Featured post

Fixed Income Trading: New venues

A simple question came up recently in a conversation – how many new Fixed Income trading venues are there?  I could not think of anywhere th...

Thursday, 6 March 2008

What Do You Want From A Hub? (March 6th 2008)

Another post from John Greenan …. anyone want to volunteer something that fits the bill?
What do you want from a hub? Let’s clarify what we mean here.  In a classic equities buy-side connectivity world the institutional asset manager would generally choose between point-to-point connections using BTRadianz/TNS/similar or a hub such as Thomson Order Routing, Reuters Order Routing, FidessaExpress, UL Net, NYFIX etc.

In the past the measure of value of a hub was seen in pure techie network terms – that the value increases as the square of the number of nodes.  That worked fine in a world where buy-sides were busy trying to add as many brokers as possible in the initial stages of the move to FIX based rather than telephone based trading.  But what are the dynamics in motion now?
Unbundling means that you do not need to deal with a broker in order to pay for their research.  So why bother with multiple brokers?  If you look at the execution capabilities of a firm such as Goldman Sachs or Morgan Stanley then it’s hard to see what else you need.  For simple risk management reasons it makes no sense to put all your eggs in one basket but the optimal number of brokers is perhaps now a great deal smaller than the consensus opinion.
Let’s put forward a new model of broker relationships – core and satellite.  Have a small number of core brokers with whom the buy-side maintains deep and broad trading relationships, multi-asset class, algos, programs and so on.  Have a number of satellite brokers that provide specialist execution capability – illiquid German mid caps, access to smaller markets where the tier one firms may not have a seat.  In this core satellite world how many brokers do you need for a global institutional asset manager?  Deal out of the US, EU and Asia Pacific.  Have say 6 core brokers and say 8 satellite brokers per region.  A total of thirty connections.
In this case it’s never mind the width, feel the quality.  Rather than a massive number of connections, lets go for the smallest number but have in depth trading relationships.  What can the hubs offer in this quality, rather than quantity, driven world?
What about a hub that offers 24×5 uptime based upon the hub having two connections to a broker and the buy-side having one connection to the hub?
What about a hub that offers crossing capabilities by default based upon being able to see all flow routed down the pipe?
Another issue, end of days.  In a global trading environment it’s hard to fit in end-of-days for sequence number resets.  Either you have a five minutes slot where a broker is unavailable or else you build two or more connections such that there are regional connections so an end-of-day can run after the business day has finished.
Intelligent management capabilities.  Monitor DKs and rejections and try and add value – notify people and let an investigation begin.
Management statistics.  What percentages of orders go where, which algos are used.  Can any patterns be determined?
Real time and post trade TCA.  How well is the firm performing in executing client orders?
I don’t think that a single hub exists at the moment with all these capabilities but it’d be nice to see it soon…
 
[Original post http://mostly.wordpress.com/2008/03/06/what-do-you-want-from-a-hub/]

No comments:

Post a Comment