StockFetcher Forums · Filter Exchange · The Crock Pot Lives!<< >>Post Follow-up
JoeyVinyl
85 posts
msg #150619
Ignore JoeyVinyl
modified
1/31/2020 5:31:11 PM

I was going for a horror movie sequel kind of title there. I'm not a writer, obviously, since that's the best I came up with. Anyway, I love reading through old forum posts to see what people have been doing and to see if there's anything I can use. When I came across TRO's original Crock Pot post I really liked the idea of fairly steady though not flashy gains. 5 to 10 percent is more than most banks will pay you for your savings, but even if you're getting that it's probably not every month, let alone every week. Of course, there's no guarantee you'll get that using this filter too, but I like these odds better.

I searched the forums and read a lot of comments and variations on the original. I want to start by saying that I agree with Kevin's assessment. If you're going for a 5% gain with a 10% stop, one loss can wipe out two wins. That's why my version looks for a 10% gain, though in reality I'll take anywhere between 5 and 10 percent if I can get it. The other problem I found with all the CP filters I saw is a technical one. The Count function counts days, period. It doesn't matter if you're trying to count a weekly thing or not. If you say "count how many times the weekly close was over $2" (for example and I know that's not the correct syntax) and you put 52 as the length, it's going to count 52 days no matter how much you want it to count weeks. Go ahead and email SF support about this. I did and I think they're getting tired of explaining that.

Another thing about the Count function is that, according to the official guide, it has a limit of 100 days. Now I have seen some filters where people are saying it works for more than 100 days, and maybe they've changed it but haven't updated the guide. I don't know. If you're getting results you're happy with and your filters are working for you then keep doing what you're doing. For my needs with this filter 100 days is good. Assuming 5 trading days per week and 4 weeks to a month, 100 days lets you look back 5 months. Is it interesting to see what happened a year ago? Maybe. Is it useful? Again, maybe. But for this filter recent history is more useful, I think.

This filter is based on the work Kevin in GA did in this post: Link: https://stockfetcher.com/forums/Filter-Exchange/5-A-WEEK-FILTER-BASED-ON-TRO-S-CROCK-POT/81139

Here's an overview of what I changed. I kept many of his section headers but changed the goal from 5% to 10%. I expanded the historical frequency check to include a full 100 days, breaking them down into months of 20. I subtracted the previous months results so I'd get a more accurate monthly count of both ups and downs. One of his weighting factors came from a way to calculate interest charges called the Rule of 78. I don't know where he got the other one. Since I'm looking at five months I made a "rule of 5" and used those to weight the different months results. The idea is that I wanted the more recent numbers to be considered more important. That's another reason I didn't care if I was looking back more than 5 months. Other than that it's pretty much the same filter. The actual screening criteria is very basic. The close numbers are based on how little money I have to trade with and that my broker is only commission free on stocks over $2.

I sort on the Reward number rather than the Score. The stocks with the highest score number often have a few ups and zero downs, but their reward numbers are horrible. I'm including the ADR5 column and the Momentum, ADX, and RSI displays because I use them for more info, but you can remove them or put in ones you like better without changing the guts of the filter. Feel free to change the guts of it too, if you want to use it. That's what I did to Kevin's filter and why I'm not claiming this is my own original work.

Finally, when I previewed this post I saw the "Your StockFetcher filter exceeds filter performance restrictions" notice, but if you click on the filter and hit the Fetch Stocks button a couple of times you'll see charts.

ETA: It seems the sorting on column works differently in the post here than it does in MyFilters, probably because the column separators aren't showing here. Also, yes, this filter, like TRO's original, should be run on a Saturday or Sunday (since it's based on weekly events) for an entry on a Monday and exit whenever you hit your target or just feel like it, but I ran it Wednesday night when I got it where I wanted it and bought into CLSD Thursday morning. I sold it about a half hour before the market closed today for a 7.99% gain. So if you feel like running it and acting on the results during the week, go for it.

Fetcher[/* MEASURE THE DIFFERENCE FROM WEEKLY HIGH TO WEEKLY OPEN */
set {wkhiop, weekly high - weekly open}
set {wkgain, wkhiop / weekly open}

/* DETERMINE THE FREQUENCY THAT 10% IS ACHIEVED OVER MULTIPLE TIME FRAMES */
set {20up1, count(wkgain > .10,20)}
set {40up1, count(wkgain > .10,40)}
set {60up1, count(wkgain > .10,60)}
set {80up1, count(wkgain > .10,80)}
set {100up1, count(wkgain > .10,100)}

/* SUBTRACT FOR ACTUAL MONTHLY VALUES */
set {40up2, 40up1 - 20up1}
set {60up2, 60up1 - 40up1}
set {80up2, 80up1 - 60up1}
set {100up2, 100up1 - 80up1}

/* APPLY WEIGHTING FACTORS (BASED ON RULE OF 78 BUT USING 5) */
set {20up, 20up1 * 0.33333}
set {40up, 40up2 * 0.26667}
set {60up, 60up2 * 0.20000}
set {80up, 80up2 * 0.13333}
set {100up, 100up2 * 0.06667}

/* CREATE A WEIGHTED SUMMATION OF GAIN */
set {up1, 20up + 40up}
set {up2, up1 + 60up}
set {up3, up2 + 80up}
set {up4, up3 + 100up}
set {reward, up4 / 5}

/* DETERMINE THE RISK OF TRIPPING A 10% STOP LOSS */
set {wkoplo, weekly open - weekly low}
set {wkloss, wkoplo / weekly open}

