Printer Friendly
Ciaran began his career as an electronics engineer in South Wales. After becoming very interested in software development, he joined Salomon Brothers in London and began his foray in the financial arena. Focusing a good deal of his career on supporting derivatives and fixed income, he is now embarking on a new landscape of uniting the equities and fixed-income technology groups at Credit Suisse First Boston where he has been for six years. Formerly global head of fixed-income technology, he is now overseeing client technology for the securities business, which is comprised of sales, research and client connectivity across equities and fixed income. I sat down with Ciaran to discuss his priorities.
WS&T: What do you see as your top three issues for your trading department right now?
Henry: Top three issues--one is handling volume. We are seeing lots and lots of smaller trades. We have automated systems that generate from a single order, several hundred executions and just handling that through put end to end is a big issue.
One of the other issues is that with more client business when clients want to trade direct to market--straight on the exchange via our infrastructure--we are having to put a lot of thought into controls so the wrong kind of orders don't end up in the market and get executed. That's kind of a new thing that we've had to put in place
Third, I think that whether your talking about fixed income or equities --over the years of the bull market-- like many firms, we have built up a lot of separate systems for a particular product, and sometimes for a particular region. So, getting the cost base down by consolidating those systems to fewer and more global solutions is the other thing that we are focusing on.
WS&T: How many systems do you have?
Henry: (He laughs) I don't know, I don't know. It's a scary number. It depends how far you look. If you look at the whole front to back world of securities, you're talking in the hundreds because there are often middle subsystems that go up to reconcile links between two independent systems and those multiply very quickly. So if you look front to back, across equities and fixed income you get quickly into the hundreds.
WS&T: How many would you like to have? Is there an ideal number?
Henry: Well, we know it can't be one. It's not achievable (laughing). I think what we are trying to do is converge on front-to-back solution for what we see as the major product cluster. So a front-to-back solution for derivatives, both fixed income and equities. A solution for prime banking that uses a lot of technology for other areas. Cash fixed income and cash equities we would like to try to leverage as much as possible, what's common across both product groups, but it is unlikely that it will be one single solution. It will be more of a hybrid. That's enough for now.
WS&T: OK, now taking a look at those three issues, how will you use technology to achieve your goals?
Henry: The volume one has been interesting because we had to apply a new approach. Previously you would expect hardware would save you. ...Computers keep getting faster all the time, if you have an unlimited budget--just keep throwing hardware at it. That doesn't work this time around because the jump of volume of orders is several orders of magnitude and we don't really see an immediate plug in for that. So the way we've approached it is, we've actually used a performance engineering approach where we use a tool with a specialized product that allows you to create a model of your trading system.
You can think of it as a wind tunnel almost, because on your screen, you create a representation of your system and you calibrate your model so that the amount of transactions flowing through it, and how you respond, is very close to your real world performance. But the next step is the key one. Within the model, you can make changes. You can increase capacity in different parts of the system. You can break the system down. You can change the hardware. You can do all kinds of things and then rerun your test and see what that does to the through put and then find the next bottle neck, and eliminate it in the simulation. So, to deal with volume, one of the key things we've had to do is to apply this performance engineering approach to the key parts of the system where we think there will be the most impact, and try to predict the next three bottlenecks we might face and engineer around them before we get there.
WS&T: Does anyone else do this?
Henry: It's common in software development away from Wall Street. It's a common approach in engineering circles. As far as I know, no other Wall Street firm is using this technique as we are right now -- to point out bottlenecks in trading systems.
The other thing that we have done on the volume side -- is it has spurred our interest in blade servers in the Linux platform as a way to get more parallelism through our infrastructure. So in conjunction with the performance-engineering work, we've been expanding our use of blade servers in different areas to see what that yields in terms of overall capacity and through put.
So the hope is that through meticulous engineering and a preemptive approach to volumes, and making sure the platform can withstand the demands on it, we hope to stay ahead of the demand curve.
WS&T: When did you start using the performance measurement technology?
Henry: We started using it in June 2002. We use a vendor called HyPerformix. There are other vendors that do it, too.
