Introduction

The average costs of a Summer and Winter Olympic Games are $5.2 billion and $3.1 billion, respectively, with some event price tags running well over $10 billion (Flyvb-jerg, Stewart, and Budzier, 2016). Furthermore, they overrun their budgets by 156% on average. These costs have large and important implications for local governments hosting the games as most recently noted in the popular press regarding the 2016 Summer Olympics in Rio de Janeiro (Worstall, 2016; Kennedy, 2016).

In defense of such expenditures, boosters and governments argue the Olympic Games bring economic benefits to the community, as well as jobs to the area. The argument for such large government outlays is often an ex-ante one put forth during the bidding process using macroeconomic multipliers or input-output models (Centre for South Australian Economic Studies and KPMG Peat Marwick, 1993; Papanikos, 1999). However, ex-post academic studies provide evidence of only a small economic gain (Baade and Matheson, 2016; Baumann, Engelhardt, and Matheson, 2012).

To provide a better understanding of the effects of hosting the Olympic Games, and more specifically to weigh in on the debate on the benefits of hosting the Games, we test for whether the prospect of hosting the Olympics impacts a host country's stock market. We evaluate this impact in the context of the International Olympic Committee (IOC) host city/country bidding process, where cities/countries submit bids and compete to host an Olympics, culminating in the IOC announcement of who has won (and lost) the right to host an Olympic Games. In our analysis, we conduct an "event study," as described by MacKinlay (1997) among others, with the triggering "event" being the announcement by the International Olympic Committee. Our analysis examines market-level stock price data of the prospective host countries around the time of the IOC announcement to see if abnormally positive or negative returns occur around the time of the announcement. Using the returns results, we can empirically test whether a host country's stock market rises, and by extension, whether hosting the Olympics increases the profits of its companies and the wealth of its citizens, consistent with the typical economic arguments made in support of hosting an Olympics.

The results provided below extend the current literature on the topic of IOC announcements on stock prices. Our results are in line with Berman, Brooks, and Davidson (2000), who use an event study approach to find no general impact of the announcement of the Sydney games on Australia. In contrast to the earlier study, Veraros, Kasimati, and Dawson (2004) find a positive impact of winning the bid for the summer 2004 Olympics on the Athens Stock Exchange, using an event study approach. However, both early studies focus on a single event. In the literature looking at multiple announcements, in particular Dick and Wang (2010) and Mirman and Sharma (2010), they find statistically significant effects. The former study finds it for winners and the latter for losers of the bids. All the studies use standard parametric event study techniques in finding their effects.

We extend the event study literature on the impact of IOC announcements by (i) adding additional data (through the Chinese win of the 2022 Winter Olympics bidding process), (ii) analyzing stock market effects prior to the announcements to check for leakage of the announcement, and (iii) running non-parametric tests in addition to the standard parametric tests. Although our study contains more data, larger windows of analysis, and more robust estimates, we find little to no impact of the IOC announcements on the stock exchanges of the host countries. In other words, our results are in line with Berman, Brooks, and Davidson (2000), but in general fail to replicate the remaining literature even when taking a larger, longer, and more varied look at the data.

That being said, we see several rather large abnormal returns in the stock markets of countries who either won or lost their bid to host the Olympics. However, only one was large enough to reject a hypothesis of no effect after controlling for the fact so many tests were run. To test for an aggregate effect, we run on the order of 150 tests across multiple event windows using parametric and non-parametric approaches. We find only three could potentially reject the null at a 5% type 1 error threshold. Overall, given the event study approach, our results suggest a weak to non-existent effect that the IOC announcement affects a bidding country's stock market.

Model

To test for whether the Olympics have an economic impact on the country who hosts the games, we analyze whether the Olympics increase profits of publicly traded companies in the host countries. To measure whether there is an impact on profits, we exploit the theory that stock prices increase in value when expected future profits of firms increase. In other words, if firm ABC suddenly has an increase in expected future profits of $1 billion, then the value of the firm increases on the order of $1 billion, after adjusting for a variety of factors, such as the discount rate of the future earnings. Given the theory, it is key to pinpoint when expected profits of firms would likely increase as a result of hosting the Olympic games. If those times can be pinpointed, then measuring changes in the value of the firms in a host country, or their market capitalization as measured through their stock prices, provides evidence of an economic impact from hosting. In the case of the Olympic games, prior research suggests the point in time when companies are affected is when the IOC announces a country will host an Olympic games. It is at that point that firms can expect to get future business and profits by providing goods and services to the Games. As a result, we analyze the change in stock prices around the time of the IOC announcement.

To put it differently, we are using the event study approach following Berman, Brooks, and Davidson (2000), Veraros, Kasimati, and Dawson (2004), Dick and Wang (2010), and Mirman and Sharma (2010).

Event and Estimation Window

We take our notation and discussion of the event study model we use from MacKinlay (1997). In the context of event studies, our "event date" is the day of the IOC's announcement of who won the bidding process and will host the Olympics. To ensure we do not miss the potential impact of the announcement, we analyze a battery of event windows around the event date, investigating whether the announcement could have affected the stock market prior to the announcement, during the announcement, and after the announcement. In particular, we investigate eight different event windows, [[[tau].sub.1], [[tau].sub.2]], including [0,1], [0,2], [0,5], [0,9], [-2,2], [-5,5], [-5,-1], and [-2,-1] where t = 0 is the day of the announcement, [[tau].sub.1] is the start of the event window, and [[tau].sub.2] is the end of the window being analyzed. The days are counted as the change relative to the previous day and all windows are inclusive.

We analyze a wide range of windows for a variety of reasons. First, we are attempting to replicate previous studies. As a result, we include their windows. In particular, Dick and Wang (2010) include [0,1], [0,2], [0,5], and [0,9]. However, we fail to replicate their results. Therefore, we investigate windows that start prior to the announcement date, or [-2,2] and [-5,5], which is fairly common in the literature where information may be leaked. For completeness, we also provide estimates for windows [-5,-1] and [-2,-1] to be able to consider pre- and post-announcement effects separately and to ensure the effect isn't lost to changes in expectations right before the announcement.

Whatever the window, the objective is to capture whether host countries had abnormal returns during these event windows. In other words, do we see a country's stock market index, or a their firms' profits, increase unexpectedly around the time of the announcement? Central to our analysis is the assumption that the news contained in the IOC announcement is unexpected. Otherwise, the announcement would be priced into the market in advance of the announcement. However, as noted above, we utilize eight different event windows, to help in identifying and isolating market effects, even when there is leakage of the otherwise unexpected announcement result. Furthermore, we note the announcements are often a surprise to the general public, as documented in the news; refer to Magnay (2011) and BBC (2013) for the Pyeongchang and Tokyo games, respectively. Note that in a scenario where a city is expected to win, such as Beijing's winning bid for the 2022 Winter Olympics (Rauhala and Birnbaum, 2015), this would bias against finding a country-level stock market effect.

To estimate what is normal, we estimate returns of a country's stock market over an estimation window of -241 [less than or equal to] t [less than or equal to] -41. For comparison, we use the same estimation window as Dick and Wang (2010). All measures of t are trading days.

Estimating Abnormal Returns

In determining what is abnormal, we use the estimation window to calculate normal returns using a market model. Specifically, we use ordinary least squares (OLS) to estimate the linear relationship

[R.sub.it] = [[alpha].sub.i] + [[beta].sub.i][R.sub.mt] + [[epsilon].sub.t] (1)

for each announcement date, or observation, i, where [R.sub.it] is the one-day return on the stock market for country i. As noted above, t represents the date relative to the announcement. [R.sub.mt] is the market return as measured by the MSCI World Index. All returns are calculated using closing prices. The estimated parameters from OLS are [[??].sub.j], [[??].sub.j], and [mathematical expression not reproducible] by announcement i. Note in our context, the country level index is the individual announcement return, while the global market is used as the overall market return. Normally, a country's stock market index is used as the overall market return relative to a specific firm's stock price return.

Given the OLS estimates of the linear relationship and error by country and announcement, we estimate whether a stock market has experienced something abnormal by calculating the error in the linear prediction, or

[mathematical expression not reproducible] (2)

where [mathematical expression not reproducible] is the sample of abnormal returns for firm i on day t.

Note the sample of abnormal returns for country i stock market, or [mathematical expression not reproducible], is calculated relative to the global market in period t. The market term controls for systemic risk. The abnormal returns are normally distributed given standard assumptions with a mean zero and variance [[sigma].sup.2]. The normality assumption will be used and then relaxed when testing for an effect of host announcements on stock market indices.

Estimation of the Cumulative Abnormal return

Our objective is to test the effect of the IOC host announcement on a bidder's stock market. To increase the power of the test, we aggregate abnormal returns over several days and announcements. The cumulative abnormal returns by event are calculated as

[mathematical expression not reproducible] (3)

and across events as

[mathematical expression not reproducible] (4)

where N is the total number of announcements/events and [[[tau].sub.1], [[tau].sub.2]] is the event window. Under the standard assumptions as discussed in MacKinlay (1997) and elsewhere, the variance of the statistics in equations 3 and 4 are estimated as

[mathematical expression not reproducible] (5)

[mathematical expression not reproducible] (6)

respectively.

