Thanks fan27, I use NinjaTrader for developing and testing systems and CQG as my data feed. With NT, I can only backtest/optimize one system at a time. Unless I use more then one computer, but I only have one computer. What do you ask "test multiple systems concurrently"? Thanks
Hello Humpy, Please keep the following in mind when it comes to buying a trading system: 1. Confirm X years or Y number of trades with actually verifiable live trading results. Anyone can sell you an over-optimized (curve fitted) back tested trading system. I can sell you one now by just showing you the back tested trading performance. But this will be unethical and wrong of me because I do not have actual live trading results to show you. There is nothing wrong with buying a trading system and I recommend it for sure, but make sure you do your due diligence.
Not sure what you mean here. I have developed multiple strategies in Ninja and had them running concurrently with no issues, all on the same computer. Let me know know what is the issue and I will try to help.
Hello RedDuke, I apologize. Yes, I can run multiple strategies in NT8 and NT7 concurrently live and simulated. In NT8, I can use strategy analyzer to backtest/optimize multiple strategies at once. One the issues I have is 1) Deciding the amount of historical data to use to backtest/optimize intraday strategies. Currently I am using 2006-2011 as out of sample and 2011 to 2017 as in sample for walk forward. Because its intraday (estimated 600 trades per year), this takes along time. Sorry OP to be off topic. Perhaps I can PM you RedDuke.
@globalarbtrader has published a couple of excellent books. Implementing his ideas would likely get you a Sharpe ratio of ~1: you won't get rich but do better than the vast majority of retail traders. With no investors to stop you out at the bottom, you can run e.g. 20% vol and compound quite nicely overtime.
What most new retail traders don't understand is that averaging 20% per year consistently is very good. The problem is that does not reconcile with them wanting to quit their day job and trade full time with an underfunded trading account. I would not even attempt to primarily live off of trading income unless I had 7 figures in my trading account.
Multiple uncorrelated systems are obviously much better. The more you have, the better your equity curve will be. The difficult part is finding things that are uncorrelated and managing margin.
You don't want to trade 2 systems that do the same thing on a correlated instrument. Say if system1 tells you to go long SPY and system2 tells you to go long DIA, you won't achieve any benefit to the equity curve because they are correlated instruments, if the market falls your drawdown will be effectively doubled or multiplied. But if system1 goes long SPY, system2 goes long UNG (nat gas), the instruments have no correlation and therefore the results will improve the equity curve as the drawdown for both systems is unlikely to happen concurrently.
Thank you d08, I understand now. Very interesting. Would you agree the following are example correlated instruments somewhat: ES , NQ, YM, TF / RTY Noncorrelated instruments example : NQ - CL ES - GC ES - ZN