ROUNDTABLE PANEL
 | Mike Roberts, head of European operations, ITG Europe |
 | Chris Sims, head of investment operations, Gartmore Investment Management |
 | Tony Whalley, head of trading, Scottish Widows Investment Partnership |
 | Tim Wildenberg, head of execution services, UBS |
Paul Scott, director, Fix City
Carl James, head of portfolio services, Henderson Global Investors
Lothar Cerjak, head of advisory and order fulfilment, Crédit Suisse
Moderator:
Henry Smith, editor, FT Mandate
Henry Smith: What is the difference between an Order Management System and an Execution Management System?
Mike Roberts: To me it’s where they sit on the trading continuum and the functionality that they give to the user. The OMS really is the engine. It’s where you keep your books and records, it’s your integration to other systems, it brings in compliance and it serves a much broader base of users than an EMS. It sits in front of the EMS in terms of the trading continuum. EMS to me is much more ‘best of breed’, and it’s really targeted at the front desk, at the trader – the individual who interacts with the liquidity in the market and has to work on orders directly. Sales are a much smaller community in terms of user types.
Henry Smith: There’s a lot of talk at the moment about the capabilities of OMS and the pressure that’s being put on them. Tim, what are the main factors putting pressure on the performance of OMS?
Tim Wildenberg: I think the point about EMS is that it has arrived because there’s been a change in the business model and the requirements. Over the last 10 years, technology has essentially enabled the buy-side to become more empowered, and effectively as the power has moved towards the buy-side, they’ve needed more tools to be able to manage the information they have and to be able to control their trading technology. Growth in the use of algorithmic trading and direct market access products are clear manifestations of that. The EMS is really the marketplace’s answer to the buy-side requiring tools to be able to access those extra services.
Paul Scott: Historically the organisation of the buy-side has changed from what used to be a more traditional, fund-manager-dealer to a set-up where the fund managers would trade on their own funds. Now we’re centralised and have created internal specialists who are responsible within the organisation for actually executing those trades. Because of that, you’re seeing some of those tools that were perhaps broker-based tools, now moving onto the buy-side. In doing so, you’re seeing a proliferation of software that allows the buy-side dealer to do more, to make more decisions themselves. Also, the increase in the number of executions that have been done, the volumes that the order management system used to have to deal with, have gone up considerably.
To continue to provide that level of information to buy-side dealers, we need the software to move on.
Lothar Cerjak: The profile of the people doing those kinds of tasks has also changed. On the buy-side, the execution people face the task of still having the same job to do as before, and we try more and more to give them some advisory functionality rather than just pure trader and execution functionality.
Tony Whalley: I think that EMSs seem to be changing on a very gradual uphill basis and we are getting continual improvement in what they offer and the way they offer it and, more importantly, the speed with which they offer it. I think with the OMSs, what we’re tending to see far more is step changes. You’ll go from one phase to the next one, maybe it’ll be six months, maybe even longer down the road, whereas the EMS is constantly changing.
Tim Wildenberg: I think the EMS is a much simpler, much lighter piece of infrastructure than an OMS, which touches many more people, and it’s a much bigger enterprise-wide infrastructure, which clearly has to move for logical reasons. My personal opinion is that EMSs only exist as a market vacuum. Over a period of time,
I don’t think they’ll exist. I think the OMSs will take over.
Henry Smith: Right, and why do you think that is?
Tim Wildenberg: Well, I think the reason EMSs exist is because the OMSs couldn’t
deliver, basically.
Tony Whalley: I disagree with that. I think the OMSs can deliver, but it takes longer than some people would like to wait for those OMSs to deliver what is required. We’ve been through it,
I think everyone’s been through it, you sit there and you get told the new phase is going to do this, and you think it’s fantastic and you want it. It’s due in October, then it’s Q1, and then it’s Q2. Whereas the EMS says: ‘You want it, you got it,’ and it’s done. It’s far more focused on individual client needs than the broader needs of the market.
Lothar Cerjak: That’s part of the problem, that the EMS is advancing at a faster pace than the OMS can. More than that, in the daily business process, the outcome from the EMS and the OMS, that’s part of where they don’t fit properly together now.
Paul Scott: Actually, now that we can do what we can do, some of the traditional functionality from the OMS we could take on board, but
I think the issue with that is that if you try to turn your EMS into your OMS, once your OMS or your EMS is touching your asset allocation or portfolio modelling, or downstream settlement confirmation and talking to your accounting agent, you can’t take it out and put the next version in.
Tony Whalley: Exactly, that’s entirely the point. It’s very much because we’re in a culture where it’s: ‘I want this and I want it now.’ Now whether it’s good to have it now or not is a moot point. The fact is everyone wants the latest technology now. It’s like you with your Blackberry and me with a pay-as-you-go mobile phone. The difference is that everyone thinks that new technology is going to help them execute better. That’s the question really, isn’t it? Does it help you execute better? Or should we actually go back to basics and look at the skills that traders actually bring to the market and enable them to enhance those skills by having the flexibility to do what they want in the market?
Henry Smith: How critical is the lack of functionality in the traditional buy-side OMS around some of the more exotic investment instruments and across asset coverage?
