Algo Trading

Tabb Group believe that the spend on low-latency technology is set to rise over the next three years. Not sure that is such a revelation, even in these troubled times.

Why do I say this? Well, below are the positives and negatives.

Project Fusion: having direct dealer ownership sures up liquidity in the face of the likes of LiquidityHub
– The no. 1 FI e-trading ECN
– Thomson-Reuters merger allows for greater resource access (esp. via Reuters salesforce in Asia and other EM’s)

– lower volumes in Govt bonds, their bread and butter, due to the credit crunch
– these volumes are under further threat from Hedge Fund interest in joining MTS
– in over a year TW’s credit offering has failed to impact MarketAxess or Bloomberg
– despite the no. 1 status, TW’s functionality and architechture has failed to advance to any great degree and it is difficult to see this happening in 2008 (see next point). This ranges from the simple (market depth) to the more complex (API to accomodate algo trading)
– Thomson-Reuters merger means TW could get caught up in big firm bureacracy

Overall positive but not without a few things to watch out for.

A brief piece here about access to inter-dealer markets for Hedge Funds. Now widely practiced in FX, via Prime Brokerage offerings such as on ICAP’s EBS and Reuters TBS, but resitance remains in the band market.

Cantors makes some odd comment about access for non-bank clients on eSpeed using Algos but not Hedge Funds……

In some way connected to my previous posts on the effects of recent market volatility on Fixed Income e-Trading and Hedge Funds on MTS, Bloomberg have a piece today stating that US Tresury volatility is at 3-year highs.

“The retreat by so-called black-box traders and hedge funds caused orders for Treasuries to drop as much as 80 percent, said Mark Ficke, senior managing director at ESpeed Inc., the second- biggest interdealer broker.”

Now they measure liquidity here in terms of size, normal order size of $500m is now $100m in 2-year notes. However, it looks like my view that market volatility should mean increased volumes may not ring true on eSpeed in the States…….but the article mentions that Fed data indicates August averaged $767 bn a day in trading, a record. Then a trader from Mizuho in The States screams “I JUST WANT LIQUIDITY WITH MY VOLATILITY!”.

So volatility is high, volumes are high (so I was right, just eSpeed is down due to algo seizure) but liquidity is low.

We are aware that MTS are still looking into ways to allow Hedge Funds/Prop Houses direct access to the MTS markets.

A question for the readership. I assume the proposal is “direct”, as in they trade as Citadel, for example, rather than a Prime Brokerage arrangement where the trades take place as Bear, Goldman, Morgan Stanley etc, etc? Much as Reuters and EBS do in their FX PB offerings, including an API to plug in algo models.

I assume this because the banks could slam the door shut quite easily on a PB/Third Party model.

What is the eSpeed model in the States? Do the Hedge Funds trade in UST under their name or under their Prime Broker/s?

I’ve just had a hard copy of the Bearing Point Electronic Bond Market 2007 survey come across my desk. No idea when it came out, other than it was this year. I can’t say I was aware it was out beforehand but had seen their 2005 version.

I won’t go through it in detail, as much of the detail is going over old ground, however there were some interesting findings on the share of e-volumes in B2B and B2C markets as well as some potentially provocative points in the conclusion. The survey respondants, as far as I can tell, were sellside only.

* B2B Volumes (where the only Cash Bonds are involved in the transaction, i.e. no bond/futures basis trades): MTS 80%, HDAT 15% (Greek market that big?), Eurex Bonds 4% and Senaf 1%. Volumes stagnated between €16-18 billion a day (voice ex. basis trades estimated at €17 billion a day) with little scope for volume growth in the next few years.

* B2C Volumes: accounts for 40% of Euro Government Bond Trading (grew by between 7-10% in 2006). TradeWeb accounts for 52%, BondVision 21%, Bloomberg (ex. single-dealer) 21% and Other (RTFI etc) 2%. Doing some rough maths this should mean approx. €10 billion per day (sounds about right) in B2C Euro Govt e-trading, however the 40% doesn’t quite add up. I can only guess they mean 40% of customer flows rather than all turnover.

* The survey expresses some surprise at tighter spreads in B2C vs. B2B. However that is easy to explain as banks are happier to price for clients than competition, especially when they are obligated to quote at a certain min. spread (MTS) in B2B and that min. spread is no longer reflective of the broader market. In saying that, good luck to MTS if they chose to tighten it!

* Longer-Term Conclusion:
– US Treasury market still dwarfs Euro Govt market (approx. $500 bn a day versus equivalent of $90 billion a day) due to European fragmentation. One solution is for EU Debt Agencies to form a Super Debt Agency (call it the EUDA) where all Govt debt is issued by one issuer (say the EU) which will provide more liquidity and lower the funding costs of member states. Hmmmm, I’d guess that has as much chance of happening in the next 10 years as peace in the Middle East. I like the idea but how do you get agreement on how to split the €€€€€’s between the nations? Anyway it would solve the problem of members breaching the Stability and Growth Pact! 🙂
– Banks should utilise Algorithmic Trading along the lines of UST market and introduce non-bank players – much as MTS have thought about – thus beggining to merge B2B with B2C and also bringing in voice (via a Bond version of Swapswire) to form an exchange model.

The conclusion offers major changes to the structure of the European Bond Market, however it is a bit short on detail as to how this would actually work in practice. Not sure they’ve given it enough thought.

From today’s FT regarding MTS allowing Hedge Funds into the inter-dealer market.

Must be a slow news day as this says little we didn’t already know…..

Next Page »