StockFetcher Forums · Filter Exchange · New here: What is your screen you bet money on?<< 1 2 3 4 5 ... 6 >>Post Follow-up
judgetrade
107 posts
msg #35349
Ignore judgetrade
3/14/2005 3:14:12 PM

Hi, I am new here, nice site, nice screens!

I am trading two systems right now: mid term relative strenght (www.pitbull.com)
and short term realative weakness (a system that I have programmed with fasttrack that ranks the most oversold stocks every day, buys them, hold for 4 Days an then reranks.)

So I base my strats on stuff that has been researched by akademics (the newer onces, not the guys who still defend the efficient market hypothesis):

(Korajczyk, R.O., Sadka, R. (2003), http://www.afajof.org/Pdf/forthcoming/Korajczyk_Sadka.pdf):

"There is a growing literature on the predictability of stock returns
based on the information contained in past returns. At very short
horizons, such as a week or a month, returns are shown to have
negative serial correlation (reversal), while at three to twelve
month horizons, they exhibit positive serial correlation (momentum).
During longer horizons, such as three to five years, stock returns
again exhibit reversals."

From the billion screens I went through I like the following the best so far:

offset is 5
show stocks where price is above .2 and 3
and RSI(2) is below 1
and volume > 1000000

__________________________________

offset 15
Show stocks where
the High is less than 25 percent above the bottom linear regression line(60,1.25) and the 21 day slope of the close is above .25 and
close is above close 1 day ago and price is between 3 and 12 and
offset 0 days ago


_________________________________________________


So here is my question:

What is the screen you bet sucessfully bet your money on?
Thank you.

Best Regards

Andreas



judgetrade
107 posts
msg #35363
Ignore judgetrade
3/15/2005 4:36:45 PM

I tested the following screen with a similar code on fasttrack from 1997 - today:

offset is 1
show stocks where price is above .2
and RSI(2) is below 1

Results (no trading costs applied, but usually such strats are very
robust to trading costs):

LastValue 1405727, AnnRet 88.7%, Mdd 74.0% on 03/18/03 [allways in the market]
LastValue 651169, AnnRet 70.9%, Mdd 32.2% on 09/17/97 [only in the market when
the nasdaq trends up]


Since the original screen adds the following line

"and volume above 1000000"

the results should be even better, if you assume that high volume
are prefered by mutal funds.


Here the Fasttrack (www.fasttrack.net) Code:
It ranks the stocks by the lowest RSI and buys the lowest 5 holds 4 days and then reranks.

[Expression]

fam = |0|.fam

STTABUY = (Rsi(fam, 2) *-1)
Save (STTABUY)

[ExprFamilyTrade]
Family = STTABUY
StartDate = 06/01/97
StopDate = 06/01/05
MMarket = RYMXX
Positions = 5
BuyMinRank = 5
HoldMaxRank = 5
HoldMinDays = 4
Signal = buyandhold
FnuFile = 5554|0|BH

[ExprFamilyTrade]
Family = STTABUY
StartDate = 06/01/97
StopDate = 06/01/05
MMarket = RYMXX
Positions = 5
BuyMinRank = 5
HoldMaxRank = 5
HoldMinDays = 4
Signal = momentumft
FnuFile = 5554|0|MFT

Some Data from the result file:

[...]

4 USTR 12/17/04 12/21/04 4 32 3 -0.8 70040
5 GREY 12/17/04 12/21/04 4 17 1 6.9 87996
2 WSFS 12/17/04 12/22/04 5 60 1 -0.5 96730
3 SONS 12/17/04 12/23/04 6 1580 2 4.3 76360
2 ROG 12/22/04 12/28/04 6 606 2 1.1 97780
3 JILL 12/23/04 12/28/04 5 47 3 -2.0 74805
4 ALX 12/21/04 12/28/04 7 19 3 -1.2 69223
5 EAGL 12/21/04 12/28/04 7 599 5 2.5 90223
1 AFP 12/20/04 12/30/04 10 684 1 -1.4 71134
1 CBMD 12/30/04 01/03/05 4 172 1 -2.1 69675
2 WTFC 12/28/04 01/03/05 6 371 2 -2.8 95091
3 DCO 12/28/04 01/03/05 6 1103 3 4.5 78153
4 WST 12/28/04 01/03/05 6 1134 4 8.5 75133
5 IDEV 12/28/04 01/03/05 6 1492 5 -4.0 86614
1 INSU 01/03/05 01/07/05 4 1312 1 -2.0 68260
2 AVTR 01/03/05 01/07/05 4 624 1 -2.7 92510
3 LAVA 01/03/05 01/07/05 4 1472 1 -3.7 75293
5 HGIC 01/03/05 01/07/05 4 1223 4 -3.6 83481
1 DEL 01/07/05 01/11/05 4 699 2 4.0 70972
2 LFG 01/07/05 01/11/05 4 577 3 1.8 94159
3 AXE 01/07/05 01/11/05 4 534 3 -0.9 74600
4 FCNCA 01/03/05 01/11/05 8 140 1 -7.0 69858
5 TMP 01/07/05 01/11/05 4 357 5 4.8 87482
1 SWKS 01/11/05 01/14/05 3 726 1 0.6 71398
2 KDN 01/11/05 01/14/05 3 632 2 1.8 95863
3 PCH 01/11/05 01/14/05 3 367 3 -0.6 74178
4 SURW 01/11/05 01/14/05 3 29 4 -2.6 68063
5 QSFT 01/11/05 01/14/05 3 1148 2 4.0 90981
1 RYMXX 01/14/05 Holding 59 1672 1672 0.3 71580
2 RYMXX 01/14/05 Holding 59 1672 1672 0.3 96109
3 RYMXX 01/14/05 Holding 59 1672 1672 0.3 74368
4 RYMXX 01/14/05 Holding 59 1672 1672 0.3 68237
5 RYMXX 01/14/05 Holding 59 1672 1672 0.3 91214

