Sponsored DMA
-
zerohair
- Posts: 1
- Joined: Thu Jan 01, 2004 12:00 am
Sponsored DMA
On the US equities side, there's no such thing anymore as naked DMA (theoretically anyway). The way it works now, DMA has to have certified risk checks that meets all the requirements of regulation 15c35. Every PB is required to make sure each sponsored client has some sort of hardware or software solution sanctioned by them which hits every check before orders go out. As a result, there are many 15c35 solutions out there either offered directly by the PB or pre-certified with the PB via third party. With some PBs, you also have the option of rolling your own and certfying directly with them. As a result, it's not nearly as wild west as it once was, so the assessments are much less likely an issue... unless you're asking them to get a little frisky with the intraday PMA.
- Nonius
- Posts: 0
- Joined: Thu Jan 01, 2004 12:00 am
- radikal
- Posts: 0
- Joined: Thu Jan 01, 2004 12:00 am
Sponsored DMA
What market are we talking about?
In Asia, sponsored DMA is pretty common. The exchanges place the burden of risk checks on the brokers, who are generally pretty lax in their application in an effort to attract the low latency business.
In the US, on the derivatives side, it's pretty similar. You have to convince your broker that you are doing all the necessary risk checks or you have to let them do them for you.
I think Zerohair's reply covers this exactly actually. One nuance is that generally margin stuff is only checked post on drop copy.
In Asia, sponsored DMA is pretty common. The exchanges place the burden of risk checks on the brokers, who are generally pretty lax in their application in an effort to attract the low latency business.
In the US, on the derivatives side, it's pretty similar. You have to convince your broker that you are doing all the necessary risk checks or you have to let them do them for you.
I think Zerohair's reply covers this exactly actually. One nuance is that generally margin stuff is only checked post on drop copy.
There are no surprising facts, only models that are surprised by facts