Excerpts from a roundtable discussion sponsored by ADP, held on June 19 New York City.

MODERATORS

Ed Berko, senior vice president for strategic processing, Automatic Data Processing Joe Kolman, editor, Derivatives Strategy

PARTICIPANTS

Ben Golub, managing director and head of derivatives risk management, BlackRock Financial Management Ron Tannenbaum, director of marketing, GlobeOp Financial Services Rafic Dahan, CEO, Arcordia Michelle McCarthy, managing director, Deutsche Bank Christine Berthet, consultant Sanjay Mithal, vice president for financial products and services, eCredit.com Sendhil Revuluri, director in equity trading, UBS Warburg Mike DeAddio, chief technology officer, Cygnifi Scott Stirling, vice president of sales and marketing, Kronos Software

Ed Berko: The late 1980s and early 1990s was a period of tremendous technology-driven innovation in the capital markets. Whoever had the better mousetrap for pricing and managing risk and processing transactions got the business. The market-makers able to warehouse risk within and across asset classes got the bulk of the trading volume.

These days the Internet is driving innovation, but the key challenge is not developing new financial instruments. It's aggregating disparate data from around the world so you can measure and manage risk globally.

At ADP, we've been helping clients outsource their back-office operations for almost two decades. We believe that outsourcing is one of the best ways of dealing with middleware, relational database management and other capital-markets technology issues.

A talk I had a few weeks ago with a potential client is instructive. This company gets 256 data feeds every month in ASCII file format. Some have individual positions, some have aggregated positions, some have mark-to-market or position values. It takes about a month to get all the data to one location. Then the company has to run some basic sanity checks on the data to make sure they're clean. They check three basic things: Did we get a file from XYZ branch in Hong Kong? Is that file empty? Is that file the same one we got last month from XYZ branch? Then the data is put into the risk management engine, and management up and down the line in this bank waits to see what the value-at-risk number is, and what the stress-test results are for positions that are now one month old. This is the way it is done at many large institutions.

"In some firms, it costs $50 a trade to process a spot or forward transaction. A centralized service provider can do this at a fraction of the cost.” —Ed Berko

I think Internet-based outsourcing will prove to be the best way to manage this type of complexity. The ultimate goal is efficiency and cost reduction. In some firms, it costs $50 a trade to process a spot or forward transaction in the back office. A centralized service provider can do this at a fraction of the cost. Most financial institutions don't care how the back office gets done, as long it's done and there are no mistakes.

A lot of new startups are providing these services and don't have much overhead or infrastructure weighing them down that older, traditional, financial institutions have in office space, hardware, software and staff. You'll have a chance to hear from three of them today: BlackRock, GlobeOp and Arcordia. I think most institutions in five to 10 years will be outsourcing the bulk of their plain-vanilla transactions—either to regional service providers that will be aggregated across a few different hubs, or to one or two service providers in the same geographical region that focus on specific instruments and specific asset classes.

Ben Golub: Some people may not know that BlackRock has been outsourcing risk management services for over five years, in addition to managing money. Our risk management folks like to talk about "assets under risk management.” We've crossed over the $1 trillion mark.

Since this conference, BlackRock announced the formation of BlackRock Solutions, which makes risk managment enterprise sysems available to third parties.

The way we got into this activity was almost by accident. BlackRock is a very successful and fast-growing money manager. When we talked with prospective clients, we always talked about the infrastructure that governed our investment management activities. To our surprise, quite a few companies said, "I wish I could have what you have for our other accounts.” As a result, we decided to unbundle our services, and now we offer risk management services for more than 26 companies separately from our asset management service.

"The consequence of even the appearance of impropriety is so large that it simply can't be a possibility. There's just too much reputation at stake.” —Ben Golub

When you think about these services, it's important to ask a number of critical questions, such as "What's behind the screen?” Risk management is akin to a dialysis machine. If it doesn't work, you might have a noble obituary, but you're dead. In our view, the highest-value-added parts of the service are innovation and responsiveness to change. What's going to keep that screen or report current? How do you know the models work? And there are certainly academic questions such as Are these methodologies analytically precise? Are they implemented efficiently and reliably?

Pragmatically speaking, the most important thing is for the methodologies to be practical and useful. While you may have something that is academically very elegant, if it can't deliver in a decision-relevant time, or if it takes too long to get that model implemented, what are you going to do in the meantime? How do you bridge the gap between perfection and the fact that you're long that position today? Risk management outsourcers have to grapple with this central question constantly—and provide a suitable answer.

Ron Tannenbaum: I'm happy to be here today to explain what we do at GlobeOp. GlobeOp is a total outsourcing service for middle and back office fund accounting, and fund administration.

We built our business model in many respects coming out of the Long-Term Capital Management experience. Notwithstanding what happened from an asset management perspective, the people who ran the middle and back office and accounting infrastructure were excellent. LTCM was a firm that went from $1 billion up to $7 billion, back down to approximately $5 billion, and then virtually to $400 million. When they were at about $400 million, they were supporting a balance sheet of roughly $120 billion, and off-balance-sheet transactions of $1.3 trillion, with just shy of 40 different counterparties in six different legal jurisdictions. And all those had mark-to-markets on both sides. LTCM never missed a payment on any of those transactions, never lost track of its collateral, never restated net asset value, and had its audited financials finished every year in February.

