StockFetcher Forums · Filter Exchange · Yet another W%R filter<< >>Post Follow-up
graftonian
1,089 posts
msg #128916
Ignore graftonian
6/3/2016 2:47:37 PM

A while back, on a another thread papagatorga wrote:
"I use
Williams%R (30) has been above -20 for the past 3 days
and price is between 7 and 13
and Williams%R(30) is above -20
and volume is over 300,000
and stock is not ETF

This has performed very well in market bullish trends, one week, two weeks and one month
It works well with all prices but I find more higher percentage gains in these price ranges.
There seems to be some positive correlation with stocks who have declined in price the previous day.
The W%R(30) comes from Big Trends, they have a complete trading strategy.
If you combine this with stocks who have very recently given a Parabolic SAR Buy signal, it works very well
I also use directional Index as a Confirmation.signal
GOOD LUCK"

Well, this got me thinking And I don't have enough to do, so I've tried to incorporate his ideas into one filter. But never being one to leave well enough alone, I did have to add a few personal touches. I've always said if someone would hold the mirror, I could do my own brain surgery.

papagatorga, comments?

Fetcher[
/*PAPA GATOR*/
draw ema(13)
/*IFT scaled to fit W%R (Y=AX + B)*/
SET{I2, IFT(5,9) * 60}/*the a*/
SET{I, I2 - 50}/*and the b*/
DRAW Williams %R(30)
draw Williams %R(30) line at -20
draw Williams %R(30) line at -80
DRAW I ON PLOT Williams %R(30)
SET{W, WILLIAMS %R(30)}
ADD COLUMN Wcount
sort on column 5 ascending
draw Parabolic SAR(0.02,0.2)
draw adx
add column adx
draw adx line at 20
set{di_diff, +di - -di}
draw di_diff line at 0
add column di_diff
/*the filter*/
Williams %R(30) greater than -20
set{Wcount, count(Williams %R(30) above -20, 4)}
Wcount between 1 and 4
di_diff > 1
adx > 20
psar_signal_days between 1 and 3
price is between 1 and 21
average volume(30) > 123456
market is not OTCBB
set{PSAR_SIGNAL_DAYS, days(parabolic SAR(0.02, 0.2) greater than price, 10)}
add column PSAR_SIGNAL_DAYS
do not draw PSAR_SIGNAL_DAYS
/*market info*/
set{spy_ema10, cema(ind(spy, close), 10)}
draw spy_ema10
set{spy_IFT, ind(spy, IFT(5,9))}
draw spy_IFT
draw spy_IFT line at .50
draw spy_IFT line at -.50
/*B&W*/
set{x, close}
set{10pct, x *1.1}
set{3pct_stop, x* .97}
draw price line at x
draw price line at 10pct
draw price line at 3pct_stop
]




c1916
77 posts
msg #128918
Ignore c1916
6/3/2016 3:30:43 PM

This is interesting...worth a bit of paper trading time and effort. One question I have out of the box is: What is the exit strategy? What is the counter move that would suggest the filter's conceit is no longer in effect?

graftonian
1,089 posts
msg #128921
Ignore graftonian
6/3/2016 4:32:42 PM

c1916 wrote:
This is interesting...worth a bit of paper trading time and effort. One question I have out of the box is: What is the exit strategy? What is the counter move that would suggest the filter's conceit is no longer in effect?

1) Exit Strategy? I am sure others out there have better ideas. 2) I am not sure what you re asking. papagatorga stated this was a fair weather filter, so I add the SPY data to keep up with the market trend.
If the market is trending down, I switch between SXPU and UPRO.


papagatorga
124 posts
msg #128922
Ignore papagatorga
6/3/2016 4:35:14 PM

I have no clue whatsoever as to what your filter does. Not a criticism, just I don't know coding. Seems to back test well for 1-3 weeks. A month or so ago when the market was performing very well, my filter produced many returns of over 20%. That is what really peaked my interest. At that time I scanned prices between 3-20. I was getting scans with over 100 stocks. Studying the results I found such a large percentage of the over 20% came from the 7-13 price range. The exit strategy as set by BigTrends is when the price closes out of the green (below -20) exit if the close is lower the next day. Your scan has a high percentage of winners but not a large list of high percentage winners. Of course my scan is also not showing a high percentage of big winners, except in the two week back test where it does a little better. I will be happy with any improvements, as long as we all can learn and make more money.
. .

graftonian
1,089 posts
msg #128925
Ignore graftonian
6/3/2016 6:50:02 PM

Papa Gator (or should that be Mr. Gator),
Thanks for your interest. What are your thoughts on sorting the results; I am leaning towards sorting on ADX descending, and limiting results to those stocks that have been above W%R for two or three days only.

