The conventional wisdom is that derivatives dealers haven't had an easy time selling their wares directly to pension fund managers. Even when they point out that outside money managers are far from averse to listed futures and options, it's still a tough sell. But not an impossible one. According to the experts we've recruited for this roundtable, the rejection of derivatives by funds is far from uniform. Moreover, there are signs that the logic of derivatives is gaining some converts, and will gain more in the years ahead.

By Portia Richardson

George Oberhofer
Frank Russell

Over the past several years we have seen a modest increase in the use of derivatives among pension funds. In general, as plan sponsors become more acquainted with risk management and how derivatives work, they are in turn becoming more willing to utilize derivatives in order to boost returns.

Futures and options, which one might define as the most basic of derivatives, are now prevalent in plan sponsor portfolios. Currency forwards and options are also increasingly common as pension funds purchase non-dollar bonds.

But the biggest change in recent years is not plans's increased use of derivatives but rather their growing attention to managing the risks associated with derivatives use. Much of this increased attention is the result of recent, highly publicized derivatives-linked losses at various corporations and funds. The most proactive funds are tightening controls by revising their investment guidelines.

In our experience, the most ineffective guidelines are those that provide "do" and "don't" lists according to instrument type. They do not work for two reasons: first, financial innovation typically renders the guidelines obsolete within a matter of months. Second, even if a particular derivatives structure is on the verboten list, a clever money manager looking to boost returns can instead build his own hybrid structure using several simple cash instruments which, when taken together, behave exactly like the prohibited derivative in question.

Rather than zeroing in on specific instruments, we encourage our clients to focus more on portfolios' sensitivity to various risk factors. In particular, we feel that worst-case scenario analysis is especially well-suited to plan sponsor guidelines. Thus, for any given "catastrophic" move in any market, the plan can identify its maximum tolerable loss. These figures can be expressed in terms of minimum yearly return or maximum yearly loss.

For pension plans new to risk management, we feel risk-source-by-risk-source scenario analysis is a more effective tool for measuring risk than a single summary statistic such as value-at-risk. Although VAR is quite elegant, providing a single "risk number" which can be used to evaluate a portfolio's risk, it also involves myriad assumptions, which effectively make the VAR measurement process a black box for all but experienced quants. In addition, VAR analysis, which summarizes expected risks over all possible outcomes, can mask the presence of scenarios so dire that even a small probability of occurrence would be unacceptable.

For most plan sponsors a change to risk-based guidelines will not make it necessary for them to invest in extensive and costly risk management systems. The guidelines serve to establish a standardized vocabulary and set of expectations between manager and plan administrator with respect to analysis and risk reporting. The plan sponsor's main job is to review the manager's risk analysis for reasonableness and spot-check confusing or suspicious results

Managers looking at risk-based investment guidelines at first worry that they impose an additional client servicing burden. On closer inspections, most realize that such guidelines provide greater flexibility to use the full range of financial instruments available, while giving the client better assurances that unacceptable risks are not being taken.

Leo de Bever
Vice president, Research, Ontario Teachers' Pension Plan Board

The history of our fund forced us to use derivatives on a big scale. In 1990 we had about $20 billion in assets; 100 percent was invested in non-negotiable Ontario debentures. Our current asset base of $40 billion Canadian has been diversified into a mix of one third fixed-income, one third Canadian equity and one third foreign equity. We could not sell the original debentures, so we used swaps to gain exposures to the TSE, TOPICS, S&P, and various European exchanges. We also use a variety of derivative instruments to modify the duration profile of the remaining unswapped debentures.

The other major use of derivatives arises in our currency management. We have a one third foreign exposure, and hence significant foreign currency risk. We use forwards and futures to hedge some of that risk. We used to be 100 percent unhedged, on the assumption that this would give maximum diversification and a marginal incremental return at a tolerable risk. Since the Canadian dollar has mostly declined in value relative to the US dollar, and the US dollar has historically been weak against the yen and deutsche mark, that was a profitable strategy compared to hedging. In the last year we have successfully used selective currency hedging, when there seemed to be a compelling logic for doing so.

