Expiry Fridays

Weeks with Expiry Fridays - Bullish or Bearish?

I admit that fspeculator cajoled me into doing this little bit of research and curiosity got the better of me and so I went off to see if it was true: Are weeks with Expiry Fridays more bullish than your average week in the markets?

Data

I thought that the most logical data set to use to confirm (or not) this question would be the DJIA cash index. This index isn't affected by rollover to a new contract as futures are and is constantly adjusted by a divisor to keep it representative of its constituents over time.

I took 990 weeks of data starting from 27 June 1986 to 10 June 2005. Of those 990 weeks, 75 of them were weeks with expiry Fridays.

Assumptions

I assumed that the third Friday of each quarter was an expiry Friday and did not check that against any expiry calendars.

Testing

The testing was very simple and I did it quickly by using Excel and a number of formula which I then converted to values in the spreadsheet.

The testing measured the gain (loss) made from the open on the first trading day of the relevant week to the close of the last trading day of the same week. The gain (loss) was measured as a percentage of the opening price. The price movement over the weekend (if any) was ignored. Another advantage to using the cash Dow index is that the opening price the following trading day is very close to the previous trading day's price because of the delay in updating the constituent prices when the market first opens.

Results

The results are very straight forward and simple.

The average gain for each week in the data set if ALL weeks are included is 0.205%

The average gain for the 75 weeks that ended in Expiry Fridays is 0.326%

The average gain excluding the 75 Expiry Friday weeks is 0.195%

Spreadsheet and Data

The spreadsheet with the results is in this zip file: ExpiryFridayWeeksTest.zip

Conclusion

As mentioned by fspeculator in the trading channel, weeks which have Expiry Fridays are generally more bullish than weeks that do not.