- Proposed revisions to public sector #pension plans passes #financial liabilities to #future #generations
- Future #taxpayers would have to honor #putoptions (guaranteed financial product) at the time they are least able to
- More #investment risk does not make a pension better funded - simple increases how much tomorrow will have to pay for today
Mr. Chairman and Members of the Board,
Thank you the opportunity to testify with regard to the Board's guidance concerning accounting for public sector pension plans. My comments are my own and do not represent the views of the American Enterprise Institute or any other entity.
The defined benefit pension benefits accruing to public sector employees are intended to be guaranteed and in practice are protected by law, state constitutions, political power and the simple moral case that a benefit once earned should be paid. Sooner or later, these benefits will be paid.
One purpose of accounting rules, however, is to ensure that benefits accruing to today's public employees are paid for today. This standard is referred to as inter-period or inter-generational equity. In GASB's terms, this means that "taxpayers of today pay for the services that they receive and the burden of payment for services today is not shifted to taxpayers of the future."
GASB illustrates this concept with terms such as "living within our means" and "fairness."
"Future generations are unequivocally financing part of the government services that we are receiving today."
I submit to you that the current accounting standards utilized by public pensions clearly fail to measure the degree to which plans are satisfying the criterion of inter-period equity. This failure allows today's taxpayers to pass on significant liabilities to future generations.
The proposed revisions that we consider today are superior in that, at least as written in English, they accurately describe the full range of costs and liabilities associated with funding pension benefits. However, as implemented, they would fail to fully capture these costs and would continue to allow plans to violate inter-period equity by passing large contingent liabilities on to future generations.
A simple example may illustrate the shortcomings of the current accounting approach. Imagine a plan that owed a guaranteed lump sum payment of $1 million in 15 years time. Assuming an 8 percent expected return on plan assets, the plan could consider this liability "fully funded" if it invested roughly $315,000 today. That is, assuming a constant 8 percent annual return was achieved, this present value would equal $1 million 15 years from now.
But no investment can produce a guaranteed 8 percent return. In practice, any investment in risky assets has a range of possible values. If our investment falls short of $1 million, which it will a majority of the time, we need insurance against that outcome. We can purchase it through a "put option," which is a derivative financial product that will make up the difference between the market price of our assets and the "strike price" of $1 million.
If we purchase this put option, then we have truly fully funded the liability while maintaining inter-period equity. That is, the $1 million payment can be made with certainty without resorting to future taxpayers.
If we do not purchase the put option, then the guarantee against assets falling short of liabilities is provided by future generations as a contingent liability. The value of this implicit put option represents the degree to which a pension plan today, funding guaranteed liabilities with risky assets, violates the principle of inter-period equity.
And, to a very close approximation, the value of the implicit put option imposed on future generations equals the difference between plan liabilities measured under the current accounting standards and the so-called market valuation approach in which liabilities are discounted at a low rate of return to reflect their risk.
Importantly, the value of the contingent liability imposed on future generations is not some worst-case scenario. The cost of the implicit put option is what future generations would willingly pay in order to rid themselves of the chance of an outcome that could in fact be considerably worse. The put option protecting against assets falling short of liabilities is expensive because bad outcomes in financial markets are correlated with bad outcomes in the rest of the economy, meaning, when unemployment is high, tax revenues are low and budgets are under pressure. This means that future taxpayers would be required to honor these implicit guarantees at precisely the time they are least able to do so.
This example shows that market valuation is not an ivory-tower, academic approach that ignores the actual investments public pensions make. Rather, it is superior to the current standards precisely because it includes valuable information regarding the risk of pension investments that the current standards ignore. The current standards do not assist us in planning for or ensuring inter-period or inter-generational equity in public pension financing. Future generations are unequivocally financing part of the government services that we are receiving today.
The proposed revisions implicitly recognize the value of these contingent liabilities. In the plain-English description of the proposed rules, the overall responsibility for accrued benefits is described as the sum of three sub-liabilities:
- First, the pension plan's liability for funded benefits, that is, for benefits that could in expectation be paid through assets already set aside;
- Second, the government's liability for unfunded benefits; and
- Third, the government's secondary, contingent liability for funded benefits, should the plan's dedicated assets prove insufficient.
While this is not precisely how an economist would frame the problem, it does fully encapsulate the liabilities associated with pension provision.
"...while the standard deviation of investment returns falls over longer holding periods, the actual investment outcomes become more divergent."
According to the proposed revisions, funded liabilities would continue to be discounted at the expected return on the assets backing them. Unfunded liabilities without assets backing them would be discounted using a lower interest rate derived from high-quality municipal bonds.
Note that there is something missing: while the proposals in English describe the government's secondary responsibility for covering funded liabilities should the plan's assets fall short, nowhere in the math is the value of this contingent liability calculated. But, as I show in a recent article in the Financial Analysts Journal, once you do calculate the value of this contingent liability-again, using the options pricing approach described above-it becomes clear that the total liabilities associated with a given level of guaranteed future pension benefits is best described simply by discounting the future liabilities at an interest rate commensurate with the risk of those liabilities.
In other words, the plain English supplement effectively describes market valuation; it is when the English is translated into mathematics that it goes wrong.
Both the current standards and the proposed revisions share the same failing: creating a false equivalence between risky assets and guaranteed liabilities. In both cases, a plan that takes more investment risk appears to become better funded. But the options-pricing example illustrated above clearly shows why this conclusion is false: the cost of a put option guaranteeing against adverse market outcomes rises along with the volatility of the investment whose returns it guarantees. Taking more investment risk simply lowers the contribution today while creating an equal and opposite increase in the value of the contingent liability imposed on future generations. The total cost of guaranteeing the future liability remains the same. Thus, taking more investment risk does not make a pension better funded; it merely increases the degree to which a plan violates the principle of interperiod equity.
The options pricing example also shows why government's supposedly longer time horizons do not make market risk go away. If stocks truly became less risky over longer holding periods, the cost of a put option guaranteeing against adverse outcomes would fall as the duration of the option increased. In fact, the opposite is the case: put options grow more expensive over longer time horizons. This reflects the fact that, while the standard deviation of investment returns falls over longer holding periods, the actual investment outcomes become more divergent. And it is the actual dollar values that are needed to pay benefits.
The proposed revisions include a number of salutary changes that will assist policymakers and the public in understanding the effects of the choices they make with regard to public pension financing. But the central issue can be summed up in the following question: does a plan that takes more investment risk become better funded? Both the current accounting standards and the proposed revisions answer yes, while economic theory and the practice of real-world financial markets say no. Until and unless this central question is answered correctly, both the size of pension liabilities and the steps that could address them will be misunderstood.
Andrew G. Biggs is a resident scholar at AEI