Why defined benefit sponsors should diminish return expectations
I began my actuarial career in the mid-1990s, a time when the only things larger than expected investment returns were eyeglasses and mobile phones. I can still recall developing spreadsheets based on the lofty return expectations of the era projecting that my rapidly compounding 401(k) would make me a multi-millionaire by, well, about now.
Of course, we now realize those heady days of double-digit annual returns were something of an anomaly driven mostly by irrational exuberance for websites worth nothing. Sure, I am well on my way toward a successful retirement (my employer is really into that whole concept), but I am far from being a multi-millionaire.
As I have come to terms with the real possibility that I may need to continue working into my 50s, so too has the world of defined benefit plans gradually accepted that expected investment returns just aren’t what they used to be.
Based on capital market assumptions back then, many plan sponsors assumed returns around 8%-9%. (I’ve seen old filings with assumptions as high as 11%.) These hefty expected returns were used to offset the annual accounting expense, masking the true cost of the plan with returns that were soon never to develop.
(At the time, the expected return rate was also used to determine the minimum contributions. So higher expectations led to lower funding, which ultimately contributed to the pension crisis of 2000-2002. It should be noted that multi-employer, governmental and some church plans still use this approach.)
Ever since the mortgage crisis of 2008, the world has grudgingly come to terms with the “new normal” of lower returns across virtually all asset classes. As macroeconomic and demographic factors reduce global growth to a slow walk, economists have had to reassess investment return assumptions.
Every few years, the Retirement Actuarial Services group at my company compiles a report on long-term capital market assumptions for our use by our defined benefit clients. (Voracious readers can find the full report, authored by my talented colleagues Elena Black and Yubo Qiu, here.)
If you’re more of a scanner, I’ll just tell you that expected returns today are significantly smaller. (Though still proportional to eyeglasses and phones.) According to the 2016 update just released, the long-term expectations for practically every major asset class (with the exception of high yield bonds) is markedly lower than the 2014 version.
Based on this new information, a typical defined benefit plan allocation would experience a reduction in its long-term return rate of 50 to 100 basis points. Other analysts may draw slightly different conclusions, but generally “lower for longer” is the prevailing sentiment when it comes to expected asset returns.
(Note: Legal wants me to pause here and mention that these are used for actuarial assumption setting only, and should not in any way be interpreted as financial advice.)
Pension accounting — American style
For single employer ERISA plans in the U.S., the expected return of plan assets is used as an offset against the operating expense attributable to the pension plan from benefit accruals, interest on liabilities and amortization of past losses. So to keep the accounting expense of a defined benefit plan lower in the current year, one needs to merely expect returns rather than actually experience them. (Recognition of the difference between actual and expected returns is delayed until future years through amortization of gains and losses.)
This has historically given an incentive to the finance departments of plan sponsors to push for higher expected return assumptions. Sometimes, it has even led to imprudent use of aggressive allocations to support the higher expected return. In response to some of these issues, international accounting rules ceased using the expected return concept several years ago.
With lower expected returns across the board, U.S. plan sponsors could face a choice for the upcoming fiscal year:
- Reduce their expected return assumption, which increases operating expense
- Reallocate their assets to support the current expected return, which increases risk
Unless there are other strong fiduciary arguments in favor of the second option, which seems fairly unlikely, the vast majority of plans will choose the first. As such, they should probably update their budget estimates accordingly.
For other types of plans that still use the long-term return assumption for funding purposes, a more tangible cash consequence may arise. Government plans, certain church plans and multi-employer plans reducing their valuation assumptions by 50 to 100 basis points will increase their funding liabilities by 5-15%, an amount likely to be felt in the annual contribution requirement relatively soon.
Whoever said “the key to happiness is low expectations” obviously didn’t sponsor a defined benefit plan. The world of long-term slower growth may indeed be upon us. And auditors will clearly be focusing on this matter sooner rather than later (because as we know, they live for this stuff). I imagine many of them are already peering at reports through reasonably sized eyeglasses and consulting colleagues on incredibly thin phones.
Before they come calling, single employer plan sponsors should quickly assess the impact of lower returns on their U.S. operating expenses. Sponsors of other plan types may need to plan for significantly higher contributions.
I can sympathize with them. After all, I should have been a multi-millionaire by now!
A version of this blog originally ran on the Principal blog.