Вы находитесь на странице: 1из 7

CFA Institute

Does Delisting from the S&P 500 Affect Stock Price? Author(s): William N. Goetzmann and Mark Garry Source: Financial Analysts Journal, Vol. 42, No. 2 (Mar. - Apr., 1986), pp. 64-69 Published by: CFA Institute Stable URL: http://www.jstor.org/stable/4478919 . Accessed: 14/09/2011 02:16
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

CFA Institute is collaborating with JSTOR to digitize, preserve and extend access to Financial Analysts Journal.

http://www.jstor.org

Tchnical

Note

Does Delisting from the S&P 500 Affect Stock Price?


by William N. Goetzmann, Graduate Student, Yale School of Management, and Mark Garry, Associate, McKinsey & Co., Inc.

The Evidence
The break-up of AT&T, and the concurrent public offering of securities of the spin-off companies, was expected for about a year prior to the event. The exact nature of the change in the S&P 500, however, was kept confidential until the close of the New York Stock Exchange on November 30, 1983. At that time, subscribers to S&P's telephone notification service were contacted simultaneously and told of the changes in the index. Although some trading on the Pacific Stock Exchange was possible, most of the trading in these stocks occurred on December 1. The following companies were dropped from the S&P 500: * Continental Telecommunications (CTC), * Centel Communications (CNT), * Wisconsin Electric Power Company (WPC), * El Paso Company (ELP), * Brooklyn Union Gas Company (BU),

On November 30, 1983, seven stocks were dropped from the Standard & Poor's 500 index to make room for the stocks of the seven new telephone companies. This event offers an unusual opportunity to study the effects of delisting, independent of changes in investor expectations of future performance. Efficient market theory suggests that the market prices of the delisted securities should not have changed, because the delisting would have no effect on their expected future returns. However, the evidence indicates an apparently significant and long-term drop in the prices of the delisted stocks, suggesting a possible measurable negative price effect is associated with delisting from the S&P 500.

Figure A

Daily Prices of the Delisted Stocks

44 42 40 38 36 -

3
t

34 32 30 28 26 24 IP

22 i 20 10/3/83

10/21/83

11/10/83 Date

12/1/83

12/21/83

FINANCIAL ANALYSTS JOURNAL / MARCH-APRIL 1986 E 64

Figure B

Average Daily Percentage Returns of Delisted Stocks

-2 10/3/83 10/21/83 11/10/83 Date 12/1/83 12/21/83

* United Telecommunications (UT), * New England Electrical Systems (NES) and * AT&T (old stock). The old AT&T was replaced by the new AT&T; the remaining seven stocks were replaced by the new telephone stocks. Standard & Poor's used no rigorous rules to select the stocks for delisting, although communication and utility stocks were clearly chosen in order to maintain the balance of the index. Outsiders could guess, but not know with certainty, which stocks would be replaced.' We obtained the daily volume figures and closing stock prices for the seven delisted companies and examined them for significant changes around the time of the event. We omitted El Paso from our analysis, because it was acquired by Burlington Northern on December 14, 1983, an event reported previously and expected by the market. Figure A graphs the stock prices of the six remaining companies from October 3 through December 30, 1983. The prices exhibit no apparent radical movement on December 1, 1983. However, the average daily percentage return of these six delisted stocks, seen in Figure B, did exhibit common movement on that day. On December 1, the average price of the sample set fell nearly 2 per cent-the largest percent1. Footnotes appear at end of article.

age drop in the 63-day study. Figure C shows that two and one-half million shares of the six sample stocks traded on that day-a volume three times larger than the next largest volume day in the period examined. NYSE volume on December 1 was not otherwise unusual, nor was the average market return. Using the S&P 500 index as a surrogate for general price movements, we calculated residual (abnormal) returns for the six delisted stocks during the period to correct for market price movements. We estimated sample betas from 16 months of daily returns from a period extending a year before the event to four months following. The average }beta for the delisted sample was 0.381. Figure D graphs average abnormal returns for the six-stock sample over the period.- After correcting for general market movement, we found a large negative return of 1.9 per cent coincident with the extremely high volume on December 1.3 Figure E shows a scatterplot of the sample daily returns over the same time period for which the sample beta was estimated. The line drawn through these plots reflects the sample returns that would be expected, given a market return on a specific day and the sample beta of 0.381. The December 1 return lies significantly distant from its expected value. The price behavior of the delisted stocks was statis65

FINANCIAL ANALYSTS JOURNAL / MARCH-APRIL 1986 0

Figure C

Daily Volume of Delisted Stocks

28 26 2422208 18
-

16
14 12 20 8 6 _

> 4

a~

4 2 0

~ ~~~~~~~~~~~~Dt

10/3/83

10/21/83

11/10/83 Date

12/1/83

12/21/83

Figure D

Abnormal Returns of Delisted Stocks

2.5

2.0-

1.5-

1.0
0

0.5
0

0.

