A Model for Success

With the economy in a slump, investment firms are looking to modeling as a way to increase earnings. However, don't expect to find portfolio managers strolling down the runway at the latest fashion show. Modeling in the investment world involves using technology to create investment scenarios that demonstrate how a portfolio manager can maximize return on investment, while taking the least amount of risk.

While risk management is not a new concept, especially within the last few years of a scandalous and downturned market, the idea of moving risk management out of the back office and into the front-office is picking up. Managers are finding that, by using modeling technology, they can examine what the most-probable outcome of an investment will be before the investment is made, instead of doing so retrospectively. The theory is that this insight both helps avoid costly mishaps and enhances the productivity of investments.

Despite the seemingly obvious benefits of hedging investments with analytical analysis, fundamental portfolio managers, who tend to explore stocks from a business perspective instead of quantitatively, often resist portfolio modeling, as it can restrict their examination of the market to historical or industry-standard information.

"There is an age-old argument," explains Financial Insights' Senior Capital Markets Risk Analyst Peter Keppler. "Quantitative analysts say that stock-pickers can't beat the market and, on the average, they don't. But if a stock-picker hits once, he will beat out the average quantitative analyst."

Yet, portfolio-modeling technologies are demonstrating that, no matter what method of portfolio management is being employed, modeling can be beneficial and, in fact, vital to a firm's prosperity.

Modeling in Equity
One firm that can attest to the advantage of portfolio modeling is Iowa-based Principal Global Investors, a diversified asset manager with over $90 billion in institutional and retirement assets; including equities, fixed income and real-estate investments.

Principal has always been a fundamental shop, adding value through a bottom-up stock selection, says Mustafa Sagun, managing director of quantitative research at Principal Global Investors. Any risk portfolio results were only examined after trades had been conducted, he says. However, Sagun says that two-and-a-half years ago, the firm began to more closely examine risk in hopes of reducing systematic risk and increasing the breadth of its stock examination, while continuing to focus on the firm's strength of picking stocks fundamentally.

Principal decided to bring in tools from Barra, a risk-management vendor based in California, in coordination with other customized proprietary tools to detect systematic risk and maximize portfolios. "We construct a portfolio to make sure that the majority of risk comes from stocks only, not from systematic market risk," Sagun explains. "We want underlying stock characteristics to dominate returns. That way, we know that when we outperform or underperform it is based on stock selection only."

While many firms are struggling to make sure their portfolios are meeting the industry benchmarks, Sagun says that Barra allows Principal to target higher returns. "We use the tracking-error component specifically as a target before the trade, as opposed to after, so that we are successfully beating the benchmark," he says.

Sagun says that Principal's portfolio-optimization tools work much like a huge calculator. Principal's portfolio managers input return forecasts into optimization tools. The managers then add risk characteristics of those stocks, as defined by Barra components. For example, Barra can calculate a stock's volatility, value, size or momentum relative to an index, while incorporating a percentage added or subtracted in order to outperform that index. The optimizer then advises the portfolio manager how much to buy of each stock given their particular characteristics. The manager uses this model, along with other alpha models and analyst suggestions, in order to figure out the final recommended portfolio in terms of risk and outcome potential.

Sagun notes that a modeling tool helps provide portfolio managers with a wider breadth of coverage since even fundamental information is changing daily. However, Sagun insists that a model alone should not provide a singular perception of the portfolio. "The modeling tool will provide a fundamental manager like us with a breadth of coverage, but we have global analysts that focus on depth."

Challenges of Fixed Income
Loomis, Sayles & Company, an investment firm managing $54.8 billion in assets of institutional, high-net-worth and mutual-fund clients, demonstrates the advantages of modeling within its fixed-income business. Loomis utilizes CMS BondEdge to measure potential return and risk on its portfolios on a stand-alone basis, as well as compared to industry and customized benchmarks.

Kevin Kenyon, a quantitative analyst in the core bond group based in Chicago, says that all of Loomis' fixed-income portfolios are maintained on BondEdge. "With so many companies blowing up in 2002, portfolio managers are becoming acutely aware of where the risks are," he explains. "Our portfolio managers use BondEdge to maintain portfolios and understand allocations and risks."

Kenyon says that one of the most favorable aspects of BondEdge is its ability to perform scenario analysis. With this tool, portfolio managers can construct a simulated portfolio, run it in BondEdge and see what can be gained versus what may be risked. "The scenario analysis helps us to understand what would play out in different portfolios and what the trade-offs may be," he says.

In order to conduct the scenario analysis, Kenyon says that the portfolio managers download their portfolios daily into BondEdge, along with BondEdge's proprietary databases, as well as third-party market-data feeds. BondEdge then exports a hypothetical portfolio in an Excel file. Kenyon says that this pre-trade-constructed version of the portfolio not only helps managers to ensure consistency with daily benchmarks, but also highlights the best methods of maximizing the return on the manager's investments.

Kenyon also stresses the importance of using modeling within the investment process as opposed to placing it in the post-trade domain, in order to equip managers with a preemptive view of risk.

However, Kenyon concedes that BondEdge cannot possibly arm Loomis' fixed-income portfolio managers with all of the necessary tools to avoid risk. "One of the problems is that there are no tools that help us understand risks that are unforeseeable. For example, it's not going to help you detect accounting fraud," he says. "That's where we rely on fundamental analysts who are scouring through information that you wouldn't find in quantitative-analysis tools."

Kenyon adds that BondEdge has recently added a tool to better help managers understand the possibility of default, such as in the case of WorldCom, but much of this type of risk is still undetectable by technology.

In fact, Keppler says that portfolio managers often depend too much on technology to determine risk factors and fail to take into account extraordinary events that could not have been predicted based on historical patterns. "You can be a slave to a model," he says. "When you are building technology systems, you must take into account certain assumptions in order to implement mathematical models, and they won't always hold up."