Guidelines for A Brave New Pension World

Tanya Styblo Beder, principal at Capital Market Risk Advisors, Inc. explains why pension funds need to update their investment guidelines for an increasingly complex investment environment.

Many pension funds are operating with outdated investment guidelines that can leave huge risk loopholes at their money managers. In fact, the vague, qualitative nature of many of these funds' guidelines is the antithesis of solid, quantitative risk-and-return management. Thus it is no surprise that a wide spectrum of pension funds today are re-evaluating their old guidelines and updating them for today's more complex investment environment.

Outdated guidelines typically focus on credit risk and qualitative objectives at the instrument level rather than specific, quantifiable measures of risk-adjusted return at the portfolio level. As a result, misunderstandings can occur because the guidelines can be subject to a variety of interpretations. The first questions to ask are: "How hard do I have to work to get around the letter and spirit of the law? Do the guidelines leave loopholes that could lead to intentional or unintentional violations? How much risk did the manager spend to get his or her return?"

Raised Eyebrows

We perform this exercise for many of our clients on a frequent basis. And we often don't have to work very hard to discover some eyebrow-raising violations. Intentional violations may involve a manager who purchases an ostensibly low-risk, low-maturity security that hides a much riskier payoff pattern. Unintentional violations typically involve managers who construct portfolios without sufficient regard to asking the question "what if I'm wrong?" For example, many structured note and mortgage-based portfolios were constructed to comply with an average maturity restriction of less than five years. While the portfolios complied with these guidelines in January of 1994, their enormous sensitivity to changes in the yield curve caused them to display average maturities of 15 years or more three months later.

"Market neutral" portfolios are another tricky arena. Managers here claim that they're constructing portfolios that are immune to most market changes. But a manager can be neutral only in a particular context. For example, you can remain neutral in a barbell Treasury portfolio (made up of short-term funds and 30-year bonds) as long as a certain relationship exists in the yield curve. You can go long one strong currency and short another and remain neutral as long as the two currencies remain in a particular relationship. You can be long a stock index and short a sector but remain neutral only when the market environment dictates that the sector maintain the same relative relationship to the index. Key questions to ask here are: "How wide is the zone of neutrality? Do things have to change a little or a lot for the strategy to fail? Do you understand the manager's key assumptions?

Which Widow

One fund I came across specified that it wanted "investments suitable for widows and orphans." What is or is not suitable for widows and orphans is largely a matter of individual interpretation. A portfolio of long-term Treasury strips may be appropriate if Widow Jones is 30 years old and earns a good wage. The same portfolio may be grossly inappropriate if Widow Jones is 85 and needs to live off of her portfolio income. One manager might interpret this guideline to mean that widows and orphans should be invested in US money market instruments. A second might interpret this to mean that broad diversification into venture stocks, emerging markets and CMOs is called for. A third might believe that widows and orphans won't be getting anything if the portfolio doesn't start performing - and pump up returns by spending too many risk dollars on an inappropriate investment.

Silence vs. Prohibition

How should you close your loopholes? At the policy level, you must decide where you want your guidelines to be on the scale of silent to prohibitionist. Silence leaves the interpretation of appropriate risk to your managers, and you may wind up with more risk than you intended. Prohibitionist (for example, "you cannot do it unless it's listed here") leaves little room for managers to take new opportunities, and you may wind up with less return than you intended. Neither extreme is desirable.

The most important steps to close loopholes involve updating guidelines based on risk measures and expanded risk reporting. While most pension funds have return targets designed to allow them to meet their payment obligations, few have formulated specific risk tolerance levels. Value-at-risk, sensitivity and scenario analysis can all help pension funds assess how much risk per unit of return they are willing to take. These risk measures can be used to help develop new guidelines that hinge on risk-return ratios calculated by a specific methodology rather than on vague or incomplete descriptions of the risk of particular instruments.

Keep in mind, however, that quantitative measures are only one part of an effective risk management program. Furthermore, the market risk a VAR model calculates is based on the parameters and assumptions you choose. One VAR model of a particular portfolio could show much more market risk than another. In studies we have conducted, common VAR techniques produce market risk measures that vary by 14 times or more.