Basically the filter looks for stocks that have been above W%R for 2 or 3 days
and
ADX is above 20
and
+DI and -DI are at least a plus 1

I added PSAR for your viewing pleasure, and scaled and added IFT to the W%R plot for my own grins and giggles. The SPY data is there to remind me this is a trend sensitive filter. But then, aint they all.

Will work on your suggestions for a graceful exit.

graftonian
1,089 posts
msg #128927
Ignore graftonian
6/3/2016 7:28:25 PM

Well that didn't take long.

The filter below returns stocks that meet your stated exit criterion (they fall below W%R =-20 and are lower the next day) after long run above -20.
I then put in the PSAR, and use the change in PSAR to determine the exit. It seems PSAR is almost always a better deal.

Fetcher[
/*gator exit*/
market is not OTCBB
Williams %R(30) crossed below -20 1 day ago
close below close 1 day ago
set{above20, count(Williams %R(30) above -20, 15)}
add column above20
sort on column 5 descending
Draw Parabolic SAR(0.02,0.2)
]



papagatorga
124 posts
msg #128931
Ignore papagatorga
6/4/2016 11:22:58 AM

Selecting which stocks to buy, which are more likely to post big gains, has been a challenge I have not solved. Going back and looking at the charts the day they alerted has not helped. Nothing has jumped out at me indicating one will be more successful than another. I guess news could be a big factor but who can predict that. I found a small correlation with stocks who had declined in price the prior day. As stated before I tend to select those stocks that have a very recent PSAR buy signal thinking it may be 5-10 days before a sell signal. I consider BigTrends exit strategy but look at 30 minute chart, Psar, DI, MACD, OBV and how far above the EMA 50 the price is. Remember all investment strategies are subject to the individual who chooses them. I seldom use options, but like 5-20 days movers in low price stocks. This suits me well. I back test almost every scan that is posted in these forums and very recently have come across several scans that back test as well or better than W%R. I am no expert, there are numerous posters on here that are much more knowledgeable. I just try to be helpful when I can. GOOD LUCK.

Mactheriverrat
3,131 posts
msg #128933
Ignore Mactheriverrat
modified
6/4/2016 11:33:29 AM

stocks below 25
and Average Day Range(30) above 5.50
do not draw Average Day Range(30)
add column Average Day Range(30)
and volume above 100000
market is not otcbb
market is not etf




Narrow your search!

graftonian
1,089 posts
msg #128935
Ignore graftonian
6/4/2016 3:53:44 PM

Good afternoon
ah, age old problem of sorting the results. For some time I have believed that ADR was a logical choice, but more recently I have begun to use ADX, a measure of how strong the trend is. In fact I use it as part of the filter. Conventional wisdom says below 20 (or 25) is no trend, so on the W%R filter I exclude stocks below this and sort the remainder on ADX

Fetcher[
set{a_d_x, ADX(14)}
draw a_d_x
draw average day range(10) on plot a_d_x
]



I spent all morning wrestling with the exit problem, ema, W%R, PSAR, no clear winner. Was it TRO that said "it aint what you trade, it's how you trade it"?


I feel there is some correlation of the outcome of the trade with the shape of the IFT line leading up to the trade. I like to see a deep, smooth curve starting at or near the bottom. I used the term "pregnant curve" in another post. I have tried to define this curve in SF code with poor results. Maybe some of the pros out there might offer suggestions

Enough rambling, it's lunch time.

StockFetcher Forums · Filter Exchange · Yet another W%R filter<< >>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 2022 - 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


This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.