financialDataset.txt

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. 

INVESTING  IN  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 
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 
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 
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). 

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 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. 


SAMPLE  STOCKS,  ANNOUNCEMENT  DATES,  AND  MATCHED  SAMPLE  STOCKS 

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 

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. 

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 

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  0  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 
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  0  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 

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  0  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. 

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  0  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  0  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 
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  0  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  0  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 


MEAN   CROSS    SECTION   OF   OPTION  VOLUME 

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 
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. 


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. 


MEM  CUMULATIVE  PERCENT  PRICE  CHANGES  FOR  SPLIT  OPTIONS  UNADJUSTED  AND  ADJUSTED 
TO  MATCHED  OPTIONS  ASSUMING  A  PURCHASE  AT  THE  CLOSE  OF  DAY  MINUS  TWO 

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"  Investment 

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. 

COSTS  AND  RETURNS  FOR  AN  INVESTOR  TsfHO  ACQUIRED  ALTERNATIVE 
OPTION  CONTRACTS  AND  SOLD  ON  DAY  THREE 

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  0  (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  0  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. 

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. 

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  0  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  0  with  some  tapering  off  afterwards. 

The  investment  results  indicated  very  profitable  opportunities. 
The  cumulative  percent  changes  that  assumed  purchases  on  day  0  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. 


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  0  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. 


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.