Again, as discussed in MacKinlay (1997) and elsewhere, the estimates can be normalized and assumed to be normal, or

[mathematical expression not reproducible] (7)

[mathematical expression not reproducible] (8)

under the hypothesis of a mean of zero. The hypothesis can be rejected with a specified level of confidence. These tests are referred to as parametric test statistics where normality and a constant variance in returns across time are assumed. These tests represent the standard event study approach to testing whether the announcements had an effect. If an effect had not occurred, then we would expect to fail to reject the null. However, if the null is rejected, then the alternative is that the event had a positive or negative impact on stock prices. As a result, we are able to speculate that the abnormal return was due to the event, i.e., the IOC announcement increased or decreased the future profitability of companies in those countries.

To summarize, we are testing whether there are abnormal returns over the event window. We calculate abnormal returns using parameters from an estimation window. If the abnormal returns are large, then we hypothesize it is due to the announcements. In calculating each abnormal return, we aggregate over several days around the event to ensure the impact of the information is not lost due to timing. Furthermore, we aggregate across events to improve the power of the test.

The above test for abnormal returns is a standard parametric approach, and is the approach followed in earlier work, including Berman, Brooks, and Davidson (2000), Veraros, Kasimati, and Dawson (2004), Dick and Wang (2010), and Mirman and Sharma (2010).

The potential issue with the standard approach described above is that it relies on several potentially incorrect assumptions. In particular, the data may not be normally distributed and the variance in the data may have changed between the estimation window and event window. As a result, we extend the prior literature by further analyzing our tests using two nonparametric tests. Specifically, we use a sign and sign rank test.

The sign test is used to determine whether more than half of the announcements result in a positive or negative abnormal return. In using the sign test, we are able to ignore the normality assumptions and variance estimate from the estimation window. We simply test whether the abnormal return is a "fair coin flip." To conduct the test, we count the number of countries with a positive cumulative abnormal return, calling it N+, and the total number of announcements N. Again, following MacKinlay (1997), the test statistic for the sign test is calculated as

[mathematical expression not reproducible] (9)

An issue with the sign test relates to the fact it does not account for the size of each effect. As a result, we further analyze the data using the Wilcoxon sign rank test. Specifically, we take the absolute values of all the cumulative abnormal returns, or [mathematical expression not reproducible], and rank them. Let [mathematical expression not reproducible] be the ranking and [mathematical expression not reproducible] be the rank if the cumulative abnormal return is greater than zero, and zero otherwise. Given the ranking, the test statistic is calculated as

[mathematical expression not reproducible] (10)

As before, the test statistic does not require an assumption about the normality of the abnormal returns or an estimate of the variance using the estimation window. Note, W ranges between [0, N(N + 1)/2]. Furthermore, if W is significantly large or small, then the test statistic rejects the null hypothesis that the average cumulative abnormal return, or CAR([[tau].sub.1], [[tau].sub.2]), is zero. For sufficiently large N, the test statistic becomes

[mathematical expression not reproducible] (11)

We only use the asymptotic results for the full sample of losers. Otherwise, the critical values must be determined through a counting process. We take the critical values of W for small N from Wilcoxon and Wilcox (1964). In other words, for a given type 1 error and N, which determine the low and high critical values [C.sub.L] and [C.sub.H], if W < [C.sub.L] or [C.sub.H] < W, we then evaluate the null that the average cumulative abnormal return is zero.

Data

To run the event study, we require two types of data. Specifically, the analysis requires the IOC announcement dates, as well as stock index data for each winning and losing country in the bidding process.

To acquire the dates of the announcements as well as the winners and losers, Gras-so, Mallon, and Heijmans (2015) provides information on an Olympics by Olympics basis. More recent winning bids can be found on the official Olympics.org website under each Olympic year's documentation. To ensure transparency, the complete lists of winners and losers, as well as the announcement dates, are provided in Tables 1 and 2.

The stock data of the winners and losers comes from country stock market data for the days -241 to +9, where the announcement date is t = 0. Data between -241 and -41 is used as the estimation window, and the data between -5 and 9 is used for the event window. To reiterate, the days are trading days. When available, the data was acquired from Yahoo Finance API. When unavailable on Yahoo Finance, the data was acquired through Thomson Reuters Datastream. Note several decisions had to be made regarding what index should be used to represent a country, as multiple indices exist. For transparency, the list of indices used in the analysis is provided in Tables 1 and 2. The MSCI World Index, the index used to calculate market returns, was collected from Yahoo Finance.

Results

We analyze the results along several dimensions--parametrically and non-parametrically, winning and losing bids, and varying event size windows. To allow for replication, and because the number of events is relatively small, we provide the cumulative abnormal [[epsilon].sub.i] returns and associated error estimates, or [mathematical expression not reproducible] and [mathematical expression not reproducible], for winners and losers in Tables 3 and 4.

By dividing the cumulative abnormal returns by [mathematical expression not reproducible], one can see the number of standard deviations away from the mean a particular stock market had during a particular event window. The Canadian stock market during its 9/30/1981 winning bid saw a 4.3 standard deviation abnormal return during the event window [[tau].sub.1] = -2 and [[tau].sub.2] = -1. The next highest was the USA's losing 10/17/1986 bid with a 3.76 standard deviation abnormal return during the event window [[tau].sub.1] = -5 and [[tau].sub.2] = 5. However, this later event, as well as several others with a z-statistic above 2.5, cannot be used to reject a null of no effect because of the large number of tests. In particular, one is not able to use these later individual cases to reject a null of no effect once a Bonferroni correction is made. However, the large shocks do suggest further analysis is necessary.

Rather than looking at abnormal returns by announcement, we analyze the aggregate effect. The results for the winners, first losers (those ranked in second place in the last round of the competition), and losers of the bidding process are in Tables 5-7. For the winning bids, the cumulative abnormal returns are generally positive post announcement. However, this is also true for the first losers and all losers, with the exception of the summer games. In other words, if the analysis is aggregated and one ignores the distinction between summer and winter games, then it seems countries are positively impacted no matter the outcome. Furthermore, the impacts are relatively large, at around 0.5% for post announcement windows. Note however that once the summer games are analyzed separately, the losers are generally negative and winners are generally positive. Although informative, this type of analysis ignores the statistical properties of the statistics.

Relying on the statistical properties of the estimated average effects, we fail to reject that the cumulative abnormal return is different from zero for nearly every cumulative return, as determined by the test statistic described in Equation 8. The top two average cumulative return test statistics are 1.76 and 2.06 for the winners of the summer games when using the interval [[tau].sub.1] = 0 and [[tau].sub.2] = 1, and the losers of the winter games when using the interval [[tau].sub.1] = 0 and [[tau].sub.2] = 1, respectively. However, the latter would be expected to be negative. Furthermore, they aren't particularly large. As in the individual test cases, any observed statistical significance across any of the tests must be weighted with the fact that so many tests are being run.

Relative to the closest study, Dick and Wang (2010) (DW), the estimates of the variation of the cumulative abnormal returns are very similar. For instance, we find the standard deviation of the cumulative abnormal return for the winning countries to be the same (up to three decimal places) for both the summer and winter games. Our estimates for the losers are slightly smaller, or 0.003 versus 0.004 in DW. What differentiates our results, i.e., we fail to reject a null of no effect, is our estimated impacts are significantly smaller. For instance, our estimated cumulative abnormal return for winning bids in the [0,1] event window is 0.0007 versus 0.011 in the case of DW. Therefore, our results are consistent in terms of the variation in the data. However, the estimated cumulative abnormal return statistics do not coincide. Although our results are markedly different, we note the key difference is we have additional data and have potentially used different country stock market indexes.

As our results run relatively contrary to prior evidence, in particular DW, we perform two additional non-parametric tests not in the previous literature. The results of our first non-parametric test, the sign test, are provided in Tables 8-10. For the winning bids, as in the CAR case, we find abnormal returns on average happened more than 50% of the time after the announcement date. However, only one provides sufficient evidence to support rejection at the 5% significance level following the statistic defined in equation 9. The only other test statistic above 2, and only slightly above, should arguably be the opposite sign, i.e., negative. In particular, the winter Olympic losers for the [0, 1] event window are similar to the parametric results. On aggregate, the results show little to arguably no evidence regarding the idea that the IOC announcement has a statistically significant impact on stock markets. This is especially true after the consideration of implementing a Bonferroni correction for the number of tests being run.

As our final non-parametric test, we run the Wilcoxon sign rank test described in equations 10 and 11. The results are provided in Tables 11-13. The only test statistic in the entire group to be outside of the critical values, based on a 5% type 1 error threshold, is the "all games" under the [0, 2] event window. The remaining results fail to reject the null of no cumulative effect.

Small Country Analysis

As a potential critique of the analysis, winning a bid to host the Olympic Games can have a large monetary benefit to the country, but the effect could be small relative to the size of the country's economy or stock market. As a result, our analysis might simply be trying to find a "needle in a haystack." Given this issue, we analyze the effect of the announcements conditional on the host country's size. In other words, we control for how important the bid can be relative to the country's stock market.

