The Clearest Explanation Yet of Why Public Pension Accounting Rules are Garbage

Posted by Jeffrey Brown on Nov 29, 2011

Filed Under (Finance, Retirement Policy)

Over the years, many of my academic colleagues and I have pointed out the significant conceptual flaws in the way that the Government Accounting Standards  Board (GASB) computes the value of public pension plan liabilities.  Financial economists are virtually unanimous that using expected rates of return on pension assets to discount future benefits obligations is inconsistent with basic economic and finance theory, and that it has the bad outcome of massively understating the size of the unfunded liabilities.

Our frustration only increased when GASB issued an exposure draft of new rules that were portrayed as a compromise, but which in fact were a step backwards in terms of consistency with basic financial principles.  Under these proposed new rules, public pensions would be allowed to use expected asset returns to discount liabilities for the part that is funded, and then use municipal bond index to discount the unfunded portion.

Unfortunately, we have often had a difficult time explaining why this is so, because our explanations often rely on having our audience understand difficult concepts such as “risk adjusted returns” or “options pricing.”

Finally, however, one of our colleagues – Robert Novy-Marx of the University of Rochester – has come up with a brilliantly simple way to illustrate the flaws of the proposed new GASB approach.  In a new working paper from the National Bureau of Economic Research, entitled “Logical Implications of GASB’s Methodology for Valuing Pension Liabilities,” Professor Novy-Marx highlights three logical implications of the new rules, all of which fail even the most basic idea of common sense.  I will use today’s post to highlight my favorite of the three – the idea that under the GASB proposal, a public pension plan can improve its funding status by burning money.

He gives the following example, which I will simplify a bit further than he does in his paper:

Take plans A and B.  They have identical liabilities of $175,000 payable in exactly 30 years.  The only difference between the two plans is that plan A has $10,000 of stocks, while plan B has $20,000 of assets, including $10,000 invested in stocks and another $10,000 invested in bonds.  Put differently, plan B is identical to plan A in all respects except that it has an extra $10,000 of funding invested in a relatively risk-free asset.

It does not take any formal training in economics or finance to understand that plan B is better funded.  Indeed, I daresay that 100% of my 8-year-old son’s 2nd grade class could tell us that $20,000 is worth more than $10,000.

Yet, ironically, under the proposed GASB rules, plan A is better funded than plan B!  Indeed, plan B could improve its funding status by taking the $10,000 of bonds and lighting them on fire!

How can this be?

Because in the alternative universe in which GASB operates, public plans are guided to discount future cash flows by the expected rate of return on plan assets, while ignoring risk.  Thus, if stocks have a 10% expected return, and bonds have a 4% expected rate of return, then plan A gets to discount the $175,000 liability with a 10% rate.  Plan B has to use a 7% rate (the average of 10% and 4%).  When one does this, one discovers that plan A is almost exactly funded, while plan B is under-funded by $3,000.

This is an absurd result.  It proves by example that the GASB approach to measuring pension liabilities is meaningless because it does not provide a reliable measure of anything that we care about.

If GASB wants to follow its own guidelines of providing accounting standards that improve transparency and improve public-sector decision-making, then they should immediately replace these standards with ones that are consistent with basic finance.

23 Responses to “The Clearest Explanation Yet of Why Public Pension Accounting Rules are Garbage”

  • Marcus Casey says:

    Absurd is right. Are these rules actually analyzed in a systematic manner or just thrown into the document because “it sounds ok?”

  • Jeffrey Brown says:

    Marcus, I met with the GASB board and staff for several hours last month. I mean it sincerely (not condescendingly) when I say they are a bright, engaged, and hard-working group of folks who want to “do the right thing.” But most of them lack advanced training in financial economics, and so I honestly do not think that they all fully understand the case we economists make, in part because we have not done a very good job of explaining it! That is why this example is so instructive – it does not require a deep understanding of finance to understand that something is really wrong.

