Opinions about my (simple) volatility compression filter indicator.

[M.Carlo]

Hello to all, i am new here and this is my first post. I am learning to do auto trading and have some active automatic strategies that i write. I don't use ready-made things, bot or similar.

So, like title, I'm curious to read your opinions about my simple volatility compression filter indicator. He's nothing special or complicated, but work good.

The Masters of trading teach that to identify a fairly important volatility compression we can take any volatility indicator, and calculate it on 2 values : one in a fast period and one in a slow period. When the value of the fast period reaches 50% of the slow period, yes, we have a very strong volatility compression. I have no doubt about it. This work very good.

Since I'm curious, I asked myself if there wasn't a more "logical or simple" way to filter the volatility. Not that the comparison between an short and long value was not logical, but I asked myself: "it would not be much simpler and more logical to compare an volatility indicator, calculated on X bars, with its same average? " if I do it and divide the value of indicator by its average I get an oscillator whose values move above and below the value 1 :

When the oscillator is below the value 1 it simply means that the volatility is below its average, so we have volatility compression.

When the oscillator is above the value 1 it simply means that the volatility is above its average, so we have volatility expansion.

I test it on some (brekout) strategies, with various indicators but always i have got the best results with BBW. I calculate the BBW on 20 bars and compare it with its average, always at 20 bars. I enter at breakout of a major price level only if my oscillator value is below 1 or 0.9 or 0.8 ....
I am using it on live strategy and it work good.

I am sorry for my not perfect English, i hope you understand me, i also added a photo to show what I'm talking about.​

And You, what do you think about it ?

Nice trading to all.

 

[gaussian]

It seems that what you've done is normalized the velocity of the standard deviation component of a bollinger band into the range [0, 1]. Which explains why it works well with bollinger band width, you've simply created an indicator on it's derivative (or something like that).

I'd imagine your indicator may be something like [BBW_TOP - BBW_BOTTOM]/MEAN or something similar.

I'm curious how this compares to stochastics, keltner channels, etc or any of the other several indicators. You should post comparisons which would be useful to form an opinion on.

 

[notagain]

ROKU fast and slow Bollinger Bands SMA
Slow Bollinger is 80% of bars in one week
Fast Bollinger is 40% of bars in one week
When fast Bollinger's lower or upper bands cross the slow Bollingers midline = squeeze breakout

 

[M.Carlo]

Hello gaussian, thank you for your reply, let's go to talk...

It seems that what you've done is normalized the velocity of the standard deviation component of a bollinger band into the range [0, 1].

In fact it is this, but I just thought of comparing a value with her average. If this value is above her average it has a meaning, if it is below it has another. simple and logical.

Which explains why it works well with bollinger band width, you've simply created an indicator on it's derivative (or something like that).

The concept of "value versus her avergare comparison" can applies to everything. I use BB bandwith but i test it also with ATR or other like body/range ratio or more... the reasoning is always the same : to compare value (of atr, bbw, or more) with her average. It is not mandatory to use bbw. On some instruments bbw is better but on others or on other timeframes it may be better to use ATR, just test and test and test...

I'd imagine your indicator may be something like [BBW_TOP - BBW_BOTTOM]/MEAN or something similar.

Yes, BBW is (upper band - lower band) / avg, is not "my" it is Bollinger BandWidth and he measures the distance between the bands.But i use BBW only like example.... i test also ATR and more.

I'm curious how this compares to stochastics, keltner channels, etc or any of the other several indicators. You should post comparisons which would be useful to form an opinion on.

In my trading system I don't use any indicator, so i can´t help you (but no one can forbid you from doing some tests). I have not strategy based on stochastics, rsi or more. My only traffic light are BBW ATR ADX and some volatility compression/expansion pattern.

 

[M.Carlo]

Hello notagain, thank you for your reply.

ROKU fast and slow Bollinger Bands SMA
Slow Bollinger is 80% of bars in one week
Fast Bollinger is 40% of bars in one week
When fast Bollinger's lower or upper bands cross the slow Bollingers midline = squeeze breakout

Yes, I understand, but I don't use an indicator to get an entry signal on the market, at the contrary I use it as a filter, as a traffic light.

Mine is just a different approach, but I do not doubt the goodness of your cross

 

[InAllHonesty]

Hello

It looks good on initial glance but difficult to judge given there’s a small of data (image) to view.

Have you published the script? I’d be happy to view and give you my feedback.

 

[notagain]

Hello notagain, thank you for your reply.


Yes, I understand, but I don't use an indicator to get an entry signal on the market, at the contrary I use it as a filter, as a traffic light.

Mine is just a different approach, but I do not doubt the goodness of your cross

I see, using compression as a signal to avoid trading during chop.