Our future use of derivatives is strictly a matter of convenience and suitability. We don't consider them either desirable or undesirable compared to the underlying assets. Derivatives must serve an asset mix or risk reduction objective. Rule one is not to buy what you don't understand. Most of the horror stories about derivatives involve poorly understood vehicles with a lot of leverage. Our swaps have no more risk than is implicit in the purchase of the underlying assets using cash. The main difference is the need to manage "cash at risk" or the exchange of collateral on periodic settlement dates to reflect the net change in the value of our equity swaps.

Although we've looked at the Group of Thirty guidelines, they did not seem well-suited to pension funds with long-term liabilities and hence long-term investment horizons. Still, the spirit of measuring risk and containing it is very much part of our management approach. We pay close attention to our ability to withstand catastrophic and sustained simultaneous declines of 20-25 percent in all markets in which we operate. The main difference from the G-30 approach is that we are more concerned about our ability to ride out a six-month downturn, instead of worrying about daily fluctuations in assets and liabilities.

We are also implementing a frequent monitoring of risks using a value-at-risk (VAR) approach. Because VAR depends on very specific assumptions about the distribution of risks, and correlations estimated in a very specific way, we have shied away from relying exclusively on someone else's "black box" to estimate fund risk. We also want to relate asset risk closely to liability risk. We prefer to use conservative estimates of the historical frequency of disastrous asset and liability outcomes as opposed to the elegance of estimated probability distributions which tend to predict that a 1987 crash occurs once every couple of centuries. We need to understand how the assumptions yield the estimates of risk. If something goes wrong, we cannot stand in front of our board and argue that we relied on someone else's understanding of our problems.

Our extensive use of swaps to turn fixed income exposure into equity exposure clearly crosses the asset classification lines. We're trying to break down "asset islands," i.e., the tendency to look upon opportunities as fitting narrowly defined asset classes. We want to take advantage of cross-asset opportunities, whether derivatives are involved or not.

About 90 percent of our portfolio is managed internally. Our external managers are instructed to focus on the value of the assets they select, and to let us worry about hedging whatever other risks we have. Internally, we focus on catching "unintended risks" which can arise when more than 50 people interpret policy and execute it in ways that may have unintended side-effects on risk. That particular source of risk is not necessarily higher when derivatives are involved. In assessing whether any instrument is useful, our main focus is not on whether it is a derivative. Our main concern is whether it fits the structure of both our assets and liabilities.

Duane Smith
Director, Benefits Financing, The Goodyear Tire & Rubber Co.

I think that pension funds's use of derivatives has increased over the last three or four years, and will continue to increase. Derivatives are used for various strategies or to control risk. Derivatives enable pension funds to change asset allocations, to hedge foreign currency exposure as more money as goes over into non-US assets, or to enhance index funds.

Derivatives usage is most pronounced in foreign currency hedging, bond strategies that use interest rate swaps, and for equities that are incorporated into enhanced index funds. Managers of enhanced index funds buy the derivatives rather than the equity securities.

In our case we've reviewed our guidelines on the use or non-use of derivatives, and spelled out how they can be used. We don't want any leveraging to enhance returns. We use them for risk control, to substitute the cost of buying the security, and to hedge currency risk. We don't want to use leverage and have more exposures than our assets. Also, we require managers to indemnify us against any losses for unauthorized use of derivatives.

A majority of our managers, though, won't be allowed to use them. We would not hire a manager who uses derivatives without risk controls. We ask them what controls they have in place to limit their exposure. Most managers use controls of some kind. Most are also very sensitive to the derivatives issue and have revised their internal investment guidelines to address them.

Ron Araujo
Manager, US West

As funds become more aware of the cost advantages that derivatives provide, it will be difficult to avoid using them, irrespective of how much bad press they've received recently. The pressure to outperform benchmarks is too great, and pension funds are much more aware of the varying types of derivatives and the many ways investment managers use them. Because they are now more knowledgeable, pension funds will likely increase their usage in futures. Of course, if there is another blow-up, all bets are off.

Reporting requirements have certainly been increased a great deal. Additionally the money management community now has a heightened awareness of their clients' expectations. However, as with any kind of market innovation, the controls take a bit longer to get in place. As bad as 1994 was for derivatives in general, it certainly caused the investment community to consider their control process in this area.

Money managers will always be chosen for their ability to add value to an investment program. If they can demonstrate that using derivatives helps them achieve this, either through risk control or return enhancement, many funds will see this as an edge, especially if the manager has demonstrated controls in place.