Multiple Tactics

Too many funds desperately seek a single risk measure. In our experience, however, no single approach guarantees effective risk measurement. For example, a fund with a more stringent VAR standard (higher confidence intervals, multiple time horizons, greater market moves in the historical data base, less favorable correlation assumptions, etc.) typically needs different limits and stress testing than a fund with a more lenient VAR calculation. In general, the weaker the calculation, the stronger its complements must be.

Measuring Managers

It's important to develop specific ways to measure the performance of a particular manager. You have to ask yourself, "Do I understand the assumptions that have to exist so the performance I'm expecting actually happens? If I have several managers and each produces 600 basis points over the S&P 500, who used more risk?"

It's also important to adjust risk guidelines to particular strategies. In market neutral strategies, for example, the trick is to define guidelines that keep investments within certain safety zones, and provide warnings for risk gone awry. For example, a manager may be permitted to construct a hedge portfolio strategy using IOs and POs from the same underlying collateral, but not permitted to do so from CMOs backed by premium collateral for the IO and discount collateral for the PO. One common approach is to limit portfolios to those that would be neutral to certain predefined moves in the variables that matter.

The most common mistake is limiting investments or derivatives to particular instruments. A broad rule permitting interest rate swaps may lead to swaps tied to an unintended index, for example, Libor squared when only Libor flat is contemplated. Guidelines permitting convertible securities could leave you with a complex bond linked to the spread between six different interest rates that converts to an undesirable Mexican common stock. A rule restricting portfolios to "blue chip stocks" may or may not include foreign blue chips. And what about structured instruments that are either based on a formula involving the price of a "blue chip" stock or are convertible into "blue chip" stocks?

Here's another example: "You can buy anything as long as it's two years and triple-A." This used to work extremely well in the past. You could control your credit exposure with triple-A rated investments. And the two-year maturity limit would guarantee you'd get an instrument with limited volatility because there was no such thing as an instrument whose duration was longer than its maturity. Financial engineering changed all of that. Now you can have a triple-A security with a maturity of less than two years that behaves like a 20- or 30-year bond.

The same goes for guidelines based on any kind of average measure. Average maturity measures can become very dangerous as the yield curve twists and changes shape. Some complex mortgage portfolios may have an average maturity of 2.5 years but could extend to 17 or 18 years in certain scenarios. Finding the average on a single nonrated deal, moreover, may be close to impossible.

Global managers present a particular set of problems. In recent years, many investors have been seeking diversification by running into the open arms of managers of global equities, thinking they're buying expertise in picking global stocks. But when we perform the attribution analysis to determine whether their performance was the result of good stock picking or the appreciation of the currency, it often turns out that the stock picking was pretty weak, but the currency factor was pretty strong.

Risky pieces

A lot of people think of themselves as secure just because they've done a VAR calculation. The reality is that this is not sufficient to contain risk. VAR and other quantitative strategies are important, but they aren't the only thing. In fact, you're far from done. In our risk-adjusted wheel of fortune (see chart on page 59), the quantitative pieces are actually only a third of the pie. The pieces that address the qualitative aspects of risk protection are also vital to a successful risk management program.

One of the greatest challenges facing pension funds is a lack of timely information. Unlike mutual funds and banks, which view their risk position once per day or more, the vast majority of pension funds are unable to view even their holdings more than once per month with as much as a 30-day time lag. Good limits and checks and balances are also important. The goal is to cover as many pieces of the wheel as possible. Then, at the end of the day, look at where the Achilles heels are in your risk management program. But realize, that even though you can prevent many things that have caused others pain and loss, you can't prevent a clever new form of fraud.

One final piece of advice: After you've set new guidelines, it's important to explain them to your money managers. In most situations, your money managers are not maniacally scheming to seed your portfolio with unacceptable risks. They are simply trying to maximize your returns while abiding by vague guidelines written perhaps ten or 15 years ago. If you take the time to explain why your fund has modified its guidelines and how new risk measures work, your money managers will generally be happy to go along.