    They are also being bombarded by public officials and a large share of the actuarial profession with misleading – but “good sounding” – arguments to the contrary. I look at it this way: imagine a sentient creature from another planet landed on the earth, and had no understanding of our system of mathematics. We decide to teach them math. But rather than teaching them math by standard instruction, we do it by debate. So we have a MIT-trained mathemetician patiently explain why 1+1=2. On the other side, we had a Harvard-trained lawyer (someone really smart, but with a totally different area of expertise, and, in this hypothetical example, not particularly good at math) argue that 1+1=3. This alien creature would have the alien equivalent of cognitivie dissonance, and the resolve it by “compromising” and learning that 1+1 = 2.5.

  • Dean says:


    I’m glad you were not this condescending during the user forum in Chicago, or it would not have been the useful exchange of ideas that it was. Your assertion that the GASB has not aligned itself with you and your colleagues because we don’t understand what you’re saying is uninformed. In fact, we understand financial economics quite well, and would grasp it even if we didn’t have both staff and board members who had been through doctoral programs with including substantial study of economics. What you fail to recognize, I believe, is that there may be an acceptable answer other than your own, and that what is most meaningful and decision useful in accounting and financial reporting may be different from what is theoretically correct in economics.

    I can say without qualification that the staff and board have given considerable thought to the arguments you and your colleagues have made; in fact, it may be the case that the Board has spent more time deliberating your single issue than any of the dozens of other important decisions that comprise the GASB’s proposed standards.

    Personally, I don’t agree with your arguments regarding the valuing of the liabilities, despite having taken more graduate-level economics courses than I can count. (To paraphrase a recent sitcom, that’s because I took a lot of courses, not because I can’t count very high.) Perhaps I didn’t get the message. What I did learn, however, is that demeaning and ridiculing another person or their position does not meaningfully add to public discourse, but is rather the tack that people sometimes take when their own views are gaining no traction. I would never, for instance, refer to the views of financial economists regarding measuring the pension liability as “garbage” or coming from an “alternative universe.”

    It was truly a pleasure meeting you in Chicago and getting to hear from you first-hand. You are an evidently brilliant economist and we were quite grateful that you could spend an afternoon with us. I thought you did a very good job of conveying your views, and we all came away with some very good points to chew on. I hope that we will hear from you again in the future.



  • Jeffrey Brown says:

    Dean – sorry if I offended you, but the simple fact is that I have never heard an intellectually coherent argument that supports the current GASB position. You may have taken a lot of graduate economics classes, but I am guessing you have not taken any asset pricing classes or you would see the error in GASB’s thinking quite clearly. I have yet to find a single financial economics faculty member who finds the GASB approach correct or useful. There is not 100% agreement on the “right” way to do it, but there is virtually 100% agreement in the financial economics community that the current approach is wrong. I have spent considerable time trying to understand the alternative view, and found it completely lacking in rigor.

    I do appreciate you taking the time to talk in Chicago. But if GASB decides to stick with the current rules, I will continue to criticize them, as I think the are fundamentally flawed.

  • Dean says:

    I welcome your criticisms, Jeff, because I believe that meaningful comments on our proposals improve them. You are clearly capable of commenting without sinking to name-calling, so I found this column disappointing, rather than offensive. You’ve developed a persuasive illustration that doesn’t need to be saddled with vitriol. (BTW, I never took an asset pricing course, but there are people here who have.)

  • Jeffrey Brown says:

    I did not “name call” any members of the board or staff. Indeed, in my first comment I referred to all of you personally in complimentary terms. I have also done so privately to my colleagues, many of who have been less complimentary.
    I did, however, refer to the rules as garbage, and I continue to believe that is an accurate description. As for “alternate universe” – maybe that went too far, but I will just call that a rhetorical flourish.

  • Sasquatch says:


    I am an undergraduate student that was e-mailed this article. Jeff’s points seem valid to me. Could you explain to me why his argument is invalid?

  • Frank Keegan says:

    Everybody would feel a lot better about this issue if those arguing for status quo had anything at stake. The fact is they have nothing to lose and everything to gain by continuing to defer adequate funding of state and municipal pensions. By the time the bill comes due, they will be gone, and taxpayers and public workers will be stuck with a $4 trillion tab. As nearly as anyone can tell, the economists advocating reality have nothing to gain, now or in the future, by doing so.

