The bottom line: public pension plans’ poor funding levels would be even worse if they were accounted for the way that private pension plans are, the fact that their accounting methods differ has contributed to the funding crisis, and Jeremy Gold was either a prophetic or foolish in attempting to call attention to this fact.
Let’s start with more actuary-splaining:
Actuarial valuations . . . in the corporate world
In the corporate pension world, there are two types of actuarial valuations: accounting valuations and funding valuations. The former determine what liabilities and expense are recorded on the company’s books, and the latter determine what contributions the employer will make to the pension fund, or define a range of choices.
The interest rate — or, in actuarial terminology, the discount rate (since you’re discounting to the present, the present value of a future benefit) — for accounting valuations is pretty nearly the corporate bond rate; once upon a time, it was just a generic bond rate; then more attention was paid to ensuring that the duration of the bond rate is equivalent to the duration of the plan liabilities (that is, simply defined, that the weighted average of the future payouts of the bond index match the future payouts of the pension plan); now most companies use a yield curve to determine the discount rate.
The consequence of this is that liabilities can increase substantially in periods when corporate bond rates are low in a given country. Incidental fun facts: internationally this can be a bit tricky. What do you do about countries where there are very few long-dated corporate bonds? Sometimes you use government bonds instead, and sometimes you can add a bit of an adjustment to represent what you think the discount rate would look like if there were enough corporate bonds. What about a country where there are no long-dated government bonds? To be honest, my colleagues and I never really had a good answer, but just made some reasonable estimates and tried to persuade the client that they weren’t material.
But I always preferred to go back to the original text of FAS 87, the first FASB statement on pension accounting, which says that
Assumed discount rates shall reflect the rates at which the pension benefits could be effectively settled.
that is, the rate at which an employer could purchase an annuity from an insurance company for the benefit. And that’s always felt right to me. If you have to assign a value to a piece of property, you base it on how much you could sell it for in the market. If you have to value a liability, it makes sense to assess its value in the market, by how much it would cost to “exchange” that debt, in this case, by purchasing annuities. Conveniently, group annuity rates are more or less the same as corporate bond rates, so that the corporate bond rate has become the norm.
Pension funding? That’s a bit different. In the United States, there used to be a fair bit of flexibility in funding, with a defined minimum and maximum funding level, the former to ensure adequacy and the latter to ensure that employers didn’t take advantage of the ability to avoid taxes by making (tax-deductible) contributions to their pension funds. In this idealized world employers would fund their plans more generously in good years, less so in lean years. And because the expectation was that employers were setting their annual contributions with the full understanding that, as a business, they were taking risks, and that they’d have to make up any shortfall in the future, they were allowed to use an interest rate that matched their expected return on assets. But that’s in the past — the Pension Protection Act of 2006 prescribes specific rates that must be used in funding valuations, based on government bond rates, so the “choose your interest rate” is really more of an idealized approach, or what a church plan (exempt from funding requirements) might do.
. . . and in states and municipalities
But the rules are different for government pension plans.
There are, of course, no federal government rules for how much states and cities must fund their pensions — states might instead define some funding target (“100% funding target . . . in 50 years”) and pledge to contribute the required amount each year; and those states may or may not actually do what they say they’re going to do, and the method of calculating the contributions needed to get to that funding target may be more or less fanciful depending on who’s deciding the assumptions.
But even government financial accounting rules are different, as defined in GASB 67. A plan that’s unfunded, and that never has any intention of being funded, is valued based on
A yield or index rate for 20-year, tax-exempt general obligation municipal bonds with an average rating of AA/Aa or higher
that is, not much different than a corporate pension plan.
But a plan that is funded, and whose actuaries determine that the combination of funds already in the plan, as well as contributions scheduled to be made by the state or local government in the future (whether or not they’re actually made is another story), are enough to pay out benefits in the future, has different rules, using
the long-term expected rate of return on pension plan investments that are expected to be used to finance the payment of benefits.
And if a plan is partially funded, they use a weighted average of the two rates.
Which leads to some peculiar outcomes, such as that reported at Wirepoints, in which the pension plan for the city of Chicago almost-magically is in a better financial position this year than last, not because of an increase in contributions or a decrease in benefits owed, but because the city council’s most recent budget includes a new schedule of contributions which intends, by means of increases each year in the future — which may or may not actually happen — to arrive at a funding level sufficient to shed the lower discount rate requirement.
Does this sort of manipulation sound any more reasonable to you than it does to me?
The more you think about it, the more bizarre it becomes, that the valuation of a future debt should depend on how aggressive you are in your plan to save up to pay for that debt.
Understated liabilities have consequences.
In the corporate world, asset return-based funding discount rates had been allowed with the expectation that companies could catch up out of future profits, if necessary, and that they didn’t have any particular obligation one way or the other, as long as their accounting is correct and disclosed to shareholders, regulators, and the financial community, to have greater or lesser levels of debt. But when a corporation goes out of business, it can no longer make up the difference between its pension fund and the cost of annuity purchases, and the federal government is left holding the hat.
But when a government body uses the GASB-defined asset return rates, it creates all manner of potential for misgovernment, using the device of assuming higher asset returns (either simply by declaring it so, or by investing in riskier assets) to create artificial reductions in liability and as a means of justifying inappropriately low contributions. It makes pensions seem cheap compared to actually giving employees wage increases in the here-and-now, and (see my earlier actuary-splainer) enables lawmakers to effectively borrow money to pay wages, asking the next generation to pay for the wages of this year’s teachers and toll-takers and other state workers. And it means that when, as in the case of Detroit, the municipality is no longer, in corporate-speak, a “going concern,” the money’s not there.
What would Illinois’ 40% funding rate look like if the state were obliged to use the same accounting regulations as corporations? It wouldn’t be pretty – but it would be honest.
And Jeremy Gold?
That’s where, finally, Jeremy Gold comes in. New reports that he passed away on July 6th may not have made their way much outside the actuarial world, but Gold was an actuary who watched the pension fund raids of the 1980s and then embarked on a mission, iconclastically warning for more than 25 years about the risks of pension underfunding caused by too-lax funding and accounting methods. In a Forbes article in 2015, “Where Are The Screaming Actuaries?“, he took actuaries, or specifically, the actuarial profession to task for not ensuring that professional standards required actuaries to provide more appropriate values:
The actuarial profession acknowledges, but does not fulfill, its duty to the public. . . .
The public will get the best estimates only when the Actuarial Standards Board requires the actuaries, who do have the data, to produce economically pertinent and decision useful numbers.
a refrain he elaborates on in an article at ConcernedActuaries.com. Yes, actuaries follow accounting standards and the requested scope of work by the governments engaging them, but Gold’s message is that actuaries have a professional duty, and a duty to the public, to disclose “true” liabilities whether or not the accounting standards or the legislature calls for it. If you’re even more interested, Mary Pat Campbell has even more information on the man and his life’s work at her blog.
So what will Gold’s legacy be?