In this article i’m gonna deliver you an how-to simple normalized oscillator creation method that can be adjusted by any values you wanna be parsed by the indicator itself.
Read boundaries indicator information while trading can be quiet effective, specially on momentum one to get informations about great momentum period or exhausted one. Bounded mean reverting informations like Stochastic indicator are not so effective as it relies on price centring process that need constant adjustment to the price moves. So let’s focus on momentum indicator here.
Technical indicator normalization
What we need here is to set the informations we would like to be normalized. We said we are looking on momentum information delivering, so a simple moving average can make the job.
1 |
indicator = wilderaverage[20](totalprice) |
A Wilder’s moving average is noiseless and by adding “totalprice” instead of close information give us also a not so noisy price information from period to next period. Now we have a good point of what is momentum at the moment, but not a clear information of “where” is the trend and what to do : entering, exit or hold our position(s) ?
To a better visualization of what’s happen right now, we have to know of what is the moving average information compared to its recent extreme values. Like measuring the ambient temperature, if today is colder than yesterday, we need to settle the recent measurements by looking for recent highest and lowest :
1 2 |
recentHighest = Highest[100](indicator) recentLowest = Lowest[100](indicator) |
5 times the moving average period will give us a good comparison of where are we in the recent trend.
Now it’s time to normalized our indicator value to its recent highest and lowest values :
1 |
normIndicator = (indicator-recentLowest) / (MAX(recentHighest-recentLowest,.01)*100) |
Let’s see what happen :
As you can see, normalized value of a moving average can give better clear points of time reversal and trend timing. This indicator show now, as any other bounded indicators, what we call “overbought” and “oversold” territories… But it has a major advantage : these 2 territories are periods adaptive as it determines the overbought and oversold threshold on look-back periods only (100 periods in our example).
Of course periods of calculation could be adapted to timeframe and instrument. This indicator example is attached to this post (see below).
Now you can modify inserted values to create your own normalized indicator, so let’s try by yourself !