Mostly… fixed income and cross product eTrading

September 28, 2011

What we’ve got here…is failure to communicate

I went to see a theatre adaptation of the Paul Newman classic “Cool Hand Luke”.  It got me thinking about the speech paraphrased for the title of this post.

Right, so what has that got to do with Fixed Income Trading?

Trading requires a difference of opinion or of time horizon.  If I think a bond is worth 101 and I am happy to hold to maturity and the best bid I can find is 101 then I will hold the bond.  If you think the bond is worth 101 then you will not buy it from me.  In time horizon terms, I may not be able to hold the bond until maturity as I need to fund redemptions from a fund, so I have to accept less than 101. Ok, markets 101 over.

The majority of instutional real money asset managers I have worked with or know about seem to have pretty similar investment models.  And valuation models (or just Bloomberg ALLQ). So you are unlikely to find many trading opportunities among a consensus of valuations – only trades based upon different time horizons.

So – if you were trying to build a bond trading platform in line with then how do you actually get trades?

In a previous post I suggested that the key is RETAIL+INSTITUTIONAL=TRADES!

I think that the real trick here is structural; bear with me.

A retail business will have relationship managers who keep the client relationships going.  They are the people in the driving seat in some ways.  There is a dealing desk which is often seen as an extension of the back office and a cost centre rather than a profit centre.  At the back is a custody operation which one assumes will be making a few pennies from lending out stock in the same way that an institutional custodian will lend stock.

So – if you say to the wealth management dealing desk – put your positions onto the order book at limit prices and employ some kind of maker/taker model such that the desk can generate some PnL then you have a series of heads of desk who would be interested.

So what’s the relevance of the title?  Well, the failure to communicate in this case is that buy-side market participants are so segmented that there are multiple discrete segments.

Retail money is not interacting with institutional money

“Small institutional orders” run through the platforms such as BV, TW, MA, BLP.

“Large institutional orders” are on the telephone

So – if we have a market place (or multiple competing marketplaces) where all sizes of orders can interact in an anonymous fashion then we should see more trading – provided there are enough firms who understand the model and are willing to make the changes to their models and trading behaviour.

In my discussions on this subject I keep echoing the point that the equities markets can – in this specific case – teach the fixed income people a thing or two.  Construction of synthetic order books, smart order routers, partial executions, different order types (not just market or limit), market surveillance, compliance, audit,  exchange memberships and so on are all well trodden paths for the equities world. 

Oh, and if you fancy seeing the play, follow this link

September 16, 2011

Fixed Income trading – exchanges/MTFs/SEFs/same old stuff?

I try and follow what’s going on in the world of Fixed Income trading as much as the next person in the industry who has a related day job.  After a while the whole Dodd-Franks thing starts to make my eyes spin and I need to sit down with a damp cloth on my fevered brow.

Anyway, the question that I am mulling over is what happens next?

On the one side the Fixed Income hub firms are busy rolling out SEFs and looking to put CDS electronic.

On  the other side we see at least four firms trying to push the European cash bond markets onto an exchange-like model – following on from  

The four firms I know of are:

Bondvision MTS – part of the LSE group —

Galaxy – part of TradingScreen

Bondmatch – part of NYSE

Vega-Chi – private firm with a lot of ex-Goldman Sachs guys on the team

All four seem to have similar business models.  But is there demand for the idea of pushing good old fashioned cash settled bonds onto an exchange?

Over the last few years banks have starved their credit desks of capital to the point where most banks will only work an order rather than commit capital.  So if that reading is correct then it would be an efficiency play to disintermediate the sell-side and allow buy-sides to trade against themselves in an anonymous model.  This is a similar model to that of Liquidnet “classic” where real-money buy-sides interact entirely anonymously with Liquidnet giving up all trades to JPMorgan.

But are buy-sides ready?  One analysis is based on three components: systems, people and processes.  Do buy-sides have the three components to do this?  I suggest that many do not.

The critical mass needed for an exchange in this space is not clear – certainly it’s necessary to get multiple participants on the platform but that’s not sufficient.  In order to trade you need differences of opinion in value.  That suggests to me that what is needed are different classes of buy-sides.


If I was starting up one of these exchanges I would approach the big private client businesses and ask them to participate on the platform – to provide limit orders for the book of their inventory.  Allow for multiple small clips on one side of the book to be traded against one big order on the other side in a neat way.  Allow retail and institutional folks to dance on the same platform with protection against stubbed toes…

I would also ask institutions to place limit orders for all of their hard to value/hard to sell inventory.  So many institutions find themselves holding a credit that is traded by one man in a bank in Luxembourg and he’s on paternity leave.  Stick it on the order book and see what happens.  Let’s be realistic – it’s not like you are “showing your hand” anymore than you would be if you phone the Luxembourg bank up for a price – you might find that you don’t get your face ripped off.

Get a few trades done and then you are allowing people to safely dip a toe into the water.  Make sure there is excellent post trade performance and the venue will find more order flow arriving.  It’s not going to turn into a massive amount of liquidity in the way that FTSE100 equity names trade but it will become a self perpetuating virtuous cycle.  The velocity of circulation of bonds will increase as the liquidity is concentrated and prices are published. 

Eventually the platform can move to provide a tap auction / primary listing facility.

If you ever read about Drexel Burnham Lambert and Michael Milken you see that he and that firm were credited with popularising below investment grade or junk bonds.  How?  Simple – he sold bonds and he made a two way secondary market. Investors felt secure that they could sell if needed.  I am not suggesting that the DBL model is one to follow – but the provision of the optionality of a place to sell a position makes folks much more likely to trade in the first place.

And what will the banks do if they see a move to exchange trading for cash bonds?  I know of at least two big banks that have built fixed income “MTF-like” platforms.  I cannot imagine it would be hard for them to kick the tyres, light the fires and go-live with a competitor to the four firms named above.

“Build it and they will come….”

Blog at