How to reduce ES slippage
- Thread starter bearmountain
- Start date
right, so looking to make that extra tick. If I get slippage on both entry and exit. no good. system is hosed.
I am a small retail trader.
I am certain 100's of other system designers have looked at this problem. can a small retail trader working withing the limitation of the trading platform compete with these bots?
So far today, in the first 90 min of trading. The system generated 6 trades. So a total of 12 orders.
If I was willing to 'wait' today, 11 out of the 12 orders would have been filled at limit, without any slippage. some of the waits were quite long and adverse price action.
the second problem is, what do I do with the 12th order. how long to wait, how much loss to accept as a result of exit order not being filled...just thinking aloud...
The prob with your order filled and market order filled far away is whoever wrote the script or program orders wrote them wrong causes malfunction.
CME server only recognized manual orders:
1) market order
2) stop market order
3) stop limit order
4) limit order
CME server does not recognized automated/simulated/algor orders:
1) MOO
2)MOC
3)MIT
4)OCO
5)any type of orders programed by algor
These type of orders above are held in your computer or broker server and not in cme server because CME server doesn't recognized them.
That is your prob.
Good luck
You have answered your own questions.
If you gain 1 extra tic in 12 orders, that's 4 full points net in 90 minutes of trading. Nobody can outperform that, not even the best algos.
Many, many systems that "work" in theory/testing, completely crap out after comms and slippage. Yours is likely similar.
For thick FIFO markets, it is very hard to earn the spread by joining the best bid / offer. You have low queue priority, and will frequently only get filled on your bid when the market goes offered there. In effect, you have still crossed the spread.
Trying to get time priority in a thick FIFO market such as the ES is difficult as many orders are entered well in advance, and cancelled when the inside market moves near if the firm no longer want the trade.
The reason for using limits is certainty of price, as opposed to markets which provide certainty of execution. For immediate entries, aggressively priced limits (at best or at best skewed by 1-2 ticks) is going to be best for ES.
If you are working large orders, you can stack limits in advance in the area you expect the market to turn. For orders less than 400 lots or so, I do not see a point in trying to get filled on limits vs markets or trying to catch part of the spread.
The other reason for this is adverse selection risk. You'll get filled when you are wrong, however on your correct trades the market will frequently trade at your price but without trading through it, assuming you are joining the best bid / offer when you want to trade.
Assume you trade at a low, when the market has stopped going down and has direction has changed to go up. Your signal goes "on" to buy. Instead of paying up, you join the bid and are 990th in line. Four things will happen which will mess up your chances of execution if you are correct:
1) less market orders coming in to hit the bid
2) traders who are trying to go long will be lifting the offer, sometimes in size
3) offers will be pulled as informed traders are no longer sellers
4) genuine interest bids will join yours, and while you are first in queue the bid / ask imbalance will be read by algos who will take the offer and bid the next tick higher
Assuming you are incorrect, you will save a tick on every trade you close at a loss, however you will miss execution on a large portion of your correct trades, potentially rendering your system -EV.
In thick markets, limits are useful for getting filled at a set price when you may not react fast enough to hit market (ie when you expect a spike). Otherwise I am in and out the ES at market (aggressively priced limits actually).
Trying to get time priority in a thick FIFO market such as the ES is difficult as many orders are entered well in advance, and cancelled when the inside market moves near if the firm no longer want the trade.
The reason for using limits is certainty of price, as opposed to markets which provide certainty of execution. For immediate entries, aggressively priced limits (at best or at best skewed by 1-2 ticks) is going to be best for ES.
If you are working large orders, you can stack limits in advance in the area you expect the market to turn. For orders less than 400 lots or so, I do not see a point in trying to get filled on limits vs markets or trying to catch part of the spread.
The other reason for this is adverse selection risk. You'll get filled when you are wrong, however on your correct trades the market will frequently trade at your price but without trading through it, assuming you are joining the best bid / offer when you want to trade.
Assume you trade at a low, when the market has stopped going down and has direction has changed to go up. Your signal goes "on" to buy. Instead of paying up, you join the bid and are 990th in line. Four things will happen which will mess up your chances of execution if you are correct:
1) less market orders coming in to hit the bid
2) traders who are trying to go long will be lifting the offer, sometimes in size
3) offers will be pulled as informed traders are no longer sellers
4) genuine interest bids will join yours, and while you are first in queue the bid / ask imbalance will be read by algos who will take the offer and bid the next tick higher
Assuming you are incorrect, you will save a tick on every trade you close at a loss, however you will miss execution on a large portion of your correct trades, potentially rendering your system -EV.
In thick markets, limits are useful for getting filled at a set price when you may not react fast enough to hit market (ie when you expect a spike). Otherwise I am in and out the ES at market (aggressively priced limits actually).
Thank you guys for all your replies. Starting this thread helped me make good inroads in fixing this issue. All the best.
"For system traders it is a little different, slippage in a short term active system can make or break a system. The edge is tiny...."
So one thing you can do (which i did for my systems and ended choosing limit vs market for some of my systems) is to actually test it out. Put in code in your system that will filter or test signals to see which ones are taken. The way i did it is from entry to exit i scan every bar to see if the high of the bar takes out my entry price by a tick or more; if so my limit order at entry price will get filled. Now do that and see how it compares with your market order p/l.
good luck!
-gariki
So one thing you can do (which i did for my systems and ended choosing limit vs market for some of my systems) is to actually test it out. Put in code in your system that will filter or test signals to see which ones are taken. The way i did it is from entry to exit i scan every bar to see if the high of the bar takes out my entry price by a tick or more; if so my limit order at entry price will get filled. Now do that and see how it compares with your market order p/l.
good luck!
-gariki