/* DETERMINE THE FREQUENCY THAT A 10% STOP LOSS IS TRIPPED OVER MULTIPLE TIME FRAMES */
set {20dn1, count(wkloss > .10,20)}
set {40dn1, count(wkloss > .10,40)}
set {60dn1, count(wkloss > .10,60)}
set {80dn1, count(wkloss > .10,80)}
set {100dn1, count(wkloss > .10,100)}

/* SUBTRACT FOR ACTUAL MONTHLY VALUES */
set {40dn2, 40dn1 - 20dn1}
set {60dn2, 60dn1 - 40dn1}
set {80dn2, 80dn1 - 60dn1}
set {100dn2, 100dn1 - 80dn1}

/* APPLY WEIGHTING FACTORS (SEE ABOVE FOR EXPLANATION) */
set {20dn, 20dn1 * 0.33333}
set {40dn, 40dn2 * 0.26667}
set {60dn, 60dn2 * 0.20000}
set {80dn, 80dn2 * 0.13333}
set {100dn, 100dn2 * 0.06667}

/* CREATE A WEIGHTED SUMMATION OF LOSS */
set {dn1, 20dn + 40dn}
set {dn2, dn1 + 60dn}
set {dn3, dn2 + 80dn}
set {dn4, dn3 + 100dn}
set {risk, dn4 / 5}
set {ratio, reward / risk}

/*SCREEN FOR THE STRONGEST CANDIDATES*/
Show stocks where weekly open is > 2
and weekly close is between 2 and 20
and Average Volume(50) is > 250000
and reward is > 0
and reward is > risk

draw reward
draw risk on plot reward
draw Momentum(10)
draw ADX(10)
draw RSI(10)

add column Average Day Range(5) {adr5}

add column separator
add column reward
add column risk
add column ratio {score}

add column separator
add column 20up1 {1 up}
add column 40up2 {2 up}
add column 60up2 {3 up}
add column 80up2 {4 up}
add column 100up2 {5 up}

add column separator
add column 20dn1 {1 down}
add column 40dn2 {2 down}
add column 60dn2 {3 down}
add column 80dn2 {4 down}
add column 100dn2 {5 down}

add column separator
add column industry
add column sector

sort column 7 descending
chart-display is weekly
]




Cheese
642 posts
msg #150620
Ignore Cheese
1/31/2020 6:23:53 PM

Excellent. Thank you, Joey.

JoeyVinyl
85 posts
msg #150621
Ignore JoeyVinyl
1/31/2020 11:06:13 PM

@Cheese

You're welcome.

One of the things I like about this filter is that you can easily screen for stocks that are steadily falling too, if you like to short sell or we're in a bear market. Just change the "reward > risk" to "risk > reward."

Cheese
642 posts
msg #150626
Ignore Cheese
2/1/2020 11:38:17 AM

Joey, your user ID is punk rock and your Crock Pot filter absolutely rocks. Thanks again.


JoeyVinyl
85 posts
msg #150627
Ignore JoeyVinyl
2/1/2020 12:55:28 PM

Cheese, that's funny because that's exactly how I came up with it. Some friends and I were talking about what our punk rock names would be if we had formed a band. Since records and albums used to be polyvinyl chloride discs I chose Vinyl, and Joey is an homage to Joey Ramone.

nibor100
474 posts
msg #150633
Ignore nibor100
2/2/2020 11:50:14 AM

@joeyvinyl,

What exactly are the numbers in your Up1 column telling us?
It appears from the filter code that it is summing the wkgain variable over a 20 day period only when it has a value > .10(10%).
However, when I add columns to look at the daily values of wkgain more than 4 days ago SF will not display them no matter what I try, reducting columns, commenting out code etc.

For example ONCY, and LLIT each have a up1 value of 14 for the first 20 day period, Since there are generally 5 trading days in most weeks I would expect numbers like 1,2,3,4 or 5,10,15,20 so 14 is kind of a surprise to me.

the 14 actually makes me wonder if SF is recalculating your wkgain variable each day based on the prior 5 days(overlapping weeks) or some other kind of calculation, perhaps using each days close within a given week of data as the week end value, and keeping the weekly open value constant in order to determine a weekly high inside of each 5 day period.

Im hoping your own knowledge/development testing can clear this up for me.

Thanks,
Ed S.

Cheese
642 posts
msg #150636
Ignore Cheese
2/2/2020 2:06:23 PM

Im one of the members who from time to time still have to make extra efforts when dealing with sF weekly.
sF Weekly prices seem to be a mixture of firm and fluid data, depending on when we run weekly filters.
____________________________________________________________________

https://www.stockfetcher.com/help/How-are-weekly-prices-computed/100382
How are weekly prices computed?
Weekly measures on StockFetcher are computed as:
- weekly open: open price from first trading date of the week.
- weekly high: highest traded price during the week.
- weekly low: lowest traded price during the week.
- weekly close: last traded price for the week.
- weekly volume: aggregate of daily volume for the week.

For the most recent (current) week, the high, low and close are based on the data available for the week.
____________________________________________________________________

For more information on weekly measures on StockFetcher, please see:
https://www.stockfetcher.com/sfhelp/?id=100296
https://www.stockfetcher.com/sfhelp/?id=100227&isiframe=&isiframe=
____________________________________________________________________

Hope this helps with your research.

StockFetcher Forums · Filter Exchange · The Crock Pot Lives!<< >>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 2018 - 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