StockFetcher Forums · Filter Exchange · TRO's STAT SCAN for SWING TRADERS<< >>Post Follow-up
johncoxme
12 posts
msg #124412
Ignore johncoxme
7/18/2015 7:12:04 PM

Hello SF friends,

New member

I have seen some amazing filters on this site.

I am not a coder but I like stats, and found TROís STAT SCAN for SWING TRADERS fascinating. I know that Kevin did some work with it too.

Basically it shows the counts (weeks) in a year that a stock gains 1%, 2%, 5%.

Since I use a stop of 10% I added the count of weeks the stock dropped 10%.

I also included the %gain and %loss counts for the last month, since recent performance is important.
Forgive me if someone has already written a filter that does this.

I added 2 columns ďaverage % gain on 5% sell limitĒ and ďaverage % loss on 10% sell stopĒ

Instead of simply going with stocks that get the most 5% gains in the year (like JDST this week), I think accounting for negative volatility (risk) is important. JDST, for example, is the 5% king hitting 5% a whopping 51/52 times in the year. So why not throw all your money in JDST gold and make a fortune week after week? Probably because itís not guaranteed. On Tuesday you might just wake up to find gold up big and JDST down 20% (it is overbought). Considering the fact that JDST hits -10% 27/52 or about half the time, for someone like me who uses a 10% stop, this would not be a good pick, since I might get stopped out before it makes its habitual 5% gain.

A typical stock in this amazing filter gives 5% gains a median of 60% of the time yielding (.05*.60) 3% average weekly gains.

A typical stock shows 10% losses a median of 15% of the time (the standard deviation in loss count is nearly double for 5% gain counts!) for an average weekly loss of 1.5%

(If your stock doesnít hit those stops letís assume 0 gain to be conservative).

Thatís a 3%-1.5% = 1.5% weekly gain, and on a $100000 investment, a steady $1500 per week (excluding commissions and exponential gains)!

Not bad. (Is there something wrong with my reasoning?)

So this is how I would use the filter since I use a 10% stop loss and 5% stop limit. Simply buy the stocks where the average weekly gain/loss difference is greatest (should be easy adding this code). I definitely would stay away from stocks like JDST since its average weekly gain (4%) is less than its average weekly loss (5%).

Take care traders,

John


/* TRO STAT SCAN for SWING TRADERS - use only on Saturday and Sunday */
set{whiop, weekly high - weekly open}
set{whilowop, weekly open - weekly low}
set{Long_Profit, whiop / weekly open }
set{wkProfitPct, 100 * Long_Profit }
set{Long_Loss, whilowop / weekly open }
set{wkLossPct, 100 * Long_Loss }
set{onepct, count(Long_Profit > .01 , 52)}
set{twopct, count(Long_Profit > .02 , 52)}
set{fivepct, count(Long_Profit > .05 , 52)}
set{triggered, count(Long_Profit > .02 , 1)}
set{minus10pct, count(Long_Loss > .10 , 52)}
set{five%, fivepct/52}
set{expgain, .05 * five%}
set{loss%, minus10pct/52}
set{exploss, .10 * loss%}
DRAW TRIGGERED
set{onepcttriggered, count(Long_Profit > .01 , 4)}
set{twopcttriggered, count(Long_Profit > .02 , 4)}
set{fivepcttriggered, count(Long_Profit > .04 , 4)}
set{minus10%triggered, count(Long_Loss > .10 , 4)}
add column wkProfitPct
add column onepct {hit 1% target over last 52 weeks}
add column twopct {hit 2% target over last 52 weeks}
add column twopct {hit 2% target over last 52 weeks}
add column fivepct {hit 5% target over last 52 weeks}
add column minus10pct {hit -10% over last 52 weeks}
and add column separator
add column onepcttriggered {hit 1% target last 4 weeks}
add column twopcttriggered {hit 2% target last 4 weeks}
add column fivepcttriggered {hit 5% target last 4 weeks}
add column minus10%triggered {hit -10% target last 4 weeks}
add column expgain {average % gain on 5% sell limit}
add column exploss {average % loss on 10% sell stop}
and add column separator
and add column weekly open {wopen}
and add column weekly high {whigh}
and add column weekly low {wlow}
and add column weekly close {wclose}
and add column separator
close is above 5
average volume(90) above 500000
TWOPCT ABOVE 44
sort column 7 descending
chart-display is weekly






compound_gains
152 posts
msg #124417
Ignore compound_gains
modified
7/19/2015 8:25:44 AM