-1.0

Figure E

Sample vs. S&P 500 Daily Returns (beta

0.381)

30

U~~

-3 -3 -2 -1 0 S&P 500 Returns (per cent) 1 2

tically significant on the trading day following the change in the composition of the S&P 500. But was the price drop due merely to temporary price pressure? If so, we would expect the price to rise after the initial drop. Figure F, charting the cumulative abnormal returns of the sample, shows a steady downward trend following the delisting date. This suggests that the drop was permanent. Because the six delisted stocks were all utilities, however, they may have been subject to common price movements beyond the delisting effects. To test this possibility, we also charted the cumulative abnormal returns of the S&P 40 utilities. Figure F shows that the sample moved in unison with the utility index during the four months following the event, but while the S&P utilities remained above or around the market return, the delisted sample trended lower. That the delisted stocks were all utilities does not account for their lower price after delisting.

Implications
The observed delisting effect becomes more interesting when viewed in the context of a broader range of research that indicates that the market values smaller, lesser known or unlisted securities different-

ly from larger, well known securities. Chief among these observed anomalies is the "small firm effect" the tendency for the stock of smaller firm to outperform that of larger firms.4 Explanations of the phenomenon vary widely. Some argue that smaller firms are priced lower because of an associated risk premium that is not captured by existing pricing theory. Others claim that a year-end tax effect exists. Another explanation is that the higher returns represent a premium for the relatively limited amount of information available about smaller firms.5 Our analysis does not conflict with the theory that the market pays a premium for information availability. The immediate drop in the price of the delisted stocks adds another dimension to the question, however. When the six stocks were dropped, there was no immediate "loss" of information. We might expect that, as months went by, analysts would focus less attention on these securities, hence less reliable predictions would be available about their future performance. Intuition thus suggests that the loss in value would follow the decrease in information, rather than precede it. The evidence suggests that prices dropped in anticipation of future decreases in available information. In other words, the delisting was a clear signal to the

FINANCIAL ANALYSTS JOURNAL / MARCH-APRIL 1986 U 67

Figure F

Cumulative Abnormal Returns

5 4 S&P Utilities 3

2-2
5,-3
-4

Delisted Stocks

-7
-8 -9 - 10 11/30/83 12/29/83 1/27/84 Date 2/27/84 3/26/84

market regarding the amount and quality of future analysis and reportage. The "label"or S&P "seal of approval"seems to carrywith it broadly understood implications. Removal of that seal has an adverse effect. A recent study of "popular"versus "generic"securitiesuses a measureof institutionalinterestin securities and also the standard deviation of analysts' earnings projections as a proxy for the degree of research a security attracts.6Not surprisingly, firms "neglected" by analysts outperformed those more thoroughlystudied. The analogy to the product mar-

kets-where brand names carry clear "price premiums"-seems appropriateto our findings. Our findings suggest that listing and delisting represent a particulartype of information-in marketing terms, a "productattribute"-that is unrelated to other characteristics.Delisting from the S&P 500 apparentlycarriesclear, unambiguous signals to the market about such seemingly uncertain issues as future standard deviation of analysts' earnings predictions as well as the future quality of information about a company.7

Footnotes
1. If investors knew with certainty,then an arbitrage the positive effects of securities remaining on the opportunity might have existed. To test this hyS&P 500 index. pothesis we formed a portfolio of the most likely 2. We calculated the abnormal returns by multiplycandidates for delisting, given informationavailing the sample beta by the percentage return on able before the event. Besides the two telecommuthe S&P 500 and subtracting the product from the nicationsstocks and El Paso, we included the eight sample's percentage return. lowest capitalized companies in the S&P utility 3. Even including the extreme abnormal returns in index. The adjusted abnormalreturnsto this portearly October resulting from our use of closing folio, held over the event day, were not signifiprices, the standard deviation of the sample set cantly different from zero, suggesting that the abnormal returns beginning a year before the negative effects of delisting were compensated by event, and extending over the first quarter of 1984
FINANCIAL ANALYSTS JOURNAL / MARCH-APRIL 1986 O 68

