UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN ' BOOKSTACKS Digitized by the Internet Archive in 2011 with funding from University of Illinois Urbana-Champaign http://www.archive.org/details/investinginoptio800reil »i' FACULTY WORKING PAPER NO. 800 investing in Options of Stock Announcing Splits Frank K. Reilly Sandra G. Gustavson y. yj" i. College of Commerce and Business Administration Bureau of Economic and Business Researcli University of Illinois, Urbana-Champaign FACULTY WORKING PAPER NO. 800 College of Commerce and Business Administration University of Illinois at Urbana-Champaign September 1981 Investing in Options of Stock Announcing Splits Frank K. Reilly, Professor Department of Finance Snadra G. Gustavson University of Georgia Abstract This study examines the profit potential for individuals investing in the options of stocks announcing stock splits. The results indicate very little downside risk and substantial upside potential for those who get the information from the Dow Jones broad tape and invest the day before it is announced in the Wall Street Journal. juj-y, X50J. IN\^ESTING IK OPTIONS OF STOCKS ANNOUNCING SPLITS* Frank K. Reilly Sandra G. Gustavson** INTRODUCTION A previous study by one of the authors indicated that the stock market was generally quite efficient in adjusting to the announcement of a stock split. Specifically, the results indicated that trading volume and stock prices reacted rapidly to the announcement of the stock split on the Dow Jones News Service and most of the price and volume adjustment occurred prior to the report in the Wall Street Journal (WSJ) . A specific test of the profit potential for an investor who acquired the stock after the announcement of the forthcoming stock split in the WSJ indicated a small positive price movement, but there were no excess returns after allowing for commissions. While it apparently is not possible to derive excess returns from investing in the stock of the firm, one may question whether one might enjoy excess returns from investing in the options of the firms that announce stock splits. There are two main *The authors acknowledge the data collection of K. S. Subash, the extensive computer programming of Wenchi Wong, the use of the computer facilities at the University of Illinois and the comments of participants in the Finance Seminar at Illinois especially Paul Fellows, John Gilster, Steve Sears, Kevin Waspi and David Whitford. **The authors are Professor of Finance, University of Illinois and Assistant Professor of Risk Management and Insurance at the University of Georgia. Frank K. Reilly and Eugene F. Drzycimski, "Short-Run Profits from Stock Splits," Financial Management , Vol. 10, No. 3 (Summer, 1981), pp. 64-74. -2- reasons why such a strategy might be viable. First, the added leverage from options could transform a small profit on the stock into a much larger percent profit on the option. Second, the Black-Scholes option pricing model implies a positive relationship between the variability of 2 return on a stock and the value of an option for that stock. A study by Bar-Yosef and Brown indicates that stock returns become more volatile during the period surrounding a stock split. Thus, one might expect an increase in the value of these options because of the increased variance of returns on the stock caused by the announcement of a stock split. Therefore, the purpose of this study is to examine the return per- formance for investors who acquired exchange-listed call options for NYSE stocks that announced splits during the period 1974-1979. The sample and tests are discussed in the next section and the results are reported and discussed in section three. The final section contains a summary and conclusion wherein we discuss the implications of the results. TEST PERIOD, SAMPLE AND TESTS Test Period The test period was similar to that used in the prior stock split study. Specifically, rates of return were examined for the period 2 Fisher Black and Myron Scholes, "The Pricing of Options and Corporate Liabilities," Journal of Political Economy , Vol. 81, No. 3 (May/June, 1973), pp. 637-654; Robert Marton, "Theory of Rational Option Pricing," Bell Journal of Economics and Management Science , Vol. 4, No. 1 (Spring, 1973), pp. 141-183. Sasson Bar-Yosef and Lawrence D. Brown, "A Reexamination of Stock Splits Using Moving Betas," Journal of Finance , Vol. 32, No. 4 (September, 1977), pp. 1069-1080. -3- from 15 days prior to the announcement of the stock split in the WSJ to 20 days following the annoiincement . Split Sample The sample of split stocks includes every stock listed on the NYSE that had a stock split during the period 1974-1979 where the firm had options listed on one of the options exchanges at the time of the split . There were several instances where the split stock currently has options listed, but did not have listed options when the split occurred. In these instances the stock could not be considered. Based upon these criteria, the split sample consisted of 35 stocks. Matched Sample Each split stock was matched with a similar non-splitting stock that also had a listed option at the time of the split. The matching was based upon three factors. The first matching criterion was comparable system- atic risk (beta) , which is necessary because of the importance of return volatility