Market Data Services: an Outsourcing Target?

More firms are considering outsourcing their market data services. Is this a good thing? By Brian Slater, principal, The Duncannon Group

When you pick up a technology magazine or newsletter it’s easy to find an article about yet another outsourcing agreement between one of the big IT service providers and a financial institution. Today, these agreements are creeping ever closer to a new area: market data services.

Outsourcing is not new, but it has come to the fore over the last two years as institutions respond to economic circumstances. This is despite the fact that there is no evidence that outsourcing delivers the economic benefits that are promised over the full life span of the agreement, especially if economic circumstances change.

But if it is used in a strategic approach to reducing fixed cost structures, outsourcing provides better control over cost drivers--and sometimes flexible business capacity--without the costs of resourcing internally.

I share the belief of a friend of mine who was one of the progenitors of "aggressive sourcing"--that the shape of the financial industry in this era will be defined by the effective use of outsourcing. (For those who haven’t come across the term "aggressive sourcing," it is not negotiation with fisticuffs but rather applying corporate leverage and competitive positioning to the acquisition process.)

Generally speaking, the outsource model works best with functions that are largely commoditized--that is, those that vary little from enterprise to enterprise and as such can be leveraged in scale. Until recently, market data services didn’t apply. They were thought to be mission critical and specific to each organization and thus largely remained a proprietary silo.

But these days, market data is less isolated. Most of the outsourcing agreements to date have focused on specific slices of the IT pie, most typically data center, network or desktop, either severally or jointly. In a lot of cases the internal trading technology organizations, which is where responsibility for market data predominantly lives, were excluded from these agreements. But recently, some firms have outsourced their entire IT organizations--not just infrastructure but also IT management, system development and maintenance, including trading technology.

So as more and more areas are outsourced, market data functions are beginning to be affected. In some ways, this makes sense. The process model for managing and operating market data is the same regardless of the enterprise, and although there is choice in the management tools, they by and large do the same things.

Operational functions associated with market data, such as monitoring feeds and distribution networks, which were once thought to be totally unique, can be aligned with other infrastructure outsourcing efforts that are managing many business critical applications.

Interest is currently focused on outsourcing the administration of market data: service acquisition and permissioning, bill reconciliation, cost allocation, inventory and contract management, etc. This function manages the large expense outlay that market data represents. It is ripe for outsourcing because the overall effectiveness of internal market data management organizations is not high; more often it is driven by operational exigency rather than efficient process. And in many businesses there is a level of frustration about time to respond or about costs that cannot be managed or explained.

Firms perceive transferring the market data administration function to an outsourcer as not only providing cost reductions over internal departments, but also improving efficiency and turning a fixed, people-driven cost into one driven by work units performed.

Another attraction that customers associate with outsourcing is that the providers build their offerings around an "industry best practice" approach, using process and tools to generate the margins necessary to attract customers (usually a 30 percent-plus reduction over the current cost base) and still make them a profit.

But the cost of moving an enterprise from its current ineffective model to one representing "best practices" can be outside of current expense tolerance as it involves re-engineering the function, automating the process, installing tools, and organization and job restructuring. So it’s not as simple as going straight from A to Z.

In any case, firms shouldn’t move too quickly. As the saying goes, "The devil is in the details." Unfortunately, that level of detail is generally not built into outsourcing agreements (which are often not well-managed afterward, either). Firms thinking about this should establish expectations in terms of cost and service levels, and syndicate them to ensure there are no surprises.

Note that early indications suggest that the major outsourcing providers (IBM, EDS, CSC, etc.), at least, do not have best practices in place when it comes to market data functions. Some issue Requests for Proposal to specialist firms, most of which have the tools but not the proven "best practice" process model necessary to meet industry expectations. This will change rapidly when providers fail to meet profit and service benchmarks initially.

Firms can take the middle ground by outsourcing only certain functions in order to gain some improved effectiveness. Discreet functions can be dealt with separately with existing service providers. There is already some progress in areas like expense management and bill reconciliation, although to date they have mostly been short-term efforts. Streamlining and automating the management process can also yield improvements.

And don’t forget that most trading floors have a dedicated server complex. Because the trading floor is often in a firm’s headquarters, some of the most expensive space is dedicated to electronics. Do we need to own the whole problem? Could the back-end be elsewhere? Operated by someone else?

Clearly there are hurdles to outsourcing, but the benefits may make them worth overcoming.