At least in my experience finding order from past chaos by tuning parameters until curve fitted and expecting it to follow in futures is not working long term. I think its fundamentally wrong to optimize because stock market is not a sine wave from signal generator. There are extremely low frequencies present that in some cases may not even have done single phase shift from companys ipo up to rare but quick spikes. If algo idea comes to mind that involves parameters that have close to direct effect on output signals im probably going to skip the idea. Also dont think common university stuff and ai techniques that gets taught to millions of people can work out of the box. If too many people with similar base knowledge and ideas enter market are they is going to have advantage or are they going to reduce inefficinecys to 0? In my case best result came if using parameters that change some fundmental way of looking at input data and keeping flexible code instead of parameters in places where parameter has chance to interpret data or overoptimize to past. Also may be good to have and run multiple variations of the same algo with dif parameters , equalize asset allocations(as same algo with dif parameters proabably will contribute more or less) or even use multiple algos with each having multiple parameter combos. At the end the end result could be generated into single graph and it probably will look smoother than using 1 set of parameters. So far i have slightly over half decade of research experience and most of it half time. I think many trading algo devs goal is to get past the awful repeating cycle of idea-develop-test and replace with with modify-test + fully automated but it might best to take it as process not goal. Even if you get it live market competition also evolves and adapts constantly and at some point the system might not work and could step back into idea-develop-test... In my case the market inefficiencys will probably vanish profitability at some point as i see gradual almost linear decline in profitability in over 5 year historical out of sample tests. I think current market conditions hold good potential for algo trading because historical tests over 2008 and current times show similar peaks with higher returns than normal. For someone else their experience and knowledge may be completely different as there are infinite ways to interpret data, limited by imagination.
This is the bot live trading window. I took this picture a few minutes ago. Shortly after the two positions were closed.
I've had the same experience. I tend to set up my algos with some fundamental logic in mind, and parameters that match that logic. If backtesting doesn't show profit, I throw it out, because optimizing from there would likely overfit in ways that depart from my intended logic. If it shows profit, I may tune the parameters a little, but only if the adjusted net (removing outlier trades) shows additional gains. Usually, I don't bother tuning though.
Hey, been lurking this thread for a few weeks and this is actually my first post on the forum! Congrats on the success with the algo in the last few months, impressed by the performance. I have a few questions if you don't mind. 1. You posted a screenshot above with your GUI. What software is that? Or is it custom designed. 2. What way do you recommend to get from designing to running algos live? I'm currently on TradingView and I have no problem making strategies -- I'd like to say I'm great at it, as I've been on PineScript for years. But there's no algo support on TV. Where would you say is best to move for execution? 3. How does your algo do in a choppy market?
Hi 1. It's a Microsoft Access 2000 app on a custom designed DLL. 2. I wrote my own platform because I'm a coder, so this is my "recommended" way for running algos live. There are several trading platform out there that require little or no programming experience, such as MetaTrader, TradeStation, AmiBroker, NinjaTrader, etc. but I can't really recommend any of them. Maybe, if I had to choose, I'd go with MetaTrader. 3. My current algo tries to stay away from volatility and choppy phases in general.
I forgot to mention that the new algo works on 5min bars. Last entry: yesterday 12:55 long @10017.50.
Idk if its correct but it seems like you run it on single instrument. Could it benefit from multiple instruments? Can you vaguely describe what sort of algo you are operating?
As for risk management, what is your philosophy about using stops/profit targets say vs ATR trailing stops or no stops with an indicator exit? I have been backtesting a good deal with a variety of the above with mixed results. I agree with 931's position on optimization, it's too easy to curve fit with some of the software tools. I am thinking that more care in trade/money management is a better 80/20 use of system designer time.
Yes, I currently trade MNQ only. Likely so, but my estimation is that the Nasdaq 100 has more potential compared to other stock indexes and instruments that exhibit more erratic behavior. I believe in technology. Define "vaguely".
You may have answered this before, but do you work with profit taker/stop loss levels, or do you constantly move from long to short to long (e.g., buy above a certain level, then sell when it falls below, then buy when it goes above, etc.)?