The analysis found a combination of reduced-risk/better-matched investment policy and reduced-duration plan design was most effective at reducing volatility.
In its Pension Finance Watch, Towers Watson noted the baseline investment portfolio underlying its Pension Index represents a 60% equity/40% fixed income allocation, with the fixed income component mostly invested in aggregate/core bonds. This policy, fairly typical historically, potentially adds volatility in two ways: (1) through the relatively high percentage of equity investments and (2) through the relatively low level of interest rate sensitivity, which results in a duration mismatch versus the typical plan’s liabilities.
Towers Watson reviewed the impact of an alternative policy that reduced the equity allocation to 40% and pushed the duration on the fixed income component to more effectively match the duration of plan liabilities. The impact on plan volatility was that year-to-year volatility — defined as the average (absolute value) percentage change in funded status— is cut almost in half versus the Pension Index baseline through this change in investment policy.
According to the report, another option for suppressing financial volatility is to reduce the sensitivity of plan liabilities to interest rate changes. Plan sponsors can do this by moving to account-based plan designs and by increasing their program’s utilization of (non-interest-rate-sensitive) lump sum payments.
Towers Watson identified two options — the first involving a one-third reduction in interest rate sensitivity (liability duration reduced to roughly eight versus the benchmark program’s duration of 12), the second implying the complete elimination of interest rate sensitivity. Account-based plans are typically designed to be much less sensitive to changes in interest rates (i.e., when account balance accumulation rates are linked to market rates and lump sums are paid at retirement), the report said. However, account-based plan designs are generally implemented with long transition timelines, which implies that liability durations are reduced but not eliminated for decades.
The impact of reduced liability duration was only mildly beneficial, with year-to-year funded status volatility reduced by only 5% to 10% versus the Pension Index baseline in each case. And, totally eliminating liability duration was no more effective than simply reducing it, likely because a “no duration” liability remains out of sync with the benchmark asset portfolio, which has at least some effective duration.
Finally, Towers Watson looked at what would have happened if the reduced-risk/better-matched investment policy and the reduced-duration plan design changes were both implemented. This approach was most effective at reducing funding volatility, cutting it essentially in half versus the baseline.