WS&T: How about the other two issues--compliance/controls?
Henry: That's been interesting. We've actually had to build a whole new layer of functionality in the transaction cycle -- very far up in the cycle -- right out to where the client FIX connection comes in. We always had rudimentary controls there but, it has just become more and more sophisticated. So we had to build, essentially, this whole new layer we called the compliance engine. That is something we had to build to do various checks on orders coming in: Are they suitable for the market? Is the size too large? Is there a credit risk issue? Does the price on a limit order make sense given where the market is trading? Could someone have transposed the size and price field?--It's happened, not here, but it's happened in practice.
WS&T: Did you build that yourself?
Henry: Yes, we had to build that inhouse, but we do use a lot of third-party middleware and market-data feeds. The usual workflow products you find on the street are not fast enough for the real-time transaction rules. The business rules are complex and changing all the time.
WS&T: Is that something that falls under your risk management dept or trading?
Henry: That's under the trading side.
WS&T: Do you find compliance technology is becoming more important with these issues, anti-money laundering requirements and other initiatives--
Henry: Yes, definitely. Trade surveillance is an interesting area in which we are investing in a big way and I think the trade-surveillance systems that are now coming into the market are very sophisticated. Not only do they allow you to have a complex set of rules, but they look at certain patterns in trading that are inappropriate. And by their nature, they sit on top of your trading systems and see every transaction that is going on. These systems have potential in other areas, which could be calculating best execution for example or even consolidation of positions and so on. We are making a big investment in that area.
WS&T: How much of any investment?
Henry: I can't give you those numbers.
WS&T: How about number three--consolidating systems.
Henry: Yeah, that's a challenging one. We have a number of projects under way that are looking at this. We used to be organized with fixed income and equities as very separate, self contained units, and we have broken up the organization. So now we have a derivatives organization that functions back to front and that group is in the process of implementing an entirely new infrastructure for derivatives based on a combination of third-party platforms. The Panorama product, it's no secret, and some inhouse technologies we have built for equity derivatives. That project has been underway for several months and we think that we will see some big dividends there in 2003. That world has depended very much on spreadsheets. That is the way it has developed. We need to eliminate the spreadsheets and improve controls and increase the volumes.
On the fixed-income, equity, cash-trading side, we have a very fast order-management system for equities and we are right now looking at how we can extend that into fixed income further. It is a complex task because the products look quite different, with the exception of futures and foreign exchange. But we think we can get there over time. That effort is in the design stage right now. We have done a number of things to physically start the process of consolidating.
WS&T: What is the impetus for doing this--is it to cut costs, to improve efficiency, both?
Henry: It varies a little. In derivatives we know that with the growth in the credit derivatives markets that has happened over the past few years, we feel that we need to have the right scale in our process and infrastructure to be able to be a player in that business of the size that we want. In derivatives we know it's about making sure that we can take advantage of the business environment, but it's also about controls. When you're running a business that relies heavily on spreadsheets a lot of things can go wrong, so that is something you want to get away from.
So, I think in derivatives it is about capacity increase because of demand and control.
In the cash, fixed-income and equity side, I think that in response to what has happened in the market, we are really looking at cost reduction and consolidation and that is the primary objective. Although it is also true that over time, with electronic trading becoming gradually more prevalent, they (cash and fixed income) start looking more like equity markets as well. So, some of the products are seem to be converging.
WS&T: How open were your traders to trading electronically and how are they now about embracing new technology?
Henry: We started in fixed income in 1990 with the Bloomberg GovTrade product. So that channel has been available. But it wasn't until TradeWeb came available in 1997, and 1998 when it gained traction, that fixed-income electronic trading was a reality of significance. I think the adoption and the reaction is different for each market.
I would say for the beginning of TradeWeb it was hard for people to get used to that style of working. When we started on TradeWeb, we were in competition with five or six other dealers, We hadn't done that with Bloomberg. We had systems for producing prices, but they weren't heavily automated, and people wanted prices back very quickly. So we discovered, very quickly, that we had to automate the production of pricing. The first thing that really had to change in the middle of 1998 was putting in auto quoting. That was a big change for the desk and that took some getting used to and fine tuning to get right. As it stands today, we can only use it for trades of a certain size--smaller trades. We don't feel that we can use it for large trades. Large trades move the market and you need to have a sense of liquidity and so on.
