As financial executives are well aware, the recent economy showed that managing risk can be a devilish proposition. Although defined benefit (DB) pension plans could not have been completely sheltered from the crash of 2008, it did demonstrate that event risk was not being properly managed and that the plans were highly exposed to risky assets, even though they were regarded as well funded.
While DB plan sponsors cannot control the market, they can temper their pension fund’s exposure to it. They first need to understand how bad “bad” can be and then take action to mitigate the impact of extreme outcomes.
By examining the inherent risks and economic consequences of possible future events, plan sponsors can determine the degree to which risk is acceptable and what must come “off the table.” In essence, DB plan sponsors need to understand the answers to three questions: What is the plan’s current risk profile? How might that profile change as the funded status of the plan improves/deteriorates? What actions can be taken to change that profile and what will it cost?
Plan sponsors can generally use three approaches to mitigate the financial risk of DB plans:
Plan Design. Normally, unacceptable risk arises not from the accrual of additional benefit liabilities but from benefit promises that have already been made and the assets accumulated to support them. While plan design changes may be viewed as an application of the advice “when you find yourself in a hole, stop digging,” they often have little short-term impact on the risk profile.
Moreover, changing the plan design is not viable for plans that are frozen.
Funding. Funding the plan to cover 100 percent of the cost of annuities to pass all risk to an insurer is usually cost-prohibitive and an inefficient use of capital. A less-costly approach might be to develop a policy to contribute a level annual amount that exceeds the minimum legal requirement and create a buffer to mitigate future volatility.
While this may lessen minor volatility, extreme shocks can deplete the buffer in a year. Further, using this approach without changing the investment policy will probably be financially inefficient.
Investment Strategy. This is a key consideration for all plan managers, and is the focus of the remainder of this article.
Surplus Management
Volatility of funding requirements, P&L and balance sheet metrics are driven less by absolute changes in assets or liabilities than by the relative change in each. That is why the recent financial period was a “perfect storm.” Assets declined because of market conditions that affected essentially all asset classes while declining interest rates drove up liabilities.
With current accounting and funding rules requiring quicker recognition of these changes, an efficient de-risking strategy needs to focus on assets and liabilities by setting portfolio risk in accordance with the actual funded status of the plan, using a holistic approach that marries risk management in the asset portfolio with liability risk management.
Liability-Driven Investing
The primary reason for amassing pension fund assets is to meet the benefit obligations of the plan. Liability-driven investing (LDI) views risk by considering both assets and liabilities and their combined impact on contributions, pension expense and the pension balance sheet. LDI introduces a conscious consideration of liability risk in developing asset allocations and strategies. The primary liability risk for most pension funds is interest rate risk.
In its most general form, LDI looks at liabilities in tandem with assets. As a result, any investment strategy that accounts for liabilities in any way can be considered an LDI strategy. In a typical LDI strategy, a pension fund’s asset portfolio can be split into two types of portfolios:
Return-generating portfolio (RGP) is typical of the asset-only perspective where the focus is on getting the most return with the least amount of volatility for an asset portfolio (without regard to liability).
Liability hedge portfolio (LHP) focuses on hedging liabilities (i.e., investing assets in a manner that mitigates the variability of the total plan’s funded status). As the LHP consists largely of bonds and like assets, its return potential is less than that of a typical RGP.
The core principle of LDI is to develop the asset allocation strategy and examine its impact on the plan in a holistic framework. The total portfolio can be thought of as a combination of the LHP and RGP, where the LHP is the portion of the portfolio that attempts to hedge liability shocks and the RGP is the portion that maximizes asset return. Individual risk appetites will determine how much of the portfolio is in the LHP and the RGP.
Traditional asset allocation strategies that focus on RGPs only typically have fixed policy targets for asset classes that are rebalanced in a cost-efficient manner based on market movements and not liabilities. With LDI, the LHP is utilized to hedge the plan against movements in a plan’s liabilities due to market movements.
An LHP typically consists of longer duration fixed income, as well as synthetics such as interest rate swaps and Treasury futures. Although conventional LDI portfolios are an improvement over the traditional RGPs, the allocation between RGP and LHP is either static or changed on an ad hoc basis, which is often sub-optimal in hedging the overall plan risk.
By contrast, in a dynamic asset allocation strategy, portfolio risk level responds to the plan’s funded status in a systematic manner. As a pension plan’s deficit is reduced (from investment returns, contributions or favorable liability movements), the risk posture is reduced as assets are moved from the RGP to the LHP.
In other words, when a plan is significantly underfunded, the sponsor over-weights the RGP portfolio because the gap must be closed. But as the gap is closed, weight shifts to the LHP to avoid backsliding once better funding levels are achieved.
