«Edwin D. Maberly Professor of Finance University of Canterbury Raylene Pierce Senior Lecturer School of Business Christchurch College of Education ...»
The Purchasing of Naming Rights for Sports Stadiums: A Harbinger of Bad
Corporate Governance or Just Bad Timing?
Edwin D. Maberly
Professor of Finance
University of Canterbury
School of Business
Christchurch College of Education
Department of Economics
University of Canterbury
The Purchasing of Naming Rights for Sports Stadiums: A Harbinger of Bad
Corporate Governance or Just Bad Timing?
The literature on corporate governance and the market’s delayed reaction to news events proliferated over the last two decades. This paper examines return patterns surrounding the event date for firms purchasing naming rights for North American sports stadiums. One argument appearing in the financial press is that such acquisitions are a harbinger of widespread corporate mismanagement and hubris at the highest levels of corporate governance. Purchases of stadium naming rights provide sidebenefits to executives such as “being in the limelight” and the use of supplementary corporate boxes.
Thus, management has a strong incentive to undertake such investments even if their decision is not value enhancing to shareholders. The extent to which these agreements are associated with negative risk-adjusted returns is an empirical question, which this study addresses. On average, negative riskadjusted returns are observed over the three years following the event date, and these results are significant at standard levels of significance. The efficient market hypothesis suggests that these results are not due to a cause and effect relationship but represent data snooping or just bad timing.
The Purchasing of Naming Rights for Sports Stadiums: A Harbinger of Bad Corporate Governance or Just Bad Timing?
Call it the stadium-naming jinx. As soon as you put your name on something, you might as well go to the lawyers and draw up the bankruptcy petition [MSNBC, “100 Million Strikes and You’re Out,” December 4, 2001] Recent news associated with the collapse of several high-profile United States (U.S.) firms including WorldCom, Enron and Adelphia question the effectiveness of corporate governance, and more so after various senior executives of these firms were served with federal felony changes. A recent article entitled “Distress Call” appearing in the August 19, 2002 issue of Barron’s states that “the failure of highfliers like WorldCom, Enron and Adelphia has struck fear in the hearts of many investors, raising worries about the stability of corporate Americaas a whole.” Besides being in financial distress, WorldCom, Enron and Adelphia have the distinction of having recently purchased the naming rights to sports stadiums. Articles appearing in MSNBC and Barron’s suggest that a relationship exists between the purchase of sports stadium naming rights and subsequent corporate distress. The implication is that the purchase of stadium naming rights is associated ex post with weak corporate governance or other value destroying managerial activities. With the advantage of hindsight, the purchase of stadium naming rights was a signal for shareholders to unload their shares or to engage in appropriate risk reducing strategies—buy puts or sell in-the-money call option on the underlying firms.
This paper examines the signalling effect of firm’s purchasing naming rights for North American sports stadiums covering four sports—football, baseball, hockey and basketball. With the exceptions of Wrigley (Chicago Cubs) in 1926, Busch (St.
Louise Cardinals) in 1967 and Rich Foods (Buffalo Bills) in 1973, all other naming rights were purchased after 1990, which is this study’s starting point. Daily return data for the forty-two corporations that purchased naming rights after 1990 are collected for the four years surrounding the event date.
Several companies including America West Airlines, XCEL Energy and WorldCom are excluded from the data modelling section due to data problems associated with mergers. In particular, pre-event data associated with mergers involving these companies is suspect as to its accuracy. However, many of these firms experience major price declines in the post-event period and inclusion actually strengthens this studies empirical findings. It can generally be argued that a purchase of naming rights should produce a non- negative effect, since there are numerous benefits of such transaction, e.g., a broader name recognition, an improved reputation of corporate responsibleness, etc. The results of this paper show the opposite. There are negative results observed for 3 years and beyond past the contract date. It is argued that such effects are due to value-diminishing corporate governance for the shareholders, with the stadium deals serving as a credible signal. Mana gers know more about the prospects of the firm than outsiders, and need to send signals to the investment community in order to raise the corporation’s stock price, thereby raising the value of their own shareholdings and enhancing their managerial reputations (Bhattacharya (1979)).1 Generally such signalling is performed through increased dividend and capital expenditure announcements, as well as stock splits. It would appear that naming rights should achieve the same objective. After all, the company is committing itself to pay large sums of money on an annual basis for 5-20 years, which should serve as a signal of corporate strength. However there is a caveat involved. The companies’ executives are enjoying side-benefits such as supplementary corporate boxes brought by such transactions, which provide an incentive for them to enter into these agreements, even if they are not in the best interests of shareholders. Extending the concept, such transactions are projected as only the tip of the iceberg of weak corporate governance, with other, less visible but still value-destroying projects that managers undertake to enhance their interests at the expense of shareholders (e.g., Enron, WorldCom and Adelphia are recent examples).
An exception to the rule discussed above is when management is also substantial owners of shares as in the case of Microsoft.
The second notion pursued here is in regard to underreaction. More recently, the process by which new information is incorporated into stock prices has come under scrutiny. A wide body of research has argued that markets tend to initially underreact to a set of specific news announcements. The most common area studied is stock splits. Ikenberry and Ramnath (2002) have reported a drift of 9% in a year following the split announcement; examining a shorter time window, Ikenberry, Rankine and Stice (1996) have reported excess returns of 3.38% in a 2-day announcement period around the split. Using a similar methodology, this study documents the existence of statistically significant negative abnormal returns over the post-announcement for the companies purchasing naming rights. This result has two possible explanations. The first hypothesis is of underreaction. The phenomenon of negative abnormal returns continuing for a significant period post-event indicates that investors appear to be underreacting to the news of stadium sponsorship. Such underreaction would be consistent with new theoretical articles such as the model by Daniel, Hirshleifer and Subrahmanyam (1998), which analyses how analysts might overweight their own priors when valuing firms and thus underweight new information such as split announcements. For pedagogical purposes the second hypothesis is referred to as the Marginal Negative Effect (MNE) hypothesis. Under the MNE hypothesis, stadium name purchase is only one symptom of a larger problem of corporate mismanagement, and it takes several years for the mismanagement to grow sufficiently large to be significantly reflected in stock prices.
