If you move in certain circles in the gold investing community, I bet you've heard it said: "Gold Gets Taken Down 95% Of The Time On Unemployment Numbers Day".
Well, I decided to figure out if it is true. If it is, its definitely tradeable and you could easily pull down a 6-figure income trading gold options just one day a month. Nice work if you can get it.
Here's what I did to test this theory. I:
- Downloaded a couple years of the daily results for the Gold ETF, GLD (open, high, low, close).
- Got the Non-Farm Payroll Dates for the last 2.5 years from the Bureau of Labor Statistics.
- Computed the results of a "going short GLD at the close on the day before" and exiting:
- At the close of Non-Farm Payroll Day (a Friday).
- At the close of the next trading day after Non-Farm Payroll Day (usually Monday).
- At the low of the Non-Farm Payroll Day to see what the result would be if you were a day-trading genius.
- At the open (9:30, about an hour after the numbers are released) of the Non-Farm Payroll Day.
Here's the results I got in tabular form (click here: Download 20100805gldunemploymenttrade for the spreadsheet):
The results really don't look that good. Please let me know if I made a blunder in the spreadsheet. Here's what I see in the tabular results:
- They don't look tradeable to me. The average gain is way to small relative to the maximum loss.
- Actually, for the close-to-close trade results in a winner only 52% of the time so at best you can say "THEY TAKE IT DOWN 52% OF THE TIME ON UNEMPLOYMENT DAY". That is not going to impress the CFTC or the SEC or anybody.
- To my eye, the close-to-open trade looks the best. The winnings are meager, but there aren't really many substantial losses. Here it is in graphical form.
The maximum loss is only around 2% (which is acceptable). That was in early September of 2008 when all hell was visibly breaking loose (AIG going belly up after Fanny, Freddie and Bear Sterns). The problem is that the average gain is 1/10th that at .2%. That really doesn't look tradeable.
Now, looking again I see that this trade was a winner for every month in 2009 except one. At that point in time I guess you could say that gold got hit 95% of the time. Before then and after that it has been a mixed bag. I'd say its a complete crap-shoot for tomorrow.
Well, another gold community bit of folk-wisdom turns out to be pretty questionable when looked at empirically. Do you think this calls into question the whole gold intervention theory? I would say it doesn't for me, but it makes me a little less convinced than I used to be and makes me want to back-test more of that kind of folk-wisdom.