Chapter 8
The Secular Decline in Interest Rates and the Viability of DB Plans
As we said in our discussion of the accounting rules applicable to DB plans, the market-based elements in FASB’s corporate pension accounting regime have, in effect, translated the secular decline in interest rates into losses on corporate financial statements.
So, let’s take a look at interest rates. Figure 8.1 (from the Federal Reserve Bank of St. Louis) charts the yield on Moody’s Seasoned Aaa Corporate Bonds from 1950 to the present. The yield here is a little lower than the “high-quality” corporate bonds used to value pension liabilities for financial statement purposes, but it will serve to illustrate the point.
Interest rates went up until 1981—to be precise, for this data series, they went up until September 1981. And then they went down.
For example, as of year-end 1999, the average discount rate used to value corporate pension liabilities (S&P 500 firms) was 7.51%.10 As of year-end 2016, the average of the spot second and third segment rates published by the IRS (a “good enough” proxy for the 2016 discount rate), was 4.13%.
Let’s go back to our example participant again—the one with the benefit worth $200,000 25 years from now. At a 7.51% discount rate, the present value of their benefit is $32,720. At a 4.13% discount rate, the present value of their benefit is $72,717. The cost of providing the exact same benefit has more than doubled. Since the year 2000.
Whatever the design flaws of the traditional DB plan (and they are serious), whatever the regulatory burden, for most employers these plans are simply no longer affordable.
If we ignore the inflation-induced spike in interest rates in the early 1980s, then the real nosedive in interest rates hit in the 1990s, when they started to drop below their 1970s levels. Chief Financial Officers—in an earnings-sensitive company, the officers responsible for worrying about this sort of thing—started to worry. And then started to look for ways to get out.
What About the Other Two Sponsor Risks—Investment and Mortality?
While they still matter, the effects of changes in investment return (and investment return volatility) and unexpected increases in mortality have had less of an effect on DB finance.
Investment Returns
It is true that there were two spectacular equity crashes in the last 17 years. At the beginning of the century, when the “tech bubble” burst, the S&P 500 had three bad years, losing 10% in 2000, 13% in 2001 and 23% in 2002. And in 2008, during the Global Financial Crisis, it lost 37%.
Nevertheless, the return to a 60/40 equity/fixed income portfolio over any 25-year period since 1950 has been above an annualized rate of 5%. A vindication, of sorts, of modern portfolio theory.
Ironically, interest rates were such an overwhelming factor in the increase in the cost of DB plans that if these plans had, in the year 2000, simply been invested so as to hedge the plan’s liability-interest rate risk, they would have done better than they did pursuing returns on assets.
Many sponsors found themselves in the extraordinary situation that their plans were, in some years, earning double-digit returns on their assets, and yet they were in a worse funding position at year-end. Because they had lost more—as a result of declines in interest rates—on their liabilities.
Mortality
To connect the obvious dots, the longer a participant in a DB plan lives, the greater the value/cost of their benefit. Along with mandating the use of specific interest rates, the PPA mandated the use (with certain exceptions) of IRS mortality tables to value liabilities.
In 2014, the Academy of Actuaries published new tables—updating tables from 2000—and a new mortality improvement scale, showing significant and unexpected decreases in mortality/increases in life expectancy. The IRS finally got around to adopting these new tables/improvement scale (in modified form) in 2017 (effective in 2018). The new mortality tables/improvement scale (and the actual longer lifespans they anticipate) are expected to increase the value of a typical plan’s liability portfolio by around 4%. And, of course, that will also increase the cost of providing DB benefits.
The Academy’s 2014 mortality improvement projections were controversial and, thus far, have not been borne out by subsequent experience. But an unexpected 4% increase in liabilities due to mortality improvements is an “experience loss” that the DB system can survive.
The catastrophic losses because of the long-term decline in interest rates—doubling or in some cases tripling the cost of providing $1 of retirement benefit—were for most companies, unsustainable.
Take another look at that chart of interest rates since 1950. Interest rates went up—and the whole time they went up it got cheaper and cheaper to provide $1 of DB benefits. And then they went down, and the whole way down—right down to the present—DB benefits just got more and more expensive.
And—actually—the problem with interest rates reflected a more profound problem with DB plans.
A Fundamental Lack of Transparency
If a sponsor knew that it would have to value the DB promise it made to its employee-participants using, e.g., a 4.13% discount rate, then it could adjust its plans and make financial arrangements accordingly. That’s a steep price—relative to, say, 1980 interest rates. But if that is what the benefit costs, or at least that is what FASB’s valuation model says it costs, then, as they say, it is what it is.
More to the point—if the employer could know today what a DB benefit payable 25 years in the future would actually cost, then it could make a rational decision about how much of an employee’s compensation to devote to retirement savings.
But, as we have been saying, that actual cost cannot be known. The only way for the sponsor to bring any measure of certainty to this transaction is to settle that liability today. Either by buying an annuity from an insurance company—in effect outsourcing this issue—or by buying a zero-coupon bond of the same amount and duration as the liability for the participant’s benefit. Avoiding having to do that was, however, why DB plans were created in the first place.11
Put yourself in the shoes of the sponsor here. It makes a promise to our 40-year old employee-participant in 2000 that it thinks is only worth $32,720. That corresponds, presumably, in some way to what that employee’s services were worth. But it turns out, now, that delivering on that promise is actually going to cost $72,717.
For most corporate DB sponsors, who have to (after a fashion) book that $40,000 loss to income, that result is simply unacceptable.12
* * *
As a result, most sponsors simply wanted out of the DB business.
The question then became—how do you do that?