On the equity side, although FIX has been around since 1992 with Fidelity and Salomon, it really became something we focused on in the middle of 2000. It was only at that time that we saw enough client demand to provide that service. I think introducing FIX in equities was more of an add-on. Trades were coming in through electronic pipes but traders didn't do anything differently. They went to the sales trader or they went to the market. So there hasn't been the paradigm shift that there has been in fixed income. And that might be why between middle of 2000 and today we have gone from 0 to 500 clients connected via FIX in equity trading.
We regularly do, every week, about 65% of our orders via FIX and about 25% of our shares.
WS&T: What about the other 35%?
Henry: Most products, with the exception of structured products, have some sort of electronic system out there. We trade eurobonds on Bloomberg for example. We have our PrimeTrade system that does futures, foreign exchange and also some bond markets. We have a convertible-bond electronic-trading platform. We use companies like MarketAccess that cover the credit markets and multi-dealer systems, and TradeWeb has expanded beyond treasury. So, I think a lot of the products that are essentially commoditising products. There is already some form of electronic-trading channel available and we are present on all of those.
For the structured products, things like OTC derivatives, we are part of Swaps Wire but that's more of a post-trade documentation rather than a pretrade effort. I'm not sure if we are going to see anything pretrade for OTC structured products any time soon because the process is very complex.
WS&T: What ECNs do you use?
Henry: I don't know the exact mix. We have a system called pathfinder that will execute in 20 or more liquidity pools or ECNs. So I know we are present in all of the popular ones.
WS&T: What is your view of Supermontage.
Henry: I don't have a view. I'm waiting to see what happens.
WS&T: In your view how long will it take FIX to catch on in the fixed income markets?
Henry: Look at the history in equities, what does that tell you about some of the take up and challenges, and take a look at some of the current landscape in fixed income and the appetite for further investment in fixed income and the appetite for further investment in technology for electronic trading. When you add those things up, it looks to me to be further out than we would like.
If you think about when FIX first came out in '92. The bull market spurred on the investment. It took a while. The current version is 4.3, and most of our connections are still running 4.0. We'd like it to happen more quickly because it is an efficient way for us to reach our clients. We've invested in a lot of infrastructure for it. We're ready for it. But the buys side has to be ready to make the investment in an order-management system, or some sort of interface of their own to be able to communicate with us via FIX. And considering the current market climate, it is hard to see that happening any time soon.
The other thing that needs to happen is FIX 4.4 -the fixed income piece--is a very rich set of functionality--it's a non trivial upgrade from 4.3 to 4.4 So that's going to take more time than previous upgrades. It's implemented the BMA's (Bond Market Association's) business rules.
If you add to the picture, FIX grew up in equities. And now FIX is emerging in fixed income and at most firms, and they don't get anything out of 4.4. I'm not sure if they are in a position like we are where we have both equities and fixed income merged together. FIX is owned by the equities side. It will be difficult for a lot of sell-side firms to mobilize quickly.
WS&T: Someone at the conference (TradeTech for Fixed Income in London) said it would be about 15 years before it gains any real traction.
Henry: I think the first four or five years, was before it really took hold. Yeah, in order to reach that inflection point where it is in equities--where you are not leading edge anymore, "you're a laggard if you aren't doing this"--all those factors I described will take three years or possibly even more to get there.
WS&T: You mention that it is up to the buy side to take part in order for this to be successful. Will the sell side help the buy side to implement order management systems?
Henry: What needs to happen is we need to see some real demand from the buy side--that's when I need to do things. If, in order to make that happen, you have to somehow share the cost -- the investment-- that's a possibility, so I wouldn't rule that out.
WS&T: Is anyone doing that?
Henry: I think firms do it, I know we've done it even in equities when it comes to a situation where it makes sense to both parties to get someone started. It's not a normal practice but I think it is an accepted practice that happens from time to time.
