True if I knew some programming. I can change few parameters but a good programmer with trading knowledge would probably sniff it out or him & his few buddies together could figure it out.
There's of course no guarantee, but for you, it might improve progress. What would a hedge fund do with your algorithm? There's also the route that you make the system configurable so it won't work without the right parameters, but then you start limiting your potential again. Or perhaps learn programming yourself.
I had the same feelings 10 years ago. I am really not a programmer, but decided to try Tradestation and EasyLanguage. Even if at I think this language is very logical, that was not that easy for me I still use EasyLanguage. I used few times, services like Freelancer but never gave more than 20% of a system to any programmer. For example, I couldn't come up with an automated position sizing (Volatility, standard deviation, etc...), so I asked a Freelance Guy and also I bought a simple strategy (with a nice position sizing code), with open code. After that, I was finally able to code my own position sizing, thanks to them. Do you all think that programmers will not use your systems?
I can program but won't use someone's system for the simple reason that the system most likely doesn't make money.
A coded system has no sustainability unless it's personally configurable. Trading ability goes way beyond coding itself which is just a necessary tool, otherwise all the programmers would be driving around in lamborginis.
If you're a good software developer you're making enough to drive around in a Lamborghini if you choose, regardless of what you're working on. No need to risk any money trading! Most just buy a house in San Francisco instead; the personality type the blows money on a Lamborghini generally doesn't have a lot of overlap with the personality type of that makes for a great software developer.
Agreed, however it is often better to reconfigure in code as in practice parameters are often limited. Especially for trading systems one should avoid optimizing through multiple parameters, so as to avoid overfitting and selection biases. The more parameters being optimized the worse problems with over-optimization and dimensions becomes. So wise to be good friends with programmers or be one yourself.
As I said before I have no knowledge of programming & don't have time to learn it. But system does use traditional indicators & since a template is created in an unsophisticated software that I am using, it can be easily duplicated in platforms like Meta Trader, Ninjatrader & Tradestation. There might be more but I am not aware of it. On that template there is a price window with traditional indicators & drawing tools. Under the price window there are 8 sub windows each again containing traditional indicators. I have designed a set of rules for entry, stop loss & exit points. I also use 12 monitors ( but one CPU ) for one future ( Say ES ) with different time frames in each monitor. There are times when I get more than one signal on those 12 monitors. e.g. I might get signals on 3 minute, 8 min & 21 min charts & sometimes entry price is different on those 3 signals. I want to execute all the signals given by the system. Now in layman's term, can you guys suggest what is the best way to totally automate the system where computer does everything ? i.e. use one of the above platforms or a better platform OR write a code for the entire system ? Also can you educate me what would you call this kind of trading ? Is it algo, a quant, HFT, a robot, a mechanical system, a black-box system or something else ? Also any idea on the kind of cost involved for the entire project ? Thank You
Sounds like a multiple timeframe system. Do you have mechanical rules where you take ALL signals on all possible permutations of allowed inputs (entry & exit), or is there some discretion where you ignore or break rules? To me such complexity sounds discretionary. Rules might be easy to duplicate but performance not. Have you backtested rules thoroughly and know expectancy of the system? The more mechanical the system, the harder you follow it and research it, the easier to automate. Automation isnt the hard part.