Truth Or Consequences
Second in a Series on Healthy Pension Fund Governance
Why do some defined benefit pension plans work well while others do not? The difference arises from choices made by pension fund boards. One critical choice is illustrated below.
Footnote 21 of Berkshire Hathaway’s most recent annual report discloses the following about its pension plan assumptions:
The “Discount rate” is used to derive the present value of Berkshire’s pension obligations. That rate is supposed to reflect the creditworthiness of the obligor (the better the credit, the lower the discount rate). The “Expected long-term rate of return on plan assets” is used to derive the amount of pension assets Berkshire should have set aside in order to meet future obligations. That rate is supposed to reflect a reasonable judgment about future investment earnings on those set-asides.
For example, assume Berkshire obligates itself to make a $100 pension payment in 20 years. Using the 4.1% discount rate, it would record the initial present value of the obligation as $45 (rounded). The difference — $55 — would gradually accrete into (i.e., be added to) recorded liabilities over the next 20 years. E.g., the recorded obligation would grow to $47 after year one, $67 after year ten, $82 after year 15, and $100 as of the due date.
Then, using the 6.5% expected rate of return, Berkshire would set aside a minimum of $28 in assets today. You’ll notice that $28 in assets is less than $45 of obligations. But that works because Berkshire’s pension obligation would grow at a slower rate (4.1%) than the rate at which assets are expected to grow (6.5%). Eventually the difference is extinguished by the time the payment is due. E.g., assets would be expected to be $30 after year one, $53 after year ten, $73 after year 15, and $100 on the due date. As a result, Berkshire can safely hold less in pension assets than pension obligations, though the difference must evaporate over time. (But also, Berkshire chooses to hold additional assets in case it doesn’t achieve its expected rate of return. In reality it was 83% funded as of 12/31/15.)
In contrast, compare the approach taken by the California State Teachers Retirement System (CalSTRS), which posts this statement in its most recent annual report:
You will notice that CalSTRS chooses to employ the same rate for both purposes. Using the same example as above ($100 due in 20 years) CalSTRS would record the initial obligation at only $24, 47% less than Berkshire would record the very same obligation. The balance ($76) would accrete into recorded liabilities over the next 20 years. Then CalSTRS would use the same 7.5% rate as its expected rate of return to initially set aside assets of $24. This has the following consequences:
Hidden liabilities. Even though teacher pensions are guaranteed by the State of California and a state may not discharge obligations through bankruptcy, CalSTRS’s board chooses to employ a discount rate more appropriate for a junk-level credit and one that is roughly twice the State of California’s discount rate. This has the perverse consequence of initially valuing a promise guaranteed by the State of California at a lower amount than a less valuable promise from a private-sector corporation and even below the value of other promises made by the same state. (The same promise of $100 due in 20 years in the form of a bond issued by the State of California would initially be valued closer to $50.) The choice of that discount rate masks the true size of pension promises until it’s too late. Citizens only learn the truth later when liabilities explode through growth at the high discount rate. Yet — as if living in a world scripted by Kafka —at that point citizens are also told that it’s too late to do anything about those promises.
Unsafely funded at anything less than 100%. Unlike Berkshire, CalSTRS may not claim it is safely funded at anything less than assets equal to 100% of liabilities. That concept applies only when the discount rate is less than the expected rate of return. This explains why the so-called “80% rule,” a frequently mis-cited rule of thumb that pension funds are safely funded so long as assets equal or exceed 80% of pension obligations, does not apply to pension funds that discount obligations at the same rate as they expect to earn on assets. When discount rate = expected rate of return, only 100% funding is safe. (And even then, unlike Berkshire, CalSTRS has no cushion in case it fails to earn its expected rate of return. As a result CalSTRS and any other pension fund that discounts at the same rate as it expects to earn really needs more than 100% funding to be safe.)
The use of unreasonably high discount rates by California’s public pension fund boards constitutes theft.
The use of unreasonably high discount rates by California’s public pension boards constitutes a form of theft from California’s citizens by masking the true size of pension promises at the time they are made. To address this form of theft, California’s pension fund boards must choose a lower discount rate reflecting the creditworthiness of the obligor. Few choices by public officials matter more to Californians. While changing the discount rate won’t help to address existing unfunded pension liabilities, new unfunded liabilities are stealthily being enabled every day through the use of a deceptive discount rate. That is nothing more than financial chicanery deliberately employed to hide compensation for government employees at the expense of students and other citizens. If state pension fund board members — you may see a list of them here and here — continue to reject truthful recording of liabilities, they should be liable for the consequences.