock Exchange, and to contribute to a better understanding of the impact of selected determinants on the daily tracking error. The sample synthetic ETFs are found having higher tracking error than the sampled physical ETFs. The synthetic-type ETF managers may be difficult in using derivatives to replicate the benchmark performance. A panel regression model with cross-section fixed effects indicates the tracking error of the sampled physical ETFs is negatively related to size but positively related to expense ratio, dividend yield, trading volumes, market risk, and number of constituents in the target indexes. The results conform with the hypotheses that the expense, delay in receiving dividends, the trading cost and the market risk may erode the tracking ability; on the other hand, the economies of scale will improve the tracking ability. This study may help to raise a broader discussion of potential tracking error determinants and to provide some new insights.