# Autoregressive Covariance Oscillator by Tenozen

Well to be honest I don't know what to name this indicator lol. But anyway, here is my another original work! Gonna give some background of why I create this indicator, it's all pretty much a coincidence when I'm learning about time series analysis.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Well, the formula of Auto-covariance is:
E{(X(t)-(t) * (X(t-s)-(t-s))}= Y_s

But I don't multiply both values but rather subtract them:
E{(X(t)-(t) - (X(t-s)-(t-s))}= Y_s?
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

For arm_vald, the equation is as follows:
arm_vald = val_mu + mu_plus_lsm + et

val_mu --> mean of time series
mu_plus_lsm --> val_mu + LSM
et --> error term

As you can see, val_mu^2. I did this so the oscillator is much smoother.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

After I get the value, I normalize them:
aco = Y_s? / arm_vald
So by this calculation, I get something like an oscillator!

(more details in the code)
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

So how to use this indicator? It's so easy! If the value is above 0, we gonna expect a bullish response, if the value is below 0, we gonna expect a bearish response; that simple. Be aware that you should wait for the price to be closed before executing a trade.

Well, try it out! So far this is the most powerful indicator that I've created, hope it's useful. Ciao.

(more updates for the indicator if needed)

Fixed:
- line plots NaN/Incomplete line.
- Some timeframes plot NaN.

Usage reminder:
- If the resolution is lower than 10 Minutes, it won't plot anything. Please use a resolution equal to or above 10 Minutes.

Updating the Indicator Preview