Impact of Stock Split on Stock Return

Proceedings of ASBBS Volume 16 Number 1 THE IMPACT OF STOCK SPLIT ANNOUNCEMENTS ON STOCK PRICE: A TEST OF MARKET EFFICIENCY Garcia de Andoain, Carlos Longwood University carlos. [email  protected] longwood. edu Bacon, Frank W. Longwood University 2O1 High Street Farmville, VA 23909 [email  protected] edu Phone: 434-395-2131 Fax: 434-395-2203 ABSTRACT The purpose of this study is to test whether the investor can make an above normal return by relying on public information impounded in a stock split announcement. Using risk adjusted event study methodology, this study tests â€Å"how† and â€Å"when† public announcements of forward and reverse stock splits affect stock price. Stock split announcement samples include 38 two for one, 39 three for two, and 10 reverse splits. A total of 36,714 observations for the announcement samples and the corresponding S&P 500 stock index were analyzed using standard risk adjusted event study methodology. Results suggest that the firms’ public stock split announcements did not affect stock price on the announcement day. Rather, for the two for one and three for two forward split samples, stock price exhibited a significant positive reaction up to 27 days prior to the announcement. For the reverse split sample, stock price exhibited a significant negative reaction up to 30 days prior to the announcement. Results support the semi- strong form efficient market hypothesis since stock prices adjust so fast to public information that no investor can earn an above normal return by trading on the announcement day. Investors greet forward stock split announcement with a positive sign, whereas they view reverse splits as bad news. Management may be using stock splits to adjust stock price to a more marketable range, downward with forward and upward for reverse splits. Evidence here suggests signs of insider trading activity up to twenty-seven days prior to the announcement of the stock split. INTRODUCTION Stock split announcements have always been very common phenomena among firms and continue to be one of the least understood topics in finance. A stock split announcement increases the number of shares of a company while decreasing the price per share. The two for one split is most common, for example a company with 500 shares at $10 per share will issue 500 additional shares bringing the total to 1000 shares theoretically dropping the stock price to $5 per share. A stock split usually takes place after an increase in the price of the stock, and it carries a positive stock price reaction. (Asquith) This phenomenon has not yet been fully understood, regardless the numerous studies in the field. ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS Volume 16 Number 1 BACKGROUND Barker (1956) presented one of the most popular theories to explain stock split behavior. Barker findings failed to consider the split action itself. Barker’s study concluded that price changes occurred because of the increase in cash dividends and not from the split action. (Johnson). According to the â€Å"signaling hypothesis†, managers use stock split announcements to convey positive information about the firm (Ikenberry, Rankine, Strice). Investors see a stock split announcement as a positive thing, whereas a reverse split does not convey favorable information. Fama (1969) suggests that the stock market is â€Å"efficient†, meaning that stock prices adjust very fast to new information. The theory of market efficiency is concerned with whether prices reflect all the public available information or not (Fama 1970). Efficiency implies that it is impossible for the investor to earn an above normal return from public information. PURPOSE The purpose of this event study is to test market efficiency theory by analyzing the impact of three samples of stock split announcements on the firm’s stock price. Stock split announcement samples include 38 two for one, 39 three for two, and 10 reverse splits. Specifically, how fast does the market price of the firms’ stock react to the samples of regular and reverse stock split announcements examined? The study tests whether the investor can make an above normal return