Chris Sims: From our point of view I would suggest that the lack of functionality isn’t necessarily critical at this stage. The majority of the volume in our house would be equity-based, moving onto simple derivatives-based. Once you get onto the more complex stuff, it’s of a lower volume and most of the functionality that’s put in is to aid automation – to make sure you can get through volume, whatever asset class you’re actually in.
Paul Scott: On the asset class side of things with regards to the role of the trader and actually sourcing liquidity, is it the case that again, in a centralised dealing desk, in quite
a lot of houses we’re still split between fixed-income, equities and derivatives? Even in
some instances, they would be using separate OMSs, so the bonds guys are talking to the fixed-income dealers as opposed to the equities and fixed-index fund managers talking to theirs. It’s not the case that there’s more consolidation on the centralised dealing desk where the order management systems are able to cover more of the asset types. Therefore we’re seeing a move for the dealers to be able to cover more of the asset classes rather than being segregated by instruments. With that happening, you might get a case where your dealers – where they can’t get the underlying equity – may look for one of the more lucrative instruments to try to cover that position for the fund manager.
Tony Whalley: I think that’s very true, and the natural split on the dealing desk is between equity and fixed-income. I think you will find there are a large number of houses – ours is one in particular – where the two are totally separate. You then get the area of where the derivatives sit. Derivatives tend to sit, for historic reasons, more with the equity side than the fixed income side. FX as well seems to sit with the equity side rather than the fixed-income side. Once you get over all that, I think what you’re really looking to achieve is a platform where the trader on the equity side needs to know what is going on in the derivative markets, in the convertible markets, so they know where they can source the liquidity required. It all comes together.
Lothar Cerjak: I think the important factor is that it depends on the size of the firm. In a smaller asset management firm you’d like to have this one-stop shop in one machine, but in the larger firms you have these separations and I don’t think they will fade away.
Paul Scott: Coming back to the difference between the OMS and the EMS, is it fair to say that a lot of functionality that exists in the execution management systems actually existed in the broker’s own products? Are the brokers leading the quest for DMA and so on? Are they creating tools?
Tim Wildenberg: There’s no doubt about that. I’m looking forward to getting on to the topic of broker-neutral EMS versus the multi-broker EMS. This technology effectively enables the guys at Gartmore and Crédit Suisse to get close to market and see the market depth. The types of tools that existed on the sell-side dealing desks for 10 years – the order book depth technology, the mouse trading technology – were tools that were never required by the buy-side, because the buy-side was essentially a broker-facing function. What’s happening is that as the client side is becoming a market-facing function they need the tools to be able to face the market. Therefore much of the technology we’re now seeing is because they built sell-side technology and market-facing technology for some time and expanded it.
Henry Smith: On the question of developing technologies, should the traditional OMS take on more execution functionality or hand off the execution to the EMS?
Mike Roberts: Both. It really depends on the function of the user. There will be instances where the OMS is able to handle the interaction with demand with no question. Tony’s EMS has integration to algorithms and to fixed networks and to liquidity. It depends on the nature of the order that they are working versus the market. That will continue to exist and there’ll be instances where some of that transactional business needs to be routed to an EMS because of the sophistication required to work that market against the order.
Currently levels of integration between OMS and EMS are what I refer to as ‘staged’ integration. It’s fixed messaging, the EMS doesn’t actually know the OMS and the OMS doesn’t know the EMS. They’re operating in isolation with risks and workflow complication as a result, so the integration that the vendors are undertaking now between EMS and OMS is to increase the sophistication of that integration. It is being exposed to the compliance requirements and being exposed to the portfolio modelling. It’s really a question of the level of sophistication of integration. As it stands today, it’s not sufficient and it needs to be more sophisticated.
Lothar Cerjak: And who’s going to run the EMS or the OMS part? It’s a question of who you’re going to give the EMS to.
Chris Sims: The preference for us would be to be able to write our own integration at the moment with what I would refer to as our EMS. We want to pull data in when our orders are brought in. It tends to be a pull, rather than a push, apart from the execution side, but the preference would be for us that we can do the integration or we can change and respond to the dealers’ needs rather than have to go back to a vendor of any description. Heaven knows how many years I have been playing with OMSs or EMSs now, but the turnover of staff in vendor companies is absolutely atrocious and the quality of products that we actually get from vendors is very, very poor.
Henry Smith: Why is it so poor?
Chris Sims: Generally speaking, I would say if we do an implementation product for an OMS or an EMS, the percentage of time spent on regression testing for an OMS for any implementation can be as much as 75 per cent of the total project effort. With an OMS, you’ve got it to the extent where the first time you put it in you know it’s perfectly straightforward. The second time you put it in you’ve already got it integrated with your back office, you’ve probably made efficiency savings which meant you lost the headcount that could have sorted things out for you manually. You’ve got to make sure that every possible workflow will go through to your accounting systems so the fund managers and the dealers have got that exact same view when you do an upgrade. We find that vendors tend to be a bit lax in their testing.
Tony Whalley: Is it the case, though, that when you look at the budget available to people on the buy-side compared with the budget available from the sell-side or the vendors, there is a massive difference? We have a pretty substantial budget, whereas a lot of the smaller organisations are run on very, very tight guidelines and find it very difficult to spend a lot of money on first of all looking for the right OMS and secondly implementing it. They would far rather have someone come in with a plug-and-play type system which is very simple, very easy and has very few frills and actually get that put in place quickly and for as low a cost as possible.
