EDM: panacea or palliative?

Data management, arguably among the most frequent cause of operational headaches at buy-side firms, remains a less than thrilling yet fundamental issue for many in the industry.

As more and more managers move beyond standard long-only equity investments to more complex asset types and trading strategies, problems of cleansing, storing, distributing and reconciling data across multiple systems have become commonplace.

Heightened awareness of the links between data management and everything from performance and analytics to compliance and risk management has prompted traditional asset managers and hedge funds to consider implementing enterprise data management platforms and strategies, whether devised internally or by a growing number of third-party technology providers on the market.

Indeed, enterprise data management (EDM) has been touted in some quarters as a veritable 'cure-all' for whatever operational ailments afflict a particular manager. If successfully implemented and maintained, EDM platforms and strategies can certainly improve efficiency and reduce risk; there is no arguing against the fundamental benefits certain managers stand to reap via EDM. But the cost and operational impact on a firm's business brought about by implementing an EDM strategy – whether that entails a single platform or several interconnected systems – combined with considerable and ongoing maintenance requirements can prove not only prohibitively expensive and impractical for managers lacking budgets and resources, but also dangerous for those lacking a clear understanding of what EDM can and cannot provide their business.

Like any major IT undertaking involving a hefty price tag and implementation effort, EDM can have obvious benefits for those managers large and complex enough to experience significant data management problems. But industry observers indicate that the upfront and ongoing cost and maintenance required to make EDM worthwhile can deem this issue too expensive and impractical for firms with less pressing data issues.

Sam Auxier, director at Deloitte & Touche's capital markets practice, explains that many of his firm's buy-side clients consider EDM systems and strategies to drive down data-related errors and consolidate infrastructures, but he emphasises the long-term requirements necessary to keep these implementations viable.

"The question is: What's the payback? Does reduction of errors or consolidation of systems really justify this expense?" Auxier says.

"Second, what is the effort required to maintain data quality on an ongoing basis? Among firms we've seen that have successfully implemented this strategy, they've actually put a data quality team in place, which could be just one person for a smaller firm or up to five or six for larger organisations, dedicated to ensuring that data is correct and resolving discrepancies."

Adam Sussman, senior consultant at financial IT consultancy Tabb Group, reports that issues of fiduciary and regulatory responsibility along with operational complexity can drive managers to justify EDM implementations.

"Mutual funds are more likely to be implementing EDM systems because they're larger and have bigger regulatory and fiduciary responsibilities to make sure all the data across their organisation is protected and stored adequately, whether it be for investment-decision purposes, or reporting, or dealing with SEC audits," says the consultant, adding that smaller hedge funds and asset managers more often than not have other more pressing IT and operational concerns.

"Smaller funds with less than $1 billion in assets [under management] are operating on shoestring budgets. Some of them don't even have order management systems, so first on their list will be an OMS, not an EDM platform," Sussman notes.

According to some observers, however, the evolution of outsourced EDM capabilities will eventually make these implementations more practical not just for larger managers, but also for lower-tier firms seeking more efficient data environments.

Adam Honore, senior analyst at Boston-based financial consultancy Aite Group, expects such a scenario to develop within the next two years. "If you're a smaller shop right now, EDM is probably impractical for the most part, but I'd say in the next 12 to 18 months that won't be true," Honore predicts, adding that currently many early adopters with ample budgets and staff are providing the primary demand – and paying top dollar – for EDM projects.

"With outsourcing opportunities, those prices are going to come down, and once they do and a few customers come on board you're going to see very affordable outsourcing, but right now it's impractical for a lot of tier-two and tier-three customers. They probably don't have the operational or IT staffs to support the infrastructure of an EDM solution," Honore continues.

Beyond the practicality of taking the EDM route, firms must also keep in mind what exactly EDM means, and how dynamically it can affect their operations.

Peter Ellis, Catherine Doherty and John Robertshaw, principal consultants at UK buy-side advisory firm Investit, warn against the tendency among many managers to equate EDM with data warehousing, which can potentially further complicate their data management headaches.

"Firms make the mistake of equating EDM with a consolidated data management platform or data warehouse. That's a serious misconception," Ellis cautions, adding that in many instances EDM can and should entail a best-of-breed approach possibly involving more than one data management platform. In other words, Ellis describes a component-based approach to data management overlaid with an EDM-based architecture.

"When you try to build a single data warehouse, you end up with a serious data analysis problem. A much better approach is to take a best-of-breed architecture and look at each of your businesses and the systems they require, and then identify the data management requirements for each of those areas and build a data management platform to meet them."

Robertshaw adds that in many buy-side firms, data has historically been more often maintained than actually managed. Now, some Investit clients have begun taking a more long-term, proactive approach to the issue.

"A number of firms we're aware of are addressing data management as an ongoing, long-term programme, and for each of their development activities they've got extensions to functionality development allowing them to correctly address and organise the data components of their projects as part of their overall programmes," Robertshaw observes, adding that instead of maintaining central data projects, these managers have "devolved" data management to their business units' individual development activities.

