Hi Matthew,
at Thursday last week (April, 17th) an error occured because a very profitable long trade at my system ‘Arabica intraday only’ (85074712) was not executed. I immediately sent an error ticket, but nothing happened up to now.
Therefore I ask you for help. The following happened:
At about 8:20 ET I entered the following orders, confirmed by emails from C2:
87109169 4/17/14 8:20 ET Buy to open long position 1 @KCK4 at stop 190 day order
87109171 4/17/14 8:20 ET Sell to close position 1 @KCK4 at stop 187.5 Good Til Cancel
87109191 4/17/14 8:21 ET Sell to go short 1 @KCK4 at stop 187.45 day order
87109193 4/17/14 8:21 ET Buy to cover short position 1 @KCK4 at stop 189.95 Good Til Cancel
The order execution happened in the following way:
First the price went below 187.45 and the short-position was opened correctly.
Then the trend changed and the price went up quickly, above the stop loss level to cover the short-position at 189.95 and also above the buy stop level at 190.0 to open a new long-position.
But the short-position was not coverd by the (buy to cover) stop loss order at 189.95, it was closed by the (buy to open) order at 190.0, designated to open a new long position.
The consequence of this was, that no long position was opened, although the price went up straightforward to 200.60 untill the end of the trading session.
Because I have installed a time-exit at 13:27 for all positions, this missing long-position had a profit of about 10.6 points or about 3975 $.
Dear Matthew, please could you correct this error and add the missing long-position to the trade list and the stats.
Thank you very much in advance for your cooperation.
Michael
[LINKSYSTEM_85074712]
Hi, Michael:
I think everything worked correctly. I have responded via private email, explaining the time sequence, etc.
Dear Matthew,
Thank you very much for sending the email and describing exactly what happened when C2 software filled the STOP orders.
I completely agree with our description. But describing what happened does not solve the problem I mentioned.
Unfortunately the same problem occured now the 3rd time since I released this system here at C2 and I suspect it will occur again in the future. The first 2 times I did not complain because I thought this did happen accidentally and besides this the missing trades did not have much impact of the stats. But now while this is happening the 3rd time I suspect there is a persisting problem with the sequential arrangement how STOP orders are filled by C2 software.
Please let me try to explain the problem again. In your email you write:
"
Now, here is what happened.
Signal #191, Sell To Open, filled at 8:30:11. Now you were short.
Your Buy Stop @190 (signal id #169) was filled at 10:25:19. This flattened the short position.
"
I agree comletely. This happened. But it should not happen in this way! Because when the system was short, there were 2 Buy Stop orders:
1) BTC stop @189.95. This was signal id # ending in 193.
2) BTO Stop @190. This was signal id # ending in 169.
And the big question is now, why was the BTO Stop @190 filled first and not the BTC stop @189.95 although the stop price @189.95 is lower than the stop price @190.
Obviously the software did decide to fill the Stop signal with the higher price (@190) first, but in real trading, the signal with the lower price (@189.95) should have been filled first.
I am sure, if there are 2 Buy Stop orders at a real broker’s Futures account the sequence of filling this orders is, that first the order with the lower Stop price is filled and and not the order with the higher Stop price.
Therefore the core question is, why did C2 software decide to fill the signal with the higher Stop level first? This caused all the problems and it seems to happen more often.
To make this message not too long I will stop writing at this point and only ask, if you could agree, that the sequence of filling Stop orders should be determined by the Stop price level and not by anything else (for example the time when the orders were posted).
Thanks again for taking time to research the reason for this problem.
Best regards,
Michael
[LINKSYSTEM_85074712]
Hi, Michael:
Now I understand your complaint, and I agree. We’ll change our Fill Engine to insure fills get reported in the right price sequence, which in turn will insure conditionals and OCA groups are processed correctly, too. I’ll let you know by email when the changes have been completed, hopefully shortly.
Dear Matthew,
I’m glad to hear that you agree now and the problem could be understood.
If the announced changes in your software are implemented, I’m confident that this problem is solved and will not happen again in the future.
Because the problem is realized now, I would like to ask you again to add the missing Long-position from Thursday last week (April, 17th) to the trade list and the stats.
Here again the dates about this signal:
87109169 4/17/14 8:20 ET Buy to open long position 1 @KCK4 at stop 190 day order
For further details to this missing trade please see my first posting to this subject.
So I think, we can bring this discussion to an good end now.
Thank you very much for discussing this problem in depth and your willingness to make changes in your Fill Engine.
Best regards,
Michael
[LINKSYSTEM_85074712]
What if you had subscribers at the time , what will happen in this case ?
