TheRumpledOne
6,411 posts
msg #36077
- Ignore TheRumpledOne |
5/22/2005 1:58:18 PM
There is NO FINAL VERSION.
There are many versions that work.
Which one YOU use it up to YOU.
|
corsino
259 posts
msg #36078
- Ignore corsino |
5/22/2005 2:49:41 PM
I agree that the RSI(2)<1 criteria seems to be the best single indicator, but it, in itself, does not guarantee winning results. Backtest results depend a lot on the other parameters in the filter. There are filters using RSI(2)<1 that show losses in backtests. But that doesn't take away from its value.
|
corsino
259 posts
msg #36079
- Ignore corsino |
5/22/2005 3:34:53 PM
Actual Results.
In the short time I've been trying out RSI (2)<1 filters, I've completed six trades. Four winners, two losers. The winners were ALSC, MZ, ARDI and GTW. The losers were PHMD and PRCS.
Actually, I'm using just a few hundred dollars, making the trades thru BUY AND HOLD to keep commission costs down. There are some definite disadvantages to BUY AND HOLD though. There's only three "windows" during the day in which you can make trades, and you can only buy certain stocks they deal in. So sometimes you can't buy/sell what or when you want.
|
judgetrade
107 posts
msg #36146
- Ignore judgetrade |
5/30/2005 1:50:37 PM
I am pretty new to the comunity, but let me say:
There is no reason why somebody should stop posting about a great indicator, that catches a great principle (overreaction, panik selling!) that has worked and will work! Only if you belive in (by heart and brain) you system (or indicator) you can trade it!
It encourages me a lot and I say thank you!
I took me some weeks to find my filter based on the rsi indicator and backtest it with SF. But I know how long it usually takes to design a system!
Rump and others are laying out a potential living here(!) If you take it as a start and make a lot of homework! It saves you years of work!
So be nice!
Andreas alias Judgetrade.
|
wallman
299 posts
msg #36147
- Ignore wallman |
5/30/2005 10:05:36 PM
when developing my method i spent over 2500 hrs over a 9 month period charting every under $10 stock in the universe i could find to try to come up what would work for me,mind you,this WASN'T using SF for bascktesting,this was as it was happening,using SF for backtesting stuff on various filters can look real good or bad until you actually break each stock down individually,a ton of work but if you want it as your living,you gotta want it bad enough to put some work into it
|
burzin
3 posts
msg #36183
- Ignore burzin |
6/3/2005 4:22:00 AM
I was playing with the Weekly RSI(2) filter and fine tuning it simply blasts away any previous returns I have seen so far on any other system, and this isnt with micro cap stocks either. Whoever came up with RSI(2), major props to you.
Just short and close after 5 days.
I get a 220% annualized return and thats not even trading microcaps and an 80% reliability rate!!!
The problem with the RSI above 99.7 is that it only gets a few picks a year.
I was studying Weekly RSI(2) some more, and I notice that whenever the indicator does a sharp dip after a rise, it serves as a very effective early warning system. I am unable to get the filter to pick the exact spots I want though. This is the closest I have come up with.
If anyone can help improve it, it would be a major plus. I want it to pick up stocks as soon as their Weekly RSI(2) breaks the uptrend and starts going down.
This is a great site!
|
burzin
3 posts
msg #36185
- Ignore burzin |
6/3/2005 4:55:14 AM
Found the original post. It was mary4ever who came up with the Weekly RSI(2) filter here - http://www.stockfetcher.com/stockdb/fetcher?p=forum&sub=view&fid=1002&tid=36081.
I fine tuned it so it gives me exact entry points and it works pretty well.
Now if I can just figure out what criteria is to pick up Weekly RSI(2) as soon as it enters a downtrend...I tried using slope of, but that doesnt work either.
As an example of what I mean, look at the chart of MRVL for 25 April 05. Even though the price went up that day, the Weekly RSI(2) dropped, and the stock gave some good gains for shorts over the next 3 days. How do I pick that point using Stock Fetcher?
Thanks for your time people. Any help is much appreciated.
|
brokedown
21 posts
msg #36188
- Ignore brokedown |
6/3/2005 8:04:01 AM
I sugjest you read how the weekly rsi is calculated. See the thread "Weekly indicator flaw" under the comments and bugs section. Weekly indicators are a floating calculation and are only "weekly" on saturday. If you are going to use them on a daily basis, I would urge extreme caution.
Tom
|
shelupinin
120 posts
msg #36189
- Ignore shelupinin |
6/3/2005 8:38:00 AM
Hi !
I really like RSI... Just an example of good filter which using RSI:
Try to backtest it with 1% "stop loss", no "profit stop" and "Maximum holding days" equals 1. You'll see about 300% ROI... and it's for filter which uses stock over 5 USD price, so you can use margin at you broker...
|
burzin
3 posts
msg #36191
- Ignore burzin |
6/3/2005 1:11:53 PM
Brokedown, thanks for the post. I am going to paper trade for a few weeks just by eye balling charts with Weekly RSI(2) set ups right now to see how it fares in the real world. My broker offers a real time simulation with commision costs, slippage etc, so it should give a pretty real vision.
I still havent figured out how to program SF to pick up the Weekly RSI(2)downtrend on the first day though. Any ideas on that are appreciated.
Shel, I backtested your indicator and it certainly gives the return promised. Very impressive. I guess the actual returns are going to be lower due to commission costs since you are going to get stopped out most of the time because of the 1% threshold. I am going to play with it to see if I can get the reliability up. I will let you know what I come up with.
|