Standard Deviation and ATR

[Rognvald]

I want to compare the merits of using Standard Deviation of price against Average True Range in the construction of an indicator band for the purpose of assisting in calculating risk of entry.

Does anyone have any observations that might be helpful please?

 

[BBB]

I take it you are obviously aware that Bollinger Bands use SD, so I take it you mean constructing ATR bands and comparing to BB's?

How about constructing BB's with the upper band being 2 SD's from the highs, and the lower BB being 2 SD of the lows? May be interesting.

You may also want to look at Keltner Channels, which I believe are ATR bands

http://www.chartfilter.com/reports/c19.htm

 

[Rognvald]

thanks for the response
I know about Bollinger and Keltner but have been looking at Stoller which is similar to Keltner but uses a simple MA of ATR on an MA of Close rather than Keltner's MA of average price on average range. Stoller also uses a shorter primary avrerage.

My interest is in the difference between using ATR and SD as a means of covering price range and in the different interpretation that is placed on price touching/exceeding bands depending on construction

 

[Rognvald]

I supose I need a mathematician/statistician to tell me the difference btween Standard deviation and average ATR if any? Also need to think through volatility/momentum relationships

 

[TheBramble]

Standard Deviation is a measurement of volatility. It's a simple MA of the data (whichever data you're applying it to) of n periods summing the squares of the difference between the data and its MA over each of the preceding n periods; dividing the sum by n and then finding the root of this result.

BBs do exactly this - they add/subtract the SD from the price.

High SD values imply high volatility; Low SD values imply low volatility.

Low SD values 'normally' occur before large upward moves. I don't know why this doesn't occur as often statistically before large downward moves.

Consensus is that high SD occurs at major tops and low SD at major bottoms.

In terms of your specific focus (risk evaluation on entry) it would depend on your timeframe I guess. But in principle, Low SD (compared with recent historical SD) might be good support for a long position and High SD for a short.

ATR is from the same volatility 'stable' as BBs and SD so other than to suggest playing around with them on an experimental basis, I can't see what additional benefit they would have in relation to establishng risk entry criteria.

 

[Rognvald]

Bramble
Thank you. I'll ponder a bit more. Its the volatility/momentum bit I'm trying to tease out in my mind

 

[BBB]

I think of momentum as directional volatility. That why, even though price may be making new highs, if the latest highs are not moving to relative extremes of their standard deviation as the last high (ie the last lower high was out side the bands, the recent higher high is inside the bands) then this is more meaningful in terms of the idea of divergence.

? any help?