by relying on public information imbedded in a stock split announcement, as well as if stock price is affected by a stock split announcement. This study investigates if acting on public information is enough to have an unusual return, or if there must be an illegal action such as inside trading to be able to â€Å"outperform† the stock market. Which form efficiency is the market? Research shows that the market is semi-strong form efficient. An above normal return can only be gained from inside information, and not when acting in public information. LITERATURE REVIEW Fama defined market efficiency in terms of how quick the stock market reacts to the information and suggested three kinds of market efficiency: Weak form, semi-strong and strong form efficiency. If market is weak for efficient, then stock price reacts so fast to all past information that no investor can earn an above normal return (higher than the market or the return on the S&P 500 index). This study shows how investors will not earn a high return from acting on public information (stock split announcement), while investors having access to inside information will make an abnormal return. A second kind of market efficiency is semi-strong. It states that stock price reacts so fast to all public information that no investor can earn an above normal return (higher than the market or the return on the S&P 500 index) by acting on this type of information. (Fama 1970). Splits usually result in high market valuations, but study done by Fama (1970), Dodd, Patell and Wolfson, found that there is no evidence of abnormal return after the release of public information. They concluded that the market assimilates and takes into consideration public information very fast, within 5 to 15 minutes after the disclosure (Malkiel). This supports the idea that an investor acting on public information will not earn an above normal return. When this happens the market is said to be semi-strong form efficient. If the market is strong form efficient, then stock price reacts so fast to all information (both public and private), that no investor can earn an above normal return (higher than the market or the return on the S&P 500 index) by acting on this kind of information. Studies made by Friend, Brown concluded that profit can only be gained by having access to private or inside information, which is illegal. Fama ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS Volume 16 Number 1 presents evidence supporting that efficiency is not met in the strong form and that the semi-strong form is more accurate. This study agrees that stock split announcement are affected in a company stock price according to the semi strong form efficiency which states that stock prices reacts so fast to all public information that no investor can earn an above normal return after the announcement is made. An example would be information concerning a merger. If an investor would buy shares on the announcement day of the merger, the semi strong market efficiency believes that the investor would never be able to earn an above normal return, because adjustments had already been done in the stock price. The market has already been adjusted, so therefore the only way to outperform the market in this case would be by using inside information. METHODOLOGY: This study includes samples of companies that announced a two for one, three for two or reverse stock split announcement. These companies trade their stock in either the NYSE or NASDAQ. The Data for this study was collected from http://finance. yahoo. com/. The announcement date (Day 0) is the day that the stock splits are announced. Every stock return from the companies and from the S&P 500 index was also collected. The Event Study proceeds as following: 1. Historical prices for both the firms and the S&P 500 were collected from day -180 to day +30, being the event period -30 to +30 and Day 0 the announcement day. 2. Holding Period Return was calculated for all the companies as well as for the S&P 500 on the event period days (-180 to +30). HPR was obtained from the following formula: Current Daily Return = (current day close price – previous day close price) / prev. Day close price 3. A regression analysis was performed, being the current firm return the dependent variable and the S&P return the independent variable. The data that was used was the one belonging to the pre-event period (from day -181 to -30). The alpha and the beta were obtained from the regressions. 