Fetcher[/* TRO STAT SCAN for SWING TRADERS - use only on Saturday and Sunday */
set{whiop, weekly high - weekly open}
set{whilowop, weekly open - weekly low}
set{Long_Profit, whiop / weekly open }
set{wkProfitPct, 100 * Long_Profit }
set{Long_Loss, whilowop / weekly open }
set{wkLossPct, 100 * Long_Loss }
set{onepct, count(Long_Profit > .01 , 52)}
set{twopct, count(Long_Profit > .02 , 52)}
set{fivepct, count(Long_Profit > .05 , 52)}
set{triggered, count(Long_Profit > .02 , 1)}
set{minus10pct, count(Long_Loss > .10 , 52)}
set{five%, fivepct/52}
set{expgain, .05 * five%}
set{loss%, minus10pct/52}
set{exploss, .10 * loss%}
DRAW TRIGGERED
set{onepcttriggered, count(Long_Profit > .01 , 4)}
set{twopcttriggered, count(Long_Profit > .02 , 4)}
set{fivepcttriggered, count(Long_Profit > .04 , 4)}
set{minus10%triggered, count(Long_Loss > .10 , 4)}
add column wkProfitPct
add column onepct {hit 1% target over last 52 weeks}
add column twopct {hit 2% target over last 52 weeks}
add column twopct {hit 2% target over last 52 weeks}
add column fivepct {hit 5% target over last 52 weeks}
add column minus10pct {hit -10% over last 52 weeks}
and add column separator
add column onepcttriggered {hit 1% target last 4 weeks}
add column twopcttriggered {hit 2% target last 4 weeks}
add column fivepcttriggered {hit 5% target last 4 weeks}
add column minus10%triggered {hit -10% target last 4 weeks}
add column expgain {average % gain on 5% sell limit}
add column exploss {average % loss on 10% sell stop}
and add column separator
and add column weekly open {wopen}
and add column weekly high {whigh}
and add column weekly low {wlow}
and add column weekly close {wclose}
and add column separator
close is above 5
average volume(90) above 500000
TWOPCT ABOVE 44
sort column 7 descending
chart-display is weekly
]



Hmmm...simplistic approach, but if you traded the top three stocks/ETFs each week you'd be up some 35% YTD (before slippage, commissions and all of the other little things trading live). The syntax below is a 1-week look back and it's sloppy coding as I didn't take the time to clean it up properly.

Fetcher[
/* TRO STAT SCAN for SWING TRADERS - use only on Saturday and Sunday */
set{whiop, weekly high - weekly open}
set{whilowop, weekly open - weekly low}
set{Long_Profit, whiop / weekly open }
set{wkProfitPct, 100 * Long_Profit }
set{Long_Loss, whilowop / weekly open }
set{wkLossPct, 100 * Long_Loss }
set{onepct, count(Long_Profit > .01 , 52)}
set{twopct, count(Long_Profit > .02 , 52)}
set{fivepct, count(Long_Profit > .05 , 52)}
set{triggered, count(Long_Profit > .02 , 1)}
set{minus10pct, count(Long_Loss > .10 , 52)}
set{five%, fivepct/52}
set{expgain, .05 * five%}
set{loss%, minus10pct/52}
set{exploss, .10 * loss%}
DRAW TRIGGERED
set{onepcttriggered, count(Long_Profit > .01 , 4)}
set{twopcttriggered, count(Long_Profit > .02 , 4)}
set{fivepcttriggered, count(Long_Profit > .04 , 4)}
set{minus10%triggered, count(Long_Loss > .10 , 4)}