However, according to either of the hypotheses, rational investors will discount the information associated naming rights and sell the stock, which should be reflected in immediately downward adjustment in prices. However, the empirical findings of this study do not support market efficiency as a delayed reaction is observed.
It should be noted that many studies find negative post-event drifts in selfselected corporate events. Ikenberry and Ramnath (2002) provide an extensive review of such literature and cite studies by Agrawal, Jaffe and Mandelker (1992), Loughran and Vijh (1997), and Rau and Vermaelen (1998). These papers document negative long-horizon abnormal returns for merger deals. Other studies cited by Ikenberry and Ramnath include Michaely, Thaler and Womack (1995) who examine omission of dividends, and Dharan and Ikenberry (1995) who examine firms moving from one trading market to another. More recent findings by Lee and Loughran (1998), and Spiess and Affleck-Graves (1999) are concerned with timing of the issuance of straight and convertible bonds.
Finally, this study is not trying to claim a hard “proof” of bad governance or of market underreaction. Researchers tend to look for interesting results, hence often performing some form of data- mining giving rise to spurious output, and hence committing Type I errors or rejecting the null hypothesis when the null is true (see Merton (1985)). Also, there is no robust asset pricing model that could be used.
Researchers are both beneficiaries and victims of the academic progress, using ad hoc models that may be having statistical power, yet no theoretical justification.
1. Dataset and Methods
1.1 Dataset Corporate stadium-sponsorship is a relatively new phenomenon, and this study’s observations occur over the 1990 to 2001 period. As mentioned previously, three stadium-sponsorships were signed prior to 1990, as well as three others signed over the 1990 to 2001 period, but with data-collecting difficulties. Thus this paper covers the population of stadium- naming events with exception of six instances. Return data plus other firm specific data is obtained from Data Stream, Value Line Investment Survey,Yahoo-dot-com, and the annual reports or other reports filed with the United States Securities and Exchange Commission. Thus the data collection process is extensive especially the task of identifying the list of companies purchasing stadiumsponsorship, the timing of the deals, and the corresponding stock prices. Data is collected entirely from primary sources, as no ready data sets exist. Hence the time and effort associated with data collection is extensive. Entries are recorded weekly, one year prior to the event month, and three years posterior to the event month. 2 The approach for choosing this timeframe is consistent with standard methodology used in event studies (for example, see MacKinlay, (1997) and Ikenberry, Rankine and Stice, (1996)).
1.2 Methodology First a matching benchmark index is selected to reflect each stock’s risk characteris tics. The two indices selected are the S&P 500 and Nasdaq100 indices.
Next a $1,000 zero-cost portfolio is created consisting of the stock and the matching index one year prior to the event. This portfolio is carried without rebalancing to the event month, thus obtaining pre-event cumulative abnormal returns (ARs). In the event month, the zero-cost portfolio is rebalanced and held for three years, and postevent cumulative ARs are calculated based on this data. The entire set of stadium sponsorships are outlined in appendices I to IV. Several firms do not have a complete set of return observations, which include one-year prior and three years after the event date. In some cases, the initial public offering (listing) of shares on an exchange is within one year of the event date. For other cases, the event occurs after June 1999, and three years of post-event returns are unavailable. In either case only the available observations are used. Once the zero-cost portfolios are obtained, an array of tests is performed on the data set, which includes non-parametric, time series, and analysis based on Wiener processes. One of the firms in the population, Nasdaq listed Qualcomm (QCOM), is considered an outlier due to QCOM’s inclusion in the S&P 500 index after it purchased the stadium naming rights, which created tremendous buying pressure on the stock. Therefore QCOM is classified as an outlier.3
1.3 Analysis: Non-parametric approach It is well known that financial data is non-normally distributed, and hence standard parametric methods, which assume normality, produces biased results. Thus a non-parametric approach is more desirable. ARs are estimated for five subperiods surrounding the event date: (1) one- year pre-event; (2) six- months post-event; (3) one-year post-event; (4) two-years post-event; (5) three-years post-event. In analysing the groups several different tests are employed. For example the Wilcoxon Signed-Rank test is used on each of the groups to determine whether they are significantly different from zero; Wilcoxon Matched-Pairs Signed-Ranks test examines whether there is a significant difference between the pre-event returns and post-event return groups; sign tests examine randomness of positive/negative returns within each group; Kolmogorov-Smirnov Two-Sample examines whether there is a significant difference between pre-event and three- years post-event returns. All of the tests are performed with and without QCOM in order to examine QCOM’s impact as an outlier.
1.4 Wiener Process A standard procedure is to model abnormal stock returns using a generalised
where the coefficient “a” indicates the drift rate, and “b” indicates volatility. This process is also often called a Brownian motion with drift. It is a Gaussian Process with expectation and covariance functions,
Generally, a Wiener process has the following necessary condition (their union
constitutes a sufficient condition):
A process B = ( Bt, t ∈ [ 0, ∞)) is a Wiener process if:
ii) It has stationary, independent increments. While cumulative abnormal returns are not stationary, their differences (abnormal returns) are.
iv) The process has continuous paths over time that is “no jumps.” Since preand post-returns are analyzed separately, there is no rebalancing within Wiener process. Hence the path satisfies the continuity requirement.