For example: I enter a long at 900 with a profit target at 901 and a stop at 899. My profit target is entered (right after I get in the trade) as a sell limit 901 , stop is a stop market 899. My question is, what do you do when the trade reaches your profit target (901), trades a few hundred contracts at that price --but you don't get filled -- and then totally reverses on the next bar to get stopped out (899)? On a paper trade, this would be look like a successful trade .. but in reality it was a stop out. Should I exit at the market at the profit target and give up a tick everytime? Or just chalk it up to bad luck and deal with it ? Any thoughts or experience would be appreciated.
exactly, the problem of backtesting some strategies is that the realistic execution is hard to emulated accurately.
What you could do is attach a traling stop loss of 1 point, so in case you don't get filled at your target you don't exit with a loss. Personnally when scalping for 1 point i never use market orders, my experience is that it eats most of your profits.
I try to look at the market depth and T/S. If market is 'fading ' I tighten my target with 1 tick If you don't feel good with this because your R/R % changes, also change your stop -loss Hope this helps
Here are my thoughts. If I am actually trading a set-up that has a one point target and a one point stop loss, I have confidence that the set-up has positive expectancy. Also, I would be confident that using a one point stop loss resulted in more gains than using a different stop loss, either smaller or larger. In that case, I would trade based on the rules of my "system" and stay in the trade until I hit my target or was stopped out. The only exception is that you may want to have a time filter. If your target is not hit within x minutes exit at the market. I have found that the best trades usually work right away and a time filter helps eliminate some losses.
no comment, except on the situation you asked about. In your specific case, as soon as your target trades, you hit a hotkey which moves your stop up 5 ticks. It is very quick. If you're using tws, the whole thing is a bracket trade and the OCA status of the stop and limit remains in tact even though you change the stop price. Good luck trying to move any size with that system. If you're computer testing, only count trades which trade through. If your system breakseven, you may still have something due to the fact that you didn't count the trades that just touched. If you come up with something that makes money using 1 pt targets and stops I would sure like to hear about it, because last I heard they quit making those.
you can get a better place in line if you enter on a stop with your limit already working before the stop gets hit.
No you can't pretend to be able to be winning as long as you're are not sure that all the trades at 901 has been filled and when market print 901 you can't be sure so your backtest is at least flawed by principle. >On a paper trade, this would be look like a successful trade .. >but in reality it was a stop out.
PROX....... i do not wish to add any confusion to your dilemna but it is not possible to backtest a 1 point scalping system. (been there done that). I could not possible tell you the number of times that the ES would hit my price 3x , 4x , even 5 times and not fill me. Having said that though i would not adjust the stop or limit order initiallly placed because it just put you at the "end of the line" in the orders to be filled.
You have to get your order in their early, and keep it there. Since you're a one-lot trader, most likely the market will have to trade THROUGH your price in order for you to get filled. That is why alot of guys that trade 50's and 100's will pull after they get some portion of them.