add column fivepct {hit 5%} 5 days ago
add column minus10pct {hit -10%} 5 days ago
close 5 days ago > 5
count(volume > 250000, 21) 5 days ago equals 21
set{chkhgh, count(high 5 day high more than 4% above open 4 days ago, 1)}
set{chklw, count(low 5 day low more than 9.99% below open 4 days ago, 1)}
set{chk, fivepct / minus10pct}
add column chk 5 days ago
add column chkhgh {+5_%}
add column chklw {-10_%}
chklw < 1
sort column 7 descending
set{hgh, open 4 days ago * 1.0}
set{lw, open 4 days ago * .9}
draw price line at lw on plot price
draw price line at hgh on plot price
set{cl, close / open 4 days ago}
add column cl {weeks close}
]



sohailmithani
192 posts
msg #124421
Ignore sohailmithani
modified
7/19/2015 4:34:04 PM

Can any of this be back tested so to gain more confidence on this filter.

I added the following lines in the code:

set{bestbet, expgain / exploss}
set{Yourpick, bestbet * 100}

Please see if these make sense in context of the filter.

mahkoh
895 posts
msg #124446
Ignore mahkoh
7/20/2015 5:31:56 PM

One issue comes with incorporating your stoploss; JDST hit the 2% target 50 times and the stoploss 27 times out of 52 weeks.
This implies that at least 25 weeks both the target and the stoploss were hit. The problem is that there is no way to tell which was hit first.

DrJimO
35 posts
msg #124448
Ignore DrJimO
7/20/2015 10:42:07 PM

What would be the best way to use a filter like this. Would it be best to buy a few of the selected issues and reevaluate every week or would it be more fruitful to buy and hold for a period say 4-8 weeks before reevaluating? The filter includes 4 week numbers so would reevaluating every four weeks seem realistic? Also, which number(s) are best utilized as the purchase selection criteria?

sohailmithani
192 posts
msg #124452
Ignore sohailmithani
7/21/2015 10:10:00 AM

Started trading this one this week. Will on weekly basis run this filter on Sunday and buy 3-5 stocks at open Monday. Selection is based on their profitability ratio (per my code additions above).

Practical (live) trading issues noted this week.

1) Buying at market on open on Monday may mean better or worse price than system fill e.g. ANAC
2) 5% profit hit per system but may not in live trading due to low volume etc. e.g. ANAC

So far noted these. Will update if required.

sohailmithani
192 posts
msg #124472
Ignore sohailmithani
7/21/2015 10:35:20 PM

Stopped out on XNPT as bought on Open Monday at 7.74 (system open is 7.6). So in real world I got stopped out but system is in play.
Now I have only ANAC and NBIX to recover. My target is if I am lucky then to have 2 of 3 successful to get to breakeven. Not sure how to use this filter. Should I be buying top 5 and wait to see 4 of them hit the target so I make 10% (20% - 10%) if anything else I will be under water or should I close all positions when total net gain is 5% or net loss is around 10% or what??????

jemarcks
26 posts
msg #124736
Ignore jemarcks
8/24/2015 2:56:07 PM

I found this Crock Pot filter some time back and paper traded it. You have to manually back test this but its worth the effort.
I started trading this system live in my 401k using ETFs and a profit target of 2%. Not greedy but I always hit my targets on the first day or the 2nd day. 2+% per week gain in my retirement account isnt to shabby.
For example, today during the DOW dropping 1000 points I made 3.56% on DUST in 15 min. I made 2% on two others at the same time.
I usually just buy 4 positions at the open on Monday morning, enter my limit orders and im done for the week with this strategy.
Thanks to all those that posted in the various threads about this filter!!
Happy trading!

jemarcks
26 posts
msg #124737
Ignore jemarcks
8/24/2015 6:07:51 PM

mahkoh: When you see a profit target and stop loss hit in the same week, you can tell which one was hit first by looking at the daily historical data. That usually breaks the tie. On the rare case that doesnt solve the mystery, you will need to look the intra day data.



StockFetcher Forums · Filter Exchange · TRO's STAT SCAN for SWING TRADERS<< >>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