in the option pricing model. The systematic risk measure was derived using daily returns during the calendar year prior to the split and also the beta during the year of the split. For example, when looking for a match for a stock that split in 1978, we considered the betas derived from daily returns for the split stock during calendar 1977 and also 1978. For cases in which the betas changed dramatically during the two years the beta that best reflected the period prior to the announcement was used (i.e., if the stock split announcement was early in the year we used the prior years beta; if the announcement came late in the year the current year's beta was chosen). -4- Given several reasonably similar betas, the firm that was in the same or a closely allied industry as the split stock was selected. The final criterion was that the options for the two stocks have the same expiration cycle. A list of the split sample and the matched sample is contained in Table 1, Option Data Option prices and volume were collected for the 36 day period from 15 days prior to the split announcement to 20 days following the anno tine ement. A major question concerned how many option contracts to consider, in terms of striking price and term to expiration. While there are typically several different options for each stock, the option prices will generally move together, based upon price changes for the underlying stock. Because the returns on the various options are highly correlated the options analyzed were limited to one "in the money" issue and one "out of the money" issue using the option prices as close as pos- sible to the stock price on the announcement date. For example, suppose the stock was selling for 7A at the time of the stock split announcement. We would attempt to find a 70 option (in the money) and an 80 option (out of the money) for the stock. The decision was always made based upon the prices prevailing at the time of the announcement , since the study assumed an investor would purchase the appropriate options on 4 this day. No difference in option selection would have resulted had the decision been based on prices prevailing on either the day before the announcement or two days before. -5- TABLE 1 SAMPLE STOCKS, ANNOUNCEMENT DATES, AND MATCHED SAMPLE STOCKS Anno unc emen t Split Stock Options Date Matched Options Gulf & Western Ind, Inc. 7/17/75 ITT Dow ChPTTiical 2/06/76 DuPont Halliburton, Inc. 2/20/76 Asarco United Technology Corp. 4/1A/76 AMP Exxon Corp. 5/21/76 Texaco Levi Strauss & Co. 6/11/76 Weyerhaeuser Atlantic Richfield 6/29/76 Gulf Oil Diamond Shamrock Corp. 7/16/76 Hercules Deere Co . 7/28/76 IBM Digital Equipment 8/17/76 Sperry Corp. Bristol Meyers 2/08/77 Corning Glass Works Sears, Roebuck 2/08/77 Kresge (K-Mart) Owens Illinois Inc. 2/11/77 Revlon Pepsi Cola, Inc. 2/25/77 Beatrice Foods Coca-Cola Co. 3/03/77 Norton Simons Phillips Petroleum 3/15/77 - Mobil Oil Raytheon 3/24/77 AKP Boeing Co. 8/02/77 Fleetwood Enterprise McDermotts Co., Inc. 11/09/77 Hewlett Packard Santa Fe Int'l 11/30/77 NCR Abbott Labs 3/13/78 Sobering Plough Baker Int'l Corp. 4/27/78 Corning Glass Works Northwest Ind, Inc. 5/10/78 Evans Products Tandy Corp. 5/15/78 Itel Signal Cos, Inc. 9/18/78 Avnet Hilton Hotel Corp. 11/17/78 Holiday Inns IBM 12/20/78 Pitney Bowes DuPont 1/16/79 Herciiles General Dynamics 1/25/79 Boeing Mobil Oil 1/29/79 Phillips Petroleim Bally Mfg. 2/21/79 Fleetwood Enterprise Union Oil Of Calif. 2/27/79 Gulf Oil Philip Morris 3/1/79 Pepsi Cola, Inc. Hewlett-Packard 5/21/79 Black and Decker R. J. Reynolds 10/19/79 Heublein -6- Because the most active option issues are those with relatively' short expiration dates, only options with less than six months to ex- piration were considered. As a result, four option contracts per stock were examined when possible — a short and longer term "in the money" option, and a short and longer term "out of the money" option. For example, assume that the previously discussed stock announced the split in March and that the stock options were on a February-May-August-November cycle. Ideally, the following four options would be examined: (1) 70 May, (2) 70 August, (3) 80 May, and (4) 80 August. In addition to the option analysis, the price and volume of trading for the underlying stocks was also considered. Tests The analysis of the price and volume for the options involved the examination of the time series plots of the average prices and trading volimie for the underlying stocks and options of the split stocks and the matched sample stocks. The trading rules based on this analysis are tested for abnormal profit opportunities. Price Changes . The average cross-sectional prices for the stocks and options for both the split and matched sample stocks were computed for each day during the 36 day period surroimding the split announcement as follows: N 35 MOPSS = Z SOP. /N and MSP = L SP . ^/35 t .^1 x,t t .