WS&T: Is it mostly the smaller firms--who have smaller budgets-- that need assistance? Or larger ones, who you do more business with, that you would be concerned about?
Henry: Research says that groups that are most FIX enabled are the larger ones, but smaller ones are catching on quickly too. There are platforms out there that are a relatively low cost of entry. They are paid for on a screen by screen basis and are even hosted so its not something that they need to build. It has been trickling down. But it is the larger firms that are in the 70% category, not 100%.
WS&T: What is CSFB doing to achieve STP in fixed income?
Henry: I know that we have done a lot internally in creating a seamless front to back process and I know we've gone a long way. One of the things we got from putting our electronic-trading system on a sales desk, and mandating that the sales force uses it for price inquiry and trade entry for certain products and trade sizes, is that it took our cancel and correct rate down from about 17% to 8% over the course of the year. We want to do more of that.
Certainly internally, we do as much as we can because other than the error rate reduction there are benefits for us as well.
WS&T: Would you say, when it comes to technology, that CSFB is a pioneer or a does it take the wait and see approach?
Henry: I don't think we can afford to take the wait and see approach. We're in a competitive industry and I don't think we can be called laggards. I think we are market leaders, early adopters in general. If you go back to the history as I said we first started by putting bonds on Bloomberg in 1990, and TradeWeb was actually created out of CSFB as well as Brokertec. Along the way we've seen development of the FIX protocol as part of what BMA is doing. I think we've tried to drive technology so I would say we are more in the pioneer category than in the market-laggard category. One of the things we've gotten better at is judging the right level and pace of investment. In the boom years it really didn't matter--time to market, spend everything, go out try to do everything at once. Our philosophy now is different, we want to be market leaders, but we're judging where we are really going to be a leader and where we are going to hold off a bit, if it makes sense. We passively decline to lead on all fronts.
WS&T: Any areas where you have designated that you want to be a leader?
Henry: I think FIX for fixed income.
Others will give you a rather pessimistic view of adoption. We have clients today transacting fixed-income products via FIX with us and we have been able to create that within the variance of the FIX specification with a small number of clients. We want to continue to do that because that type of thing helps to gradually help with the broader industry adoption. Because we have taken the equities infrastructure and fixed-income electronic-trading infrastructure using fix protocol internally we are ready for that, perhaps more ready than some of our competition is.
WS&T: What firms do you view as your main competition as far as keeping up with technologically?
Henry: I don't know if I want to answer that. You know who our competition is, the same peer group is giving us the competition technology. The usual suspects.
WS&T: Who do you look to determine where you go from here? Vendors, consultants, internally --how do you monitor where you are and where you should be headed?
Henry: I think we have two things. We have a very forward thinking business group here whether you are talking about banking equities or fixed income. Everyone is very creative and that always generates, very quickly, new business ideas and concepts so that is something that has been a catalyst for the work we have done.
On the tech side we have an organization that does research of technology trends -- what platforms are gaining traction, what standards are beginning to emerge and so on. That's a powerful group internally. We use that as a resource.
Outwardly they are explaining to the tech department what is coming up next. They also act as an internal consulting group. We can go to them with something brand new or a problem and they will have some very smart technology resource to apply to it. So we have this chief technology officer group to do research on what is happening in the technology market and bring that in to the department and work closely with the business units to think of the next big ideas.
WS&T: Is it a group made up of CTOs?
Henry: No. It's a group--party of a strategic architecture and engineering department focused on mapping tech trends and providing consultancy.
This group is focused on pure technology not business. They are the group that introduced Linux to the firm. They do things like that. They focus on horizontal technologies.
WS&T: How do you share ideas among different product lines and departments throughout the firm?
Henry: We are coordinated internally. The management team is structured in such a way that all the product lines are represented as well as the infrastructure groups and we talk every day and we meet formally once a week. And we have a strategy, which is written down and we review it together and share our plans. All of that is pretty much out in the open.
In terms of the business units, there are multiple levels of engagement. There are steering committees that drive all of the technologies for equities, all of the technologies for fixed income. There is so much dialogue going on it is hard for me to put it into a nutshell.