  • Jeffrey Brown says:

    Thanks for your comment, Frank. There is little question in my mind that one of the reasons (though not the only one) that pensions are so underfunded in so many states is because legislators place insufficient weight on future generations. For a wide range of issues – entitlement underfunding, climate change, etc – legislative bodies have consistently been unable to make optimal long-term decisions, so the problem is not unique to public pensions.

  • Mike Bishop says:

    It’s clear that highly trained professional economists have done such a fine job with the U.S. economy, not to mention that of the European Union, that we should rely only upon their expertise to determine the present or future value of assets.

    It seems to me that people like Friedman and Laffer were/are highly trained professional economists. Look where their conclusions have led us.

  • Jeffrey Brown says:

    Whether one likes the economics profession or not, and whether one likes the economics discipline or not, it is difficult to excuse a methodology used by GASB that utterly fails to account for risk. Those who disagree with us have yet to put forth any coherent rationale for the use of expected returns that is logically consistent. This is not an ideological issue – it is an issue of measurement.

  • I was one of five west coast financial analysts that participated in the “Users Forum” GASB had in San Francisco a few months ago to discuss their proposed pension accounting/reporting rules.

    First – I also found them very committed – smart – trying to do the “right” thing – very experienced in their professions. Perhaps my most important impression was they will not be turned around by political pressure from defenders of the status quo standards.

    I made an argument in favor of lower discount rates – but wasn’t successful in getting my idea across.

    But the more I’ve thought about it, the more I think the most impactful part of the proposed standards isn’t the value of Net Pension Liability to be put on the Balance Sheet – it’s the pension expense going on “Income Statements”.

    Today governments are allowed (not exactly this but in effect just about this) to recognize the pension expense associated with unfunded pension obligations as they make payments to reduce the unfunded value. This is like saying the payment of a debt created the debt. My County is spreading reporting the pension expenses that created its current unfunded obligation over the next 30 years. They would be reporting the last of the past pension expense that created today’s unfunded obligation in 2040!

    When the new rules take hold governments with large unfunded obligations – and large Pension Obligation Bond debt – are going to be drenched in red ink – a veritable blood bath because GASB will force them to flush all those unreported past pension expenses out of their hiding places and onto the Income Statements. The rules about when pension expenses must be reported in the future are complex – but the result will be that governments will have to report specific causes of pension expense anywhere from the year following events or over as many as – maybe 12 years – depending on the nature of the cause. Today they report generic “pension expense” over as many as 30 years.

    I really believe this is the most valuable change in GASB’s proposals.

    The biggest change to their current poroposals I hope GASB adopts regarding reporting the Net Pension Liability isn’t the discount rate. I REALLY REALLY hope they require governments with Net Liabilities to report the expected payment schedule as part of the required footnotes – adding them into all the other debt payments so that people will see what’s in store.

    After all – if there’s enough theoretical justification to put Net Pension Liablity on the Balance Sheet – then there is equal justification to put the projected payment schedule in the footnotes.

    I think that would have more real impact on the public than whatever the value of the debt is reported to be.

    I’ve also concluded that – although I know the rules should be “solidly grounded” theoretically – bottom line I’m more interested in a radical change in behavior in financial management.

    If allowing the assumed rate of return to be the discount rate significantly reduces the political push back (which could possible include states excusing local governments and themselves from having to abide with that part of the new rules if lower rates were specified) and we get all the rest of the reforms (with the change in the projected payments schedule) – then as a practical matter it would be well worth it.

  • Jeffrey Brown says:

    Thanks John. I told them they could render this entire controversy irrelevant if they would simply require disclosure of future benefit cash flows. I really hope they do it! Then we can all discount or otherwise use however we think is appropriate!

  • Ed Ring says:

    I emphatically agree with Jeffrey Brown and applaud his choice of words to describe the state of public pension accounting rules. Using a weighted average return projection based on asset composition does yield perverse portfolio management incentives, as Brown points out. But beyond this difficulty is the overly optimistic rates of return being used by the pension funds in their projections. The probability that pension funds, who are market movers insofar as they collectively now have literally trillions of assets under management, can earn 7.75% per year or more is virtually nil. The broader market has scarcely budged over the past decade, yet pension funds are still going to earn 7.75%, year after year, decade after decade?

