I'm looking for a clever way to calculate a noise trading indicator, preferably based on standard stock indicators, and would be grateful to any more versed in technical studies for ideas on how to approach. I'm certain many have done similar systems and don't want to reinvent the wheel.
I'm developing a mean reversion system that looks for instances where an indicator (it's a simple dataseries, not an OHLCV object, if that matters) deviates from its 50 day moving average, and then takes a position assuming convergence of the indicator. For simplicity, let's say it was the 3 day EMA vs the 50 day EMA, and the strategy seeks convergence back to the 50 day.
What I'd like to build into the system is a measure of the recent tendency for the short term EMA to mean revert relative to the tendency of the longer period EMA to trend (since that's when the model fails). My ideal trading opportunity would be one where there was lots of volatility in the indicator but generally going sideways. My nightmare symbol would be one that is low volatility or in a steady up/down trend.
Thanks in advance
I'm developing a mean reversion system that looks for instances where an indicator (it's a simple dataseries, not an OHLCV object, if that matters) deviates from its 50 day moving average, and then takes a position assuming convergence of the indicator. For simplicity, let's say it was the 3 day EMA vs the 50 day EMA, and the strategy seeks convergence back to the 50 day.
What I'd like to build into the system is a measure of the recent tendency for the short term EMA to mean revert relative to the tendency of the longer period EMA to trend (since that's when the model fails). My ideal trading opportunity would be one where there was lots of volatility in the indicator but generally going sideways. My nightmare symbol would be one that is low volatility or in a steady up/down trend.
Thanks in advance
Comment