If you have a simple screen, let me know if you want me to backtest it.
Only simple stuff, because you can not programm everything in
fasttrack.

Regards

Andreas (forgive my bad english, its late in germany so I do not spell check it).




marine2
829 posts
msg #35374
Ignore marine2
3/16/2005 8:48:48 PM

Ask that same question to TheRumpledOne and I am sure he will tell you all about his RSI(2) filters' abilities to do wonderful things for you.


corsino
259 posts
msg #35568
Ignore corsino
4/11/2005 12:46:33 AM

As I posted elsewhere, I ran the same backtest on about 25 filters I had compiled from various sources. My purpose was to eliminate some of them from my scans and save time. Although other filters may perform better at other times, these two backtested the best at this time :

RSI(2) Filter :

Stocks where RSI(2)<1 and
average volume (10) is above 300000 and
close is between 0.5 and 3 and
close more than 10% below close 1 day ago

>20% DECREASE :

Stocks where close dropped more than 20% over the last 1 day
and average volume (10) is above 100000 and
close is between 0.5 and 3

Of course, this was an entirely mechanical backtest, buying the complete scan, using no judgement, and holding for one-week periods. It is entirely possible, that other scans in the hands of skilled traders would perform better.




corsino
259 posts
msg #35569
Ignore corsino
4/11/2005 1:17:24 AM

Perhaps I should have explained the backtest again.
1. For each of 25 filters, I used offsets of 68, 58, 48, 38, 28, 18, and 8.
2. At each offset, I recorded the performance for a 1-week holding period.
3. I then added the total Percentage gain or loss for the seven offsets.

The RSI(2)<1 scan total was approx + 57%
The >20% loss scan total was approx + 46%


leaddog
38 posts
msg #35571
Ignore leaddog
4/11/2005 7:13:44 PM

corsino,

Did you just use the stock fetcher performance feature, no entry criteria?


corsino
259 posts
msg #35572
Ignore corsino
4/11/2005 11:56:48 PM

leaddog

Yes, I just used Stockfetcher. I took one filter at a time, offset it 68, 58, 48, 38, 28, 18, and 8. At each offset, I recorded the performance percentage after 1 week. If there were no stock picks at an offset, I just recorded a 0. I then added the percentages recorded for that filter. I did the same for 25 filters.
The offsets I used were arbritary, as was the 1-week performance period. The price range of 0.5 to 3 was also arbitrary, although I was interested in low-priced stocks.
I wasn't trying to prove that one filter is better than another, it was out of curiosity to eliminate some from my list. I'll probably do the same periodically, to see if there are any filters that consistently do well.


corsino
259 posts
msg #35573
Ignore corsino
4/12/2005 12:22:22 AM

leaddog

I just realized that I didn't answer your question fully. No, I did not use any additional entry criteria . At each offset that I got a list of scanned stocks, I just clicked on the "Performance" header to see the results for the entire list.



shelupinin
120 posts
msg #35575
Ignore shelupinin
4/12/2005 5:16:12 AM

Hi, judgetrade !

I bet my money on following filter
Fetcher[
volume gained more than 200% over the last 1 day
and price dropped more than 5% over the last 1 day
and price is above 5
and volume is above 100000
]


that filter not far from you RSI(2) < 1 ... I add stocks which matched my SF filter to "watchlist" at my online broker optiosxpress.com and then monitor price movement at the beginning of trade day, if price is going up then I quickly buy that stock...


leaddog
38 posts
msg #35582
Ignore leaddog
4/12/2005 9:57:46 PM

corsino,

I took your exercise another step further with the RSI(2) <1 filter.

I went back 70 days and using the stock fetcher performance feature checked everyday up to today buying and holding for a week. It came out to 129.59% gain!

I wonder if there is something I am missing....


StockFetcher Forums · Filter Exchange · New here: What is your screen you bet money on?<< 1 2 3 4 5 ... 6 >>Post Follow-up

*** Disclaimer *** StockFetcher.com does not endorse or suggest any of the securities which are returned in any of the searches or filters. They are provided purely for informational and research purposes. StockFetcher.com does not recommend particular securities. StockFetcher.com, Vestyl Software, L.L.C. and involved content providers shall not be liable for any errors or delays in the content, or for any actions taken based on the content.


Copyright 2016 - Vestyl Software L.L.C.Terms of Service | License | Questions or comments? Contact Us
EOD Data sources: DDFPlus & CSI Data Quotes delayed during active market hours. Delay times are at least 15 mins for NASDAQ, 20 mins for NYSE and Amex. Delayed intraday data provided by DDFPlus