Our clients are, in the main, asset managers. We put our company together toward the end of last year to address common concerns of the asset management community: namely, managing overheads, maximizing margins and state-of-the-art risk management reporting. GlobeOp is there to make sure the customer's database is as clean and stable as it can be. We centralize the customer's data feeds back to one single-source entry. We have a fully developed capability for derivatives settlement processing, so we're flexible for any new products that asset managers want to trade.

Rafic Dahan: Arcordia started in April of this year. It was a response to many issues, but particularly the increased volume and complexity of interest rate derivatives products. The aim was to build something that could adapt to market changes fairly quickly—changes in regulation, in products and in client requirements. We've found that the biggest savings have come from transitioning away from legacy infrastructures. Going forward, we'll be concentrating on the post-trade activities of capital-markets clients by processing OTC derivatives using a web-enabled IT infrastructure. This will include providing settlement services via an application service provider, and other complementary services such as confirmations, accounting summaries and document repositories.

Joe Kolman: I believe one of the earliest efforts in outsourced risk management was a Bankers Trust effort, managed by Michelle McCarthy. Michelle, am I right?

Michelle McCarthy: We were in a business very much like the one Ben Golub is in. It used to be called RAROC 2020; now it's db RiskOffice because Bankers Trust was bought by Deutsche Bank. The original effort took some of the risk management people from Bankers Trust and some of the people from the custody area to do risk measurement for pension plans. We still do those, but now we've migrated more toward asset managers.

Pension plans were natural outsourcers. They had no reason to have a team of five people doing risk management. But asset managers are on the borderline: Should they hire those five people or shouldn't they? As a result, they're much more demanding customers.

One issue that's sticky for both types of clients is secrecy. We've had to point out over and over again that we're independent from the derivatives group. And even so, there are entities that want to encode their account names and want us to be shot at dawn if anyone ever finds out about their trades.

Kolman: To me, that's the Achilles' Heel of outsourcing. I've heard people say, "You want us to move data off-site to your centralized data warehouse? Are you nuts?”

Berko: That's a good question for the whole panel—how do you deal with that?

Golub: Certainly, you can't have a business in which you're getting confidential information without having credible and ironclad ways of dealing with it properly. Obviously, we have to do the right thing, and we're done that throughout our history. We've handled portfolios from more than a dozen of our competitors. The consequence of even the appearance of impropriety is so large that it simply can't be a possibility. There's just too much reputation at stake, particularly if you have a public company with a real balance sheet and a real market value—$2 billion in our case. So it is essential to have technology, secure phone lines, segregated databases and rigorous internal procedures to manage the information flow.

Kolman: I'm curious what the people in the outsourcing business think will happen in five years. At, say, Merrill Lynch, what percentage of the back office will be outsourced? Will Merrill have a back office? My guess is that it will.

McCarthy: The smaller-tier institutions are the more natural ones to go for this at the beginning. And there are some negatives to outsourcing, even for the small institutions. Companies that outsource may have less control of the piece of their business they outsource; they may find that it's hard to get a couple of third parties to cooperate to get things done. Some companies use us for risk management, and use somebody else for accounting, and somebody else for another piece. Unless they are very clear with each provider about what its responsibility is, they'll be frustrated trying to get the whole bundle together. So even for the small guys, it's better to have something tailor-made to their exact needs. Some of those parts that need to be outsourced are not going to be very profitable for the outsourcing firm. We've staked our business on the fact that some things should be outsourced. But it can be frustrating for both sides as well.

Kolman: Just because it's economically feasible to outsource doesn't mean that, politically, it can be pulled off.

Berko: The boutique firms tend to outsource first, and on simpler instruments like spot and forward foreign exchange, money markets and exchange-traded transactions. When they start outsourcing and their cost per transaction goes down, that changes the whole cost paradigm for the larger institutions. You see this happening with the older, traditional full-service brokers. All of a sudden, on-line brokers entered the business, and the whole cost-and-service paradigm shifted. The traditional full service brokers were then compelled to get into the on-line brokerage business. Many firms are on-line being pushed by competitors to outsource. And outsourcing typically starts with the simpler and higher volume instruments and moves to the more complicated ones.

Christine Berthet: The real issue is, What kinds of new products are companies creating with their data and their systems? If you're processing your own information and not using it to create any new value, there's no point in doing it yourself. Merrill Lynch, for example, would never have created the CMA account 20 years ago if it didn't have the flexibility to manipulate its system to support new products. It would be more valuable for the firms to focus their technology and processing on creating new offerings for the clients.

Kolman: Outsourcing is a lot closer to Schwab's culture and the way it does business—low-tech, high-volume—than it is to Merrill's. So what should Merrill be doing? Simulating e-Schwab? Or can Merrill create another value proposition, where its high-tech culture gives it a competitive advantage? I believe a Merrill Lynch needs to find a different paradigm to compete with Schwab, rather than just trying to use an outsource provider to be more cost-effective.

