IT Improvements Could Be Key to Hedge Funds' Survival

Date: Tuesday, April 28, 2009
Author: Chris Kentouris, Securities Industry News

York Capital CTO details technology initiatives

As scandals and shrinking assets take their toll on hedge funds, strong operational controls and technology will be key to survival, according to Gary Maier, chief technology officer of York Capital Management.

For most firms, of course, improvements will need to come amid budget cutbacks. Global IT spending by hedge funds will fall 20 percent this year, to $1.35 billion from $1.7 billion in 2008, estimates Celent. Unless they are faced with a collapsing platform, funds likely will postpone acquiring or replacing large-scale systems in 2009, says the research firm.

But York Capital, a global, multi-strategy fund with $9 billion in assets under management, is in the process of launching just such a system-York for Information, or Y-FI, a proprietary portfolio management platform that ties together functionality from both third-party and in-house software. That initiative, along with a new Hong Kong office, telecommunications efforts and data center upgrades, have kept Maier busy since he joined the firm in July 2007.

Previously, Maier, who has 20 years of IT experience, spent about four years as CTO of Five Mile Capital Partners, a fixed-income alternative asset manager in Stamford, Conn. He has been head of financial services at middleware systems provider Iona Technologies and a managing director at KPMG, where he focused on enterprise integration. As VP and head of trading systems development at BlackRock in the 1990s, he was a critical contributor to Aladdin, the firm's highly regarded technology platform.

Maier recently spoke with Securities Industry News about his projects at York and his thoughts on the technology issues confronting the hedge fund industry as a whole.

Have there been more questions about your IT and operational due diligence due to recent market events?

There's clearly an increased interest in the efficacy of operational and compliance processes, valuation methodologies--especially for more complex asset classes and level-three holdings--and counterparty and issuer risk controls. This last point is probably most representative of the current climate. Previously, counterparty risk was focused almost exclusively on the sell side's oversight of the buy side. With the large, well-publicized failures in the industry, however, that dynamic has clearly shifted and the little guy can now have as much concern about counterparty risk as the big guy. Other questions increasingly revolve around portfolio construction, as investors dig deeper and seek greater transparency into a firm's holdings.

How are you using technology to improve transparency?

Like many firms in our space, York had historically relied on a collection of off-the-shelf vendor products, tied together with a lot of "sneakernet," to support front-, middle- and back-office requirements. In the early days, York was principally an event-driven equity manager with perhaps a small credit overlay. Off-the-shelf systems in the equity space, especially for small-to-midsized value- and catalyst-oriented managers, are abundant, functional and relatively cost-effective, so this was a prudent course. As the firm grew, and its organization, products and transactions become sufficiently diverse and complex, it was discovered that many of these off-the-shelf systems have considerable functional gaps and don't quite scale.

When I arrived, we performed considerable diligence on a number of third-party systems. We wanted to own our data, however, and not be held hostage to a given product's functional limitations, data accessibility or cost structure. One of our critical needs was to create a multi-tiered integration framework, backed by our own enterprisewide data warehouse, on top of which all functional and data components, whether third-party or internal, would sit.

This is where several of the third-party systems we were evaluating began to fall off the wagon. Integration, in many cases, seemed to be an afterthought. In one case, the audit trail was lost when transactions were fed from external sources, rather than through the system's own application screens. In other cases, product coverage simply wasn't there, especially for asset-backs and bank debt. And access to underlying data was often sufficiently proprietary as to require a consulting engagement just to create a custom report.

What was the end result?

We opted to take a hybrid approach. I call the concept, which leverages a highly dynamic service-oriented architecture, "just-in-time integration." Our platform and core application suite, called Y-FI, offers a collection of real-time, multi-asset-class, portfolio management functionality. It's in beta-testing and scheduled to go live within a few months.

What other infrastructure projects have you worked on?

When I joined, we were in the midst of rolling out a new Hong Kong office. I literally had about two weeks to design a data center and outfit other infrastructure for this facility. Legacy plans called for what I would call a "mini me" version of our New York headquarters. Latency was a concern, but creating two separate data centers with their own servers, data storage and communication systems presents its own set of challenges--not to mention is incredibly expensive.

