StockFetcher Forums · General Discussion · Edgerater Backtesting, Graphing and Scanning Program is starting a 4 part video series on its capabilities<< 1 2 3 >>Post Follow-up
nibor100
919 posts
msg #155042
Ignore nibor100
12/21/2020 1:45:15 PM

For all in search of backtesting filters here's the link:

http://academy.edgerater.com/tradetestguide/

Ed S.
PS His simple scripting language can do a lot more than SFs

styliten
162 posts
msg #155044
Ignore styliten
modified
12/21/2020 7:30:48 PM

@nibor100

The MORE I became accustomed to SF, the MORE I appreciate its simplicity in syntax. Currently, my biggest grievance against SF is the lack of support for "chart-display is MONTHLY" but I'm willing to live with SF even with that grievance.

The things I liked especially about SF are:

1. ability to draw horizontal price lines (say day position(0.2, 30) and day position(0.8, 30)). This can NOT be accomplished very easily on Stockcharts.com AT ALL. And the advantage in doing this is not only HUGE but also next to impossible to be over-exaggerated.

2. built-in support for linear regression intercepts/linear regression bands, time series forecast, etc., etc. The discovery of the power of DMA is very recent and accidental. But when combined with #1 above, it seems to me that the need for backtesting has all of sudden evaporated.

3. yes, a few well-known indicators such as the KAMA/adaptive MA, Hull Moving average, etc. are noticeably missing. But the combination of DMAs and #1 above can make every other missing indicator a very minor inconvenience at most, except the monthly charting capability.

4. with such powerful built-in tools as cema, cma, clri, etc., etc. SKY alone is the limit to one's imagination.



nibor100
919 posts
msg #155047
Ignore nibor100
12/22/2020 2:03:43 AM

@styliten,

That's an interesting response to my Edgerater post....

a. In your point 1, to me it seems that you actually accomplished the "impossible", which iwas over-exaggeration of the impact of 2 horizontal lines drawn on a SF stock chart.

Also I found it curious the use of the 2nd parameter being 30 days when you seem to have a desire to look at monthly charts which would be somewhere around 22 trading days.

b. Regarding your point 2. my limited imagination is not seeing how I could replace a backtest of say an RSI(2) trading strategy combined with Wilder's Directional Movement Indicators with your 2 suggested lines and the use of SF's DMA indicator..

Perhaps you can provide an example of a replacement for a 3 year backtest of some strategy, so I can better visualize?

c. Regarding your point 3, I suspect that many traders who consistently use an indicator, that is not in SF, are even more inconvenienced by the missing indicator than by the missing monthly chart capability; especially, since even the SF weekly chart capability seems to only be usable on a visual basis due to the wonkyness of using weekly parameters as filter criteria.

Clearly you and I see things a bit differently,
Thanks,
Ed S.



VirginTrader
73 posts
msg #155048
Ignore VirginTrader
12/22/2020 9:10:52 AM

Thank you Ed for sharing! That is one impressive Excel based tool. Have still to figure out how to backtest your own idea but its got some cool templates included.

styliten
162 posts
msg #155057
Ignore styliten
12/22/2020 1:41:47 PM

@nibor100

The reason that horizontal price LINES based on day position formula are the holy grail is because they can be viewed as the dollarized volatility benchmarks. In effect, dp(0.2, X) and dp(0.8, X) are very similar to the 30/70 lines in RSIs. When X = 30, that's 6 weeks, or about 1/2 of the 13 week quarterly trading cycle.



nibor100
919 posts
msg #155063
Ignore nibor100
12/23/2020 2:17:55 AM

@VirginTrader,

1. If you go to the Edgerater website Blog page and search for this blog(just search on the word "simple")

"A Simple Long Term Timing Signal For The Markets: BackTest"

he explains a quick and dirty script, with how to get entries and Exits and then run the backtest which he calls 'Trade Simulations'.

2. He also has some old videos on his Edgerater Youtube channel that cover running simulations in more depth. and of course there will be an explanation coming up when he finishes his 4 part video series he's currently working on.

3. To give everyone an idea of the complexity level of the Edgerater scripting language here is a script for the standard MACD with Histogram:

MACD : EMA(CLOSE,12) - EMA(CLOSE,26);
SIG : EMA(MACD,9);
HIST : (MACD-SIG), COLORSTICK;

Ed S.





nibor100
919 posts
msg #155071
Ignore nibor100
12/23/2020 1:11:00 PM

Below is a SF filter from a 2016 post in · Filter Exchange · Breakout Trading Filter - GREAT RESULTS! followed by an Edgerater Chart Script I wrote to emulate the filter in Edgerater Pro and to also backtest it.

This is a more complex comparison of the 2 filter code languages.
Note: I tried to keep variable names common between the 2 programs.

