Date - Model Output - Closing price of TNA (my etf of choice) when the signal was generated
Oct 01 - Short - 47.82
Oct 04 - Long - 45.91
Oct 05 - Hold
Oct 06 - Hold
Oct 07 - Hold
Oct 08 - Short - 50.79
Oct 11 - Hold
Oct 12 - Long - 51.34 (miss)
Oct 13 - Hold
Oct 14 - Hold
Oct 15 - Hold
Oct 18 - Short - 54.47
Oct 19 - Long - 50.32
Oct 20 - Short - 52.57
Oct 21 - Hold
Oct 22 - Short
Since Weds, we've had 2 short signals w/ no new buy signal, so I added more shorts Friday. I dislike holding shorts through the weekend since Mondays are usually positive, but I'm a slave to my system. Back-testing my model back to 2008 shows a 74% winning percentage overall. September had 7 winners out of 8 for a very nice 87% winning percentage. It is 5-1 this month (83%). The 1 loss was minimal. The short trade from Weds is currently wrong (TNA closed fri @ 52.74). We'll see what Monday brings. Good luck all, and thanks for reading!
Hey, I followed a link to your site. I'm curious about your setup. What software are you using to generate the signals? What did you use to run the back-test? Also, why did you only back-test to 2008? Data should be available to back-test further...do the results change?
ReplyDeleteskyace,
ReplyDeleteI'm simply using an open office spreadsheet which ingests RT market data every 60 seconds. The model is based heavily on SPX price action but also on relationships between the SPX, RUT, VIX, and COMPQ.
Backtesting is all done in excel...er, open office (I dislike MSOffice'07, but that's another post). I could only find free daily data back to 2008. I have not transitioned to a decent trading platform yet. I plan on switching to tradestation in 2011. Until then, I'm at the mercy of what I can find on the internet. Obviously, I haven't had much luck finding data on my own. Can you offer any suggestions? I would love to back test further. Thanks.
Ah, that's understandable. Sorry, I don't know of any free sources. A paid one is backscanner, but I don't know if that platform is suitable for your needs.
ReplyDelete