We take two approaches to answer the critique. In the first, we follow the approach of Dick and Wang (2010) by analyzing the correlation between a country's [bar.CAR](0,5) (and [[theta].sub.0](0, 5)) and its size using a simple linear regression, or

[y.sub.i] = [[beta].sub.0] + [[beta].sub.1] share of GDP + [[beta].sub.2] Summer Games + [[epsilon].sub.i], (12)

As in Dick and Wang (2010), we focus on the winners, and where the size variable is "taken as the percentage of the individual country GDP relative to the world GDP in the announcement year." The GDP data was taken from the United States Department of Agriculture ERS International Macroeconomic Data Set. The results are in Table 14. As before, we fail to find any meaningful impact of the bid on a country's stock market even when controlling for size.

The second approach bifurcates the dataset into a group of the large cities and a set of the small cities as defined by the size of the cities hosting the games, where a "win" by a large city within a country might be viewed as a "win" for the country as a whole, resulting in more extensive national stock market effects. Given this split, we re-run the parametric results. The results are provided in Table 15. The results fail to reject a null hypothesis of no announcement effect under any reasonable significance level with the exception of the [0,2] event window for the winners (as seen before).

Difference in Winners and Losers

As an additional robustness check of the results, we calculate the difference in the stock market returns between winners and losers using both the parametric and non-parametric approaches. In other words, we work to improve the power of the test under the assumption that winning the bid to host the Olympics would positively affect the future profits of that country's companies, while it would hurt countries who lose the bidding process.

Under standard assumptions, the difference in the cumulative abnormal returns, or the parametric approach, is estimated as

[mathematical expression not reproducible] (13)

where L = 1 if observation i was a losing bid, and zero otherwise. Its estimated variance is

[mathematical expression not reproducible] (14)

and the test statistic can be estimated as

[mathematical expression not reproducible] (15)

under the hypothesis of a mean of zero. In terms of the non-parametric test, we calculate the test statistic as

[mathematical expression not reproducible] (16)

where N is the number of winners and losers, [mathematical expression not reproducible] is the number of winners with a positive abnormal return in the event window, and [mathematical expression not reproducible] is the number of losers with a negative abnormal return in the event window.

The results of the difference between winners and losers are provided in Table 16 for the parametric test and Table 17 for the non-parametric test. Unlike before, none of the test results show a test statistic above 2. In other words, we fail to reject the hypothesis that the difference in the returns, which would be expected to be positive if a winning announcement was good and a losing announcement was bad, is zero.

Conclusion

Cities who host the Olympic Games spend substantial sums of public funds to prepare and host the event. Boosters and governments of bidding countries argue the economic benefits of hosting outweigh the costs. However, the argument for such large government outlays is often an ex-ante one put forth during the bidding process using macroeconomic multipliers or input-output models.

To analyze the actual benefits rather than predictions, researchers have used a variety of methods based on measured outcomes in the host city and country after the Olympics have taken place. We take one of these approaches by testing for the impact that hosting the games has on the profits of the firms in the host country. To test for an impact on firm profits, we follow the literature by using an "event study" approach, which measures stock prices right around the time of the IOC announcement of the winning bid, and if stock prices (expected profits) rise during that period, then researchers can conclude there are substantial economic profits for the country hosting the games.

The literature we refer to and follow includes Veraros, Kasimati, and Dawson (2004), Dick and Wang (2010), and Mirman and Sharma (2010). In these studies, the researchers have found some statistically significant evidence that announcements impact stock markets, and by extension, hosting the Olympic Games has an economic benefit.

In following the earlier work, we have failed to find any statistically significant impact of the announcements on stock prices. In other words, we find that an argument for a positive economic impact from the Games cannot be justified using an event study approach.

Given our findings run contrary to previous results, we have extended the event study approach along several dimensions. In particular, we have (i) used additional data, (ii) investigated a larger range of event windows (time around the announcement), (iii) tested for an effect using two additional non-parametric tests in addition to the standard parametric tests, and (iv) considered the difference in impact between winners and losers using both the parametric and non-parametric approaches. Furthermore, we have worked to make our findings transparent. In particular, we have provided the announcement dates, stock indexes, and individual cumulative abnormal returns used in determining our findings. As a result, researchers can analyze and replicate our work.

The debate regarding the economic benefits of the Olympic Games is important. As part of the debate, we hope our results provide a transparent and empirically driven analysis that informs the public on the benefits of hosting, and by extension, whether they should or should not host an Olympic Games.

References

Baade, R., and Matheson, V. (2016). "Going for the gold: The economics of the Olympics." The Journal of Economic Perspectives 30(2), 201-218.

Baumann, R., Engelhardt, B., and Matheson, V. (2012). "Employment effects of the 2002 Winter Olympics in Salt Lake City, Utah." Jahrbucher fur Nationalokonomie und Statistik 232(3), 308-317.

BBC. (2013, September 7). "Olympics 2020: Tokyo wins race to host Games." Retrieved from http://www.bbc.com/sport/ olympics/24002795

Berman, G., Brooks, R., and Davidson, S. (2000). "The Sydney Olympic Games announcement and Australian stock market reaction." Applied Economics Letters 7(12):781-784.

Centre for South Australian Economic Studies, and KPMG Peat Marwick. (1993). Sydney Olympics 2000: economic impact study. KPMG Peat Marwick.

Dick, C. D., and Wang, Q. (2010). "The economic impact of the Olympic Games: evidence from stock markets." Applied Economics Letters 17(9):861-864.

Flyvbjerg, B., Stewart, A., and Budzier, A. (2016, July 5). "The Oxford Olympics Study 2016: Cost and Cost Overrun at the Games." Said Business School WP 2016-20. Retrieved from SSRN http://dx.doi.org/10.2139/ssrn.2804554

Grasso, J., Mallon, B., and Heijmans, J. (2015). Historical dictionary of the Olympic movement. Lanham, MD: Rowman & Littlefield.

Kennedy, M. (2016, June 18). "Rio's governor declares 'state of calamity' ahead of Olympic Games." NPR.org. Retrieved from http://www.npr.org/sections/thet-wo-way/2016/06/18/482593048/rios-governor-declares-state-of-calamity-ahead-of-olym-pic-games

MacKinlay, A. C. (1997). "Event studies in economics and finance." Journal of Economic Literature, 35(1), 13-39.

Magnay, J. (2011, July 6). "Pyeongchang named as host city for 2018 Winter Olympics." The Telegraph. Retrieved from https://www.telegraph.co.uk/sport/olympics/8620658/Pyeongc-hang-named-as-host-city-for-2018-Winter-Olympics.html

Mirman, M., and Sharma, R. (2010). "Stock market reaction to Olympic Games announcement 1." Applied Economics Letters, 17(5), 463-466.

Papanikos, G. T. (1999). "Tourism impact of the 2004 Olympic Games." Tourism Research Institute, Athens, Greece.

Rauhala, E., and Birnbaum, M. (2015, July 13). "Beijing quietly celebrates Olympic win." The Washington Post. Retrieved from https://www.washingtonpost.com/world/ a-more-sub-dued-celebration-in-beijing-after-nabbing-2022-winter-olympics/ 2015/07/31/f89b0e2c-36f5-11e5-ab7b-6416d97c73c2_story.html

Veraros, N., Kasimati, E., and Dawson, P. (2004). "The 2004 Olympic Games announcement and its effect on the Athens and Milan stock exchanges." Applied Economics Letters 11(12), 749-753.

Wilcoxon, F., and Wilcox, R. A. (1964). Some rapid approximate statistical procedures. Lederle Laboratories.

Worstall, T. (2016). "Hosting Olympics bankrupts another place: Rio De Janeiro declares financial disaster." Forbes. Retrieved from https://www.forbes.com/sites/tim-worstall/2016/06/18/hosting-olympics-bankrupts-another-place-rio-de-janeiro-de-clares-financial-disaster/

Bryan Engelhardt, (1) Victor Matheson, (2) Alex Yen, (3) and Maxwell Chisolm (2)

(1) University of Wisconsin--Oshkosh

(2) College of the Holy Cross

(3) Stonehill College

Bryan Engelhardt is an associate professor of economics at the University of Wisconsin--Oshkosh. He earned his PhD in economics at the University of Iowa where he studied macroeconomics and labor. Dr. Engelhardt previously worked at the Federal Reserve Bank of Cleveland and the College of the Holy Cross.

Victor Matheson is a professor of economics at the College of the Holy Cross. He is president of the North American Association of Sports Economists and a coauthor of The Economics of Sports, 6th ed. He has worked as professional referee in top soccer leagues in the US.

Alex Yen is an associate professor of accounting in the Leo J. Meehan School of Business at Stonehill College. He earned his PhD at the University of Texas at Austin, and previously worked at Price Waterhouse LLP, Suffolk University, and the College of the Holy Cross.

Maxwell Chisolm is a sales and trading analyst at UBS. He earned his BA in economics at the College of the Holy Cross where he was a research assistant in the Department of Economics & Accounting Summer Research Program.