Sanjay Mithal: Strategic outsourcing and partnering is integral to our business. The immediate impact of incorporating technology-based solutions on an ASP basis is cost savings and process efficiencies. The less tangible benefits depend, of course, on the business model, but data capture and data synthesis will lead to significant developments in product innovation. What will happen over time is that best-of-breed providers of third-party solutions will apply sophisticated data-based strategies in creating new products, presenting themselves to the market as innovators. In many cases, successful entities will, in essence, reinvent traditional businesses, but from the outside—that is, not in the mainstream organization where the threat of direct competition can prevent such dynamic change. This is the paradigm that organizations such as JP Morgan, Lehman Brothers and others with e-finance arms have adopted with great urgency.

"Successful entities will reinvent traditional businesses, but from the outside. This is the paradigm that JP Morgan, Lehman and others have adopted with great urgency.” —Sanjay Mithal

Golub: If you think about the history of outsourcing, there have been successes and failures. There was a company called IBM that outsourced an operating system to a little startup vendor and its chips to a commodity chipmaker. At BlackRock, we've thought about that issue in terms of the potential impact of making available our analytics and technology on our competitiveness in the asset management business. Ultimately, you have to ask, what is core?

I think our initial hesitancy to provide outsourcing in a pure form was exactly because of that concern. Because in terms of innovation, it's not clear where all the value-added in our management is coming from. In our case, we came up with one way to deliver portfolio, analytics and transaction processing four years ago that was, at the time, very new.

You can look at a company like BlackRock as doing one of two things: Either we are asset managers with a clever middle office, or we've got a bunch of people who write trade tickets that then get processed in an extremely highly-efficient manner. But the truth is, we're something in-between. Yes, you need to have a good investment management process, but is that enough? Are our clients buying alpha or are they buying the servicing of pools of assets to specified levels of volatility? I believe that innovation comes from the intersection point between markets, financial theory and technology. I think the market leaders are the ones who are pushing that point to the limit.

Sendhil Revuluri: It seems like innovation is already being outsourced. Cisco is buying from networking companies. Schwab is buying from companies that offer people tools to manage their own personal finances on the web. Big companies are outsourcing innovation by letting other, smaller companies come up with something and then buying it from them. You should look at the long term and see where you want a company to be. On the supplier side, as an outsourcer, the way to look at it isn't to try to figure out where your client's base competency lies, but to look at who's running the company. If traders run the company, it's never going to outsource trading. In a sales-driven culture—for example, Merrill—it would be difficult to outsource that kind of function. But they may very well be able to outsource something else.

Mike DeAddio: This comes back to the core proposition. Ben, is BlackRock an asset management firm or is it providing risk management capabilities? Would you provide these capabilities to one of your competitors? Would you allow one of your competitors to use your capabilities on behalf of its clients—which could potentially hurt your competitive advantage on the asset management side? If you say yes to that, then you've truly moved into a situation where this is your real core competency, as opposed to something that's just going to benefit your main business. So is this what you're doing or is it an adjunct to your main business?

Golub: I think you've asked an important question. It depends on the type of competitor. In our case, there are some companies we perceive to be major competitors but most are not. The number of companies we consider real competitors is quite small. Would we sell to that small list of five or six entities? Probably not. Would we sell to everyone else? We would. These are exactly the issues people think through as they figure out where their value-added comes from. Over time, of course, this can change. I think all companies are learning that you're going to have multifaceted relationships—that you're going to do business with companies that are partially competitors. It's not a black-and-white situation.

The second question a realistic company has to ask is Are you prepared to eat pieces of your own business, or is somebody else going to do it for you? While in an ideal world you might think you had the tremendous luxury of deciding whom you sell to, I think realistically the advantages that companies have are often small and transient. And it's an illusion to think that your company has the power that can keep a technology out of the market for any serious period of time—especially if you think it's good.

Scott Stirling: Which is the best kind of company to run these services? As a medium-sized bank, I would feel comfortable with certain types of institutions running certain activities—perhaps back-office processing, maybe middle-office analysis or even nonstandard trade-capture and pricing. However, I might not feel that comfortable sharing such information with a software company, which perhaps didn't have the understanding inherent within the day-to-day activities of the marketplace. What is the panel's general view about who is the best within this space?

"As a medium-sized bank, I would feel comfortable with certain types of institutions running certain activities. However, I might not feel that comfortable sharing information with a software company.” —Scott Stirling

Tannenbaum: I don't know the answer, but in our marketing efforts it's been very important to our customers that GlobeOp is totally independent. We have studiously avoided any ownership interests from the Street. For trade entry, we don't care too much what software people use because we can import data from any other package. It's a constant effort, but its our roal to be vigilant for all conflicts of interest.What GlobeOp is to our customers is a pure middle office, back office and accounting service.

Golub: On this issue, I can't speak for BlackRock as an institution. My own gut feeling, as an individual who's observed a lot of activity over quite a few years, is that markets teach you about risk. And, ultimately, innovation comes out of responding to market activity.