S&P500 Relative to Observed Mean Reversions, page-4

This posts may be of interest if you develop proprietary indicators.

I've had some interest in Renko charts over the years however have never traded them as I find the concept of a fixed block size renders them inappropriate for a market with dynamic phases.
An advantage of Renko is that the removal of the time dimension removes a complexity that impacts time based momentum indicators - convergence as momentum slows which triggers 'whipsaw' trading when the market moves again after the indicators have converged.
When markets slow, they often become volatile which compounds this issues.
Using a fixed block is also not much different to running a trailing stop depending on how you use the Renko.

To get around Renko's late, early & whipsaw signals instead of using a fixed block size, I created a Renko using a dynamic block size based on recent daily volatility.
The end result is a Renko that is tight in a trend and loose during uncertainty.

Another aspect of Renko that I dislike is the lack of visibility of the short term price action hidden inside the blocks and longer term time based price action which has value for finding levels and pattern set-ups.
For this reason, I plotted the 'Volatility Renko' on a time based chart.

Backtesting as a Long/Short/Cash indicator simply based on its value relative to its 5D SMA the results were good both with in and out of sample data, across instruments & without changing the initial settings (SMA period & volatility measure) that i chose at random.



A shorter time frame for better visuals:


As you can see it triggers early exists to reduce drawdown and has very few unprofitable signals or whipsaws.

From the lack of responses I'm not sure if anyone's reading this thread so this may be the last post here.