4. The expected return for each firm as well as for the S&P 500 was calculated: Expected Return = (Alpha + Beta) x S&P actual return 5. Excess Return was obtained from the difference between Actual and Expected Return. Excess Return = Actual Return – Expected Return 6. Average Excess Return (for the Event period) was calculated as: Average Excess Return (AER) = Total Excess Return / n (number of firms in the sample) 7. Cumulative Average Excess Return for the event period (Day -30 to Day +30) was calculated by adding the AER for each day in the event period. 8. A correlation test was done with AER and CAER. The graphs represent AER and CAER plotted against Time. ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS Volume 16 Number 1 Table 1 describes 38 companies that split their stock on a two for one basis between December 1, 2006 and May 14, 2007, along with their respective alphas and betas. TABLE 1: TICKER AFAM ACLI COMPANY NAME Almost Family Inc. American Commercial Lines Inc Selective Insurance Group Inc. ZOLL Medical Corp Trimble Navigation Ltd. Albemarle Corp Guess? Inc. Cooper Industries Ltd Jacobs Engineering Group, Inc GameStop Corp Sealed Air Corp. Carlisle Companies Inc CarMax Inc. Harsco Corp. Amphenol Corp Cabot Oil & Gas Corp Nike Inc Cummins Inc Greif Inc DATE ANNOUNCED Dec 11 Feb 06 TRADED INDEX NASDAQ NASDAQ ALPHA 0. 01665915 -0. 000394377 BETA 0. 08530878 2. 602491516 SIGI ZOLL TRMB ALB GES CBE JEC GME SEE CSL KMX HSC APH COG NKE CMI GEF Jan 30 Jan 25 Jan 25 Feb 07 Feb 14 Feb 14 Jan 26 Feb 12 Feb 16 Feb 08 Feb 22 Jan 23 Jan 17 Feb 26 Feb 15 Mar 08 Feb 26 NASDAQ NASDAQ NASDAQ NYSE NYSE NYSE NYSE NYSE NYSE NYSE NYSE NYSE NYSE NYSE NYSE NYSE NYSE -0. 000319706 0. 004077614 -0. 000187534 0. 0022 37728 0. 001589658 0. 000761731 0. 001074342 0. 000477979 0. 00085897 -0. 001167829 0. 003087277 -0. 001056001 0. 000467862 0. 000826123 0. 001079523 -0. 000720045 0. 02203648 1. 38328513 1. 207411999 1. 321541131 1. 327988752 2. 246784079 1. 308635864 1. 946533548 1. 721660362 1. 172042857 1. 346601558 1. 240366727 1. 658082593 1. 86971211 1. 568927816 0. 553921446 1. 980439113 1. 880200397 ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS VLGEA AZZ ATR TSO GEO TSBK VSEC MRO GIL NRG CROX AGN PMFG MIDD SJR PVA GILD PBR STR Village Super Market Inc AZZ incorporated AptarGroup Inc Tesoro Corporation Geo Group Inc Timberland Bancorp Inc. VSE Corp Marathon Oil Corp. Gildan Activewear NRG Energy Inc. Crocs, Inc Allergan Inc PMFG Inc Middleby Corp Shaw Comm CL Penn Virginia CP Gilead Sciences Petroleo Brasileiro Questar CP Mar 21 Apr 09 Apr 18 May 01 May 01 Apr 25 May 01 Apr 25 May 03 May 02 May 03 May 02 May 04 May 04 May 10 May 08 May 08 May 11 May 14 NASDAQ NYSE NYSE NYSE NYSE NASDAQ NASDAQ NYSE NYSE NYSE NASDAQ NYSE NASDAQ NASDAQ NYSE NYSE NASDAQ NYSE NYSE Volume 16 Number 1 0. 00054113 0. 002118906 0. 00174286 0. 00160687 0. 002825174 0. 000615586 0. 001278324 0. 000144992 0. 003089016 0. 00241574 0. 00282982 -0. 000453038 0. 002024817 0. 02028334 0. 001186211 -0. 00050926 . 000009116 -0. 00064373 -. 000142796 1. 351096108 0. 681656728 0. 033542167 0. 973844695 1. 578867077 0. 107464578 2. 457597999 0. 986395517 0. 000111517 0. 316285515 1. 783171812 0. 952984111 0. 039990601 1. 964415725 0. 938731083 1. 1695925 1. 517629839 1. 817825121 . 706466451 Table 2 describes 39 companies that split their stock on a three for two bases between August 23, 2006 and May 15, 2007, along w ith their respective alphas and betas. TABLE 2: ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS TICKER NGA EPIQ BAM WMS VIVO IEX ATLS VSEA BWS WCN RSG JCTCF MDCI PFBC CMCSA SWS BKE VOL SSI FMD CRVL GBCI AFG SPAR COMPANY NAME North AM Gav Epiq Systems Inc Brookfield Asset MGT V M S Industries Inc Meridian Bioscience IDEX Cop Atlas America Inc Varian Semicond Brown shoe corp Waste connections Republic SVCS Jewett Cameron Inc Medical Action IND Preferred Bank LA Comcast Cp A SWS Group Inc Buckle Inc Volt Info Science Inc Stage Stores Inc First Marblehead Corp Corvel CP Glacier Bancorp American Financial Group Spartan Motors Inc DATE ANNOUNCED May 15 May 10 May 02 May 07 April 19 April 04 April 27 April 24 March 08 Feb 12 Feb 01 March 13 Jan 09 Jan 25 Feb 01 Nov 30 Dec 12 Dec 20 Jan 09 Nov 10 Nov 13 Nov 29 Nov 15 Nov 02 TRADED INDEX NASDAQ NASDAQ NYSE NYSE NASDAQ NYSE NASDAQ NASDAQ NYSE NYSE NYSE NASDAQ NASDAQ NASDAQ NASDAQ NYSE NYSE NYSE NYSE NYSE NASDAQ NASDAQ NYSE NASDAQ Volume 16 Number 1 ALPHA -0. 