Names and Conventions

Understanding my language

Like I highlighted in my previous post What am I offering?, my trading strategy is built on top off a fairly basic regression model that is able to show fairly good performance trading ES/NQ futures long & short. The regression model builds a fair value index and we use the difference between the value of the actual index and the regression model to make a determinant on overbought or oversold.

I call this difference as the Regression Curve Metric where higher values generally lean on conditions being overbought and lower values lean on conditions being oversold. Since I work on daily candles, I use daily closing values of the curve metric to make judgements in my trading strategy. Note that my trading strategy uses a combination of the curve metric with other oscillators and moving averages.

I’ve found giving the model the highest preference works really well historically. So if the model is at a state of being overbought or oversold, then closing existing positions and switching to the other side works really well. However, these are not the only situations in which my strategy may elect to switch positions and working with oscillators and moving averages augments the model’s performance substantially.

I also make sure to calculate changes between daily closes of the regression curve metric (1-delta if you will) as well as the maximum and minimum values of the regression curve metric intraday. In my backtest, I’ve found this to be useful in scaling/re-entering positions but not so much in switching from longs to shorts or vice versa.

The regression curve metric uses a combination of factors including economic data, some of which updates daily and other updates weekly. I also monitor a forward looking version of the curve metric that updates its values based on the forward looking data becoming available. In my backtest, I’ve found this to be useful in scaling/re-entering positions as well.

I hope this provides context on the weekly posts when I say Regression Curve Metric, Regression Curve Metric (Daily Difference), Maximum Regression Curve Metric, and Minimum Regression Curve Metric.