Thanks for the bug report, and good job on the clear description and screenshots, good catch on the indicator never going negative.
The indicator calculates the value as: ATR - MovingAverage of ATR
Since your FST never goes negative, also the curve has fewer dips than FSB, I suspect either:
1) the ATR is bigger than FSB
2) the Moving Average of the ATR is smaller than FSB
A couple of simple things to try first:
1) from screenshot, FST has only 400 bars of data; try to increase the number if possible
2) reduce the periods so they are a smaller percentage of the number of data bars
See if that helps the indicator go negative. But that is not the solution, would require changing your strategy, hopefully we can avoid that.
Another step, if you are interested in providing more info, use this custom version of the ATR MA Oscillator. It will show the values of the ATR and the Moving Average of the ATR, used in the oscillator calculation. It will verify if my guess is correct. Let me know after you download it, I will remove the attachment because it's not useful for any other purposes.
If the ATR and MA of ATR values match FSB with the same data, that will be very useful info, it will help point to the bug in a different place.
For screen shots, please be sure checkbox 2 "show traces" is ticked. It will help me figure out which one is different and causing the problem.
I don't know if I'll be able to fix it, but might be able to make a workaround for you.
edit - removed attachment