hammertime17
1 posts
msg #59510
- Ignore hammertime17 |
2/6/2008 7:51:09 PM
Hi there,
I am new to the Stock Fetcher forum and am posting this short stock filter to see if anyone has any suggestions for improvement:
Upper Bollinger Band(20,2) below Close
and RSI(15) above 80
and Average Volume(90) above 50000
and Close above 1
I have back tested this bearish strategy every year since 2002 and the annual ROI averages around 35% per year (not bad!) during a great bull market from 2002 to 2007. The concern I have is that this filter shows ~2500 trades per year, and even with a deep discounted broker, these commissions can add up significantly!
---Can anyone suggest any improvements to this short filter (5% stop loss, 10% profit exit, or 20 days) to reduce the number of trades and increase the win/loss ratio?
Thanks in advance!
Mike
|
nikoschopen
2,824 posts
msg #59515
- Ignore nikoschopen |
2/6/2008 9:50:30 PM
Here's the result of what you would see after 3 weeks or 15 days. Feel free to change the number to suit ure needs.
|
nikoschopen
2,824 posts
msg #59516
- Ignore nikoschopen |
2/6/2008 10:36:19 PM
Okay, perhaps this is more pertinent to ure question. The date offset of 10 is included to see how the stocks fared after 10 days. However, it appears that you would do well by buying at the close and selling at the following day's open.
Assuming that you have ample experience with previous bear markets, have you actually held a position for 3-4 weeks? You must have some real nerve of steel to sit through a volatile environment like now, Moreover, there's no guarantee that the damn stock(s) will dive for the next 3 weeks.
BTW, nice nick. When will we have our own "Milli Vanilli"? {8^)
|
maxreturn
745 posts
msg #59524
- Ignore maxreturn |
2/7/2008 10:25:16 AM
Mike, my guess is you would see improvement in the win:loss ratio if you only shorted stocks that were below the 200 period ema. Another thing I like to use in my filters is to only short stocks with a 1 year relative strength below that of the SPX. The following code will ensure that you only return stocks which are weaker than the SPX on a one year basis:
set{x, ind(^spx,close)}
set{y, x divided by x 252 days ago}
set{z, close divided by close 252 days ago}
set{rs, z divided by y}
show stocks where rs below 1
Regards
Max
|
makabootycall
11 posts
msg #59525
- Ignore makabootycall |
2/7/2008 11:07:45 AM
I have found that adjusting and experimenting with different stop losses give nice results, you could try raising it, I know in theory you may not want to lose more than 5%, but there may be a substantial number of stocks who lose 5% and then move into profit territory. If it raises your ROI, and holds over different time periods, then who cares?
|
nikoschopen
2,824 posts
msg #59528
- Ignore nikoschopen |
2/7/2008 12:00:32 PM
I know in theory you may not want to lose more than 5%, but there may be a substantial number of stocks who lose 5% and then move into profit territory. If it raises your ROI, and holds over different time periods, then who cares?
The problem with ure approach is that some may never come back. Witness the 2000-03 bear market. Traders who neglected the simple principle of risk management ended up losing their shirt because they had no willpower to cut their losses early. If they can't handle a 5% loss, what makes you think they'll bail out when they're down 20%?
|