Kevin_in_GA
4,599 posts
msg #122830
- Ignore Kevin_in_GA |
2/13/2015 9:16:52 AM
Yes - the Pangolin was an oddly compelling animal (sort of the offspring of an illicit affair between an armadillo and an anteater). When I was trying to create a unique name for the system, this just popped into my head.
Hopefully folks are reviewing this and tracking its progress over the next few months.
|
agawfi
6 posts
msg #122838
- Ignore agawfi |
2/13/2015 1:09:44 PM
Kevin, just wondering since you have tried and traded ConnorsRSI pullback system and your Pangolin Z system, which of the two systems did you find more robust ? As I noticed that the Pangolin system having higher drawdowns lately.
Thanks !
|
jemarcks
26 posts
msg #122845
- Ignore jemarcks |
2/13/2015 9:59:29 PM
I have been testing several versions of this system and may have some results to post soon. I like the position sizing part a lot! Back testing the position sizing has been a challenge. Any idea?
Jeff
|
Kevin_in_GA
4,599 posts
msg #122846
- Ignore Kevin_in_GA |
2/14/2015 11:03:14 AM
The Van Tharp position sizing is essentially impossible to backtest properly - I have yet to find a software platform that can do variable position sizing like this. One alternative is to keep trade sizes the same and increase the number of positions as your equity grows, but that can't be done in SF (or SS) at this time either.
|
jemarcks
26 posts
msg #122848
- Ignore jemarcks |
2/14/2015 11:27:36 PM
It looks like you can backtest different position sizing algorithms in Tradestation. I have Tradestation and Ninja Trader platforms up an running. I use tradestation for my trading platform and I have been evaluating Ninja Trader which has the market replay feature that allows me to run live trades when I have time. Not just when the market is open.
http://www.tradestation.com/education/events/on-demand-webcasts/analysis-concepts/strategy-impact-pt-2
Jeff
|
mahkoh
1,065 posts
msg #122854
- Ignore mahkoh |
2/16/2015 12:39:57 PM
When using SF for backtesting there is an option to include ATR as an entry indicator. After running the backtest you can export the trades into excel and then calculate the position size.
|
jemarcks
26 posts
msg #122859
- Ignore jemarcks |
2/16/2015 3:05:43 PM
mahkoh,
That's fine, but I would rather have my strategy compute the position size and place the trade for me.
What position sizing algorithms do you consider usable?
Jeff
|
mahkoh
1,065 posts
msg #122860
- Ignore mahkoh
modified |
2/16/2015 5:42:17 PM
I know of only one approach: Decide your stoploss and decide the amount you're willing to risk. From there you calculate the number of shares to trade.
There are several ways to determine the stoploss, I personally like to use 2.5 * weighted atr(20).
What I meant with my suggestion is that it is possible that one of your versions consistently selects winning trades for low volatility securities. When applying position sizing this strategy would outperform significantly.
|
meisher
2 posts
msg #122900
- Ignore meisher |
2/21/2015 11:06:19 PM
Thanks for sharing this scan.
Quick question.... I wanted to try and reduce the number of stocks returned and try and find the ones that have the best chance of getting a win. I changed the limit entry to reverseRSI(2,1) instead of 5. I also set it to scan for RSI below 5 instead of 15. This will get me into a stock at an RSI(2) of 1 instead of 5. Do you agree that this would help improve performance? Or does it effect some fundamental principle of this strategy? Thanks!
|
Kevin_in_GA
4,599 posts
msg #122927
- Ignore Kevin_in_GA |
2/24/2015 9:03:53 PM
It will reduce the number of candidates - sometimes fewer trades looks better but a larger chunk of your money may sit idle if you only put 10% of your capital into any trade, and your actual equity return over time will be less.
|