^^ i,t MOPSS = mean closing price of the options for the split stocks on day t. -7- SOP. = closing option price for split stock i on day t. MSP = mean closing price of the split stocks on day t. SP = closing price of split stock i on day t. N = 35 options (if available) for split stocks. An average cross sectional series for each of the four option groups is computed: in the money short term (I,S); in the money longer term (I,L) , out of the money short term (0,S); out of the money longer term (0,L) . Series based on closing prices of the matched stock and options were computed in a similar manner. Finally, mean relative option price (MROP) and mean relative stock price (MRSP) series were derived as follows: MROP = MOPSS /MOPMS and MRSP = MSP /MMP . where MOPMS = average closing price of the options for the matched stocks on day t and MMP = mean closing price of the matched stock on day t. Based upon the prior split study, relative series were expected to increase over time and experience a major increase during the several days sur- rounding the stock split announcement. The important question in this study is whether the increase in the option series is enough to cause abnormal returns for an investor. Volume Changes . The average cross sectional volume series were computed as follows: N 35 MOVSS = I SOV. /N and MSV = I SV . /35 t .^1 i,t t .^^ i,t -8- MOVSS = mean volume of trading for the options of the split stocks on day t. SOV. = volume of trading in the option of the split stock i ' on day t. MSV = mean volvime of trading in the split stock on day t. SV = volume of trading in split stock i on day t. N = 35 options (if available) . The mean volume series for the matched sample were com.puted simarly, and the mean relative option volxome (MROV) and stock volume (MRSV) series were calculated as follows: MROV = MOVSS /MOVMS and MRSV = MSV /1>1MV . where MOVMS = mean volume of trading for the options of the matched sample stocks on day t. MMV = mean volume of trading in the matched stock on day t. Assuming that the transmittal of information to the public does not occur until the announcement date, trading should increase on that day and for a few days following the announcement. Investment Tests . The investment tests performed indicate whether an investor could take advantage of the price movements in the options of stocks announcing splits. The analysis is in two parts. The first considers the cumulative percent change for the various cross sectional series assuming a purchase on the announcement date (day 0) , the day before (day -1) , and two days before (day -2) . Day is important be- cause it is the day the announcement appears in the WSJ . The announce- ment will appear on the broad tape of the Dow Jones News Service some- time on day -1. Both of these tests assume purchase at the closing price -9- of the day indicated. Purchase on day -2 would imply some inside infor- mation. But as noted later, the results of this test are important in bracketing the expected returns available to investors who are able to purchase sometime during day -1 — not just at its close. Because this analysis indicates some substantial relative percent changes for the options, the second set of tests considers the specific returns to an investor who acquires the "typical" option during this period (day 0, day -1, -2) and sells it on day +3, paying commissions both to buy and sell. These returns are compared to a similar investment in the matched sample options. RESULTS The results are presented in four subsections: (1) the relative price and volume results for the underlying stocks, (2) the option price series (i.e., MOPSS; MOPMS; and MROP) over the 36 day period, (3) the option volume series, and (4) the investment tests. Underlying Stock Results Exhibits 1 and 2 contain the time series plot of the relative price and volume series for the underlying stocks. The stock price results are very similar to those in the prior study on stock splits. Specifically, the relative stock price ratio was about 1.80 at the beginning and re- mained in that area until day -1, when it increased from 1.7953 to 1.8369 (a 2 percent increase). On day it went to 1.8455 and increased each day until it peaked on day +3 at 1.8644. Subsequently, the rela- tive ratio declined slowly and ended the sample period at 1.8407. As before, most of the price reaction occurred prior to the announcement (_) GD cn CD >- I QZ LU H M DZ P W rs:: (_> o 1 — CO ll 1 LU QZ o o D O -in CM o o o o o pO ^ g: 1— K— 1 D _j ° D_ nO~' O o > a _i Sci: d o o o " I 05 "I BB •! SB '1 EB '1 OB M 8^ 'l Diiyy aoibd Nb3k Si M £ I— cc _J uu 0= CE LU Dt- -Z OS •; DlibU oe -i 3Wn"lDA 06 "0 Nb3W 00 "0 -12- in the WSJ and the increase after the announcement was not adequate for excess returns after commission. Prior to any announcement the relative volume series ratio was about .60 to .70. On day -1 it jumped to 1.04, on day it went to about 1.16, and peaked on day +3 at about 1.20. Subsequently it gener- ally declined and ended about where it began. In summary, although the sample is much smaller than the original stock split study (35 vs 130) and the control sample is different (a matched sample vs the market), the results are very similar. Clearly the major price change occurred on day -1 and it is not possible to derive excess returns after commissions, in the absence of inside infor- mation. Likewise, the volume increases are tightly grouped in the time period from day -1 to day +3. Relative Option Price Results Table 2 contains the average cross section results for the 36 day period for the 12 series (three sets of mean prices for each of four option types). Exhibits 3, 4, 5, 6 contain the time series plots of the mean relative option prices for the various types of options. The mean relative option price series for the in the money short option started the sample period at a value of 1.371 and varied between 1.40 and 1.28 prior to day -1. On day -1 the ratio went from 1.310 to 1.558 — a 19 percent increase. It peaked on day +3 at 1.768, which is about 13 percent higher than the close of day -1 and 12 percent above the close on day 0. Subsequently the ratio declined somewhat to an ending value of 1.71 which was substantially above the pre-announcement values. -13- CM CI. tn c o •H ./ 4J / to B. / OJ 4-1 O, / a y p^ / J-l V c 0) E / 0) 0) ''^ > o •H 0) c 4J x: 3 cn n) j-> >,^ c >, Q « 4-1 ^ to Q 00 t^ o m On r>- in LO OD o-j r^ 00 03 t^ a> 1— l\Dco'.oa^mr^ooo^r^vDoo'X>u~icNir~roi— iCTi mrsir-)CNiOi-ivDOnr^co-3-cNioo-^^Jr-^o^o^ mm iH'HtH^THi-H.