While you try to outfit satellite offices with a fairly basic IT support team to keep costs down, you're also confronted with a whole range of issues, like backups, compliance monitoring, shared resources and common execution, settlement and risk platforms, as well as establishing and enforcing standards across all offices. What I really wanted to do was make all satellite offices feel as if they were part of the same firm. I decided to establish an MPLS backbone between all our primary worldwide offices, serving as a dedicated wide-area network. At each endpoint, we installed WAN accelerators from Riverbed to reduce data latency.

Other telecommunications projects?

Planning for the Hong Kong office also gave us the opportunity to review our phone systems and turrets. I wanted to reduce telecommunication costs, both between our offices and with the outside world. VoIP was the obvious choice, but we still faced the hurdle and expense of outfitting our traders with turret systems, none of which seamlessly integrated with the core VoIP PBX systems we were evaluating. The specialty nature of turret systems from market leaders like IPC and BT command a pretty high price tag.

I identified a new platform from Mitel Networks that was plug-and-play with the company's core PBX platform. While it provided about 80 percent to 90 percent of full turret functionality, it was more than adequate for our needs and was priced at a fraction of the cost. By running VoIP between all offices over our MPLS backbone, we were able to leverage least-cost call routing. If I make an international call from New York, the phone system is smart enough to grab dial-tone from the nearest global branch. The global deployment paid for itself on day one.

With all this newfound interoffice connectivity, we needed to supplement our core data center and disaster recovery [DR] facilities and ensure adequate scale and redundancy across servers, storage, power and networks. To that end, we migrated from DAS [direct-attached storage] to an iSCSI SAN [storage-area network] and collocated a fully replicated SAN at our offsite DR facilities. We further introduced a combination of blade servers and virtualization, to increase our compute density, allow seamless server failover and make ourselves more "green."

What are you doing in risk?

Y-FI allows in-depth drill-downs and stratifies aggregate exposure within the corporate hierarchy, across all trading counterparties, issuers and even other peripheral participants, like agents and servicers. Beyond risk management, this provides much greater oversight and control of operational, compliance, management, financing, and execution and commission functions.

On the market risk front, pricing policies, procedures and stress tests are most critical. Our valuation committee, with representation from top management, compliance and operations, regularly meets to review such issues, consults with our auditors and sets guidelines. For our more thinly traded holdings, we try to be conservative and supplement any model pricing with multiple broker bids and recent trades in comparable securities.

Do you feel that the proposed hedge fund legislation will solve problems like the Madoff fraud?

The Hedge Fund Transparency Act and comparable legislation introduced in the House largely focus on trying to close existing loopholes that were ultimately responsible for the D.C. Court of Appeals' decision to vacate the mandate that hedge funds register with the SEC. York is already registered, as I expect are most other larger managers. Some of the smaller firms opted not to register or de-registered after the appellate decision, given the associated economic, legal and management overhead. It would certainly introduce another level of oversight that these smaller firms may have been able to previously bypass, but it's doubtful that registration by itself would prevent the potential for fraud.

What is the solution?

I think a third party probably needs to be involved in the monitoring process, but it's still unclear as to who that should be and how much information a hedge fund should disclose. The industry is also taking steps to mitigate risk. For example, the standardization of credit default swaps helps normalize these OTC products and enhances liquidity, reduces operational backlogs and increases transparency. In the meantime, investors themselves could require that their hedge funds provide more transparency.

How would you define transparency?

From an investor's perspective, transparency is typically defined as what you own and where you own it. It's how you stratify and report on your portfolio holdings, which could be as simple as top-ten holdings, winners and losers. It can also be bucketed differently--by regional or currency exposure, industry sector exposure and issuer exposure. Or, as in the case of at least one recent investor inquiry, it could be by duration weightings.

How can smaller hedge funds improve their IT and operational health?

There will be continued consolidation in the hedge fund industry and smaller funds will face many obstacles, beyond just technology. Many prime brokers have stated that they will shift their focus principally to larger and more stable, revenue-producing clients. The offers of "hedge fund hotel" services, operational support and free technology appear to be a thing of the past, though some niche prime brokers and administrators may ultimately seek to fill this need.

As for access to robust but cost-effective technology, it in part depends on the products in which a given firm transacts. If the products are fairly vanilla and homogenous, there are a number of offerings, especially in the equities markets. However, in the fixed-income, credit and multi-asset-class categories, I have yet to see a rich enough third-party system that is at once manageable and cost-effective for a smaller manager.