Williams %R

D

dac8555

I trade off 2 lists.

1. Is sort of a CANSLIM derviative. For longer term holds, but with more technicals and tigher stops. Strong EPS grwoth, smaller companies, "buy the dip" type stuff. for twhen the market is strong.

2. Based purely on technical backtesting that i do on my Bloomberg. I can it my "ace int the hole" list. the backtesting results vary from 30-226% annualized return..and probably average around 100% per year.


In regards to #2...I have forund that many of the best results come from manipulating Will %R periods. I love this indicator...really simple...battle tested.

yet, i dont really see this as widely used as some other indicators.

Who else is a Will%r fan?
 
.


In regards to #2...I have forund that many of the best results come from manipulating Will %R periods. I love this indicator...really simple...battle tested.

so what periods or settings have you found that give the above results ?:D
 
it varies per instrument...you have to play around...sometimes it works...sometimes it doesnt.

GLD for instance..13 period willr is best yielding about 70% before comm and slippage.
 
Saying 70% before commision and slipage means absolutely nothing.... you can have a system yielding 1000% and once slippage and commisions are figured in you may have a negative expectancy....

Also, unsure what you are using for your backtesting, but watch out for the "using future data" problem, and watch out for using the OPEN price of a day when your signal was generated and not the close of that day or open of the following day after the signal.... Failure to do these things will show GREAT systems that have no PRAYER of comming anywhere near expectations...

If you are manipulating W%R periods.. then there is a good chance you are simply curve fitting past results to the W%R.... you can take almost ANY indicator and if you tweak the parameters enough you can make it "FIT" the past data but fail horribly in forward tests...
 
Quote from dac8555:

Who else is a Will%r fan?
More...

Psst, I am. But it is just a few secs and Thunderdog will show up bitching about Larry... :)

Today was a slow day, but still excellent for W%R. On sideways days like this one can play the extremes without even looking at the pricechart...
 
Quote from frostengine:

Saying 70% before commision and slipage means absolutely nothing.... you can have a system yielding 1000% and once slippage and commisions are figured in you may have a negative expectancy....

Also, unsure what you are using for your backtesting, but watch out for the "using future data" problem, and watch out for using the OPEN price of a day when your signal was generated and not the close of that day or open of the following day after the signal.... Failure to do these things will show GREAT systems that have no PRAYER of comming anywhere near expectations...

If you are manipulating W%R periods.. then there is a good chance you are simply curve fitting past results to the W%R.... you can take almost ANY indicator and if you tweak the parameters enough you can make it "FIT" the past data but fail horribly in forward tests...
More...

Yes and no...before comission and slippage is the only way to make that statement becuse My comission and slippage is defferent from yours...so i use a gross vs. net number.

you are right on "open vs. close" price data...i agree.

I disagree on manipulation if the data...If an indicator has a default...said default may or may not reflect the behaivor that that particular group of people that are trading the stock. I think changing variables is the only way to see if there is lght at the end of the tunell on that particular indicator. There are pleantly of things that you can tweak until the cows come home and positive results never take place.

good point that backtesting can ALWAYS have poor forward results...which is why is is only one part of the equation...i normally use charting, trendlines, and stirck risk control as well....no amount backtesting can do all the work for you. there are always other factors.
 
Pekelo,

Would you be kind enough to share how you use the Williams %R
indicator?

I've used it a couple of ways. (1) Setting is a 6 more like a turbo.
(2) Setting is a 10 or 13. I'am not sure the best way to use Williams %R, but I like the indicator.

I have found that it is pretty accurate as a entry/exit indicator.
Pekelo, what have you found using Williams %R? What settings do you use? Apparently you feel at ease using it!

Thanks for your help and insight...

snarlyjack
 
Quote from snarlyjack:

Pekelo,

Would you be kind enough to share how you use the Williams %R
indicator?
More...

Sorry for the late answer. On sideways days like this Friday, you can simply play the extremes of W %R, like going long at -100 and shorting at 0.

My estimate is that at least 40% of the trading days are sideways. Sorry but I am not sure what set up my chart uses...
 
Here is the chart of yesterday's Dow, out of the 7 signals, only 1 failed. Depending on what the stop loss was, even that one wasn't a loser...
 

Attachments

You must log in or register to reply here.
Back
Top