-HtHi--l'H.HiH'--l'HrSCNl<>)rvimcNlCNlCSrOCNIcSCNlCSCNCS!NCNCNCNICSI m?? aNONO^oicrvoor^cOi-Hvo— lcSLOo^L^lnr^lcor^l-3■-3■a^csla^Ln^£>mo^ \Doo-3■\DOoooo^ao L^|Lnl^l-^u-l~3■~^■^^r«.r^a^cooocooot^^^^^^~-^^^~I^^^oor^r^ i1 00 CO i-I.H.-liHr-lT-:iHTH ,H r-l f-H rH H 1— lO-l1— ^o»3■Ooo.-^a^-*oo■— l"^OOOvO\OOOCNcsicM\OOOOrOCNivOOLnO>u^i-H Or-I m-^rommmmmmmcMtNCNimu^LnovDr^oi^i — i — voin^vOvo^r^^vOvovo r^r^ iH r-l t-) rH .H rH rH iH .H ■-I THiHiHi-lrHiHT-lr-liHiH o^oo^o^<^mm^Ol-lo^r~»oma^oooo^or^lno^co^^4lnooocT^mmo^u^moo\omcr>aN'-(iJ~imiri\ocrir~.msTrsiirim<3- mmmmcNjt-g^^r-j,— lOOo^a^o^o^o^a^ooo^o^oOI^ccoooocoooooo^c^ooo^oo^^^^ .HiHr-lrHiHrHr-I.HiHiH'-itHOOOOOOOOOOOOOOOOOOOOOOOO CT\ooooOl^lOu^o^mo^3-inmorOi— icNi oomt^ocr\OcNOr~-\Dr- la^oo^a^o<^0'^f^od-'ni5Ln\D^in -3-vD-3-ooooor^ocNcsiovOCT\iO'>D— ir^3-v£3r^va't— ^u^vOcr~Js3■r^cNi^iHCNicNic-viro£>^nlnL^l^^ 00 CO 00 r^ C0l^r~.00000000000000O^O^O^0^Oa^O^0^O^O^a^a^O^O^O^O^0^O^O^O^O^C^ i-l -3•lnoot-^a3a^Lnc^l1— icNif-^LniHCM^r^mrMmr~->— lc^o^o^moOl— lmr~.moo^o^O^^ r-l^^.-lOcslr-^r^^-^^MmI-lc^o^O<-lcs^.-^^or^.3■r^<■c^lrs^^^^cs<-vd-^OI^oor-^^nLnLn 0\ >^o\0'>0\DvD>X)0\OvDvDinLnvor^t^r~r~t^r~.r^r^p~r^r^t^r^r^p^r~-i^r~-r^r~-i — r^ Ln<•^ocNI--{oo^oo^^Oln-a■mcSlHOr-(cso^•-(rHl-(rH,HlH.HCsl I I I I I I o o az en o o -in LU ro 1 1 I LU H C ) ca » > M cc X Q_ o O LU LU CC CE LU 0V2 02 -2 00 -2 09 •! DlibU 03 'I 0» M 02 3Diyd 30by3Ab .Nd3W 00 'I a o H M CQ M X X o I >- LU O LU X LU O az O O LU LU CO LU ov -2 02 -2 00 -2 DUbU 33IU°d 30tiy3°AV'NH3W°^"' 00 OB "0 o o CO o X CO UJ o 3 1 lO 1 H 1 UJ O X ^— * w oc Q_ o o UJ LU CC LU 00 ■» 09 ■£ 0? ■£ 09 -Z Ofr Diiby 33iyd •2 00 "2 09 30by3Ad Nb3W 02 M 08 "0 o o o o >- UJ o o ID O I i I IxJ (_) d O O LU UJ 01 CE LU 09 •£ OE -E 00 •£ Oi '2 OV ■£ 01 "2 09 '1 DliyU 33iyd 3{)by3A0 NU3W OS -I OS M -18- It appears that these prices increased after the announcement, and generally maintained the relative price gain. The mean relative option price series for in the money long options started at 1.569 and declined to 1.471 on day -2. During Day -1 it increased to 1.645 (12 percent) and peaked on Day +3 at 1.905 (16 per- cent above day -1) . Subsequent values were lower but were consistently above any of the p re-announcement ratios. Again, the big increase oc- curred on day -1 and continued until day +3, followed by a small decline. The mean relative option price series for short-term out of the money options (Exhibit 5) increased from 1.631 to 1.886 on day -2. On day -1 the ratio jumped to 2.645 (a 40 percent increase), followed by a small decline on day and twin peaks on day +3 (3.05) and day +7. The ratio ended the period at 3.238 which was the second highest value. Clearly, the relative price performance of these options was quite superior for the period following the initial announcement on day -1 and after day 0. The time series plot of the mean relative option price series for the long out of the money options (Exhibit 6) indicates that the big increase occurred on day -1, with the overall peak ratio on day +3. The increase on day -1 was 26 percent, while the increase from day to day +3 was an additional 18 percent. The ending values were lower, but still above any p re-announcement values. In summary, all the relative option price series indicate that the options of stocks that annoimce a split clearly outperform the options for a matched sample of stocks beginning on the day prior to the public -19-. announcement in the WSJ. Typically, the peak relative option price oc- curs on day +3, with subsequent ratios above pre-announcement ratios. Relative Option Volume Results Table 3 contains the mean relative option volume (MROV) for the four types of options. The relative volume pattern for the in the money short options indicates that the major trading impact occurred on day 0. Note that the big price impact on day -1 was on lower volume than on day 0. These results indicate that the option trading activity was a consequence of the public announcement in the WSJ . The relative volume pattern for long in the money options likewise indicates a large increase in volume on day followed by several days with comparable relative volume. In this case, it apparently took longer for the activity to return to the normal range. Notably, the volume on day -1 was above normal relative to prior days. The relative volume for short out of the money options also likewise shows a sharp spike starting on day -1, peaking on day 0, and generally declining toward the pre-announcement volume by the 20th day. Finally, the relative volume for the long out of the money