[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [EquisMetaStock Group] How to apply ATR?



PureBytes Links

Trading Reference Links

Hi Eric
 
 
It's no surprise that Wilder used Wilders Smoothing to create an Average True Range (ATR). Wilders Smoothing is an exponential moving average and typically an EMA can require up to 5 times more data than the Periods parameter would suggest was necessary. To get around a potential problem at the start of any smoothed value Wilders Smoothing is seeded (started) with a Simple Moving Average. Therefore, the first value produced my MetaStock from ATR(20) would be Mov(ATR(1),20,S). That is, a 20 period simple moving average of True Range. Notice however that ATR(20) responds to price moments more slowly than Mov(ATR(1),20,S), which is a result of the different smoothing method used for every bar after the first valid bar.
 
If you want to see the difference that loading additional data makes to an exploration (when using ATR) then list the results of using ATR(20) in an exploration set to 20 periods across 100 or so stocks, and then compare the results of that same exploration when 100 bars are used for the exploration. I think you'll be surprised by differences that may be up to 20% or so.
 
If you're unhappy with the inbuilt Wilders Smoothing for ATR then you could apply one of the other MetaStock smoothing methods to ATR(1). Just don't forget that DEMA, TEMA, EMA and any other form of exponential MA all benefit from extra data.
 
 
Regards
 
Roy
 
 
----- Original Message -----
Sent: Wednesday, November 12, 2008 5:36 PM
Subject: [EquisMetaStock Group] How to apply ATR?

Wilder started with a concept called True Range (TR) which is defined
as the greatest of the following:

The current High less the current Low.
The absolute value of the current High less the previous Close.
The absolute value of the current Low less the previous Close.

Does anyone have any suggestions on how many sample of data for ATR
[20]? Is there any rule of thumb on selecting the size of data sample?
If I use ATR[20], should I include 20 of data sample?
If I use ATR[50], should I include 50 of data sample?

Does anyone have any suggestions?
Thanks in advance for any suggestions
Eric



__________ NOD32 3607 (20081112) Information __________

This message was checked by NOD32 antivirus system.
http://www.eset.com
__._,_.___

Your email settings: Individual Email|Traditional
Change settings via the Web (Yahoo! ID required)
Change settings via email: Switch delivery to Daily Digest | Switch to Fully Featured
Visit Your Group | Yahoo! Groups Terms of Use | Unsubscribe

__,_._,___