have you ever tried backtesting with your bot taking opposite positions that what it is currently set up to?
I will only offer this advice one more time. Today the ER2 was up huge and was in an uptrend right from the opening. (I know cause I trade it heavily). I believe it closed up almost 10 points as it ran from the opening. And your bot lost 11 ticks (assuming 1 contract)... These are the days when an intraday trend system, as yours appears to be, should make good money. The ER2 was uptrending on the tick charts, 5 min charts and Daily charts so any program should have gone long, I perhaps naively assume. So when your bot loses money on a bullish day like today, I implore you again to stop the program and re-evaluate the entry and exit signals. I have posted on 4 seperate days when the er2 was up on average 7-8 points and your bot lost money. That should be more than enough of a signal to think about what can be adjusted so that you are in on these kind of trend days when the trend is established on various time frames. Good luck and this is why I still like to keep things in human hands lol..
Yes, the ER2, although in a strong up-move, was wild today. I actually was short in the afternoon, and despite being up late in the afternoon, I did not cover in time and was handed a loss with the final strong move up. I totally agree. The bot is not taking advantage of the opening move and morning trend for some reason....and then proceeds to get chopped-up during the 11 am -> 2 pm period of time.
ER2 is a beast. I definitely don't like it. You find some random system with a 1500 + ROA and 1.6 + Sharpe with say 1200 trades , and the minute you're going live , It just seems that you're trading the biggest losing system ever, straight down. That's the experience I had with ER2. The only system I stopped with some big losses was on ER2. ER2 is constantly gaining popularity , evolving with new traders, new noise... At one time , last 2 hours breakouts where outstanding ... until i decided to trade it. I don't know if it was related to noise but this pattern was working for 4 years... A symbol that is gaining popularity is perhaps the worst thing to trade. IMHO, unliquid symbols are easier to trade even with 1.5 tick slippage per side. Stocks are more stable too. There are no sudden volume change on it ... To trade ER2 you need some tough real edge ...
Well, Frost actually defined his system as anti-trend, so it makes sense that it didn't take advantage of today's move. I would also say that an anti-trend system is as legitimate as a trend system, and perhaps even more legitimate for intra-day trading. What makes commenting difficult is that this system has been described in very loose terms so far. The biggest ambiguity comes from the fact that it apparently considers two trends that have opposite directions, and bets against the trend of the shorter duration. That's as much as I could glean from this long but otherwise soapy blog. We don't know how the trends are determined, what their length is, and what the entry/exit rules are. This is like designing an airplane and describing it as "Well, it has two wings and a propeller. Over the previous 3 month of flight simulation, it has flown pretty well, beating the average speed for the airplane of its size. Now it flies for real, but it doesn't perform quite as well. I hope that it would improve (maybe the weather will help), but I am afraid I would be dead if it makes a few more crashes. What's wrong with my design (which shall remain secret)?"
nonlinear, I like your analogy. Only difference is I am not expecting anyone to tell me what is "wrong" with the system. If I wanted in depth analysis of the system's flaws I would go into more detail. Currently I am more interested in advice that pertains to how to handle the situations the bot is going through currently... how to know when to shut it down, how to know when to tweak and what tweaks to think about.. basically anything dealing with the difficulties faced when actually bringing an automated system live.... The base system is here and done.. that is why I took it live... I don't want to get into large discussions about the specifics of the strategy itself.. I will leave that to another thread sometime in the future perhaps depending on the outcome of how the bot continues to perform... With that said, I appreciate all the advice all of you have given thus far and look forward to continued dialog. I also welcome discussions about a "generic" strategy based on the principles this one works on... just not an indepth discussion about this specific strategy.
The reason the other traders keep making these suggestions Frost, is because they represent good trading principles, and they're trying to get you to apply them to your automated trading system. Aside from the mechanics of what it takes to make an automated system (which have been explored here to some degree), the fundamental concepts on which your bot trades aren't consistent with what most traders would do to make money in the markets. Here's a thought, try trading a less expensive contract (like say ES), and trading multiple strategies on the same financial instrument ... there have been some pretty good traders at this site who have recommended that methodology as being optimal. Happy Holidays, JJ
+$165 today..... Was up close to $335 about 15 min ago, then a spike down took some of the profits away before bot got out.... Anyway, its a green day.. so something to cheer about I suppose