Table 1. Data Description: Winning Bids

Olympics     Announcement  Country Name  Stock Index

Winter 1988  09/30/1981    Canada        S&P/TSX Composite
Summer 1988  09/30/1981    South Korea   KOSPI Composite
Winter 1992  10/17/1986    France        CAC General Index
Summer 1992  10/17/1986    Spain         Madrid SE General (IGBM)
Winter 1994  09/15/1988    Norway        OSEBX.OL
Summer 1996  09/18/1990    USA           S&P 500
Winter 1998  06/15/1991    Japan         Nikkei 225
Summer 2000  09/23/1993    Australia     All Ordinaries
Winter 2002  06/16/1995    USA           S&P 500
Summer 2004  09/05/1997    Greece        Athens Index Composite
Winter 2006  06/19/1999    Italy         FTSE MIB Index
Summer 2008  07/13/2001    China         SSE Composite
Winter 2010  07/02/2003    Canada        S&P/TSX Composite
Summer 2012  07/06/2005    UK            FTSE 100
Winter 2014  07/04/2007    Russia        RSF EE MT (RUR) INDEX
Summer 2016  10/02/2009    Brazil        IBOVESPA
Winter 2018  07/06/2011    South Korea   KOSPI Composite
Summer 2020  09/07/2013    Japan         Nikkei 225
Winter 2022  07/31/2015    China         SSE Composite

Olympics     Source of Index

Winter 1988  Yahoo Finance
Summer 1988  Thomson Reuters Datastream
Winter 1992  Thomson Reuters Datastream
Summer 1992  Thomson Reuters Datastream
Winter 1994  Yahoo Finance
Summer 1996  Yahoo Finance
Winter 1998  Yahoo Finance
Summer 2000  Yahoo Finance
Winter 2002  Yahoo Finance
Summer 2004  Yahoo Finance
Winter 2006  Yahoo Finance
Summer 2008  Yahoo Finance
Winter 2010  Yahoo Finance
Summer 2012  Yahoo Finance
Winter 2014  Thomson Reuters Datastream
Summer 2016  Yahoo Finance
Winter 2018  Yahoo Finance
Summer 2020  Yahoo Finance
Winter 2022  Yahoo Finance

Table 2. Data Description: Losing Bids

Olympics     Announcement  Country Name  Stock Index

Winter 1988  09/30/1981    Sweden        AFFARSVARLDEN GENERAL INDEX
Winter 1988  09/30/1981    Italy         ITALY-DS Market
Summer 1988  09/30/1981    Japan         NIKKEI 225 STOCK AVERAGE
Winter 1992  10/17/1986    Sweden        AFFARSVARLDEN GENERAL INDEX
Winter 1992  10/17/1986    Norway        OSEBX.OL
Winter 1992  10/17/1986    Italy         ITALY-DS Market
Winter 1992  10/17/1986    USA           S&P 500
Winter 1992  10/17/1986    Germany       DAX 30 PERFORMANCE
Summer 1992  10/17/1986    France        CAC General
Summer 1992  10/17/1986    Australia     All Ordinaries
Summer 1992  10/17/1986    UK            FTSE 100
Summer 1992  10/17/1986    Netherlands   AEX INDEX (AEX) DS-CALC
Winter 1994  09/15/1988    Sweden        AFFARSVARLDEN GENERAL INDEX
Winter 1994  09/15/1988    USA           S&P 500
Summer 1996  09/18/1990    Greece        Athens Composite Index
Summer 1996  09/18/1990    Canada        S&P/TSX Composite
Summer 1996  09/18/1990    Australia     All Ordinaries
Summer 1996  09/18/1990    UK            FTSE 100
Winter 1998  06/15/1991    USA           S&P 500
Winter 1998  06/15/1991    Sweden        AFFARSVARLDEN GENERAL INDEX
Winter 1998  06/15/1991    Spain         IBEX 35
Winter 1998  06/15/1991    Italy         ITALY-DS Market
Summer 2000  09/23/1993    China         SSE Composite
Summer 2000  09/23/1993    UK            FTSE 100
Summer 2000  09/23/1993    Germany       Dax
Summer 2000  09/23/1993    Turkey        BIST NATIONAL 100
Winter 2002  06/16/1995    Switzerland   Swiss Market Index
Winter 2002  06/16/1995    Sweden        AFFARSVARLDEN GENERAL INDEX
Winter 2002  06/16/1995    Canada        S&P/TSX Composite
Summer 2004  09/05/1997    Italy         ITALY-DS Market
Summer 2004  09/05/1997    South Africa  FTSE/JSE ALL SHARE
Summer 2004  09/05/1997    Sweden        AFFARSVARLDEN GENERAL INDEX
Summer 2004  09/05/1997    Argentina     ARGENTINA MERVAL
Winter 2006  06/19/1999    Switzerland   Swiss Market Index
Winter 2006  06/19/1999    Finland       OMX HELSINKI (OMXH)
Winter 2006  06/19/1999    Austria       ATX
Winter 2006  06/19/1999    Slovakia      SLOVAKIA SAX 16
Winter 2006  06/19/1999    Poland        WARSAW GENERAL INDEX 20
Summer 2008  07/13/2001    Canada        S&P/TSX Composite
Summer 2008  07/13/2001    France        CAC 40
Summer 2008  07/13/2001    Turkey        Borsa Istanbul 100
Summer 2008  07/13/2001    Japan         NIKKEI 225
Winter 2010  07/02/2003    South Korea   KOSPI
Winter 2010  07/02/2003    Austria       ATX
Summer 2012  07/06/2005    France        CAC 40
Summer 2012  07/06/2005    Spain         IBEX 35
Summer 2012  07/06/2005    USA           S&P 500
Summer 2012  07/06/2005    Russia        RSF EE MT (RUR) INDEX
Winter 2014  07/04/2007    South Korea   KOSPI
Winter 2014  07/04/2007    Austria       ATX
Summer 2016  10/02/2009    Spain         IBEX 35
Summer 2016  10/02/2009    Japan         NIKKEI 225
Summer 2016  10/02/2009    USA           S&P 500
Winter 2018  07/06/2011    Germany       Dax
Winter 2018  07/06/2011    France        CAC 40
Summer 2020  09/07/2013    Turkey        Borsa Istanbul 100
Summer 2020  09/07/2013    Spain         IBEX 35

Olympics     Source of Index

Winter 1988  Thomson Reuters Datastream
Winter 1988  Thomson Reuters Datastream
Summer 1988  Thomson Reuters Datastream
Winter 1992  Thomson Reuters Datastream
Winter 1992  Yahoo Finance
Winter 1992  Thomson Reuters Datastream
Winter 1992  Yahoo Finance
Winter 1992  Thomson Reuters Datastream
Summer 1992  Thomson Reuters Datastream
Summer 1992  Yahoo Finance
Summer 1992  Yahoo Finance
Summer 1992  Thomson Reuters Datastream
Winter 1994  Thomson Reuters Datastream
Winter 1994  Yahoo Finance
Summer 1996  Yahoo Finance
Summer 1996  Yahoo Finance
Summer 1996  Yahoo Finance
Summer 1996  Yahoo Finance
Winter 1998  Yahoo Finance
Winter 1998  Thomson Reuters Datastream
Winter 1998  Yahoo Finance
Winter 1998  Thomson Reuters Datastream
Summer 2000  Yahoo Finance
Summer 2000  Yahoo Finance
Summer 2000  Yahoo Finance
Summer 2000  Thomson Reuters Datastream
Winter 2002  Thomson Reuters Datastream
Winter 2002  Thomson Reuters Datastream
Winter 2002  Yahoo Finance
Summer 2004  Thomson Reuters Datastream
Summer 2004  Thomson Reuters Datastream
Summer 2004  Thomson Reuters Datastream
Summer 2004  Thomson Reuters Datastream
Winter 2006  Yahoo Finance
Winter 2006  Thomson Reuters Datastream
Winter 2006  Yahoo Finance
Winter 2006  Thomson Reuters Datastream
Winter 2006  Thomson Reuters Datastream
Summer 2008  Yahoo Finance
Summer 2008  Yahoo Finance
Summer 2008  Yahoo Finance
Summer 2008  Yahoo Finance
Winter 2010  Yahoo Finance
Winter 2010  Yahoo Finance
Summer 2012  Yahoo Finance
Summer 2012  Yahoo Finance
Summer 2012  Yahoo Finance
Summer 2012  Thomson Reuters Datastream
Winter 2014  Yahoo Finance
Winter 2014  Yahoo Finance
Summer 2016  Yahoo Finance
Summer 2016  Yahoo Finance
Summer 2016  Yahoo Finance
Winter 2018  Yahoo Finance
Winter 2018  Yahoo Finance
Summer 2020  Yahoo Finance
Summer 2020  Yahoo Finance

Table 3. Cumulative Abnormal Returns by Country: Winning Bids

                                           [mathematical expression
                                           not reproducible]
                                           [[tau].sub.1] = 0,
Country      Announcement  [mathematical   [[tau].sub.2] = 1
                           expression not
                           reproducible]