001032797 0. 001183339 0. 000859066 0. 002219704 0. 0011736 22 0. 000243421 0. 000488161 0. 001788461 0. 000592124 -0. 000187979 -0. 000441765 0. 000124622 0. 001559912 0. 000301413 0. 001381697 0. 000530857 -. 000641295 0. 001338437 0. 000540995 0. 004563185 0. 003763906 0. 000329484 0. 000736169 0. 003450361 BETA 1. 997738247 1. 038735222 1. 251257403 1. 094503791 1. 550013068 1. 509306631 0. 38871871 2. 207840195 2. 599167684 0. 92152423 0. 761431985 -0. 512126102 1. 004029551 0. 867741293 0. 927831638 2. 477624454 1. 602298009 2. 358292804 1. 756904894 0. 830932855 2. 113368174 1. 743070573 0. 936337426 0. 519840545 ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS SBIB AEO CTBK IRM PERY EAT AME WGNB ACAP UBSH EML MCBI CASS CCFH Sterlin Bancshares American Eagle Outfitters Inc Trico Bankshares Iron Mountain Inc Perry Ellis International Brinker International Inc Ametek Inc WGNB Corp American Physicians Cap Union Bankshares Corp Eastern Co Metrocorp Bancshares Cass Information Systems CCF Holding Co Oct 31 Nov 14 Nov 08 Dec 07 Nov 21 Nov 02 Oct 25 Sep 18 Sep 26 Sep 07 Sep 28 Aug 04 Jul 24 Aug 23 NASDAQ NYSE NASDAQ NYSE NASDAQ NYSE NYSE NASDAQ NASDAQ NASDAQ AMEX NASDAQ NASDAQ NASDAQ Volume 16 Number 1 0. 001127642 0. 003616084 0. 001058586 -0. 0000284 0. 002794647 -0. 000020642 0. 00005895 0. 00024115 0. 000317657 -0. 00058103 0. 000419721 0. 000941528 0. 003356848 0. 002118726 1. 165421403 1. 593723526 1. 432917191 0. 627633001 0. 919648907 0. 886164833 1. 31003146 -0. 00226624 0. 066171033 1. 663620313 0. 22686963 0. 121493122 0. 113211419 -0. 08732041 Table 3 describes 10 samples of companies that split their stock on a reverse basis between August 27, 2003 and September 15, 2008, along with their respective alphas and betas. TABLE 3: TICKER OPWV ERIC IWOV SIG COMPANY NAME Openwage Systems LM Ericcson Telephone Co Interwoven Inc Signet Jewelers LTD DATE ANNOUNCED Oct 09 Oct 18 Aug 27 Sept 11 TRADED INDEX NASDAQ NASDAQ NASDAQ NYSE ALPHA 0. 00680888 -0. 006696905 0. 001398048 -0. 000938713 BETA 2. 51286756 1. 949328188 1. 469236928 0. 891488791 ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS BFLY REV CNXT IACI TMTA ERIC Bluefly Inc Revlon Inc Conexant Systems Inc. IAC/ InterActiveCorp Transmeta Corporation LM Ericcson Telephone Co April 3 Sep 15 June 2 June 09 Aug 15 April 09 NASDAQ NYSE NASDAQ NASDAQ NASDAQ NASDAQ Volume 16 Number 1 -0. 00449535 0. 000925943 -0. 004900502 -0. 001442165 -0. 002052045 -0. 004006643 0. 070525685 0. 902722337 1. 73193906 0. 982384488 1. 265168622 -0. 16807384 To test for semi-strong market efficiency the following null and alternative hypotheses are used for the three stock split samples: H10: The risk adjusted return of the stock price of the sample of firms announcing stock splits is not significantly affected by this type of information on the announcement date. H11: The risk adjusted return of the stock price of the sample of firms announcing stock splits is significantly positively affected by this type of information on the announcement date. H20: The risk adjusted return of the stock price of the sample of firms announcing stock splits is not significantly affected by this type of information around the announcement date as defined by the event period. H21: The risk adjusted return of the stock price of the sample of firms announcing stock splits is significantly positively or negatively affected around the announcement date as defined by the event period. QUANTITATIVETESTS AND RESULTS: Did the market react to the announcements of regular two for one, the regular three for two, and the reverse stock splits? Was the