January 14, 2012

Bonds: Part 4 - The Future

Tags: Finance, Bonds

So after nearly 7 years on EGB technology teams I have moved on. Not out of banking (at least not yet), but when the opportunity came to try a different business area I jumped. This is despite being able to earn more money if I stayed in my EGB niche. There are two related reasons for leaving: the industry and technology are both changing to the detriment of developers.

When I started in 2005, EGB trading was already on an upswing that lasted through 2009. I’m not sure when this upwards trajectory began, but the people around me considered it normal. Business was good throughout the mortgage and banking crisis of 2008. When times are good and the desks are making good money some of it is reinvested into technology. There was a lot of optimism and this translated into banks wanting the fastest and smartest trading platform. We got to rewrite systems. Increasingly business was relying on technology - it was becoming trade-by-wire. It seemed that there was much further to go. Algorithmic trading is big in equities, why couldn’t it be big in fixed income too. There was huge potential for writing interesting code.

This environment no longer exists. It was only when the European debt crisis began to bite in 2011 that things trended downwards, but then they went down fast. There is a great deal less ambition among the fixed income desks. Most traders seem happy just to not be losing money - something many are not achieving. Some banks are shutting down their bonds desks. As a natural consequence there is less to spend on technology and little desire to do more than maintenance. Flow volume is king, and as far as technology development is concerned, that is not hard. Thus less developers are required and the work is less interesting.

Combined with this is the creeping commoditisation of bonds technology. When I started ION provided mainly gateways, little tools and their message bus. Over time they have climbed up the product stack. Now they also sell pricing engines, autoquoters, autohedgers, risk engines, position servers - nearly everything a bank needs to trade cash bonds. The quality is fine too. Their applications won’t do everything a desk requires, but it will do most of it and the desk’s technology team can add the remainder quickly. Not all banks will use vendor systems (many don’t like Ion or see advantage in bespoke systems), but enough will to affect the market for EGB technologists.

Thus, once again, less developers are required and the work is less interesting. I don’t see the situation changing until the market improves - and I don’t see that happening soon. So I thought it time to try a different business area.