WS&T: Seems like other firms have trouble in that area--communication between groups.
Henry: Well I think what you have to do is set out to be coordinated and that's challenging. And what I just described to you about combining fixed income and equities is very hard. We think we understand the risks, we've explained to the business units what the benefits are and it is up to us to implement it.
WS&T: Who do you report to?
Henry: I report to Phil Cushimoro, he is one of the co CIOs here. The other CIO is Mark Anderson. Mark's focus is on infrastructure and Phil's is on application development so I report to Phil. But really I have two bosses, at least.
WS&T: How did you get to where you are today?
Henry: Ah, whew, I don't know. I'm an electronics engineer. I used to design modems, multiflexers, etc. and I was doing both software and hardware design.
I got really interested in the software side and the financial world was investing a lot in software development, so that's why I went there. I think I've just been drawn to the trading floors and the process of business that goes on. It's likea complex machine with lots of parts. I naturally feel comfortable working in that. As an engineer, I like to make things quicker, I like to make problems go away.
I started in electronic engineer in South Wales. When I moved to London, I didn't know what a bond or a stock was. I worked for a company called Salmon Brothers, which is well known today, and that was a big eye opener. I started in the IT department on small software development projects for equity options, and eventually ended up being one of the key people on the team to build the trading floor here in New York at 7 World Trade Center, which isn't there anymore. That was a huge project for us. Because what Salomon had done is experiment with UNIX workstations in '87. We needed to move into that building in 1989 or 1990. We were tasked to find out if the new platform, which had only been tested for 20-25 machines could work for the entire building and we had a year and a half or so to figure that out. It was a bit like the space station.
We did a 25 workstation pilot in London, then another 100 workstation pilot in NY, and then a 400 workstation pilot in NY and then we knew we could make a bet on (all of them).
WS&T: How long have you been at CSFB?
Henry: Six years.
After having spent about 5-6 yeas in the industry I wanted to understand trading sales better. So, I went joined Lehman brothers and worked in fixed-income research. That got me very involved in the business side.
WS&T: You've never been a trader, has that hurt you?
Henry: I've learned to look at the markets to value trade. Not having traded was never a set back. I've been close enough to it. I've worked with a lot of good traders. I learned a lot along the way, and I even took my FSA and Series 7. I don't feel that I've given up anything. I understand what the trader does. I respect it. I personally would not be able to do it myself because I feel that technology gives me more scope for doing different things so it really hasn't been a big draw back.
WS&T: What are your thoughts on separating trading floors--is that still a concern?
Henry:I think the whole industry has gotten a wake up call about being concentrated in one area. In London, we've always had what we think is a workable situation. It really has been in the U.S. that, for various historical reasons, expansion or whatever, we didn't keep pace with what is needed. We still have a lot of work to do before we consider ourselves at the level where we want to be. We are a lot further along than we would be. We have three buildings here all close to each other. I personally can't comment on whether that is being rethought. I would think that with regard to where we should co-locate--we have a facility in NJ, and what goes over there I would think that people who built these campuses should be wondering how much concentration they have, and really rethinking where their employees are located and spreading that around a bit. I can't comment on specific plans as I am not aware of any.
WS&T: Has there been any thought to actually splitting the trading floor? There was a lot of discussion around that right after Sept. 11 as it would be devastating to the firm, and potentially markets, if something were to happen to all the traders and technology. But most traders were resistant.
Henry: I don't know if anyone has given thought to splitting the trading floor, but the biggest focus is the infrastructure, the telecommunications--setting all that up. I think you're right, if you separate trading and sales, it's difficult.
WS&T: So what do you like to do when you're not working?
Henry: I like to go fly fishing but the whether is not exactly good for that. I like music. I play drums and bass guitar--that's my indoor hobby in the winter, flyfishing in the summer. If I get the chance, and if my wife lets me, I like to drive a car around a race track. I used to do some go-cart racing when I was younger. Getting back to my youth, I guess. I like boating as well--sailing. I'd like to do more of that.
