Hi,
I have developed a system that uses constant tick bars & traditional indicators & I have spent a great deal of time testing manually & seems to be working pretty good on most futures & Indices using a regular charting software. But that software is not sophisticated enough to make it completely mechanical so I can not test it with large amount of historical data.
I want to get it programmed so that it will be like an algo trading where the algo will do everything including finding the signals & execution of orders with entry, stops & exits.
Can you guys please suggest me which charting software will do this and find a programmer to do that ?
Also I am going to have to disclose the system to the programmer so how do I protect myself ?
Thank You
Neal
The following is my opinion and for educational purposes only:I have developed a system that uses constant tick bars & traditional indicators ...
I want to get it programmed so that it will be like an algo trading where the algo will do everything including finding the signals & execution of orders with entry, stops & exits.
Can you guys please suggest me which charting software will do this and find a programmer to do that?
...
The following is my opinion and for educational purposes only:I'd suggest to try Ninjatrader, it's programmable in C#, and its version non-live trading is free, so you can contrast your Algo. If you have some basic programming experience it'll take no much time to learn it well. Besides it has a great assistance forum to ask doubts and solve problems.
I'm gonna suggest a few tips for your backtesting:
- Try to work with the same data of your actually broker, don't buy any other data, you may finish with unreliable results and nasty surprises in your real account
- Consider the slippage while you trade, depending on your trading frequency
- Don't use leverage for a good level of performance rate
- Once you've backtested, try first a live simulation and watch reality performance
- Watch for real technical errors while live trading, such as: disconnections, overfilled...etc
Regarding to protect yourself:
- Don't tell your idea to anyone
- Don't upload your algo in any free web sites that promises you a brilliant future
- Set a good firewall, close all non-used ports, close services as FTP...etc
- Don't open any link embedded into a mail
- Use a good antivirus, I'd suggest Avira free version
Good luck
I'm in the system development business (I'm not fishing for a job here though). I only must smile about that permanent fear by beginners that programmers or other people "steal their system".
Systems are a dime a dozen. I have my hard disk full of systems. If a programmer wanted to steal, he could of course, since no NDA or contract can 100% prevent this. But a professional programmer is coding for money and not for the hope to find a profitable system that he can trade himself. If he trades, he's using his own systems. Maybe some ideas from his past programming jobs go into them, but I have never heard of a case that a programmer has "stolen" a system from a client.
If your system is simple, learn programming and do it yourself. If it is complex, you really need a programmer, but if you did it manually so far it can hardly have been this complicated.
If you get a programmer that knows one of the common automation platforms already - like NinjaTrader, then he will have an interest in trading.
As such, were your system to work, he'd most certainly at least keep it to trade himself and very likely sell it on to others.
I have seen this happen many times - people have ideas, find a developer on a trading forum and then 6 months later, they haven't heard from the developer for ages and they see he's got a website selling their ideas.
The best way to protect yourself is to find a developer with no interest in trading and to suck up the fact it'll take a little extra time to complete.
I'd also get him to just draw dot's on the screen, don't even tell him it'a an auto trader. You can convert the dot drawing to order commands yourself later.
Contracts are tough to enforce with this sort of thing because he can claim his idea is different from the one you presented and to prove otherwise would be costly and quite difficult.
You have to ask yourself if you're trading something as liquid as ES and it works on many different futures, what impact will it have on you if someone "steals" it? Especially if that person is someone making $5 an hour in India (which you know because that's what you're paying them!) It would be a pretty fantastical leap to imagine that guy either implementing your strategy with his own meager savings to the point that he arbitrages away your "edge" in each of those super liquid contracts, or that he could attract a hedge fund of sufficient size to arbitrage away your edge based on the super secret strategy he "stole" from you. In short, you're probably flattering yourself with the value of your strategy, but if you're good enough to come up with one you're good enough to come up with more if that one is "stolen". Pay someone to implement your strategy, see if it works, and once you've proven that you indeed can come up with valuable strategies you can use some of the money you earned on the first one to protect your subsequent ideas.You are so right & that is what I am afraid of.
With absolutely no knowledge of programming, I was thinking about getting it done on one of those platforms and that was also my question as to which platform would be best. I am using 12 monitors for one future like emini ES with different time frame on each monitor & that way I am able to find a trade on one or more monitors. I could use as many as 24 OR 48 monitors for just one instrument which would generate sometimes many trades at different entry points & exit points & I can choosy to find the best setups too.
Now I have tested this on many major futures in different sectors & also on options having a volume of 100K to 1 million contracts for just one strike price. So opportunities are endless. I tried it on ETFs & Stocks too but I could not find the reliable tick data for them & system doesn't work well with overnight gaps. 24 Hr markets are best.
So any idea as to which platform should be used OR another way of doing it ?
Thanks
You say you need to go mechanical, but also say you need to get the system backtested. These are two different concepts. The easiest path would probably be to backtest your system first. This can be done with almost any programmable charting software. I'd suggest you try that first and see how the system fares over time during many different market conditions and reactions.
Automating the trades is very easy compared to making a mechanical system that is consistently profitable. The problem with automation is that the system will take all valid signals, at exactly the levels given by the system, so may have worse expectancy than expected. Especially if you haven't really done such detailed analysis yourself, you'll be surprised how much will go against expectations.
Why shouldn't the programmer share the system? If you limit your programming, you'll be limiting your potential. There need to be trust and tight collaboration If you don't know/want to learn programming, I don't see how you can have a maintainable system otherwise.