I looked back through the log errors and found a ref to not enough bars etc but it was so far down the log and hadn't been repeated I assumed it had been resolved, apologies.
I'm hoping the indicator will give me a clear measure of trend, result greater than 1. I have developed a dynamic Ravi indicator (not posted yet) further using a derivation of ATR for the threshold which works well but being an oscillating threshold it tends to give me to many false signals as ATR is more smoothed than Ravi. The problem only seems to occur in periods of higher than avg volatility when Ravi is approaching the threshold, sort of defeats my original objective !! Presumably this is because the ATR has 1 extra level of smoothing than Ravi so lags a bit.
I was trying to create a dynamic threshold that will shrink in periods of low volatility but expand in periods of high volatility. This means that spikes in volatility around opening bell etc, have to be sustained in order to trigger an entry or if you like, be a real breakout. It has worked very well and has done away with the need to manually adjust an entry threshold based on S & R as per the original Ravi. Perfect for automated trading !!
This indicator should suggest that if prices are moving randomly then the result will be less than 1based on the theory that prices move as the standard deviation of daily log changes multiplied by the square root of the length of time over which the price change is measured.
I'm hoping this will do away with the variable threshold and some lag - Testing will see if this filters out a few more losing trades due to marginal signals in an automated system. Lets see if it works !!
Comment