RiskMetrics Builds a Better Model With Arrakis Platform

Financial advisers often use similar processes for wealth management. But similar approaches often work to cross-purposes when they're not completely harmonized within a firm.

"You'll find institutions where different advisers have different asset allocation models for the same type of objective," said Martin Nemeth, head of wealth management products for RiskMetrics Group, New York. "You'll find places where it's not very easy to put the different products they're trying to sell together into one proposal."

The opportunity to fix these discrepancies led RiskMetrics Group to acquire Arrakis, a risk management platform from JP Morgan Chase. The private banking group at JP Morgan had developed Arrakis in 1999 to support its Morgan Online offering. With the sale, RiskMetrics gains control of the development, hosting and maintenance of Arrakis, while JP Morgan Chase retains exclusive rights to its Morgan Online and Morgan View interfaces, proprietary analytics, and client data.

RiskMetrics Group, itself originally formed as a JP Morgan spinoff, will incorporate the Arrakis data structure for portfolio holdings information into its existing analytical toolbox. The new data structure represents a substantive improvement for RiskMetrics. "Before, we stopped at the position level," said Nemeth. "With Arrakis, we go down to a tax slot level."

As a result, automated planning tools can consider the tax basis that an investor has in any particular asset. "We've now enabled all of our sophisticated analytics to add tax-aware calculation capabilities into the product," said Nemeth. "If you have something that's very low basis, you wouldn't want to have the application not know it's low basis, trade you out of that position, and pick up the tax consequences."

Along with providing more detail, the data structure can handle a full range of assets: managed accounts, hedge funds, real assets within hedge funds, derivatives, foreign exchange, convertible bonds and capital protected notes. "Once we put that infrastructure into place, we handle any set of products," said Nemeth. "You think it up, and I'm sure we've seen it before, since we have to handle it from an institutional point of view."

Through partners such as Yodlee and Adhesion, assets are brought into the system via aggregation. "The idea is to bring that all together and give a complete view," said Nemeth.

From there, RiskMetrics applies its analytics capabilities to the assembled data. Using portfolio optimization and Monte Carlo simulation techniques, the system can detect situations that require intervention, such as concentrated options positions, or an overweighting in a particular sector as a result of volatility shifts in the market.

The Monte Carlo simulation model used by RiskMetrics is different from the one used in the Morgan Online portal. "It answers the same problem, and it does some things slightly differently, but I would say they're comparable," said Nemeth.

Still, garbage-in equals garbage-out. Advisers earn their keep by incorporating their views of the world into the financial models, based upon their firm's research on expected returns, asset class views, and macroeconomic expectations.

"We really allow firms to embed their research through data into the product," said Nemeth. "It just becomes a means for them to utilize the same pitches they're making through their economic research groups and other areas of the organization."

Given aggregated portfolio data and well-considered economic assumptions, the RiskMetrics platform can employ monitoring and diagnostic tools that automatically detect troublesome situations as they occur. That's valuable for detecting anything out-of-whack-either in the portfolio or on the payroll. "A monitor can be directed not just at an adviser, but up the supervisory chain," said Nemeth. "The organization can set up its own policing strategy around the groups that it creates for the organization."

Along with reactive steps, financial firms must also stay proactive by thinking through the potential effect of extreme situations. Using RiskMetrics, advisers can stress-test all of their portfolios at one time.

"If you're worried about a certain economic scenario, you can model it and propagate it against all of your households," said Nemeth. "That way, you know who potentially are the people you need to talk to."

Smaller organizations without the capability to perform their own research can switch to autopilot by using the RiskMetrics view of the world. RiskMetrics, said Nemeth, uses quantitative techniques "that are more market-neutral approaches, along the lines of a Black-Litterman asset allocation model type of approach for expected returns."

Users of the RiskMetrics service may include smaller regional firms, or custodial firms that want to round out their offerings with hands-off portfolio management service. "But they want an independent source of data," said Nemeth. "Why compete with your own wealth management team?"

Such a strategy is akin to a major airline starting up a low-cost brand with fewer amenities serving a smaller number of destinations. As such, the application of quantitative techniques can help to bring wealth management to a broader class of investors.