Description from the author:
There's a lot more to that prorealcode version but I was wondering if there's a simple version "Kumo Line" that only measures distance between Span A and Span B clouds with a setting for a look back period.In my pursuit to quantify the Ichimoku indicator, I have tried to quantify implied volatility by measuring the Kumo thickness. Firstly, I took the absolute value of the distance between SpanA and SpanB, I then normalized the value and created standard deviation bands. Now I can compare the Kumo thickness with the average thickness over 200 periods. When the value goes above 100, it implies that the Kumo is thicker than 2 standard deviations of the average (there is therefore only a 5% chance that this happens). A reading over 100 might indicate trend exhaustion and a reading below 20 indicates low volatility and Kumo twists (I chose 20 only by observation and not statistical significance). Interestingly, this indicator sometime gives similar information to ADX.
So far, the best use for this indicator is as a setup indicator for trend exhaustion or low volatility breakouts from Kumo twists. Extreme readings before Kumo breakouts looks interesting.
Code: Select all
//Ichimoku Tenkansen = (highest(high)+lowest(low))/2 Kijunsen = (highest(high)+lowest(low))/2 FutureSpanA = (tenkansen+kijunsen)/2 FutureSpanB = (highest(high)+lowest(low))/2 //Implied volatility based on Kumo thickness Kumodepth = abs(FutureSpanA - FutureSpanB) //Normalize implied volatility Indicator = Kumodepth AverageIndicator = average(Indicator) StandardDeviation = STD(Indicator) UpperBand = AverageIndicator + (2 * StandardDeviation) LowerBand = AverageIndicator - (2 * StandardDeviation) NormalizedIndicator = ((Indicator - LowerBand) / (UpperBand - LowerBand)) * 100 Overbought = 100 Oversold = 20 Return NormalizedIndicator as "Kumo implied volatility", Overbought as "Overbought", Oversold as "Oversold"
By simply looking at the "Kumo Line" we can determine if the volatility is there for a breakout.