    Using rough numbers, for every 1.0% a pension fund has to lower their projected rate of return, the required annual contribution as a percent of payroll has to go up by 10% (this sensitivity is lower in when the projection is in the high single digits, but higher once you get under 6.0%):

    Put another way, the size of the unfunded liability leaps by 10% or more for every 1.0% the projected rate of return falls:

    It was pension funds, among others, who depended on debt fueled economic growth to help them achieve their unrealistic 7.75% annual return targets. It was, and is, pension funds who are manipulating the market by investing in hedge funds in order to get over market returns.

    At this point, pension funds can either lower their projections, and the unions who essentially control the pension funds can return to the benefit plans they used to have back in the early 1990′s, or they can drive the economy all the way off the cliff. The slide has begun, and you don’t have to be an economist to see it happening. You just have to understand basic finance.

  • Mike Bishop says:

    I understand, Jeff, that econometrics has fully enveloped your discipline, but there are still enough of us who remember that it started as a social science to justify skepticism when you make such predictions.

  • Jeffrey Brown says:

    Mike – last I checked, I did not make any “predictions.” I simply pointed out a substantial logical and conceptual flaw in GASB’s thinking, namely, that they ignore risk. If you are concerned about “predictions,” then you really ought to be outraged about what GASB does because they are implicitly making a prediction that the pension plans will earn the assumed expected return – and because they ignore risk they are essentially treating it as if their prediction is perfect! In other words, if you want to be skeptical of predictions, you should be 100% in favor of using a risk-free rate.

  • Jeffrey Brown says:

    Thanks Ed!

  • Herb Whitehouse says:

    There is another aspect to the Robert Novy-Marx “burning money: illustration that is very important. We know that not all of the liabilities in a plan will be due in 30 years, but instead, that some will due in one month, others in 3 months, others in 3 years and so on.

    It may even be that a wise and prudent pension trustee had developed a funding policy that invested $10,000 in a combination of bonds and other assets at a 4% rate in order to have its investment policy rationally related to its funding policy need. That need — in this example — is to have $10,000 in less volatile investments precisely because of its shorter term payment obligations.

    Now the Robert Novy-Marx example is being used to suggest that GASB and its exposure draft call for burning the $10,000 in bonds in order to use GASB rules to artificially create a higher funding ratio. But the ED suggests something quite different. Now it is true that the current imprecise language in the EC intends that the appropriate rate at which to discount the liabilities, regardless of whether they are short term or long term, is the expected rate for a long term investment strategy, whatever that is. But, obviously, this was and is not the intent. The intent and almost all current practice is exactly the opposite of the Robert Novy-Marx illustration. No one uses the 10% return expectation for stocks (even if that might be appropriate”long term” rate if no liabilities were payable for 30 years) when there are shorter term obligations.

    In other words, the suggestion that I made on October 20th in Chicago is actually already inherent in the GASB ED; namely, that a reasonable discount rate must be one that is not only tied to existing trust assets, but to the expected return on those assets when they are invested in a way that is rationally connected to the actual funding policy needs and objectives of the plan.

    In other words,to suggest that GASB and the EC is irrational because it calls for the use of a 10% discount rate when the investment policy will only be so unconnected to rationality that it invests 100% in stocks is itself illogical. The only problem with the GASB ED is that it does not expressly state what is implicit in it already; namely, that the discount rate can be the expected return on existing assets when those assets are rationally connected to the funding policy needs and objectives of the plan.

  • Jeffrey Brown says:

    Herb – you miss the point entirely. Robert used the 30 year example just to simplify the math so it was easy to explain. And the use of the 10% and 4% are also examples. If you think they are too high, then fine, give me lower ones. It would take a bit of time, but I can guarantee you that Robert or I could still illustrate the logical flaw. You want a stream of payments every year for the next 50 years, using an equity return of 8% and a risk-free return of 3%? Fine, we could construct such an example. The point is that GASB is treating the equity premium as a free lunch, and it is not.

    By the way, you say nobody assumes a 10% return on stocks? Really? A lot of pension funds have 60% stock allocations are are assuming a portfolio-weighted return of 8%. So either they are assuming equity returns of 10% or more, or they are assuming absurdly high bond returns. Or they are justifying the high returns on the basis of higher allocations to alternative asset classes such as private equity, venture capital, agriculture, and the like – but still making the mistake of treating these higher returns as riskless.

