mahkoh
1,065 posts
msg #108225
- Ignore mahkoh |
10/1/2012 4:05:59 AM
Kevin, wondering to what extent the returns of these filters were generated by outperforming the index during the bear market. Can you provide some info on how they did during the last 3 year bull market?
As for the composite filter, it does indeed not seem like there is any obvious edge to be found in there.
|
Kevin_in_GA
4,599 posts
msg #108226
- Ignore Kevin_in_GA
modified |
10/1/2012 8:19:48 AM
Agreed - the composite filter performs at a slightly lower level than several of the individual ones. That is mostly because each individual filter exits on the same day, and that the batting averages are unusually high for these guys, even if the gains are modest.
That is why I posted each filter separately. I might want to look at divergence-based exits as well.
|
jackmack
334 posts
msg #108267
- Ignore jackmack |
10/3/2012 4:31:31 PM
Kevin - again thank you so much for sharing these - I like the returns.
I am just having a hard time figuring out the entry and exit signals.
Is there any way to make them a bit easier to understand?
There appear to be long times where neither signal is given and I am not sure what one should have done during these periods.
I am just making a simple observation so please do not read this as a criticism in any way - you know what you are looking for and at whereas I am not really sure what to do when for example the enter and exit signals occur together.
I am just wondering if there might not be a simpler entry and exit signal.
Thanks for all of your hard work and sharing here.
|
Kevin_in_GA
4,599 posts
msg #108269
- Ignore Kevin_in_GA |
10/3/2012 8:27:17 PM
There are only two signals - the entry and the exit. Right now you have had entry signals on all 10 filters, and are in trades. Once the exit signal triggers you are out the next day.
When you get an entry signal (essentially a divergence between price and the indicator) that can persist as long as the divergence persists, but the signal does not stay "on" during the whole time in trade.
|
jackmack
334 posts
msg #108276
- Ignore jackmack |
10/4/2012 8:00:55 AM
Kevin
I get it now - I see it plain as day now - sorry.
WOW that is amazing and if you load SSO in using the STOCH filter it has you in at 61.79 on 1 OCT 12.
Very interesting. Have you run a backtest on SSO/SDS using either the RSI or STOCH divergence filters
to see if performance is any better than your original SSO/SDS filter?
Thank you again - man this filter set is amazingly dynamic.
|
mdmink
8 posts
msg #108289
- Ignore mdmink |
10/4/2012 6:11:10 PM
Kevin, thank you for sharing and great work as always. I know that you mentioned that you did all the testing in Stratasearch , and from previous threads you prefer that platform for back-testing. I was wondering 1) if you had backtested either IWM or QQQ? If you think that stockfetcher could successfully backtest those using one of these filters? Would you expect much difference, or given the high correlation with the spy, expect the near-same results? Thanks
|
jackmack
334 posts
msg #108291
- Ignore jackmack |
10/5/2012 11:25:05 AM
Kevin have you looked into trying to create these divergences in Prodigio?
|
Kevin_in_GA
4,599 posts
msg #108297
- Ignore Kevin_in_GA
modified |
10/5/2012 2:55:10 PM
There is an exit triggered today. Sell at end of day or at the open on Tuesday.
Note that all 10 filters close their trades profitably.
EDIT: I am out at SPY 146.00. Looks like the market now is trading negative for the day, which removes the exit signal (today must be a 10 day high on the close).
|
mahkoh
1,065 posts
msg #108301
- Ignore mahkoh |
10/5/2012 5:56:05 PM
Didn't the exit signal trigger Thursday?
|
Kevin_in_GA
4,599 posts
msg #108302
- Ignore Kevin_in_GA |
10/5/2012 9:59:12 PM
You're right, it did (I have been traveling for work and not checking this filter until today). Ether way the filters delivered a profitable trade.
|