Imported from previous forum
Hi all,
In the FIX message model defined for Mass Quotes, a Mass Quote Acknowledgement (MQA) will follow each Mass Quote message. Any executions will be communicated as Execution Reports after receiving the Mass Quote Ack.
However, it is not clear whether this is the recommended approach when a Mass Quote aggresses certain orders on submission. From the user’s point of view is it acceptable to receive executions resulting from the mass quote prior to receiving the MQA ? Or else should MQA be received before any of the executions ?
Thanks
Chathura
The Mass Quote Ack should be received prior to any executions, i.e. the Mass Quote should have passed all validity checks so that the submitter knows which quotes are active. You may even have conflicting quotes inside the mass quote which result in the rejection of the entire Mass Quote. For example, it might be forbidden to quote the same instrument twice as there is no explicit sort order in a FIX repeating group. Which quote is to be valid? If you start allowing executions prior to validating the entirety of the Mass Quote you might run into problems, hence my recommendation of acking first and then executing.
Regards,
Hanno.
Thanks a lot Hanno ! I agree with the point that sending the ack prior to the executions helps the submitter. However, sending executions prior to the ack does not mean that the quote has not been validated prior to be getting executed. It is just a change of the message sequence received by the submitter. Anyway I guess if the ack is sent after the executions the whole point of sending an ack will be lost.
I think the example scenario you have mentioned can be handled via the implementation. Any subsequent quote sent for the same instrument can be considered as a replacement of the initial quote.
From a different point of view, can it be beneficial for the submitter to receive the execution first. After all, he will be more interested in the execution for his quote, so will he mind receiving the ack later ?
You have to imply a strict sequence of quotes in the Mass Quote if you want to consider a quote for the same instrument as a replacement. This is dangerous if the receiver uses parallel threads to process an incoming mass quote. I do not see the use case of putting two quotes for the same instrument in a single message and would advise against the replacement logic you suggest. It could be mis-used to input many quotes for the same instrument, expecting the mass quote to behave identical to a series of individual quote messages for the same instrument. The primary intention of the mass quote is to quote many different but related instruments (e.g. option series) with a single transaction, not to save on the number of messages.
I would also advise to only ack individual quotes inside the mass ack whenever the outcome is different from what the submitter expects, e.g. a rejection of an individual quote or a removal of the other quote side. In most cases you then only need to return the root level of the mass ack message.
Agree that it is not a use case to send multiple quote for the same instrument via a single mass quote message. However, if such a situation occurs then it needs to be strictly defined how the processing is done. System can either reject those duplicate quotes or consider as replacements to previous quotes (subject to consequences mentioned).
Point taken about acking individual quotes for rejections and cancellations. Thanks !