options, indicates a spike starting with day -1, a peak on day and the highest relative volume on day +3 followed by generally declining values. This is the only instance where volume on day +3 is the highest value. Strong volume on day +3 is consistent with the price series peaks on day +3. In summary, relative volume always shows a definite increase on day -1 when the news comes over the Dow Jones News Service, followed by a clear peak (in three of the four cases) on the day of the public -20- TABLE 3 MEAK CROSS SECTION OF OPTION VOLUME Mean Average Volume Days \ Types Relative \^ of to the \Options Announcement Day Split Set/Matched Set -15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -A -3 -2 -1 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Out of Out of In the In the the the money money money money short long short long term term term term option option .477 option option .884 .690 .649 .666 .424 .723 .469 1.058 .497 1.011 .589 .823 .614 .613 .484 1.179 .600 1.097 .764 1.189 .576 1.237 .985 .794 .398 1.006 .768 .692 .467 .656 .749 1.089 .491 .670 .566 .983 .788 .885 .830 .750 1.027 .749 .860 1.315 .601 1.126 .685 .834 .752 .918 .829 1.346 .595 .889 .690 1.458 1.469 1.730 1.119 2.055 1.429 2.307 1.402 1.580 1.371 1.732 1.268 1.426 .847 1.041 1.045 1.548 1.383 1.499 1.667 1.207 .944 .952 .965 .876 .716 .702 .807 1.095 1.399 1.393 .975 1.501 1.381 1.798 1.121 .983 .739 .900 .984 1.273 1.082 2.066 1.326 1.153 1.126 1.980 1.224 1.070 1.0587 1.972 1.267 .930 .765 1.337 .991 .762 .595 1.156 .902 .861 .793 1.574 .854 1.010 .823 .880 .984 1.327 1.351 1.795 1.159 .881 1.184 1.555 1.377 .801 .941 1.175 .758 .588 .891 1.212 .994 .724 .531 1.092 1.083 announcener.t. Typically, there is strong volume or. day +3 fcllowed by generally declining values somewhat above the pre- announcement values. Results of Investment Tests These tests indicate whether an investor could take advantage of the price movements in the options of stocks announcing splits. The analysis is in Ds'O parts. The first presentation considers the cumu- lative percent change for the various cross sectional series. The second part examines the specific returns to an investor who acquired the "t3.-pical" option during this period and paid commissions to buy and sell. Table 4 part A contains the ciimulative percent change figures for the four t\T?'es of options and Exhibit 7 contains the plot of the four split- minus-match series assuming an investor acquired the options at the closing price on day 0. The results for the short in the money options showed negative re- turns the first day, a peak cvmulative value on day 3 of 6.305. one sub- sequent negative value and an ending value cf 6.201. The matched sample results were not very good since all the relative values (split minus match) were better. Tiie relative series was never negative, it peaked at 11.259 percent on day 3 and finished at over 8 percent. The unadjusted long in the money options had a peak value of 4.67 percent, followed by numerous negative returns and a final value cf about 2 percent. The relative results peaked at 9.2 percent, only had two negative values and the ending cumulative value was b .^ percent. The short out of the money options experienced substantial vola- tility, numerous negative values, a peak of 9.5 percent and ended with -22- TABLE 4 >2:a:; ci-ii.^..ive ?ej^ce:,t price CR.y,-GES for split options l-jadjustzi -\-e ^j^j-s-^ .HATCHED OFTIOKS .^^wii_^ ._,b ..uj^b^i^ TO A, Purchase at Close of Announcecent Day (Day 0) I.S Split Split Split ".: Split Sclic minus natch Split ninus natch Split cinus natch Snlit minuE match Close of Day 1 -1.876 1.763 -1.346 1.315 -3.259 1.326 0.355 5.766 2 0.677 5,061 0.000 2.505 -2.421 3.038 0.665 7.637 3 6.305 U.259 4.671 9.212 9.497 19,323 8.075 16.607 U 2.736 6.331 1.465 6.122 1.676 1,458 2.174 4.567 5 6.071 9.006 2.573 6.487 6.331 6,515 3.815 7.353 6 7 3.048 8,616 -1.188 4.722 -1,117 6,088 0.444 6.375 0.261 8.152 -3.523 3.952 9.311 26.342 -3.327 5.934 8 0.234 5.013 -3.108 3.233 -6.797 7.832 -4.636 2.864 9 -0.599 0.365 -3.721 -0.864 -10.242 2,675 -6.610 -1.512 10 0.162 1.541 -2.771 0.908 -10.242 2,640 -6 . 034 0,002 11 1.902 3.217 -2.078 0.935 -7.542 2,938 -3.949 1.567 12 2.267 2.968 -3.246 -1.054 -5.214 4,174 -6.034 -2.496 13 1.407 5.133 -2.573 2.437 -7,169 -0.4C1 -2.795 0.951 14 6,.305 7.752 0.277 1.764 -0.372 -1.9C1 0.976 -3.290 15 7.243 4.920 2.631 2.713 2.048 4,232 3.194 1.009 16 7.139 7.008 1.227 3,810 1.583 14,028 1.020 4.454 17 4.429 4.912 -0.534 2.518 -6.052 7,703 -1.730 4.201 18 3.127 2.381 -0.713 2.849 -4.749 6,605 -2.351 0.874 19 4.117 7,624 0.079 6,146 -4.097 19.047 -1.642 9.076 20 6,201 8.262 2.059 6,403 -C.372 22.335 0.532 10.210 B. Purchase at Close Day Before Announcement (Day -1) ^S I.L 0,S 0^ spj.it Split Split Split ninus ninus ninus ninus Solit catch Srlit natch Sclit natch Split r-^atch Close of Day 1.599 1,423 5.668 5.629 -6,446 -4.941 -0.354 -1.512 1 -0.307 3.162 4.246 6.869 -9.495 -3.473 0.000 4.316 2 2.287 6.503 5.668 8.135 -8.711 -1.829 0.309 6.204 3 8.005 12.792 10.604 15.107 2.439 13.622 7.692 15.166 4 4.378 7.804 7.216 11.836 -4.878 -3.586 1.813 3.076 5 7.767 10.534 8.387 12.264 -0.523 3.133 3.446 5.869 6 4.696 10.096 4.413 10.286 -7.491 1.111 0.088 4.931 7 1.864 9.593 1.945 9.385 2.265 20.544 -3.669 4.541 8 1.837 6.445 2.384 8.686 -12.805 3.109 -5.172 1.459 9 0.990 1.781 