Canada       09/30/1981    0.0064           0.0035
South Korea  09/30/1981    0.0135          -0.0238
France       10/17/1986    0.0134          -0.0326
Spain        10/17/1986    0.0135          -0.0178
Norway       09/15/1988    0.0203           0.0264
USA          09/18/1990    0.0081           0.0088
Japan        06/15/1991    0.0138           0.0026
Australia    09/23/1993    0.0074           0.0118
USA          06/16/1995    0.0048           0.0102
Greece       09/05/1997    0.0176           0.0742
Italy        06/19/1999    0.0154          -0.0023
China        07/13/2001    0.0089          -0.0096
Canada       07/02/2003    0.0064          -0.0054
UK           07/06/2005    0.0046          -0.0033
Russia       07/04/2007    0.0135           0.0118
Brazil       10/02/2009    0.018            0.0317
South Korea  07/06/2011    0.0081           0.0054
Japan        09/07/2013    0.0161           0.0236
China        07/31/2015    0.0168          -0.0306

             [mathematical expression not reproducible]
             [[tau].sub.1] = 0,  [[tau].sub.1] = 0,  [[tau].sub.1] = 0,
Country      [[tau].sub.2] = 2   [[tau].sub.2] = 5   [[tau].sub.2] = 9

Canada        0.0108             -0.0066             -0.0142
South Korea   0.0073              0.0522             -0.0137
France       -0.0332              0.0064             -0.0083
Spain        -0.026              -0.0518             -0.12
Norway        0.0321              0.0405              0.0371
USA           0.0025              0.0041              0.0354
Japan        -0.0104             -0.0114             -0.0154
Australia     0.0079              0.0158              0.0365
USA           0.0093              0.01                0.008
Greece        0.0861              0.0614              0.0663
Italy         0.0072             -0.0022             -0.0319
China        -0.0123              0.0049             -0.0369
Canada       -0.0038              0.0049              0.0057
UK            0.0051             -0.0002             -0.0106
Russia        0.0122              0.0198              0.0322
Brazil        0.0126              0.0132              0.0315
South Korea   0.0076             -0.0106             -0.0117
Japan         0.0187              0.0049              0.0213
China         0.0019             -0.0123              0.027

             [mathematical expression not reproducible]
             [[tau].sub.1] = -2,  [[tau].sub.1] = -5,
Country      [[tau].sub.2] = 2    [[tau].sub.2] = 5

Canada        0.0498              -1.0135
South Korea  -0.0141              -0.0072
France       -0.0648              -0.03
Spain        -0.0327              -0.0676
Norway        0.0431               0.0542
USA           0.0048              -0.001
Japan         0.0029               0.0001
Australia     0.0058               0.0237
USA           0.0085               0.0182
Greece        0.0756               0.0428
Italy         0.0056              -0.014
China        -0.0232              -0.0044
Canada       -0.0036               0.0057
UK            0.0082               0.019
Russia        0.0167               0.0162
Brazil        0.0219               0.0362
South Korea   0.0224               0.0172
Japan        -0.0022               0.0153
China         0.0046              -0.1387

             [mathematical expression not reproducible]
             [[tau].sub.1] = -5,  [[tau].sub.1] = -2,
Country      [[tau].sub.2] = -1   [[tau].sub.2] = -1

Canada       -0.0069               0.0390
South Korea  -0.0594              -0.0214
France       -0.0364              -0.0317
Spain        -0.0158              -0.0068
Norway        0.0137               0.0111
USA          -0.0052               0.0023
Japan         0.0114               0.0133
Australia     0.0078              -0.0021
USA           0.0081              -0.0009
Greece       -0.0186              -0.0105
Italy        -0.0118              -0.0016
China        -0.0093              -0.0109
Canada        0.0008               0.0002
UK            0.0191               0.0031
Russia       -0.0036               0.0045
Brazil        0.0231               0.0093
South Korea   0.0278               0.0149
Japan         0.0104              -0.0209
China        -0.1264               0.0028

Table 4: Cumulative Abnormal Returns by Country: Losing Bids

                                            [mathematical expression
                                            not reproducible]
                                            [[tau].sub.1] = 0,
Country Name  Announcement  [mathematical   [[tau].sub.2] = 1
                            expression not
                            reproducible]

Sweden        09/30/1981    0.0086           0.0131
Italy         09/30/1981    0.0282           0.0149
Japan         09/30/1981    0.0056          -0.0001
Sweden        10/17/1986    0.0116          -0.0033
Norway        10/17/1986    0.0097           0.0112
Italy         10/17/1986    0.0199           0.0107
USA           10/17/1986    0.0055           0.0032
Germany       10/17/1986    0.0141          -0.014
France        10/17/1986    0.0134          -0.0326
Australia     10/17/1986    0.0085           0.0047
UK            10/17/1986    0.0079          -0.0006
Netherlands   10/17/1986    0.0091           0.0009
Sweden        09/15/1988    0.0148          -0.0049
USA           09/15/1988    0.0172           0.0009
Greece        09/18/1990    0.0244          -0.0576
Canada        09/18/1990    0.0056           0.0138
Australia     09/18/1990    0.0095          -0.0072
UK            09/18/1990    0.008           -0.0041
USA           06/15/1991    0.0087          -0.0027
Sweden        06/15/1991    0.0122           0.0236
Spain         06/15/1991    0.0132          -0.0051
Italy         06/15/1991    0.0115           0.0239
China         09/23/1993    0.0502          -0.0092
UK            09/23/1993    0.0061          -0.002
Germany       09/23/1993    0.0079          -0.0066
Turkey        09/23/1993    0.0246           0.0286
Switzerland   06/16/1995    0.0077          -0.0015
Sweden        06/16/1995    0.0079           0.0109
Canada        06/16/1995    0.0047           0.0058
Italy         09/05/1997    0.0101           0.004
South Africa  09/05/1997    0.0056           0.0056
Sweden        09/05/1997    0.0079          -0.0005
Argentina     09/05/1997    0.0106          -0.0057
Switzerland   06/19/1999    0.0137           0.0008
Finland       06/19/1999    0.0187           0.0374
Austria       06/19/1999    0.0148           0.0223
Slovakia      06/19/1999    0.0161           0.0072
Poland        06/19/1999    0.0264           0.0109
Canada        07/13/2001    0.0117          -0.0016
France        07/13/2001    0.0101           0.0143
Turkey        07/13/2001    0.0444          -0.0038
Japan         07/13/2001    0.0158          -0.0029
South Korea   07/02/2003    0.0203           0.0169
Austria       07/02/2003    0.0089           0.0061
France        07/06/2005    0.0061          -0.0033
Spain         07/06/2005    0.0058          -0.0136
USA           07/06/2005    0.0037           0.0011
Russia        07/06/2005    0.0141           0.0269
South Korea   07/04/2007    0.0079           0.0229
Austria       07/04/2007    0.0079           0.0056
Spain         10/02/2009    0.0133           0.0051
Japan         10/02/2009    0.0278          -0.0309
USA           10/02/2009    0.0115           0.0137
Germany       07/06/2011    0.0054          -0.0005
France        07/06/2011    0.0054          -0.0041
Turkey        09/07/2013    0.0159           0.0528
Spain         09/07/2013    0.0094          -0.0039

              [mathematical expression not reproducible]
              [[tau].sub.1] = 0,  [[tau].sub.1] = 0,
Country Name  [[tau].sub.2] = 2   [[tau].sub.2] = 2



Sweden         0.0086              0.0274
Italy          0.0013             -0.0407
Japan         -0.0025              0.0128
Sweden        -0.0153             -0.0063
Norway         0.0103              0.0083
Italy          0.0105             -0.0057
USA            0.012               0.0401
Germany       -0.0034              0.0165
France        -0.0332              0.0064
Australia      0.0124              0.0031
UK             0.0045              0.0042
Netherlands    0.005               0.014
Sweden         0.0053              0.0345
USA           -0.0057              0.0056
Greece        -0.0605             -0.1483
Canada         0.0151              0.0095
Australia     -0.0026             -0.0295
UK            -0.0202             -0.0199
USA           -0.0058             -0.0156
Sweden         0.0347              0.0373
Spain         -0.0132             -0.0056
Italy          0.0189             -0.0198
China         -0.0097             -0.0004
UK             0.0029              0.0059
Germany        0.0048              0.0032
Turkey         0.0507              0.0153
Switzerland   -0.0005              0.0066
Sweden         0.0138              0.0188
Canada         0.0025             -0.0046
Italy         -0.0032             -0.0181
South Africa  -0.0015             -0.025
Sweden         0.0045             -0.0152
Argentina     -0.011              -0.0179
Switzerland   -0.003              -0.017
Finland        0.0278              0.0111
Austria        0.0205              0.0065
Slovakia       0.0134             -0.0143
Poland         0.0082              0.036
Canada        -0.0065              0.0091
France         0.0051             -0.0172
Turkey        -0.0485              0.057
Japan         -0.0199             -0.0585
South Korea    0.0281              0.0462
Austria        0.0103              0.008
France         0.0078              0.0082
Spain         -0.0054             -0.0003
USA            0.0015             -0.0029
Russia         0.0358              0.0366
South Korea    0.0265              0.0443
Austria        0.008              -0.0091
Spain          0.0103             -0.01
Japan         -0.0434             -0.0237
USA            0.0058              0.0052
Germany       -0.0031             -0.0048
France        -0.0123             -0.0218
Turkey         0.0494              0.0783
Spain         -0.0002              0.0073