information surrounding the event significant? A’priori, one would expect there to be a significant difference in the Actual Average Daily Returns (Day -30 to Day +30) and the Expected Average Daily Returns (Day -30 to Day +30) if the information surrounding the event impounds new, significant information on the market price of the firms' stock. If a significant risk adjusted difference is observed, then we support our hypothesis that this type of information did in fact significantly either increase or decrease stock price. To statistically test for a difference in the Actual Daily Average Returns and the Expected Daily Average Returns over the event period day -30 to day +30, we conducted a paired sample t-test for the three samples and found a significant difference at the 5% level between actual average daily returns and the risk adjusted expected average daily returns. Average Excess Return (AER) graphs are shown below. Results here support the alternate hypothesis H21: The risk adjusted return of the stock price of the sample of firms announcing stock splits is significantly affected around the announcement date as defined by the event period. This finding supports the significance of the information around the event since the market’s reaction was observed. ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS Volume 16 Number 1 Is it possible to isolate and observe the samples’ daily response to the announcement from day -30 to day +30? If so, at what level of efficiency did the market respond to the information and what are the implications for market efficiency? Another purpose of this analysis was to test the efficiency of the market in reacting to the three samples of stock split announcements. Specifically, do we observe weak, semi-strong, or strong form market efficiency as defined by Fama, 1970, in the efficient market hypothesis? The key in the analysis is to determine if the AER and CAER are significantly different from zero or that there is a visible graphical or statistical relationship between time and either AER or CAER. T-tests of AER and CAER both tested different from zero at the 5% level of significance. Likewise, observation of the following CAER Charts (graphs of CAER from day –30 to day +30 for each sample) confirm the significant positive reaction of the risk adjusted returns for the two forward split samples up to 27 pre-announcement and a significant negative reaction for the reverse split sample up to 30 days prior to the stock split announcement. Two for one stock split announcements: ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS Volume 16 Number 1 Three for two split announcement: ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS Volume 16 Number 1 Reverse split announcement: ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS Volume 16 Number 1 There are three forms of market efficiency as defined by Fama, which are strong, semi-strong and weak form efficiency. Observation of the CAER graphs against time for two for one and three for two stock split announcements shows a positive reaction twenty seven days prior to the announcement date. Reverse splits are normally done in order make the stock more appealing for investors with an unusual low market price. (Lawson) Also, reverse splits might be used in order to reduce the number of shareholders of the company. As an example if a 1-10 reverse stock split is made effective, the investor will have ten times less shares than before, but at ten times the price. In the reverse split case, the CAER graph suggests that return falls from day -30 until day -15, while then increasing until day 10. After day 10 the stock starts to level off. CAER graphs for two for one and three for two stock splits show how excess return rises up to 27 days prior to the announcement day. From Day 0 until Day 30 stock returns start to level off. This evidence supports Hypothesis H10, which states that stock price is not affected by this type of information on the announcement date. The stock return has already been adjusted before the stock split announcement is made. The investor cannot outperform the market by using public information. The price has already been