March 2008

VOL. 16   ISSUE 2

Printer Friendly version

While the primary reason for creating a defined benefit (DB) pension plan is to help employees prepare for retirement, financial stewardship issues have recently eroded their importance as part of the employer's total rewards package to attract and retain talent. Study after study of DB plans has shown that the volatility — rather than the absolute level — of costs has spurred employers to act. Many agree with the adage "the misery of uncertainty is worse than the certainty of misery." Moreover, the enactment of the Pension Protection Act of 2006 (PPA'06), whose funding rules became effective for many employers on January 1, 2008, has the potential to make this worse by pegging liability and asset measures much closer to fluctuating market conditions.

All this uncertainty has prompted many portfolio strategists to energetically promote liability-driven investing (LDI) as an alternative approach for successful DB plan investing. It has even been approved by the Department of Labor.1 Although there seems to be more talk than action right now where LDI is concerned — since it is complex and technical and somewhat less attractive while interest rates are low — it is an approach that organizations that sponsor DB plans may want to seriously consider.

What is LDI?

Broadly speaking, LDI looks at pension plan asset allocation from the perspective of total plan risk and return. Specifically, it includes the plan's liability as a risk factor in setting the asset allocation. LDI does not prescribe a specific solution. Instead, it focuses on the excess return2 of the plan's assets over its liabilities and the volatility in that excess return.

By incorporating both asset and liability risk, plan sponsors can optimize the funded position of the plan3 and its impact on corporate cash flow and profit and loss (P&L). While asset changes and liability changes are individually important, their combined impact causes changes in the plan's funded status.

The concept of risk, or variability around the expected outcome, is important. In many situations, mitigating risk also mitigates reward. However, a powerful aspect of LDI is that employers can use it to maintain expected asset return with less risk to the plan's funded status. Alternatively, they can maintain the current level of funded status risk but increase the expected return on plan assets, thereby lowering future expected plan costs on a P&L and cash basis.

One often sees stories in the business press that equate LDI with investing in bonds. Indeed, in February 2008, when the Pension Benefit Guarantee Corporation (PBGC) announced it was abandoning its bond-only strategy, the headline in Pensions & Investments read "PBGC Says Goodbye to LDI." While investing in bonds is one way to implement an LDI strategy, it often comes with some cost. As the following case study demonstrates, prudent DB plan sponsors may want to consider alternative strategies to the bond-only approach.

Case Study: The XYZ Company Pension Plan

Although it is somewhat technical, because the field of pension economics is somewhat technical, the objective of this case study is:

This case study illustrates how plan sponsors can use LDI to either:

In both of these cases, risk is measured as the volatility of the plan's funded status.

The XYZ plan has assets equal to liabilities and a fairly conventional asset allocation strategy. As shown in Figure 1 below, there are two important issues:

 


A large portion of the plan's funded status risk is derived from interest rate fluctuations. With an asset duration of 2.2 years and liability duration of 12.0 years, if interest rates decline 100 basis points, the plan's funded status will decline approximately 10 percentage points, from 100 percent to 90 percent, leaving the plan with a pension deficit of $100 million, even if it achieves its desired asset return.

As shown in Graph 1 below, the impact of the various plan risk factors can be quantified by performing a risk attribution4 analysis on the volatility of excess return of 10.1 percent (from Figure 1).

The largest component of risk is the liability risk of about 5.4 percent (the circled bar). Similar to any investment process, the goal is to maximize the return per unit of risk of the plan. Therefore, the plan sponsor needs to evaluate whether the liability risk is a compensated risk.

 


In efficient markets, the liability volatility (driven primarily by the changes in interest rates) is seen as uncompensated risk — the plan sponsor receives nothing of value in return for taking this risk — and the plan sponsor should manage (reduce) exposure to whatever extent is possible. (Of course, it should be remembered at this point that "liability risk" is not a risk employers choose to take. Rather, it is given to them by the mark-to-market aspects of the funding and accounting rules.)

Graphs 2 and 3 below illustrate how various combinations of the initial set of asset classes - domestic equity, international equity, fixed income (Lehman Aggregate Bond Index5) and real estate - can be arranged in numerous allocations, each with its own risk. (Each point on the curve in Graph 3 represents a different allocation of these asset classes, as shown in Graph 2.)