1.736 4.555 -16.028 -3.770 -6.941 -2.941 10 1.784 2.970 2.740 6.381 -16.028 -1.E34 -6.366 -1.419 11 3.531 4.673 3.472 6.446 -13.502 -1.674 -4,288 0.133 12 3.902 4.429 2.238 4.392 -11.324 -0.571 -6.366 -3.945 13 3.028 6.586 2.949 7.922 -13.153 -4.981 -3.139 -0,507 14 8.005 9.279 5.961 7.410 -6.794 -6.794 0.619 -4,855 15 6.958 6.455 8.659 6.503 -i.530 -0.674 2.829 -C.539 16 8.852 6.545 6.965 9.510 -4.965 6.796 0.663 2.979 17 6.099 6.407 5.104 6.118 -12.108 2.946 -2.078 2,76^ 18 4.776 3.853 4.915 6.439 -10.889 1.800 -2.697 -0,591 IS 5.781 9.119 5.752 U.782 -10.279 14.022 -1.989 7.695 20 7.899 9.788 7.844 12.151 -6.794 17.077 0.177 8.808 o o cr Q 1 — ■z. l/l ^ LU M M o o 7~ 1 1 1 I 1 ; 1 LJ * ^ + ■o 2 ZD O •z. o LiJ CO o _J o LU to cr X (_) cc ID Q_ I I LU lT) cr nr (_) LU C_) OC LU LJ cr _j o o -rv o o -O o o o o O to D ex _ LJ tr D o o — ' en o o o o o o o o DO "DE 00 "S? 00-0? OD'Sl 00 'Ol 00 'S 00 "0 aoNdHo aoibd iN33y3d a/uiuinwno 00 "S- 00 'OI- -24- a negative percent. In contrast, the relative results were 19 percent on daj' 3, had only two negative values and ended vith a peak value of over 22 percent. The long out of the money options were likewise quite volatile and experienced numerous negative values. The relative series had only three negative values, a peak of 16,6 percent on day three, and an ending value over 10 percent. Overall, the relative results were quite good. The peak cumulative percent generally occurred on day +3 and ranged from 9 percent to over 19 percent. For the total period there were few negative values and all the series ended with positive cumulative values. Comparing the four, it appears that the short in the money series is most conservative, while the short out of the money series has the most volatile set of returns. Table 4 part B contains the cumulative return series assuming an investor acquired the option at the close on day -1. As noted, this is clearly a possibility for an individual who watches the Dow Jones News Service or has a broker who watches it and informs the investor of the split announcement. Such an individual might be described as an aggressive investor. The series are plotted in Exhibit 8. The results for the short in the money options indicated only one negative value, the series was over S percent on day 3, subsequently peaked near 9 percent and finished at nearly 8 percent. Relative to the matched sample, the peak on day 3 was almost 13 percent, there were no negative relative returns and the series ended at almost 10 percent. The long in the money option series peaked at over 10 percent, was never negative and ended at almost 8 percent. For the relative series the peak was over 15 percent and it ended at over 12 percent. -25- o o o - LD ID w J w J —^ ^ •>. •■ »• H M o o > — 1 -^ 1 1 * ¥ + ■u- >- CE Q LJ cn D ^_J cr Q_ I I UJ O C_) GC LJ Q_ UJ o o -CM o o o o o o o cn o cr — C_) D g5 CO cr LJ o o o o o o o o o o 00 "DE 00 "SS 00*02 00*51 00 "01 00 "S 00 "0 00 -s- 00 •01- -26- Again, the short out of the money option series experienced sub- stantial volatility. The unadjusted results had a negative cumulative value during 19 of the 21 days. The relative results were better, but still quite volatile with 11 negative values, as well as several large positive values ranging from 13 percent to 20 percent and ending at 17 percent. The long out of the money options experienced several negative values, peaked on day +3 at about 8 percent and ended the period about even. The relative results peaked at 15 percent on day 3, had fewer negative values (8), and ended at 8.8 percent. Again, the overall returns relative to the matched sample were good. Also, the in the money results were less volatile and experienced fewer negative results. Table 5 contains the cumulative returns for the four types of op- tions assuming the investor acquired the option at the close on day-2 — i.e., the day before the announcement was on the Dow Jones News Service. These results are implicitly unrealistic because they assume the investor apparently acts before any announcement. Still, these results are of interest because they provide a range of results for those who receive the news from the broad tape (i.e., the aggressive investor). The results in Table 5 assume the investor waits until the close of day -1 to buy. Obviously the investor would be able to buy somewhere between the close of day -2 and the close of day -1. Therefore, the realized rates of return for such an investor should be somewhere between the results in Table 5 and Table 6. The plot of the four relative series is in Exhibit 9. -27- TABLE 5 MEM CUMULATIVE PERCENT PRICE CHANGES FOR SPLIT OPTIONS UNADJUSTED AND ADJUSTED TO MATCHED OPTIONS ASSUMING A PURCHASE AT THE CLOSE OF DAY MINUS TWO 1, S I,L 0,S 0,L Split Split Split Split minus minus minus minus Split match Split match Split match Split match Close of Day -1 18.935 18.978 10.594 11.717 38.147 39.630 25.597 26.225 20.836 20.704 16.863 17.947 29.242 32.208 25.153 24.630 1 18.569 22.081 15.290 19.007 25.030 32.445 25.597 30.513 2 21.655 25.913 16.863 20.425 26.113 34.376 25.986 32.471 3 28.455 33.284 22.322 27.897 41.516 54.016 35.258 43.313 4 24.142 27.610 18.515 24.266 31.408 34.162 27.873 29.756 5 28.172 30.981 19.870 24.826 37.425 42.510 29.928 32.961 6 24.520 29.963 15.475 22.405 27.798 37.755 25.708 31.147 7 21.151 28.921 