March 14, 2009
I’ve written several times before about Bloomberg, and why they’re such a blot on the etrading landscape. Now we’re in a recession and Bloomberg’s clients all want to cut back on those grand a month terminals. More and more I hear folk say “why pay a grand a month for an email account ?”
Bloomberg’s difficulties should be opportunities for those involved in proprietary single dealer channels. The advantages of single dealer channels for dealers are obvious. And they can be attractive to clients too – they’re free to clients, and can carry innovative offerings that the multi dealer ECNs aren’t nimble enough for. But what they lack, by their very nature, is the direct real time price competition of the multi dealer RFQ.
Could there be some form of mitigation for this handicap ? Two avenues seem possible. Firstly the moribund LiquidityHub consortium. The tech infrastructure is in place for composite prices for a whole range of rate products. The consortium could agree to allow members to republish LQH composite price on their single dealer channels. Secondly, vendor driven aggregation. Single dealer channels are now common enough that vendors have sprung up to service the sector: Caplin & Lightstreamer for instance. The vendors themselves could perform the price aggregation in portals. Dealers could reuse vendor supplied aggregate or composite prices to make single dealer channel tickets MiFID compliant for the buy side by enriching them with the portal prices.
March 14, 2009
I’ve finished Joel Hasbrouck‘s Empirical Market Microstructure, and am a little disappointed with it. It’s unlike Harris’ Trading and Exchanges – it doesn’t present detailed and accessible history of market evolution and convention. It’s also unike more philosophical texts from Taleb or Niederhoffer. And it doesn’t present a sustained argument for a particular thesis on markets, like Malkiel‘s Random Walk. Instead, it’s a survey and summary of the current aademic state of the art in market microstructure, which is admittedly a very new field.
I’ll illustrate my view by reference to two of the later chapters, on topics that interest me particularly: limit order books and depth. Chapter 12, on order books, discusses limit vs market orders, and the Parlour model for them. A lot of maths is deployed to demonstrate that a lot of depth on the bid side decreases the probability that a buyer will use a limit order. Which is so obviously true it is completely unenlightening for the practitioner. Chapter 13 on depth is a bit better, and offers some worthwhile insights on bid offer price formation due to backfilling after a large order has walked the book. But again, a lot of maths is deployed to achieve trivial results.
I’m also marking Hasbrouck down for not mentioning the Penn Lehman project. Penn Lehman produced papers like this – a great blend of practitioner and academic approaches yielding lots of insight without too much over formalized maths which is only there through physics envy, and to shout “this is real science !”