frostengine, I am also using a stop and reverse strategy in the ER2. There are a couple of points I want to convey: 1) your conclusion about trailing or other profit-taking stops is correct, they will always worsen the results of daytrading. The reason I believe is that the upper end of your distribution of trades provides a substantial portion of your return. If trailing stops are used, those trades exit too early and leave too much on the table. I've found that trailing stops are more appropriate for larger macro-sized trades that enable you to stay in the trade for weeks or months at a time. Even then the stop must be placed substantially far away from the current price to stay in the trend. The best option I've found is to give the trades room to breathe, without giving into the fear and worry that the small winner of 20 ticks you have is going to reverse into a small loser. You NEED those outlier trades to make your system profitable. 2) The ER2, like was mentioned, still has liquidity issues. It tends to intraday trend the most out of all the major equity indexes, this is why on backtest micro-trend following scalpers look so damn profitable. BUT it is important to keep in mind that the number of trades you take per session should be relatively few for the strategy to be robust. If you use MIT orders to ensure all limit orders will fill, the slippage can and will kill you over a number of months. 3) Instead of worrying about the day-to-day PNL so much, trade the system everyday, day in and day out. NEVER try outsmart your system by turning it off for days you feel might give you losers. Instead, employ some form of compounding so that your position size will grow as your account does. This is only possible if your strategy follows 2) above. You can't get filled on 20 or 30 cars stopping and reversing every 5 or 10 ticks. It just doesn't happen in ER2. Make sure your SAR is discriminating the price action carefully to limit the number of trades you take. P.S. I am not a programmer by profession, but I have used EasyLanguage and Matlab for about 10 years now. I am in the ropes of learning java, and am trying to build my own execution platform that talks directly to IB to replace the mess of software apps I use now. Hopefully I can grab a few tips from you here and there if you're willing. Hope this helps, RoughTrader
I understand where you guys are coming from with the stopless trades. I will definately test the idea when it comes to research with directional bets. I remember back in my last year of university I was testing out ideas for a machine learning class with a friend who works at google now. We calculated with a 200x leverage account, you would double your working capital every 67 pips of profit. ie. going from 10,000$ to 16 million capital in 14 days (billions within the next week if you dont run into scaling issues) of trading if you made 67 pips profit / day (0.0067$). 67 pips a day seemed like a very reasonable goal back then. So if you can gaurentee a trade made between 3-6 pips, it would take between 10-20 trades a day. This is why locking in profits, limiting losses looked so important to me. My research back then was strictly in very short term scalps. trying to grab a pip profit over a few seconds of trading..
Walterjennings, Thanks for your input. Your idea back then of going for 3 to 6 pips and compounding aggressively is actually a pretty common one but unfortunately is impossible to implement. Going for 3 to 6 pips at a time assumes a very tight stop. That is, unless you are willing to wipe out the gains of 99 winners with just one loser. Playing in the backyard of even the most liquid of FX dealers, the minimum spread of 1 to 2 pips will immediately drag your win/loss ratio down to the ground. If it's currency futures, which are substantially less liquid than many FX dealers, the slippage and commish will achieve the same result. The point is, the trade must move with substantial enough magnitude to diminish or drown out the lossy noise of slippage, commish, spreads, etc. The scaling issues you mention are very real and are the one and only show stopper from making such a scalping strategy worthwhile to trade. When execution errors and the mentioned non-idealities of operating in the exchange are taken into account, it's easy to see why these kinds of rapid-file scalpers go broke quickly. Have a good Thanksgiving. RoughTrader
Noone really needs billions in profit in 3 weeks. Making a system that a shade as profitable as the ideal system would be a great success, say 1 pip profit a day. doubling every 67 days. until scaling slows it down. Why would you think 1 trade would wipe out the profits of 99 trades? say you assume a trade went peared shaped if it drops 1 pips from your opening position. and you automatically sell at 1 pip profit. with a 3 pip spread at a crappy broker, a bad trade would result in -4 pips. so you would need a 4:1 win:loss ratio to break even. any ratio better than that you are profitable. ib often has 0.5 pip spread. meaning a needed 3:2 or lower win:loss ratio. I think scaling wont be an issue until you hit 5-20+ million in lot sizes. the bonus about trading with huge lot sizes is that you move the market in the direction you want, eating up some of the loss caused by spreads.
While I personally don't like the scalping style of trading, VSTScalper has made a career of it and posses great skill and knowledge at doing it proftiably. You could look up his work to see how he executes his trades. As far as the "pushing the market around" comment, well that would be a nice trick if you can do it with a real account. JJ
RoughTrader, Great post. Some very helpful information there. It is becoming pretty obvious that stops generally hurt systems in daytrading, and your reasoning why makes a lot of sense. UncleTom, I am not sure if i posted in my last thread but I think I did. The llc is on hold until early next year.
I did not trade today, I shut my system down wends afternoon and will not be turning it on again until monday..... Hopefully I have a good week next week to make up for this horrible week
Don't see results as being good or bad, they are just results, and you're psyching yourself out, when you shouldn't be. If you aren't getting the results that you want, you have to keep studying and analyzing your system until it does produce those results. Long story short. I think you need a strategy overlaid on top of your system, because right now it's just processing signals. If you think it already has a strategy, you should keep looking until you find another one which is more profitable, which you like better and which will allow you to trade with a relaxed psychology and anticipatory trading style (yep, even though you're just turning on the BOT, you should be looking forward to it everyday). HH JJ