Questions are welcome,
Ed S.

Fetcher[

and Close is between .25 and 12.0
and Close reached a new 80 day high

and close above open

show stocks where count(high equals low, 60) < 15

set{xmsum, ema(200) + ma(200)}
set{xema_avg, xmsum / 2}

set{matop5, max(MA(400), xema_avg)}
set{mabot5, min(MA(400), xema_avg)}
set{v1a, count(matop5 less than 25 percent above mabot5, 80)}
set{v2a, count(matop5 80 days ago less than 25 percent above mabot5 80 days ago, 80)}
set{v3a, count(matop5 160 days ago less than 25 percent above mabot5 160 days ago, 80)}
set{v4a, v1a + v2a}
set{v5a, v3a + v4a}
show stocks where v5a equals 240


set{matop6, max(MA(80), xema_avg)}
set{mabot6, min(MA(80), xema_avg)}
set{v1b, count(matop6 less than 15 percent above mabot6, 80)}
set{v2b, count(matop6 80 days ago less than 15 percent above mabot6 80 days ago, 80)}
set{v3b, count(matop6 160 days ago less than 15 percent above mabot6 160 days ago, 80)}
set{v4b, v1b + v2b}
set{v5b, v3b + v4b}
show stocks where v5b equals 240

]



Here's the Edgerater Script/Filter:

cond1: BETWEEN(c,.25,12);
cond2: c > ref(HHV(c,80),1);
cond3: c > open;
fif: count(high == low,60);
cond4: fif < 15;
xmsum:= ema(c,200) + ma(c,200);
xema_avg:= xmsum / 2;
matop5: max(MA(c,400), xema_avg);
mabot5: min(MA(c,400), xema_avg);
v1a: count(matop5 < 1.25 * mabot5, 80);
v2a: count (ref(matop5,80) < 1.25 *ref(mabot5,80), 80);
v4a: v1a + v2a;
v5a: 80 + v4a;
cond5: v5a == 240;

matop6: max(MA(c,80), xema_avg);
mabot6: min(MA(c,80), xema_avg);
v1b: count(matop6 < 1.15 * mabot6, 80);
v2b: count (ref(matop6,80) < 1.15 *ref(mabot6,80), 80);
v4b: v1b + v2b;
v5b: 80 + v4b;
cond6: v5b == 240;

event: cond1 & cond2 & cond3 & cond4 & cond5 & cond6



VirginTrader
73 posts
msg #155078
Ignore VirginTrader
12/23/2020 9:27:59 PM

@Ed
Thank you and happy holidays! I've got 14 days to test drive this toy...

Cheese
1,374 posts
msg #155087
Ignore Cheese
12/24/2020 6:13:02 PM

@ Ed

Merry Christmas to you and all members who celebrate.

THANK YOU for re-posting a great "Breakout Trading Filter - GREAT RESULTS!".
I agree with Wilson that paying more for an advanced sub to see all features
from the above filter might be well worth it.

For now, I will just have to ask you questions based on my incomplete view of the filter.

At the end of Dec 23, 2020 trading day, I was able to find IWM, which has
var5a=240 and var5b=240, and the filter does indeed appear to be great at
finding breakouts.

But by the time. a pics finally appears from var5a=240 and var5b=240,
it seems like the bull run is probably pretty long in the tooth, and using other
filters like snappyfrog's systems, that pick would be more likely to drop next
or pull back.

So, is the above filter condition var5a=240 and var5b=240
just to show backtest examples of breakouts after the facts?

Do you have any suggestions for us to capitalize on the running up to the breakouts ?

If the answers were already covered in the aforementioned thread,
please kindly refer me to the specific posts or paragraphs.

Thanks again.


nibor100
919 posts
msg #155099
Ignore nibor100
12/25/2020 5:30:38 PM

@cheese,

1. When I run that filter on Dec 23 it does not return IWM even when I increase the price range.

2. No that filter was not designed as a backtest filter by Wilson, as he introduced it as a filter he had been using for the prior year, and had made about $32K paper trading in his $50K margin account. Though if you read all of his posts in that old thread you can see he varied his trade methodology quite a bit...

3. Looking at the charts it appears that a lot of the run ups prior to the "breakout result day" occur after a considerable period of the Close 80 day high being horizontal for many bars, so perhaps playing around with that might work. I never used the filter for trading, it was kind of my first test to see how hard it would be for me to backtest SF filter code by converting it to Edgerater script (I was pretty new to SF back then).

4. My backtest showed the filter being a little better than 50% winners for the prior year back then which coupled with many undefined trading rules led me to not pursue it much further.

Ed S.

StockFetcher Forums · General Discussion · Edgerater Backtesting, Graphing and Scanning Program is starting a 4 part video series on its capabilities<< 1 2 3 >>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 2021 - 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.