I'm hiring somebody that is an expert at coding. Back in my days as a real estate investor, I had several different pieces of software built. I didn't code any of them, but I came up with the concept. While it may have taken me several updates, eventually I got every one of them to work well. I'm approaching this with the same idea. I have a pattern I've been trading discretionarily that works well, and I'm currently having it built as an Algo in NinjaTrader for my initial test run since that is a simple place to get it done and lots of developers that are easy to find. However, assuming that it works well in NT8, my plan is to build it all professional grade by hiring a top level and well paid coder. I Completely understand it may lose money. I'm willing to take the risk. I view me trying to learn code to build this myself as the worst idea and a guaranteed way to make this thing not work.
can you explain a trading strategy that you would give a developer to develop? Any strategy, not even a real one.
But if the structure is in place and the only unknowns are some variables (input settings), then how hard would it be for a bright coder to figure out?
It's a risk for sure. And if he does? I certainly don't prefer that and would expect an NDA to be honored, but it's not like it would decrease the workings of my own algorithm. Less than ideal in other words, but not the end of the world.
I used to build some of these, it was something to do, most used Tradestation as proof of concept as you can easily roll forward and roll backward through the datasets plus has 1sec data, probably not fast enough for your needs, hourly rates have gone up these days though. Tick level data you roll forward in realtime as backtesting will not take account of the small errors it introduces, there are not many programmers who would know how to do this, and especially not so it works. Hyper reliable and fast I know exactly what you need but that would not come cheap, focus on in-memory, I have a SQL Server in-memory sitting around here somewhere, it's not institutional but would probably be enough, best you learn it yourself because you are going to be hard pressed to find anyone.
You still need to work through few permutations (indicator settings, sizing strategy, trade management, stops and exits, timeframe, session time, and markets). If they figured it out, they deserve a cut of it by trading it themselves lol.
Brute force attack on the number space search of those parameters. You're incapable of creating such an attack so can't even imagine it, so you feel safe with simple trick obfuscation. You're giving the idea away. Just learn to code.
All this talk of people taking your code is nonsense, sure it could happen but they won't understand the foundation of it so it wouldn't work for them for long, there are millions of strategies in the world and the only thing that makes this different is you, plus if you can't code it yourself and you're not prepared to pay the big bucks for developers who have better things to do that use your code, you have no choices, common sense in 2023 is a rare commodity. It never ceases to amaze how everyone paper trades for what seems an eternity and then when they take the the plunge find out investing real money is a different ballgame, algos are no different plus if you think your strategy is going to work for longer than months without constant updates you may need to rethink, anyone worth anything (which there are few of) already know this and also know you will have to come back to them later for updates so why shoot the golden goose, then if you go with someone cheaper in between they can charge you double for wasting everyone's time (it happens a lot which I find fun to participate in from time to time).