August 6, 2020

Report 09/19/19 How I build my small trading algorithm.

Reading yesterday, “Analysis and Forecast of Financial Time Series,” the chapter on the exponential moving average, I remembered that in this formula there is an alpha coefficient, and while I googled it for details, I came across an adaptive sliding one from Kaufman.

I understood it this way: it differs from EMA in that at the moment of the corridor it goes smoother, and in a trend it reacts faster to changes.

With thoughts that I can try to optimizeI’ve got my short strategy with it, I climbed first on TW to look for scripts, redoing for myself, as far as I can, and then into Excel, to “skip” my own tester.

The idea is to open only if the price has passed a certain% of the extremum of the AMA, or the second theory, is to swallow it with a deviation.

I estimated the results on the chart, but did not see anything worthwhile. If anyone uses it, by what methods?

After I decided to try to build from scratch, based on this one rolling strategy.

A few hours of preparation, a few quick manual hypothesis tests and reached something + potential, now a safe grid is being calculated for the best and safe input parameters for AMA.

Let’s see what it finds and try to filter out the result with something ... what do trend mechanisms usually filter for?

What exactly is obvious is that in BTCUSD short you need to work with take, because close at the return intersection, always and everything shows a loss-making result, unlike altos‍♂️

P.S. If you are interested in where I will go, I will be glad to see in the telegram channel
Report 09/19/19 How I build my small trading algorithm.