The Challenge

Prior to the economic downturn of 2008, a client with a frozen defined benefit plan was approaching full funding. The portfolio was actively managed, with 60% equity and 40% fixed income. There was talk of allocating more of the portfolio to fixed income, but equity returns were exceeding expectations, and the changes in funding legislation brought about by the Pension Protection Act were forthcoming. This delayed the reallocation, and by the end of 2008 plan assets were down by a third.

The plan sponsor was faced with the challenge of how to fund the plan under the new law while reducing the risk of experiencing another 2008. There were two choices discussed: write a big check and immediately immunize the portfolio with duration matched bond investments, or develop a lower-risk investment strategy that would provide excess returns to help reduce the funding deficit over time, supplemented by cash contributions. Milliman proposed an examination of various asset allocation strategies, considering the risk exposures of each on funding and accounting results, via FutureCost, a powerful stochastic asset/liability modeling tool.

The Solution

The asset-liability study began with a discussion of objectives. The client wanted to develop a long-term strategy to achieve a fully funded plan within the next five to 10 years, while taking advantage of any short-term market anomalies at an acceptable risk tolerance, and with the lowest possible contribution requirement within the risk tolerance level selected. It also wanted to stabilize SFAS expense.

Milliman started with a Stage 1 presentation where 1,000 stochastic projections of liabilities and assets over a 10-year period were modeled using FutureCost to estimate expected levels of contributions, funded status, and accounting values under the current policy asset allocation, and then over a wide range of basic equity/fixed income allocations. We looked at combinations of equity/fixed asset allocations ranging from 0% equity/100% fixed to 100% equity/0% fixed in 10% increments. In all scenarios, the fixed component was assumed to be in duration matched bonds. After reviewing the results in light of their financial goals and risk tolerance, the sponsor decided to focus on fixed income allocations of 40%, 50%, and 60%. Full funding and immediate immunization was also considered as a scenario.

In Stage 2, Milliman illustrated portfolios with each of the fixed income allocations selected, and reviewed the impact of diversifying the equity portion of the portfolio by incorporating alternative investments like non-U.S.-developed equity, emerging markets equity, commodities, hedge funds, and REITs.

The Outcome

Results indicated that moving the current 40% fixed income portion of the portfolio into long-duration bonds, better matched to the duration of the plan liabilities, would reduce volatility in funding and accounting results. Immediately moving a larger percentage of the portfolio into fixed income would reduce risk exposure, but would also reduce return and increase contribution requirements and SFAS expense.

Various percentages of alternative investments were then added to the equity portions of the three selected portfolio mixes. Optimization techniques indicated that the best return with the least risk occurred when the current equity portfolio was reallocated to include from 9% to 15% hedge funds and 5% to 13% of emerging markets and commodities. In particular, the study showed that hedge funds could be used to provide equity returns with bond-like risk (as illustrated by sample scenarios I and II below). This result came as a surprise to the sponsor, who was skeptical about the use of hedge funds in a corporate pension fund. Milliman prepared a presentation on hedge funds that provided the sponsor with background information, historical performance statistics, and a peer allocation comparison. The information made the sponsor more comfortable with this alternative investment. The chart below shows our results under the two optimal 40% fixed income scenarios.

From the above, you will note that the reallocation of some of the U.S. equities to alternative investments produced slightly better return in Scenario I and II with a lower standard deviation (i.e., risk).

Discussions with the sponsor led to the conclusion that the fund should remain at 40% fixed income, and a trigger point strategy should be implemented that would dictate moves into higher fixed income proportions as the plan attains certain funding levels. This strategy allows for excess return, supplemented by cash contributions, to assist with funding the current shortfall. We anticipate that we will be requested to conduct a search for an appropriate hedge fund investment. Given the client's size and the amount it plans to allocate to hedge funds, we have recommended that it consider only a hedge fund of funds.

This is just an example of how using FutureCost to perform an asset-liability study can enable sponsors to select investment allocations that offer the best return within an appropriate risk tolerance. Incorporating liabilities into the stochastic analysis offers a much truer picture than typical asset-only studies.