              [mathematical expression not reproducible]
              [[tau].sub.1] = 0,  [[tau].sub.1] = -2
Country Name  [[tau].sub.2] = 9   [[tau].sub.2] = 2

Sweden         0.0466             -0.0218
Italy         -0.0996             -0.0208
Japan          0.0147             -0.006
Sweden         0.0059             -0.028
Norway         0.0061             -0.0003
Italy         -0.0381              0.0204
USA            0.0383              0.0345
Germany       -0.0048             -0.0112
France        -0.0083             -0.0648
Australia      0.0027              0.0134
UK             0.0161              0.0146
Netherlands    0.0045              0.0105
Sweden         0.0205              0.0057
USA           -0.0021              0.0009
Greece        -0.2289             -0.0049
Canada         0.0094              0.0192
Australia     -0.026              -0.0096
UK             0.0031             -0.032
USA           -0.0111              0.002
Sweden         0.0476              0.0239
Spain         -0.0049             -0.0486
Italy          0.0018              0.0108
China         -0.0252              0.0091
UK             0.0201              0.0073
Germany        0.0314             -0.0052
Turkey         0.0493              0.0566
Switzerland    0.0161             -0.0021
Sweden         0.0291              0.0019
Canada        -0.0002              0.0063
Italy          0.0153             -0.0105
South Africa  -0.0274              0.0063
Sweden        -0.0004             -0.0157
Argentina     -0.0359             -0.0147
Switzerland   -0.0132              0.0001
Finland        0.0252              0.0267
Austria        0.0328              0.0209
Slovakia      -0.026               0.0175
Poland         0.0008              0.024
Canada         0.0082             -0.0077
France        -0.0162             -0.0084
Turkey         0.0803             -0.0953
Japan         -0.0435             -0.0152
South Korea    0.0634              0.027
Austria        0.0189              0.0087
France         0.025               0.001
Spain          0.0077             -0.0108
USA            0.0007              0.0098
Russia         0.0201              0.0415
South Korea    0.0553              0.0508
Austria       -0.0276             -0.0008
Spain         -0.0212             -0.0008
Japan         -0.009              -0.0428
USA            0.0045              0.0004
Germany       -0.0126             -0.002
France        -0.0423             -0.0201
Turkey         0.1178              0.0504
Spain          0.0159              0.0155

              [mathematical expression not reproducible]
              [[tau].sub.1] = -5,  [[tau].sub.1] = -5,
Country Name  [[tau].sub.2] = 5    [[tau].sub.2] = -1

Sweden        -0.0066              -0.034
Italy          0.0055               0.0462
Japan         -0.0003              -0.013
Sweden        -0.0404              -0.0341
Norway         0.0142               0.006
Italy          0.012                0.0176
USA            0.0684               0.0283
Germany        0.0049              -0.0116
France        -0.03                -0.0364
Australia      0.0116               0.0084
UK             0.0078               0.0036
Netherlands   -0.003               -0.0171
Sweden         0.0458               0.0113
USA            0.0123               0.0068
Greece        -0.1545              -0.0063
Canada         0.0098               0.0003
Australia     -0.0368              -0.0073
UK            -0.04                -0.0201
USA           -0.0083               0.0072
Sweden         0.0373               0.0000
Spain         -0.0333              -0.0276
Italy         -0.0123               0.0075
China          0.0221               0.0225
UK             0.0182               0.0123
Germany        0.0198               0.0166
Turkey         0.0183               0.003
Switzerland    0.0052              -0.0014
Sweden         0.01                -0.0088
Canada         0.005                0.0096
Italy          0.0071               0.0252
South Africa  -0.0323              -0.0073
Sweden        -0.0196              -0.0044
Argentina     -0.0758              -0.0579
Switzerland   -0.0148               0.0022
Finland        0.0098              -0.0014
Austria        0.0047              -0.0018
Slovakia      -0.0452              -0.0309
Poland         0.0402               0.0041
Canada         0.002               -0.007
France        -0.0416              -0.0244
Turkey        -0.1416              -0.1986
Japan         -0.0673              -0.0088
South Korea    0.0699               0.0237
Austria        0.0124               0.0044
France         0.0228               0.0146
Spain          0.0116               0.0119
USA            0.0047               0.0075
Russia         0.0663               0.0297
South Korea    0.0579               0.0135
Austria       -0.0156              -0.0066
Spain         -0.0101              -0.0001
Japan         -0.07                -0.0464
USA            0.0078               0.0026
Germany       -0.005               -0.0002
France        -0.0299              -0.008
Turkey         0.0763              -0.002
Spain          0.0292               0.0219

              [mathematical expression not reproducible]
              [[tau].sub.1] = -2
Country Name  [[tau].sub.2] = -1

Sweden        -0.0304
Italy         -0.0221
Japan         -0.0035
Sweden        -0.0127
Norway        -0.0106
Italy          0.01
USA            0.0225
Germany       -0.0078
France        -0.0317
Australia      0.001
UK             0.0101
Netherlands    0.0055
Sweden         0.0004
USA            0.0066
Greece         0.0556
Canada         0.004
Australia     -0.007
UK            -0.0118
USA            0.0078
Sweden        -0.0109
Spain         -0.0354
Italy         -0.0081
China          0.0188
UK             0.0044
Germany       -0.0101
Turkey         0.0059
Switzerland   -0.0015
Sweden        -0.0119
Canada         0.0038
Italy         -0.0073
South Africa   0.0077
Sweden        -0.0202
Argentina     -0.0037
Switzerland    0.0031
Finland       -0.0011
Austria        0.0004
Slovakia       0.0041
Poland         0.0158
Canada        -0.0012
France        -0.0135
Turkey        -0.0468
Japan          0.0047
South Korea   -0.0012
Austria       -0.0016
France        -0.0068
Spain         -0.0054
USA            0.0082
Russia         0.0057
South Korea    0.0243
Austria       -0.0088
Spain         -0.0111
Japan          0.0006
USA           -0.0054
Germany        0.001
France        -0.0078
Turkey         0.001
Spain          0.0157

Table 5. Average Cumulative Abnormal Returns: Winning Bids

Event Window
([[[tau].sub.1],
[[tau].sub.2]])        [0,1]    [0,2]   [0,5]   [0,9]   [-2,2]  [-5,5]

All Games
  [bar.CAR]             0.0045  0.0071  0.0075  0.0020  0.0068  -0.0015
  St.dev. ([bar.CAR])   0.0042  0.0051  0.0072  0.0093  0.0066   0.0098
  Test statistic        1.0698  1.3988  1.044   0.2178  1.0337  -0.1507
Summer Games
  [bar.CAR]             0.0106  0.0113  0.0116  0.0011  0.0049   0.0063
  St.dev. ([bar.CAR])   0.0060  0.0074  0.0105  0.0135  0.0096   0.0142
  Test statistic        1.7576  1.5309  1.1101  0.0813  0.5135   0.4444
Winter Games
  [bar.CAR]            -0.0011  0.0034  0.0039  0.0029  0.0085  -0.0085
  St.dev. ([bar.CAR])   0.0058  0.0070  0.0100  0.0129  0.0091   0.0135
  Test statistic       -0.1889  0.4779  0.3874  0.2228  0.9369  -0.6272

Event Window
([[[tau].sub.1],
[[tau].sub.2]])        [-5,-1]  [-2,-1]

All Games
  [bar.CAR]            -0.0090  -0.0003
  St.dev. ([bar.CAR])   0.0066   0.0042
  Test statistic       -1.3671  -0.0788
Summer Games
  [bar.CAR]            -0.0053  -0.0064
  St.dev. ([bar.CAR])   0.0096   0.0060
  Test statistic       -0.5569  -1.0630
Winter Games
  [bar.CAR]            -0.0123   0.0052
  St.dev. ([bar.CAR])   0.0091   0.0058
  Test statistic       -1.3546   0.8960

Table 6. Average Cumulative Abnormal Returns: First Losing Bids

Event Window
([[tau].sub.1],
[[[tau].sub.2]])       [0,1]    [0,2]    [0,5]    [0,9]    [-2,2]

All Games
  [bar.CAR]             0.0007   0.0012   0.0044   0.0051   0.0015
  St.dev. ([bar.CAR])   0.0057   0.0069   0.0098   0.0127   0.009
  Test statistic        0.1204   0.1758   0.4438   0.4008   0.1693
Summer Games
  [bar.CAR]            -0.0047  -0.0053  -0.0069  -0.0114  -0.0038
  St.dev. ([bar.CAR])   0.01     0.0122   0.0173   0.0223   0.0158
  Test statistic       -0.4746  -0.4362  -0.3983  -0.5106  -0.2412
Winter Games
  [bar.CAR]             0.0061   0.0078   0.0156   0.0216   0.0068
  St.dev. ([bar.CAR])   0.0054   0.0066   0.0094   0.0121   0.0085
  Test statistic        1.1299   1.1753   1.6684   1.7856   0.8012

