TLDR: How do you personally define support and resistance zones algorithmically, if at all. If you don't, why? Is it strategy specific, you've found that you simply don't need them, etc.
So I'm relatively new to algo-trading, but I have a bunch of forex trading strategies that I've already manually traded successfully. I would like to create some level of automation for them, with the primary purpose being to give me time away from my keyboard.
I keep running into the issue of defining support and resistance zones, on which my strategies almost always depend to work successfully.
Some of the things I've tried:
- Getting all pivot points for the last 10 or so years on the interval I'm trading and averaging out those within a range in order to get fewer and more concise lines. This sometimes works, but sometimes doesn't, depending on volatility and range of the price within that time. I'd need to manually configure as well as update with some regularity in order to keep the zones accurate across different pairs.
- Getting the lowest and highest price of an instrument over the last 10 years. Then for each price from the lowest to the highest, incrementing by 2% of the range every time, counting the number of times that price(in a range of about .5% around the price) serves as a local high or low, and only charting the 10 with the with the fewest touches. and 10 with the most touches. This also gave me inconsistent results leaning on the inaccurate side, and varied *heavily* across instruments.
I've been looking for other ways to determine the support and resistance zones, but I haven't really found anything that gave me anything consistently close to what I'd have determined by eyeballing the chart. I understand that zones are massively subjective, but I'm primarily here for any different perspectives on the matter. I can always tweak things where I need to to make it fit my goals. I just need a general concept that could work. Thank you in advance.
Submitted September 29, 2020 at 09:20PM by JoshuaSantore