Dynamic Asset Allocation
Dynamic asset allocation is a systematic way to change asset allocation as funded status changes. An extension of LDI, dynamic asset strategies address how much return is required and, therefore, how much risk to take. As the plan’s funded status improves, the rate of return required to achieve funding objectives decreases.
A dynamic asset allocation strategy recognizes the following:
Utility of surplus is asymmetric. This means that while some surplus is good (to buffer volatility), a great deal of surplus has limited marginal utility and might result in marooned assets that cannot leave the plan.
On the other hand, while a modest deficit is not good, a large deficit is much more painful. While it might be necessary for a very underfunded plan to have 100 percent of its assets in a RGP, which could significantly increase or decrease the deficit, placing the same investment ”bet” for a well-funded plan might not be equally wise because positive performance that meaningfully reduces the deficit might just expand the surplus beyond amounts that can be realistically utilized.
Funded status drives investment objectives. As funded status increases, the rate of return required to achieve funding objectives is lower, resulting in risk being taken off the table as the funded status of the plan approaches 100 percent.
The maturity of the plan is a key determinant of implementing the dynamic strategy. As plans mature, the relative impact of new benefit accruals on existing liabilities diminishes. Less investment return is required to finance those accruals, resulting in decreasing investment risk as the plan matures.
Dynamic asset allocation is a passive strategy. It does not involve tactical asset allocations that try to time the market. Changes in allocation are based on predefined rules. Shifts in asset allocation are based on funded status, not short-term forecasts for asset class returns.
Historically, dynamic asset allocation has had the attribute of selling asset classes that have outperformed, which has allowed plans to capitalize on favorable asset swings (i.e., the strategy has capitalized on mean reversion).
Implementing Dynamic Asset Allocation
A dynamic asset allocation establishes a series of target or benchmark allocations, commonly called a “glide path” — a pattern of decreasing risk exposure with higher funded ratios. There is no single correct glide path. The most appropriate approach for a particular DB plan should be based on the financial profile of the plan, its current and forecast maturity and the sponsors’ risk preferences.
Dynamic asset allocations do systematically what sponsors have done in the past on an ad hoc basis. The glide path maps out the de-risking strategy (i.e., as a plan’s funded ratio improves, the portfolio shifts from the RGP to the LHP — a greater portion of the assets is invested in bonds) in a predetermined way.
A dynamic asset allocation strategy goes beyond the conventional diversified portfolio and addresses issues such as enhanced security of the benefits, aging populations, exposure to wide swings in funded status and contribution requirements, and a goal to be 100 percent funded.
This can be an optimal approach for plan sponsors who are willing to give up surplus (beyond 100 percent funded status) to immunize their plans against event risks and improve the probability of achieving a funded status of 100 percent.
The design of each strategy is unique to each plan, but it should address:
Targets and path: Plan sponsors need to establish a targeted ultimate funded level, determine the initial and ultimate target asset allocations and decide on the path for getting to the ultimate target allocation starting at the initial one.
Once set, plan sponsors should reevaluate the targets and glide paths periodically to account for major unanticipated changes, such as changes in law, accounting standards, plan design or plan sponsor risk appetite.
Timing and decision authority for changes: This involves how often the funded percentage is measured and the allocation is adjusted, who makes the decision, how it is made and whether it is automatic.
Fund-specific issues: This involves monitoring results, determining when to reevaluate the glide path and when to consider changing the asset classes (e.g., due to liquidity requirements, plan changes, demographic changes and capital market assumption changes).
Although not the primary intent of the strategy, one positive outcome is that reallocation decisions are made in advance and designed into the glide path so the allocation changes automatically as funding levels increase.
Brief positive aberrations immediately result in a better-funded position and a chance to take risk off the table so short-term gains are less likely to be reversed if the positive shock is reversed.
De-Risking Outcome of a Dynamic Asset Allocation Strategy
Although the merit of a dynamic asset allocation strategy may resonate with the reader conceptually, can its merit be demonstrated as a viable approach to de-risking a pension plan? The answer is yes.
In the dynamic strategy, the funded status results are more tightly centered around 100 percent funded, with fewer results at either the positive or negative extremes. Overall, the goal of reaching full funding is met more of the time under the dynamic strategy. Just as importantly, adverse extreme outcomes in the “tails” (i.e., outcomes with large deficits and non-utilized surplus) are greatly mitigated.
The dynamic asset allocation strategy is an approach to managing a pension fund’s assets in a manner that balances the rewards of investment return opportunities with the inherent risks embedded in the liabilities. Plan sponsors must determine their willingness for risk-taking in terms of the current deficit positions of their plans, and de-risk these plans as the funded status improves to mitigate the likelihood of being in this position again in the future.
This article first appeared in Financial Executive magazine.