Event Window
([[tau].sub.1],
[[[tau].sub.2]])       [-5,5]   [-5,-1]  [-2,-1]

All Games
  [bar.CAR]             0.005    0.0006   0.0003
  St.dev. ([bar.CAR])   0.0133   0.009    0.0057
  Test statistic        0.3732   0.0673   0.0523
Summer Games
  [bar.CAR]            -0.0072  -0.0003   0.0015
  St.dev. ([bar.CAR])   0.0234   0.0158   0.0100
  Test statistic       -0.3061  -0.0178   0.1529
Winter Games
  [bar.CAR]             0.0171   0.0015  -0.0009
  St.dev. ([bar.CAR])   0.0127   0.0085   0.0054
  Test statistic        1.3497   0.1743  -0.1726

Table 7. Average Cumulative Abnormal Returns: Losing Bids

Event Window
([[tau].sub.1],
[[tau].sub.2])         [0,1]    [0,2]    [0,5]    [0,9]   [-2,2]

All Games
  [bar.CAR]             0.0035   0.0026   0.0013  0.0023   0.0007
  St.dev. (CAR)         0.0030   0.0036   0.0051  0.0066   0.0047
  Test statistic        1.1690   0.7085   0.2449  0.3472   0.1407
Summer Games
  [bar.CAR]            -0.0005  -0.0017  -0.0037  0.0002  -0.003
  St.dev. (CAR)         0.0044   0.0054   0.0077  0.0099   0.007
  Test statistic       -0.1125  -0.3196  -0.478   0.0168  -0.4212
Winter Games
  [bar.CAR]             0.0079   0.0074   0.0067  0.0047   0.0047
  St.dev. ([bar.CAR])   0.0038   0.0047   0.0066  0.0085   0.006
  Test statistic        2.0600   1.5743   1.0206  0.5466   0.7765

Event Window
([[tau].sub.1],
[[tau].sub.2])         [-5,5]   [-5,-1]  [-2,-1]

All Games
  [bar.CAR]            -0.0032  -0.0045  -0.0019
  St.dev. (CAR)         0.0069   0.0047   0.0030
  Test statistic       -0.4647  -0.9576  -0.6453
Summer Games
  [bar.CAR]            -0.0129  -0.0092  -0.0012
  St.dev. (CAR)         0.0104   0.007    0.0044
  Test statistic       -1.2379  -1.3126  -0.2747
Winter Games
  [bar.CAR]             0.0075   0.0008  -0.0027
  St.dev. ([bar.CAR])   0.0089   0.006    0.0038
  Test statistic        0.844    0.1339  -0.7004

Table 8. Non-Parametric Sign Test: Winning Bids

Event Window ([[tau].sub.1] [[[tau].sub.2]])   [0,1]  [0,2]  [0,5]

All Games
  Positive Abnormal Return ([N.sup.+]/N)       0.58   0.74   0.63
  Test statistic ([[theta].sub.1])             0.69   2.06   1.15
Summer Games
  Positive Abnormal Return ([N.sup.+]/N)       0.56   0.78   0.78
  Test statistic ([[theta].sub.1])             0.33   1.67   1.67
Winter Games
  Positive Abnormal Return ([N.sup.+]/N)       0.6    0.7    0.5
  Test statistic ([[theta].sub.1])             0.63   1.26   0.00

Event Window ([[tau].sub.1] [[[tau].sub.2]])   [0,9]  [-2,2]  [-5,5]

All Games
  Positive Abnormal Return ([N.sup.+]/N)       0.53   0.68    0.58
  Test statistic ([[theta].sub.1])             0.23   1.61    0.69
Summer Games
  Positive Abnormal Return ([N.sup.+]/N)       0.56   0.56    0.56
  Test statistic ([[theta].sub.1])             0.33   0.33    0.33
Winter Games
  Positive Abnormal Return ([N.sup.+]/N)       0.5    0.8     0.6
  Test statistic ([[theta].sub.1])             0.00   1.90    0.63

Event Window ([[tau].sub.1] [[[tau].sub.2]])   [-5,-1]  [-2,-1]

All Games
  Positive Abnormal Return ([N.sup.+]/N)        0.47     0.53
  Test statistic ([[theta].sub.1])             -0.23     0.23
Summer Games
  Positive Abnormal Return ([N.sup.+]/N)        0.44     0.33
  Test statistic ([[theta].sub.1])             -0.33    -1.00
Winter Games
  Positive Abnormal Return ([N.sup.+]/N)        0.5      0.7
  Test statistic ([[theta].sub.1])              0.00     1.26

Table 9. Non-Parametric Sign Test: First Losing Bids

Event Window ([[[tau].sub.1] [[tau].sub.2]])  [0,1]  [0,2]  [0,5]

All Games
  Positive Abnormal Return ([N.sup.+]/N)       0.44   0.44  0.61
  Test statistic ([[theta].sub.1])            -0.47  -0.47  0.94
Summer Games
  Positive Abnormal Return ([N.sup.+]/N)       0.33   0.33  0.56
  Test statistic ([[theta].sub.1])            -1.00  -1.00  0.33
Winter Games
  Positive Abnormal Return ([N.sup.+]/N)       0.56   0.56  0.67
  Test statistic ([[theta].sub.1])             0.33   0.33  1.00

Event Window ([[[tau].sub.1] [[tau].sub.2]])  [0,9]  [-2,2]  [-5,5]

All Games
  Positive Abnormal Return ([N.sup.+]/N)      0.61    0.50   0.56
  Test statistic ([[theta].sub.1])            0.94    0.00   0.47
Summer Games
  Positive Abnormal Return ([N.sup.+]/N)      0.56    0.33   0.56
  Test statistic ([[theta].sub.1])            0.33   -1.00   0.33
Winter Games
  Positive Abnormal Return ([N.sup.+]/N)      0.67    0.67   0.56
  Test statistic ([[theta].sub.1])            1.00    1.00   0.33

Event Window ([[[tau].sub.1] [[tau].sub.2]])  [-5,-1]  [-2,-1]

All Games
  Positive Abnormal Return ([N.sup.+]/N)       0.44     0.44
  Test statistic ([[theta].sub.1])            -0.47    -0.47
Summer Games
  Positive Abnormal Return ([N.sup.+]/N)       0.33     0.33
  Test statistic ([[theta].sub.1])            -1.00    -1.00
Winter Games
  Positive Abnormal Return ([N.sup.+]/N)       0.56     0.56
  Test statistic ([[theta].sub.1])             0.33     0.33

Table 10. Non-Parametric Sign Test: Losing Bids

Event Window ([[tau].sub.1] [[[tau].sub.2]])   [0,1]  [0,2]  [0,5]

All Games
  Positive Abnormal Return ([N.sup.+]/N)        0.54  0.58   0.54
  Test statistic ([[theta].sub.1])              0.66  1.19   0.66
Summer Games
  Positive Abnormal Return ([N.sup.+]/N)        0.4   0.5    0.53
  Test statistic ([[theta].sub.1])             -1.1   0      0.37
Winter Games
  Positive Abnormal Return ([N.sup.+]/N)        0.7   0.67   0.56
  Test statistic ([[theta].sub.1])              2.12  1.73   0.58

Event Window ([[tau].sub.1] [[[tau].sub.2]])   [0,9]  [-2,2]  [-5,5]

All Games
  Positive Abnormal Return ([N.sup.+]/N)       0.6     0.54   0.58
  Test statistic ([[theta].sub.1])             1.46    0.66   1.19
Summer Games
  Positive Abnormal Return ([N.sup.+]/N)       0.63    0.47   0.53
  Test statistic ([[theta].sub.1])             1.46   -0.37   0.37
Winter Games
  Positive Abnormal Return ([N.sup.+]/N)       0.56    0.63   0.63
  Test statistic ([[theta].sub.1])             0.58    1.35   1.35

Event Window ([[tau].sub.1] [[[tau].sub.2]])   [-5,-1]  [-2,-1]

All Games
  Positive Abnormal Return ([N.sup.+]/N)        0.49     0.47
  Test statistic ([[theta].sub.1])             -0.13    -0.4
Summer Games
  Positive Abnormal Return ([N.sup.+]/N)        0.47     0.5
  Test statistic ([[theta].sub.1])             -0.37     0
Winter Games
  Positive Abnormal Return ([N.sup.+]/N)        0.52     0.44
  Test statistic ([[theta].sub.1])              0.19    -0.58

Table 11. Wilcoxian Sign Rank Test: Winning Bids

Event Window ([[[tau].sub.1], [[tau].sub.2]])  [0,1]  [0,2]  [0,5]

All Games
  W-statistic ([[theta].sub.2])                107    150    111
  Critical Values
Summer Games
  W-statistic ([[theta].sub.2])                 21     38     33
  Critical Values ([C.sub.L], [C.sub.H])
Winter Games
  W-statistic ([[theta].sub.2])                 35     40     26
  Critical Values ([C.sub.L], [C.sub.H])

Event Window ([[[tau].sub.1], [[tau].sub.2]])  [0,9]  [-2,2]    [-5,5]

All Games
  W-statistic ([[theta].sub.2])                90         131    104
  Critical Values                                     (46,144)
Summer Games
  W-statistic ([[theta].sub.2])                24         26      20
  Critical Values ([C.sub.L], [C.sub.H])               (6,39)
Winter Games
  W-statistic ([[theta].sub.2])                26         45      36
  Critical Values ([C.sub.L], [C.sub.H])               (8,47)

Event Window ([[[tau].sub.1], [[tau].sub.2]])  [-5,-1]  [-2,-1]

All Games
  W-statistic ([[theta].sub.2])                95       102
  Critical Values
Summer Games
  W-statistic ([[theta].sub.2])                19        20
  Critical Values ([C.sub.L], [C.sub.H])
Winter Games
  W-statistic ([[theta].sub.2])                30        36
  Critical Values ([C.sub.L], [C.sub.H])

Critical values are determined by a two-sided test with a type 1 error
threshold of 5%. The critical values are from Wilcoxon and Wilcox
(1964) with N equal to 19, 9, and 10 for the All, Summer, and Winter
games, respectively.