Other industry sources caution against failing to recognise how an EDM installation will affect not just data management but other IT and operational facets of one's business.

According to Holly Miller and Peter Bergan, New York regional manager and senior consultant, respectively, at investment management IT consulting group Citisoft, even successful EDM implementations require significant operational reorganisations. "If you're successful with your implementation you'll have to reorganise your operations around that, and your staff will probably shift from reactive processes to more of a proactive, exception-based operational model," Bergan reports.

Miller emphasises the need to map out long-term IT plans and how to accommodate them incrementally within an EDM framework before moving on to actual implementation: "You're designing and implementing an integrated strategy and deciding which pieces go on to your EDM first, then second, then third. I've seen people implement their EDM applications to flatten one problem and then use it to flatten a second problem, in which case they have to undo much of their initial implementation because they didn't do all their long-term planning in the beginning."

For most IT-related initiatives – especially those tied directly to revenue generation – buy-side managers have long relied on return on investment (ROI) measures to determine whether or not to implement the system in question.

But in the case of EDM, with no clear link to breadwinning and often complex, drawn-out implementation schedules measured in years as opposed to weeks and months, it may not lend itself neatly to conventional ROI scrutiny.

Investit's Robertshaw explains that issues such as reputational risk mitigation through good data management simply cannot be measured appropriately. "It's difficult to apply strict ROI to something like data," he says. "There are areas where you can identify greater sophistication of data management that can provide benefits, but there's also a certain amount of preventative stuff that you gain in terms of reputational risk reduction and sideline issues you have with data. You're not going to be able to put a fixed return in a reputational problem that could be solved by good, quality data."

Deloitte's Auxier also sees shortcomings of traditional ROI when it comes to EDM, favouring risk-adjusted ROI as a more accurate gauge. "It's not just a traditional ROI consideration – you have to consider regulatory reporting aspects, trading error aspects, and the probabilities of revenue and cost impacts due to the large spend of an EDM project," he explains.

Sussman at the Tabb Group, however, contends that investment managers with well-seasoned chief technology and information officers can translate EDM projects into meaningful ROI figures, bearing in mind those projects' indirect impact on other operations.

"Larger organisations with CTOs and chief information officers on their management boards have the means and knowledge to translate IT projects into ROIs. I think they can understand that though EDM may not be a revenue generator, it's related to disaster recovery or can create infrastructural synergies," Sussman says.

Given that one of the most monumental tasks involved in EDM involves streamlining data management across multiple, legacy systems and infrastructures, especially at larger, more well-established managers most likely to implement EDM platforms, one might presume that start-up fund managers have an advantage in terms of setting up sound data infrastructures and avoiding similar headaches down the road.

Such a presumption may sound good in theory, but the situation on the playing field gets more complicated: start-up managers must first focus on getting their trading operations up and running, forming prime broker and administrator relationships, and raising capital. And, paradoxically, the very fact that these managers lack complicated legacy infrastructures and operations could even obviate the need for an EDM program – that is, until their operations become large and complex enough to warrant it.

Investit's Doherty observes that data management doesn't often rank highly on start-up managers' lists of IT priorities: "While you're starting up, you're trying to set up your prime broker relationships and marketing and work out how to track your portfolio – constructing a database is not very high on people's lists of activities."

Doherty explains that start-up managers may have the advantage of a clean IT sheet, but aren't usually skilled or inclined to expend efforts in EDM planning. "The problem is timing – there's a sweet spot after you're big enough to be stable and before you're too big and too complex," she says.

Bergan and Miller at Citisoft note that the key challenge for new managers is knowing how they want their data management infrastructures to look and function. "In retrofit situations, you understand exactly what your challenges are and what you want to improve," Bergan says, adding that many start-up firms will most likely opt for out-of-the-box data management tools from their prime brokers anyway.

"It's always nice to start with a blank piece of paper, but specifications gathering in this case becomes more challenging," explains Miller. "It's a lot easier to look at something and pick over it to see what's missing than it is to ask a portfolio manager or salesperson to outline every little data element they're going to require, particularly at a new firm."

Some start-up managers with more IT experience and substantial seed capital, however, are well-positioned to implement EDM systems and strategies with fewer complications.

Deloitte's Auxier notes that a few of his firm's start-up buy-side clients have indeed given serious consideration to taking the EDM route at the onset of their endeavours.

"We have a couple of situations where start-up clients have an opportunity to acquire the latest technology to put in some very fundamental IT practices to grow their business at a lower marginal cost," Auxier reports. "A lot of these managers are starting up because there's frustration with the companies they're leaving, or in some cases it's due to technology infrastructure and having the tools to do the type of investing they want."

Thus, Auxier sees a clear advantage here in terms of EDM implementation, but only as long as a manager has the capital and know-how to pursue such initiatives.