A rational asymmetric reaction to news: evidence from English football clubs

Van Buskirk (2014) investigates the relationship between disclosure frequency and information asymmetry.

Thus, sports clubs cannot use disclosure of their performance strategically as discussed by Bergman and Roychowdhory (2008).

While 37.5% international participation rate might seem arbitrary, it is a clear break in the number of seasons of international play, with all elite clubs participating in at least 6 years of international play, while Tottenham, the non-elite club with the highest amount of international play, only had three seasons where they qualified for international play. Additionally, none of the elite clubs were relegated from the EPL; the only other club not relegated during the sample period is Tottenham, who has never been relegated from the EPL.

International tournaments provide substantial revenue to participating teams. In 2013–2014 the Europa League paid €1.3 million to each participating club with additional payouts ranging from €0.38 million to €14.6 million depending on the club’s performance, while the Champions League paid €8.6 million to each participating club with additional payouts ranging from €11.12 million to €57.4 million depending on the club’s performance (UEFA 2015).

The PZ is a London-based security exchange formerly known as OFEX, which is a less liquid market of the three exchanges. The AIM is a part of the LSE, but is designed for small and growing companies, which results in the AIM having less restrictive listing requirements than LSE.

The Football League was restructured to allow the English Premier League (EPL) clubs to compete more effectively against the other elite clubs in Europe. The restructuring also allowed the EPL to negotiate its own broadcast and sponsorship agreements.

This is primarily the result of the PZ not always operating on the same days as the other markets, primarily the result of holidays.

One might define expectation of winning to hold if prob(win) > 0.50 but this would ignore the possibility of a draw. As approximately one fourth of all matches end in a draw, the 44% indicates the club that is expected to win does so, on average.

Birmingham City Football Club (FC) was omitted from the analysis because their annual report ends on August 31st which causes the first month of one season and 9 months of another season to be combined in each annual report, which in turn could bias the results.

Operating income is used instead of net income because the latter is affected by the club’s financial leverage. Unlike Pinnuck and Potter (2006), we do not use revenue as investors are more interested in the operating income because additional revenues are commonly tied to additional expenses.

Data were collected from ESPN going back to the 2004–2005 season. This data was used to confirm the validity of statto.com. Statto.com is then used to get match results from 1992–1993 to 2003–2004.

This is consistent with Baek (2016) who recommends using real prices to “identify the true relationship between market variables and investor sentiment by making [the] results comparable over the whole period of [the] study regardless of inflation (p. 1045)”.

This is the result of there being three possible outcomes in a football match; win, lose, or draw (tie).

D1 matches are used as the omitted category.

The first half includes league matches played from August to December while the second half includes league matches played from January to May. The season is naturally split by a holiday break.

May 13, 2012 was the last day of the 2011–2012 EPL season and was arguably the most exciting final day of the EPL. Going into the last day the championship was still undecided between Manchester City and Manchester United. The third position in the standings, the last guaranteed spot to play in the Champions League the following season, was also undecided among three clubs: Arsenal, Tottenham, and Newcastle. Arsenal won third place over Tottenham by just one point. As for the bottom of the EPL, while Blackburn Rovers and Wolverhampton Wolves were already relegated heading into the last day, the third club to be relegated was still undecided between Aston Villa, Bolton Wanderers, and Queens Park Rangers (QPR). Ultimately, Bolton was relegated by finishing one point behind QPR. This is just an example of the possible importance of the final matches of the season.

Results are similar for raw returns.

We do not report the 4-day CAR for brevity as there is no significance in club’s 4-day CAR in any cases after including our control variables. The negative adjusted R-square for Day 5 Abnormal Returns for winners (in Table 6) indicates that 5 days after a match any abnormal returns are essentially (linearly) independent of match results or club characteristics.

We estimated the models in Table 9 using nominal rather than real data. All results are qualitatively similar although parameter estimates using nominal data are slightly larger in absolute value. The results using nominal data are available from the authors upon request. We also estimated the season-level model substituting first and second half performance with aggregated seasonal performance measured in total season wins and losses. Wins were not statistically related to adjusted operating income whereas season losses were negatively and statistically related to the same. For non-elite clubs, season losses were not as expensive as for elite clubs but winning was no more valuable to non-elite clubs. Full results of this estimation are available upon request.

In the sample, during the first half of the season the minimum number of games played was 18 and the maximum number of games played was 23.

In the sample, during the second half of the season the minimum number of games played was 17 and the maximum number of games played was 21.

This may support Manchester United’s controversial withdraw from the FA Cup in the 1999–2000 season.

There was no statistically significant difference between the impact of the percentage of wins and losses on club operating income in the EPL and the Football League Championship.

Edmans et al. (2007) notes that losing in knockout style tournaments could be a stronger signal than winning as it ends the club’s run in the tournament while the win only allows the club to advance one more match.

We also used a probit-based seemingly unrelated regression estimation to test the cross-equation equality of the impact of past wins and losses. We reject, at the 5% level, the null hypothesis that the impacts of previous wins and losses are equal across the two outcomes.