Option trading began in US on CBOE in 1973 with daily volume of 6,470 contracts. Index trading began in 1982 with 41,389 contracts. By 1986 option trading volume has exploded to an average daily of 1,143,127 and peaked in 1987 at 1,206,201 following the crash of 1987. It took a decade for option trading volume to exceed the 1987 high when 353,825,118 contracts traded, or 1,398,510 on a daily basis in 1997. By 2006 annual option volume exceeded 2 billion contracts and is expected to grow by 40% to 2.8 billion in 2007.
The table below illustrates the explosive growth in option trading.
The attached table summarizes data for each OpEX trading day since 1980. There have been 323 such OpEX days since 1980.
The first column is a summation of total OpEX day averages for each year; the second column shows total average returns from the preceding Wednesday; and finally the third column shows summation of average returns from the whole week preceding OpEX day. As a total, OpEX day shows a negative return of -4.61% during the last 27 years. While Wednesday preceding OpEX shows a cumulative loss of -10.49% and the week preceding OpeX shows a small gain of 24.61%. Since this data represents 323 events, the immediate conclusion that can be reached is that there is no profitable strategy involved in holding equities during those periods. For reference only, the last column in the table shows actual return for each year , which adds up to 298.6 %. Please note that is a summation of annual averages not to be confused with compounded return on the DOW of 1287% from 1/1/1981-11/30/2007.
In examining the role that Wednesday plays in OpEX , of the 323 option expirations, 167 moved in the same direction as the OpEX while 156 moved in the opposite direction--that is about 50% each way. The schedule, Summation of Averages illustrates that regardless of how Wednesday preceding OpEX turns out, expecting OpEX day to return profit always results in break even(there was a trivial cumulative loss of -3.9% during 165 events) on OPEX when Wed was lower for the day.
Additionally, the week prior to OpEX is a stunning under performer by an impact of 3X when compared to other weeks during the same period. An average week is expected to return about 0.21%, while weeks prior to OpEX have only returned 0.08%.
While it is not going to be profitable to follow a mechanical system based on the direction of Wed preceding OpEX, there were some profitable exceptions to this rule. One of the more profitable historical returns could be achieved by simply buying OpEX during December when Wed was positive for an average return of 0.7%. It would almost be as profitable to short every October when Wed was positive.
It was also very profitable to short OpEX during bear markets, in fact, some of the biggest daily returns could be had by shorting OpEx during bear markets such as 1987 and 2002, but, I suspect that seasonality could also play a role.
Edited by linrom1, 08 December 2007 - 03:51 PM.
Thanks again.