is 0.6 per cent, indicating that the -1.9 per cent 5. See C. B. Barry and S. J. Brown, "Differential return on 12/1/83 is statistically significant. The Information and the Small Firm Effect," Journalof Financial Economics13 (1984), pp. 283-99. Another probabilityof such a large negative return is less study by D. S. Daliwal, "Exchange Listing Effects than 5 per cent (more than 3.07 standard deviations from the mean), placing a high level of on a Firm's Cost of Equity Capital," Journal of certaintyon the hypothesis that the drop in price Business Research11 (1983), pp. 139-151, examined was caused by externalfactors. Correctionfor the the effects of listing on major exchanges versus market using measures of beta that exclude the over-the-counter. His conclusions were consistant event period yields similar results. with the Barry and Brown research. 4. See R. W. Banz, "The Relationship Between Re- 6. A. Arbel, "Generic Stocks: An Old Product in a turn and MarketValue of Common Stocks,"JourNew Package," Journal of Portfolio Management,
nal of Financial Economics9 (1981), pp. 3-18, as well Winter 1985, pp. 4-13.

as R. Roll, "A Possible Explanationfor the Small 7. The authors wish to thank Dr. Stephen Ross of Firm Effect,"Journal Finance (1981), pp. 879of 36 Yale University for his suggestion of this topic for 887, and M. R. Reinganum, "The Anomalous research, Dr. Martin Shubik for his guidance and Stock MarketBehaviorof Small Firms in January: criticism and Dr. Stephen Brown for his technical EmpiricalTests For Year End Tax Effects,"Journal assistance and insights into the research problem.
of Financial Economics12 (1983), pp. 89-104.

The Case For Raising the Capital Gains Tax (undoing the damage from venture capital)
proportion to the ability of our citizens to pay the tax. Would it really be equitable for individuals whose principal financial well-being comes from earning In "The Case for Ending the Capital Gains Tax" long-term capital gains to live off these gains and not (Financial Analysts Journal, May/June1985)BruceBart- pay any taxes? lett argues for elimination of the capital gains tax on Bartlett's economic justification for eliminating capgrounds of both equity and economics. A closer ital gains seems to rest on the propositions that examination of the issues raises considerable doubt "capital needs to be mobile to be effective" and that about the equity of a zero capital gains tax and even "the more efficient the system of intermediation, the more about the economic wisdom of its removal. more benefit accrues directly to savers and capital Indeed, it would seem more efficacious for national users and indirectly to workers . . . and consumers." policy to raise the capital gains rate. By implication, he advocates an economic system Bartlettimplies that earners of capital gains suffer where capital is 100 per cent mobile. But before relativeto earnersof income. But considertwo typical reducing capital gains taxes further, let's look at the investors. The first purchases the stock of a non- damage that capital gains rates at the current 20 per dividend-paying, rapidly growing company and cent has already caused as of mid-1984 (the time of holds it for a number of years, selling out just before writing). the company starts paying a dividend. The second To use the figures cited in the article, venture investor purchases the stock of a mature company capital funds have increased from $39 million in 1977 that annually pays out all its earnings in dividends. to $11.5 billion in 1983. This boom in financial comWhereas the first investor is subject to a capital mitments for new ventures has meant not just a gains tax only upon sale, the second investor must growth but an explosion in start-ups of companies pay a tax concurrent with the income received. Be- making disc drives, personal and specialized computcause the increase in capital, as Bartlettindicates, is ers, work stations and terminals, and software prodnot distinct from the increase in income, the first ucts. The numbers of companies in each category investor is in a favored position; he is not taxed until range up to and over 100. Entrepreneurialism hasn't sale on the dividends that can only be imputed to the just grown, it has run rampant. years the company chose to plough backall earnings. The after-effects of this boom, evident for a year If putting the capital gains earner and the income now, are hardly surprising: Earnings for most of earneron an equal after-taxbasis is a goal of equity, these companies come public in the last few years as Bartlett suggests, it follows that capital gains have declined, with many recording losses and in should be taxed at a higher rate than earned income. prospect of bankruptcy. Of more concern for public The traditionof equity underlyingUnited Statestax policy, investors who have bought the venturepolicy is evident in our system of rising marginaltax backed IPOs have probably lost billions of dollars. (In rates;our society believes that the burden of support this estimate, I include not only the losses on the for the federal budget should be carried in rough initial shares underwritten at the time of the IPO but
FINANCIAL ANALYSTS JOURNAL / MARCH-APRIL 1986 O 69

by Alan J. Dworsky,Principal, Mt. AuburnManagement,Boston

Вам также может понравиться