12.746 21.224 41.276 60.767 20.988 29.775 8 21.120 25.773 13.232 20.587 20.457 37.618 19.100 26.318 9 20.112 20.946 12.514 16.425 16.005 29.564 16.880 21.482 10 21.057 22.286 13.625 18.348 16.005 31.471 17.601 23.145 11 23.135 24.320 14.434 18.499 19.495 32.630 20.211 25.232 12 23.575 24.146 13.070 16.322 22.503 34.579 17.601 20.635 13 22.536 26.136 13.856 19.896 19.976 29.510 21.655 24.987 14 28.455 29.772 17.187 19.742 28.761 30.244 26.374 21.563 15 29.589 27.130 20.171 21.139 31.889 36.974 29.150 26.431 16 29.463 29.199 18.297 21.937 31.288 46.330 26.430 29.359 17 26.188 26.540 16.239 20.342 21.420 37.734 22.987 28.427 18 24.614 23.736 16.031 20.638 23.105 37.088 22.210 24.930 19 25.811 29.191 16.956 24.041 23.947 49.371 23.098 23.349 20 28.329 30.261 19.269 24.650 28.761 53.761 25.819 35.024 o o o H M pq M S X en ^£. w hJ CO J r-^ ». >.. >. «. 2: 1 M 1 I 1 >— * ■ss; + -o- (X Q UJ CO LjJ I/) I cr ID Q_ I I UJ O ■z. cr x: LJ LU LJ or LJ or UJ LJ DD -ZL o o rvj D to o cr .^' "~ — (_) o DO O • Ll_ o LO 00 "VS DD "SS DO •Bt' DO "Ot- 00 -^E DO 'M 3oiJtiH3 30iyd iN3oy3d a^itiinwao 00 'SI 00 -f -29- The results in Table 6 are very encouraging. The short in the money option series ranges from 18 percent to 29 percent and is 28.4 percent on day 3. The relative series ranges from 18 to 33 percent and peaks at 33.3 percent on day 3. The results for the other types of options are similar, with the only difference being the valiie on day 3. The long in the money series is 22 percent (28 percent relative); the short out of the money series is 41 percent (54 percent relative) ; while the long out of the money series is 35 percent (43 percent rela- tive) . In this instance, the higher volatility is advantageous. In summary, these high return results are somewhat unrealistic since they assume prior information. Still when they are combined with the results in Table 5, they indicate a range of possible returns for an aggressive investor. Realistic Returns on a "Typical" In v estment Because the prior results indicated positive cumulative returns on day 3, this section presents the results for a investor who is assumed to acquire a typical option contract on day 0, -1, or -2 and sell the option on day +3 paying commissions on both the purchase and the sale at a rate derived from a broker. The results in Table 6 part A, which assumed the purchase and sale of one contract were initially surprising to the authors who were confident that there would be positive net returns in most instances because the cumulative percent changes on day +3 were typically in the range of 6 The commission rate used is fairly conservative, given the range of commission schedules in existence. -30- TABLE 6 COSTS AND RETURNS FOR AN INVESTOR TsfHO ACQUIRED ALTERNATIVE OPTION CONTRACTS AND SOLD ON DAY THREE A. Position = 1 Contracts (Commission = $24 plus 1% of value of Contract) Type of Option Buy Day Split Options on Orig. Price Net $ + Comm. Result % of Cost Matched Orig. Pr + Comm, Sample Options ice Net $ Result % of Cost I,S I,S I,S -1 -2 758.27 746.15 632.02 -17.00 -4.88 109.25 -.022 -.006 .173 488.60 487.59 487.59 -79.97 -78.96 -78.96 -.164 -.162 -.162 I,L I,L I,L -1 -2 990.57 938.05 851.19 -22.59 29.93 116.79 -.023 .032 .137 580.51 579.50 586.57 -83.77 -82.76 -89.83 -.144 -.143 -.153 0,S 0,S 0,S -1 -2 266.40 282.56 210.85 -30.03 -46.19 25.52 -.113 -.164 .121 119.95 121.97 122.98 -58.81 -60.83 -61.84 -.490 -.499 -.503 0,L 0,L 0,L -1 -2 514.86 516.88 415.88 -19.11 -21.13 79.87 -.037 -o041 .192 226.00 223.98 224.99 -68.83 -66.81 -67.82 -.305 -.298 -.301 B. Position = 2 Contracts (Comnission = $24 plus 1% of val ue of Contract) Type of Option Buy Day Split Options on Orig. Price Net $ + Comm. Result % of Cost Matched Sample Options Orig. Price Net $ + Comm. Result % of Cost I,S I,S I,S -1 -2 1,492.54 1,468.30 1,240.04 14.00 38.24 266.50 .009 .026 .215 953.20 951.18 951.18 -111.94 -109.92 -109.92 -.117 -.116 -.116 I,L I,L I,L -1 -2 1,957.14 1,852.10 1,678.38 2.82 107.86 281.58 .001 .058 .168 1,137.02 1,135.00 1,149.14 -119.54 -117.52 -131.66 -.105 -.104 -.115 0,S 0,S 0,S -1 -2 508.80 541.12 397.70 -12.06 -44.38 99.04 -.024 -.082 .249 215.90 219.94 221.96 -69.62 -73.66 -75.68 -.323 -.335 -.341 0,L 0,L 0,L -1 -2 1,005.72 1,009.76 807.76 9.78 5.74 207.74 .010 .006 .257 428.00 423.96 425.98 -89.66 -85.62 -87.64 -.210 -.202 -.206 -31- to 15 percent. These results indicate the importance of the commissions which seriously affect all contracts, but which are especially burden- some on the low priced out of the money options. For example, the typical short in the money option was assumed to be priced at $7.27 on day (see Table 2) causing the commission on the purchase to be $31.27 ($24 plus 1 percent of $727) for a total value of $758.27. Clearly, this commission of A. 3 percent on the purchase and a similar commission on the sale meant that although the option increased by over 6 percent during the three days, the net dollar result was negative. The results were generally negative for all purchases on day or day -1 (except for the long in the money option) . Note that, even with the heavy commis- sions, all the purchases at the close of day -2 resulted in positive rates of return. As noted, assuming the potential return for an aggres- sive investor is somewhere between the returns for day -1 and day -2, it appears that many