affected by the announcement of two for one and three for two stock split announcement. After the announcement day, from days 6 to 16 the return goes up, which is caused by investors that react favorably to the announcement by buying more shares. After this small increase, stock price decreases and levels off. The CAER graphs support the idea that the market is semi- strong form efficient. For the samples analyzed, public information does not affect stock price on the announcement day. Reaction is observed up to 27 days prior to the announcement date which suggests that to be able to â€Å"outperform† the market you must be aware of inside information. CONCLUSION: The purpose of this study was to test whether the investor can make an above normal return by relying on public information impounded in a stock split announcement. Using risk adjusted event study methodology, this study tests â€Å"how† and â€Å"when† public announcements of forward and reverse stock splits affect stock price. Stock split announcement samples include 38 two for one, 39 three for two, and ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS Volume 16 Number 1 10 reverse splits. A total of 36,714 observations for the announcement samples and the corresponding S 500 stock index were analyzed using standard risk adjusted event study methodology. Results suggest that the firms’ public stock split announcements did not affect stock price on the announcement day. Rather, for the two for one and three for two forward split samples, stock price exhibited a significant positive reaction up to 27 days prior to the announcement. For the reverse split sample, stock price exhibited a significant negative reaction up to 30 days prior to the announcement. Results support the semi- strong form efficient market hypothesis since stock prices adjust so fast to public information that no investor can earn an above normal return by trading on the announcement day. Investors greet forward stock split announcement with a positive sign, whereas they view reverse splits as bad news. Management may be using stock splits to adjust stock price to a more marketable range, downward with forward and upward for reverse splits. Evidence here suggests signs of insider trading activity up to twenty-seven days prior to the announcement of the stock split. ASBBS Annual Conference: Las Vegas February 2009 Proceedings of ASBBS REFERENCES: Volume 16 Number 1 Asquith, Paul, Paul Healy, and Krishna Palepu. â€Å"Earnings and Stock Splits. † The Accounting Review 64 (1989): 387-403. Barker, C. A. , â€Å"Effective Stock Splits,† Havard Business Review, XXXIV (January-February, 1956), pp. 101-106 Easely, David, Maureen O'hara, and Gideon Saar. â€Å"How Stock Splits Affect Trading: a Microstructure Approach. † The Journal of Financial and Quantitative Analysis 36 (2001): 25-51. Fama, Eugene F. Efficient Capital Markets: a Review of Theory and Empirical Work. † The Journal of Finance, 25 (1970): 383-417. 9 Dec. 2007 . Fama, Eugene F. , Lawrence Fisher, Michael C. Jensen, and Richard Roll. â€Å"The Adjustment of Stock Prices to New Information. † International Economic Review 10 (1969) : 1-21. 9 Dec. 2007 . Ikenberry, David L. , Graeme Rankine, and Earl K. Stice. â€Å"What do stock splits really signal?. † Journal of Financial and Quantitative Analysis 31. n3 (Sept 1996): 357(19). General OneFile. Gale. Longwood University. 9 Dec. 2007 . Johnson, Keith B. â€Å"Stock Splits and Price Change. † The Journal of Finance 21 (1966): 675-686. 9 Dec. 2007 . Lakonishok, Josef, and Baruch Lev. â€Å"Stock Splits and Stock Dividends: Why, Who, and When. † The Journal of Finance 42 (1987): 913-932. Lamoureux, Christopher G. , and Percy Poon. â€Å"The market reaction to stock splits. † Journal of Finance 42. n5 (Dec 1987): 1347(24). General OneFile. Gale. Longwood University. 9 Dec. 2007 . Lawson, Michael J. â€Å"Reverse Stock splits: The Fiduciary's obligations under State LAw. † California Law Review 63 (1975): 1226-249. Malkiel, Burton G. â€Å"Is the stock market efficient?. † Science 243. n4896 (March 10, 1989): 1313(6). General OneFile. Gale. Longwood University. 9 Dec. 2007 . ASBBS Annual Conference: Las Vegas February 2009