The Gateway to Algorithmic and Automated Trading

Conference Take-Aways: TSAM Trading Technologies 2010

There is always so much going on at Osney Media's annual TSAM conference on technology solutions for asset managers that you wish you could attend all the streams, but in the end you have to choose. As chairman of the trading technologies stream my choice was made, but it proved fortunate as delegates this year were in a serious investment mood.

Bob Giffords

Bob Giffords

Independent Banking & Technology Analyst

Most people were assessing new OMS/ EMS technologies, driven it appeared mainly by demands for multi-asset trading. However, budget pressures to streamline front office operations and the continuing gravitational pull of low-latency appeared to be important factors as well.

So the discussion around the improved release 1.1 of the FIX algorithmic trading definition language (FIXatdl) proved particularly timely. The aim of FIXatdl is to allow rapid integration of new broker or third party algos into OMS/ EMS software tools by standardizing the interface semantics and validation rules. This enables fairly automatic generation of GUI interfaces and reduces the likelihood of errors with more hand-crafted solutions.

For me FIXatdl meant that the half-life of algorithms will probably halve once again as cost and risk barriers to innovation continue to fall. It should also facilitate the next generation of algos-driving-algos, enabling trading desks to automate the selection, workflow and real-time supervision of execution algos. It may also encourage the spread of e-services in general as FIX becomes the default front-office lingua franca. I would certainly want to include FIXatdl support on my tick-list for OMS/ EMS vendors in future and perhaps for others as well.

There was also a lot of discussion about fragmentation and the confusion in the market as to how important dark pools really are. Speakers discussed how CESR's continuing intervention is generating a lot of cost and heated debate as firms juggle their systems to stay within the rules. I was therefore not surprised to hear that dark pools in total are still only around 4 per cent of European volumes and broker dark pools barely 1 per cent, but then what would you expect given regulatory restrictions to large-in-size, non-displayed orders and now on pricing to EBBO or mid as well?

More interesting perhaps was the figure another speaker put forward that London currently makes up between 33 and 45 per cent of the European order book. Given the rapid clustering of MTFs around London it is scarcely surprising, but the gravitational force of that liquidity will clearly impact trader strategies. My report on proximity trading hubs for Automated Trader explains why.

One person asked 'who is making money out of all these MTFs?" The speaker agreed that with 134 of them there would inevitably be a shake-out in the next couple of years. While initially total costs came down with increased competition, it appears they have risen back to 2008 levels. Apparently TCA and index data providers are struggling to keep up with all the data sources and diverse practice. There seemed general agreement that the regulators should focus much more on post-trade transparency standards, if MiFID is to be a success.

There was a fascinating description of a systematic trading framework which is clearly now able to scale to respectable capital sizes. That prompted the memorable observation that perhaps discretionary portfolio managers are now more agile than their quant-driven colleagues. Since systematic algos are only modified after evidenced weak performance and extensive simulations and back-testing of alternatives, the machines will clearly be slower to respond to changes in market regime than two eyes and a brain. We might think of that as model momentum.

Another bon mot was the comment that risk management is really about not being wrong. That of course is why risk becomes such a political issue.

However, the real highlight of the day for me was the brilliant insight that the industrialization of managed accounts with systematic trading algos could revolutionise the investment industry and shift power back to the original investors. I saw it as a kind of social lending on steroids, a democritisation of the whole industry. It could increase transparency and reduce the costs, risks and complexity of managing large investment funds. It should increase diversity too and with it liquidity in the market. Thus it could have a potentially favourable systemic impact. This requires much more serious investigation.

Indeed I found many interesting ideas buried in the remains of the day. For example, in a private discussion someone suggested we might see many more 'investor travellers' in the market: small firms or even private investors with highly opportunistic trading strategies that will keep 'moving on'. This is just another example of how technology could democratise the markets.

In the plenary discussion one speaker warned of potential trouble ahead with politicians and the media now intervening in detailed industry regulation in ways previously unthinkable. There are clearly now 'more moving parts' to the compliance puzzle and much more political event risk. Food for thought. [See "Regulatory Roadmap - Whad Next?" for some insight]

However, one final observation perhaps summed up the optimistic mood of TSAM 2010 better than any of these. Someone noted that the technology advantage of the sell side was being competed away as the buy side moves to large-scale, third-party vendor solutions. We had heard much about e-services that combine real-time, low-latency opportunities with economies of scale and scope. The new world beckons, and TSAM 2010 demonstrated once again that many people are listening.