 


 


By using LDI, the plan sponsor can hedge the liability risk and:

  1. Maintain the equity exposure, thus reducing risk, or
  2. Increase the equity exposure, maintaining the original risk, or
  3. Create a combination of 1 and 2 by mitigating some risk but increasing the equity exposure as well.

The current allocation has a 10.1 percent risk and an excess return of 2.3 percent. This excess return is used to offset some of the liability growth due to service cost6 accruals. The central questions are:

The answer, in both cases, is yes — and this case study looks at two strategies:

 


 


While Strategy A makes a strong case, the introduction of long-duration bonds alone does not eliminate the uncompensated interest rate risk. Graph 5 below shows the revised risk attribution analysis, with the reduction in the uncompensated liability risk shifted to an increased (but compensated) domestic equity risk.

 


To further reduce the liability risk, the plan sponsors may want to move to:

 


Under the approach of maintaining risk by shifting from fixed income to equity, the risk attribution analysis now looks quite different, with very little uncompensated interest rate risk, as shown in Graph 7 below:

 


Notice that the total risk is still 10 percent, because the plan sponsor again took the opportunity to reduce uncompensated risk and increase the equity allocation, a compensated risk. This approach is illustrated in the first of two approaches as described within the Strategy B environment in Figure 3 below:

 


Under Approach One, whereas Strategy A lowers expected plan costs from 5 percent to 3.5 percent, introducing both long bonds and an interest overlay reduces expected costs still further — to 2 percent of payroll in our hypothetical case study — by allowing the plan sponsor to increase the equity allocation from a total of 47 percent to a total of 67 percent.

Under Approach Two (see Figure 4 below), whereas Strategy A lowers the risk of the excess return from 10 percent to 8 percent, introducing both long bonds and an interest overlay reduces the risk still further — to 7 percent of payroll — while allowing the plan sponsor to maintain its current equity position and expected level of plan costs.

 


Conclusion

Setting asset allocation strategies through conventional techniques (e.g., the asset-only efficient frontier) makes sense only when attempting to balance investment return with investment volatility. However, it is more often the case that volatility of plan costs is of greater concern to DB employers than merely the return on plan assets — irrespective of whether the plan is open, frozen or closed. By focusing on assets and liabilities and developing an LDI strategy — whether it is with conventional assets or with derivatives like interest rate swaps — plan sponsors can effectively manage the uncompensated interest rate risk to which they are exposed.

This article was prepared with input from Segal Advisors, the investment-consulting affiliate of The Segal Company.




1 In October 2006, the DOL told JPMorgan Chase that pension fund executives may consider the risks of liability obligations when designing an investment strategy even if the plan sponsor reaps an incidental benefit — such as stabilizing the volatility of the plan sponsor's contribution obligations. Specifically, the DOL said "A fiduciary would not, in the view of the department, violate their duties under (ERISA) solely because the fiduciary implements an investment strategy for a plan that takes into account the liability obligations of the plan and the risks associated with such liabilities and results in reduced volatility in the plan's funding requirements."

2 Excess return is the surplus of the asset return over the liability return. Volatility in excess return directly affects the plan's funded status and, therefore, its emerging plan costs.

3 The funded position of a plan is the degree to which assets exceed liabilities. The return on the plan's liabilities is a cost to the plan. It is composed of the interest cost (carrying cost) of the liabilities plus any change in the value of the liabilities resulting from changes in the interest rates.

4 Risk attribution is a quantitative method of allocating the volatility of the plan's funded status (or tracking error) risk to individual sources. This is accomplished by analyzing the volatility and correlations of the different sources of risk.

5 The Lehman Aggregate Bond Index is a broad base index often used to represent investment grade bonds being traded in U.S.

6 Service cost represents the increase in liability attributed to employees earning another year of pension benefits.

7 A frozen pension plan is either closed to new entrants or accruals have been ceased for some or all participants.


About the authors:

Daniel Westerheide is a vice president and consultant in the Boston office of The Segal Company's asset liability modeling (ALM) practice. He specializes in investments for defined benefit pension plans from an ALM perspective. He can be reached at 617.244.6747 ext. 3 or at dwest@segalco.com

Stewart Lawrence, FSA, MAAA, EA, is a senior vice president and National Retirement Practice leader in the New York office of Sibson Consulting. He has special expertise in the design of qualified and non-qualified retirement plans. He can be reached at 212.251.5315 or at slawrence@sibson.com