Done For Day and New OrdStatus in FIX 4.4

, ,
1

Imported from previous forum

2

Hi, if a GT order has a CumQty of 0 (OrdStatus ‘New’) at the end of day on T+0 and a ‘Done For Day’ message is sent, I assume its OrdStatus the following day reverts again to ‘New’?

I didn’t see this transition explicitly mentioned in state change matrix but it appears a valid transition.

Is this the case?

Thanks

Paul.

3

Yes, the change of OrdStatus at the end of the day is the explicit version of a TradingSessionStatus or SecurityStatus message where a change of TradSesStatus(340) or SecurityTradingStatus(326) indicates that orders no longer trade, i.e. are “done for day”. It is not a terminal state for the order unless it expires on the same day. It goes back to New or Partially Filled or Suspended (e.g. for Closing Auction Only orders) on the next day.
Exchanges often only use a restatement of orders at the start of the day for reconciliation purposes (orders could expire over night or be subject to corporate actions) and not at the end of the day. There is a slight anomaly as orders can be restated at the start of the day with ExecType=Restated and OrdStatus with whatever it was the day before. It might be cleaner an the end of the day to use ExecType=Done For Day and not to change OrdStatus also to Done For Day. But the scenarios in Volume 4 of the FIX Spec define it this way. It may be for convenience, i.e. the owner of an order does not have to keep state outside of his orders, he can simply look at the order to find out whether it can still trade on the current day or not. That probably works well for buy-side to sell-side communication but with execution venues such as exchanges you have to keep overall state anyway and avoid excessive message volume. It gets worse if you split your order across venues and some still trade the (split) order and others do not.
Regards,
Hanno.

Hi, if a GT order has a CumQty of 0 (OrdStatus ‘New’) at the end of day on T+0 and a ‘Done For Day’ message is sent, I assume its OrdStatus the following day reverts again to ‘New’?

I didn’t see this transition explicitly mentioned in state change matrix but it appears a valid transition.

Is this the case?

Thanks

Paul.