The Evolution of the Buy-Side
Tue, 17 Jul 2012 06:29:00 GMT
An Interview with Rob Boardman
In this interview for HFT Review, Mike O’Hara talks to Robert Boardman, Managing Director and Chief Executive Officer, EMEA at ITG (Investment Technology Group), the independent research broker and electronic trading firm behind POSIT, the block crossing network.
Responsible for ITG’s European business, Mr. Boardman previously served as head of European algorithmic trading at ITG from 2006 to 2010. Prior to joining the firm, he spent 12 years at Goldman Sachs in various positions, including Executive Director on the Electronic Transaction Services sales team and Head of Connectivity for the equities division.
HFT Review: Rob, welcome to the HFT Review. As the markets have become increasingly electronic over the last few years, I’d like to start by asking you what you see as some of the more positive impacts of this “electronification” and what are some of the negatives, both from your perspective and from the perspective of ITG’s clients?
Rob Boardman: Thank you. I’ll address the positives first, two of which are pretty clear. The first is the general efficiency in our industry. Electronic trading has certainly improved the productivity of traders. And although it might have reduced the need for head count, which some will lament, overall it has made trading a more efficient activity.
As part of that, I would argue that it has also reduced risk. I know there are a lot of questions asked about the risks of electronic trading, but my view as a practitioner who started in the mid 90’s is that electronic trading is a much less risky activity than telephone broking.
I say that on the basis of looking at the rate of errors you get from taking orders verbally versus via a FIX connection of any type, whether via algos, DMA or even just routing to a desk, the error rate of verbal communication is at least 10 times higher. So that’s the first point.
The second one is perhaps a bit more subtle. At ITG we have a business that measures the transaction costs of very large institutional asset managers, people trading very large blocks of stock - tens of millions of dollars typically per ticket - which is both an art and a science requiring quite a lot of analysis, so it is not immediately obvious what the total cost of the transaction is.
HFTR: You’re talking about market impact cost here rather than more obvious costs such as commission, yes?
RB: Correct. And there are very subtle ways of measuring market impact. But whichever way you do it, if you plot the average and total cost of trading over a quarter for example, and then go back 10 years or so, there is definitely a very significant downward trend in the total cost.
That may be due to capital markets becoming more efficient over that time, not only in terms of transport of messages but at a much more fundamental level, i.e. in that the provision of risk has become much more effectively distributed to a larger number of players.
And high frequency trading has played a part in that. It’s not universally a good thing but it has certainly added more liquidity to some markets. So when you look at that curve, where the total cost of trading has reduced over the time that markets have become electronic, it’s hard to argue that HFT has poisoned the market in any way, because the statistics point to the opposite.
HFTR: You’ve given me a couple of positives there. What about the downsides?
RB: There are certainly more fears around systemic risks. The risks of individual things going wrong are reduced because you don’t get these communication errors, you don’t get people “fat fingering” buys instead of sells and those sorts of things, but you do get some different types of risks emerging around group behavior.
When you have the ability for firms to all trade momentum in an unrelated way very quickly and to exchange messages at bewildering speeds, the potential for group behavior to move markets very suddenly is accentuated. So markets have to learn how to adapt circuit breakers to a world where there’s not only incredibly low latency but also very highly correlated behavior amongst stocks and even asset classes, with the aim of preventing another “flash crash” type situation.
As for other downsides, we do hear anecdotal but nevertheless very consistent complaints from asset managers, backed up by statistics showing that in some circumstances, their engagement with high frequency trading can be negative relative to interacting with other types of liquidity. Our own research tends to support that as well.
That’s a very different conclusion to saying “HFT is bad” though. ITG as a firm doesn’t believe that all HFT is bad and we’re certainly not a supporter of outlawing HFT. Even though we serve mainly a buy-side audience, that’s not our view. We try to be enlightened on the subject.
HFTR: What type of evidence points to HFT being negative for the buy-side relative to interacting with other types of liquidity?
If you are not a registered user, click here to register for free