of these investors would have experienced positive returns in spite of the heavy commissions on one contract. Also these realistic returns are substantially above those available on the matched sample stocks options where all the returns were negative and ranged from -14 percent to -50 percent. Obviously, because of the heavy fixed component of the commission ($24) , the low price out of the money op- tions suffered most. To show the impact of the commission. Table 7, part E contains results similar to part A except that in all cases the investor is assumed to acquire two contracts. Although this requires an increase in capital, it reduces substantially the relative impact of the commission. -32- With this assumption, almost all the returns become positive irre- spective of when the investor makes the purchase — even at the close on day 0. Assuming one acquires an option at the close on day -1, the returns for in the money options are about 3 to 6 percent. Finally, assuming an aggressive investor buys the option prior to the close on day -1, the apparent returns are rather large for a short run investment. Specifically, the ranges are as follows: Short In the Money .026 to .215 Long In the Money .058 to .168 Short Out of the Money -.082 to .249 Long Out of the Money .006 to .257 Apparently the downside risk is rather minimal. The major concern is the upside opportunity, which depends upon how fast one can acquire the option after the announcement on the broad tape. As before, the short out of the money option loses money only because of its low price and the commission included. It can be determined that all the option types become profitable for all days if one assumes a purchase of 27 contracts. SUMMARY AND CONCLUSION Summary A prior study indicated that investors could not derive abnormal rates of return by acquiring the stock of companies that announced stock splits. Apparently the prices adjusted quite fast and the size of the price change was not large enough to provide abnormal returns after taking account of commissions. This study examined whether one can derive abnormal returns by acquiring the options of stocks when the company announces a stock split. -33- The sample included all stocks on the NYSE that split and had op- tions available on one of the exchanges. Each company was matched with another from a similar industry, comparable betas for the londerlying stock, and the options were on the same expiration cycle. When possible four types of options were examined — a short and long term in the money option and a short and long term out of the money option. The analysis considered the 36 day period from 15 days prior to the announcement to 20 daj's following the announcement. The price and volume results for the underlying stocks were similar to the prior study — i.e., the major price change occurred on day -1 and day 0. The major volume change came on day 0. The relative option price results were quite consistent. The biggest price move always occurred on day -1. The price changes on day were mixed depending on the type of option. Tj^pically, the peak price occurred on day +3. Notably the subsequent price action continued strong and the option price series always ended the test period above the pre-announcement values. The relative volume results generally showed peak values on day with some tapering off afterwards. The investment results indicated very profitable opportunities. The cumulative percent changes that assumed purchases on day and day -1 were typically profitable, assuming a sale on day 3, and they were better when related to the matched sample. The results assuming a pur- chase on day -2 were always quite good. These resiilts are important because they provide a range of returns for aggressive investors who can buy before the close on day -1. -34- The final analysis examined the potential returns for an investor who acquired the "typical" option on day 0, -1, and -2 and paid the full commission. The results assuming the purchase of one contract generally indicated negative returns for day and day -1, due to the high fixed component of the commission. Assuming the purchase of two contracts, almost all the returns were positive and the range for the aggressive investor who could buy during day -1 were quite good — i.e., the mean of the range generally exceeded a 10 percent return in four days. An analysis of individual trades in the case of a purchase on day -1 and sale on day +3 indicated that typically most of the individual trades were prof- itable, but the major difference came because of the large positive gains and smaller losses — the range of gains always exceeded 100 percent while the losses range exceeded 70 percent. V Conclusion The results are encouraging for aggressive investors with the oppor- tunity to learn of split announcements early and acquire option positions quickly. Apparently the downside risk is minimal and the short-run returns can be quite good. Given the information requirements and the time constraints, this might be considered to be evidence against the semi-strong efficient market hypothesis. For the typical investor who would not buy options until day 0, the returns may still be slightly positive after commissions, but the risk may also be somewhat higher. M/E/249 lECKMAN IX INDERY INC. I§ JUN95 ,d.To-Ptel' N.MANCHESTER,