Table 12. Wilcoxian Sign Rank Test: First Losing Bids

Event Window ([[[tau].sub.1], [[tau].sub.2]])  [0,1]  [0,2]  [0,5]

All Games
  W-statistic ([[theta].sub.2])                60     56     96
  Critical Values ([C.sub.L], [C.sub.H])
Summer Games
  W-statistic ([[theta].sub.2])                13     12     27
  Critical Values ([C.sub.L], [C.sub.H])
Winter Games
  W-statistic ([[theta].sub.2])                18     16     23
  Critical Values ([C.sub.L], [C.sub.H])

Event Window ([[[tau].sub.1], [[tau].sub.2]])  [0,9]  [-2,2]    [-5,5]

All Games
  W-statistic ([[theta].sub.2])                92          89   85
  Critical Values ([C.sub.L], [C.sub.H])              (40,131)
Summer Games
  W-statistic ([[theta].sub.2])                28         14    26
  Critical Values ([C.sub.L], [C.sub.H])               (6,39)
Winter Games
  W-statistic ([[theta].sub.2])                21         31    19
  Critical Values ([C.sub.L], [C.sub.H])               (6,39)

Event Window ([[[tau].sub.1], [[tau].sub.2]])  [-5,-1]  [-2,-1]

All Games
  W-statistic ([[theta].sub.2])                59       81
  Critical Values ([C.sub.L], [C.sub.H])
Summer Games
  W-statistic ([[theta].sub.2])                 9       13
  Critical Values ([C.sub.L], [C.sub.H])
Winter Games
  W-statistic ([[theta].sub.2])                22       28
  Critical Values ([C.sub.L], [C.sub.H])

Critical values are determined by a two-sided test with a type 1 error
threshold of 5%. The critical values are from Wilcoxon and Wilcox
(1964) with N equal to 18, 9, and 9 for the All, Summer, and Winter
games, respectively.

Table 13. Wilcoxian Sign Rank Test: Losing Bids

Event Window ([[[tau].sub.1], [[tau].sub.2]])  [0,1]  [0,2]  [0,5]

All Games
  W-statistic ([[theta].sub.2])                738    898    897
  Critical Values (CL, CH)
Summer Games
  W-statistic ([[theta].sub.2])                157    235    280
  Critical Values (CL, CH)
Winter Games
  W-statistic ([[theta].sub.2])                218    215    176
  Critical Values (CL, CH)

Event Window ([[[tau].sub.1], [[tau].sub.2]])  [0,9]             [-2,2]

All Games
  W-statistic ([[theta].sub.2])                 995              892
  Critical Values (CL, CH)                     (579.81,1073.19)
Summer Games
  W-statistic ([[theta].sub.2])                 326              218
  Critical Values (CL, CH)                     (137,328)
Winter Games
  W-statistic ([[theta].sub.2])                 190              232
  Critical Values (CL, CH)                     (107,271)

Event Window ([[[tau].sub.1], [[tau].sub.2]])  [-5,5]  [-5,-1]  [-2,-1]

All Games
  W-statistic ([[theta].sub.2])                1064    813      885
  Critical Values (CL, CH)
Summer Games
  W-statistic ([[theta].sub.2])                 300    228      258
  Critical Values (CL, CH)
Winter Games
  W-statistic ([[theta].sub.2])                 244    183      190
  Critical Values (CL, CH)

Critical values are determined by a two-sided test with a type 1 error
threshold of 5%. The critical values for the Summer and Winter games
are from Wilcoxon and Wilcox (1964) with N equal to 30 and 27,
respectively. The critical values for the All games case is determined
by equation 11 with N = 57.

Table 14. Regression of Impact vs. Size

              [bar.CAR](0,5)  [bar.CAR](0,5)  [[theta].sub.0](0, 5)

Share of GDP  -0.065          -0.06           -0.223
              (0.066)         (0.064)         (2.147)
Summer Games  -               -0.007          -
              -               (0.012)         -
[R.sup.2]      0.035           0.055           0.000

              [[theta].sub.0](0, 5)

Share of GDP  -0.062
              (2.186)
Summer Games  -0.237
              (0.344)
[R.sup.2]      0.028

Note: The regression run [y.sub.i] = [[beta].sub.0] + [[beta].sub.1]
share of GDP + [[beta].sub.2] Summer Games + [[epsilon].sub.1] where
[mathematical expression not reproducible] for a bid of a particular
host country in a particular year is represented by the subscript "i,"
[[beta].sub.0], [[beta].sub.1], and [[beta].sub.2] are estimated using
OLS, and the Summer Games is included in a subset of years following
Dick and Wang (2010).

Table 16. Difference in Cumulative Abnormal Returns

Event Window
([[[tau].sub.1],
[[tau].sub.2]])          [0,1]    [0,2]    [0,5]    [0,9]    [-2,2]

All Games
  [bar.xDCAR]            -0.0015  -0.0001   0.0009  -0.0012   0.0012
  St.dev. ([bar.xDCAR])   0.0025   0.0030   0.0042   0.0055   0.0039
  Test statistic         -0.6033  -0.0466   0.2222  -0.2216   0.3121
Summer Games
  [bar.xDCAR]             0.0028   0.0040   0.0055   0.0001   0.0034
  St.dev. ([bar.xDCAR])   0.0037   0.0045   0.0064   0.0083   0.0058
  Test statistic          0.7673   0.8735   0.8615   0.0151   0.5839
Winter Games
  [bar.xDCAR]            -0.0060  -0.0045  -0.0039  -0.0026  -0.0011
  St.dev. ([bar.xDCAR])   0.0032   0.0039   0.0055   0.0071   0.0050
  Test statistic         -1.8903  -1.1410  -0.7018  -0.3684  -0.2206

Event Window
([[[tau].sub.1],
[[tau].sub.2]])          [-5,5]   [-5,-1]  [-2,-1]

All Games
  [bar.xDCAR]             0.0020   0.0011   0.0013
  St.dev. ([bar.xDCAR])   0.0057   0.0039   0.0025
  Test statistic          0.3566   0.2855   0.5506
Summer Games
  [bar.xDCAR]             0.0114   0.0059  -0.0005
  St.dev. ([bar.xDCAR])   0.0087   0.0058   0.0037
  Test statistic          1.3141   1.0055  -0.1467
Winter Games
  [bar.xDCAR]            -0.0078  -0.0039   0.0033
  St.dev. ([bar.xDCAR])   0.0075   0.0050   0.0032
  Test statistic         -1.0427  -0.7778   1.0486

Table 17. Difference in Returns with Non-Parametric Sign Test

Event Window
([[tau].sub.1],
[[tau].sub.2])                      [0,1]  [0,2]  [0,5]  [0,9]  [-2,2]

All Games
  Expected Abnormal return
  [mathematical expression
  not reproducible]                  0.49   0.50   0.50   0.43   0.51
  Test statistic ([[theta].sub.4])  -0.23   0.00   0.00  -1.15   0.23
Summer Games
  Expected Abnormal return
  [mathematical expression
  not reproducible]                  0.59   0.56   0.54   0.41   0.54
  Test statistic ([[theta].sub.4])   1.12   0.8    0.48  -1.12   0.48
Winter Games
  Expected Abnormal return
  [mathematical expression
  not reproducible]                  0.38   0.43   0.46   0.46   0.49
  Test statistic ([[theta].sub.4])  -1.48  -0.82  -0.49  -0.49  -0.16

Event Window
([[tau].sub.1],
[[tau].sub.2])                      [-5,5]  [-5,-1]  [-2,-1]

All Games
  Expected Abnormal return
  [mathematical expression
  not reproducible]                  0.46    0.50     0.53
  Test statistic ([[theta].sub.4])  -0.69    0.00     0.46
Summer Games
  Expected Abnormal return
  [mathematical expression
  not reproducible]                  0.49    0.51     0.46
  Test statistic ([[theta].sub.4])  -0.16    0.16    -0.48
Winter Games
  Expected Abnormal return
  [mathematical expression
  not reproducible]                  0.43    0.49     0.59
  Test statistic ([[theta].sub.4])  -0.82   -0.16     1.15