    Look, many of us have tried to point out the flaws in GASB using sophisticated approaches – risk-adjusted pricing, put-call parity, etc – and a lot of people just don’t get it. So Robert came up with a simple explanation that shows why it is a fundamentaly flawed concept. Now you are criticizing it because it is simple? Okay, then let’s go back to the sophisticated approach. Those demonstrate the same point much more generally anyway. Trust me, I woud be delighted not to have to resort to simplistic examples to point out what is so obviously wrong with these standards.

  • Herb Whitehouse says:

    It is easy to have misunderstandings on quick blog posts. So let me take it slow so that you do not miss the point entirely.

    Yes; I am sure that some public sector investment advisers expect a 10% return on stocks, or perhaps a 6 to 7% real return over inflation. I further concede that this may influence an actuary even if there is no formal equity assumption in an actuarial valuation.

    But I did not intend to deny even that public sector actuaries may use expected returns that are far higher even than the capital market assumptions that the plan’s investment advisor will use in constructing the pension portfolio. So lets skip over all your comments that talk about whether or not I might think Robert’s return illustration was to high. Those comments entirely miss the point.

    with respect to your thought that I may be criticizing the example because it is too simple, that also entirely misses the point., I tried to set up the point being made by Robert by saying that if it an accounting standard is crazy if it shows a higher funding ratio on the facts in Robert’s illustration after burning half the plan’s assets, then it would be even loonier if the liability was not a single payment coming due in 30 years; but rather, actually included shorter tern liabilities for which less risky assets — even if they have a lower expected return — might be very appropriate in the real world.

    Now getting back to the key point that you missed…

    …perhaps you might pint out to everyone the public sector pension actuarial firms that use an expected equity return as the actuarial assumption for the expected return on assets — especially when there are short term liabilities that the fund trustees have decided warrant investments in lower volatility (with lower returns) assets.

    If these actuarial firms do not exist, or exist if violation of there professional practice standards, then try to follow the main point that I had tried to share. A small part of my point is that actuaries do not use equity returns assumption as the actuarial expected return for a pension fund. The reason is one that I suggested was inherent in the GASB standard; namely, that the investment policy must, if it is to be used as the basis for the actuarial return assumption, bear a rational relationship to real world pension payment needs.

    Now, I am not saying that public pension fiduciaries are very good at realizing that investment policies should be ordered to a plan’s funding policy objectives. Quite the contrary.

    But as I pointed out in Chicago, this inadequacy in public sector pension funding management does not mean that GASB intends to suggest that the connections do not need to be made. My point in the blog was to say that what I recommended making explicit in the standards, is necessarily already implied in the standards.

    The example that Robert and you have used — despite your “guarantee” that Robert or you can illustrate the logical flaws in the GASB position, pushes logic so far outside of reality and actual practice that it amounts to a reductio ad absurdum.

    You have succeeded in pointing out to everyone the need for GASB to make explicit what is implicit in its ED; but you have also pointed out how far from reality this “burning money” position really is.

  • Jeffrey Brown says:

    I am afraid you are the one missing the point. Namely, that how assets are invested is 100 percent irrelevant for determining the present value of the liabilities. If I owe my bank 1000 dollars per month for the next 30 years, the value of that liability does not change just because I invest my savings in stocks vs bonds. Basic finance. A fund may choose to invest in bonds, stocks, or apple orchards. Will effect the risk profile and expected returns, but it does not change the present value of the liabilities.

  • Herb Whitehouse says:


    This is “a” central point in the public pension accounting debate; but it is not one that everyone who does not agree with you about that point has missed.

    I just think that it is a shame that anyone who takes your position is willing to lose the opportunity to improve GASB public pension accounting standards in an all or nothing campaign to win this central point.

    I am not asking you to stop advocating for what you hold out as “truth”, “logic” and “basic finance”, however, I would appreciate any help in moving the public sector pension world toward very achievable changes to “accounting gimmicks” (c.f., Bill Gates) where we might find common ground.

    For example,we might even agree that an accounting standard that either a) moves actuaries and public sector investment fiduciaries toward the equivalent of burning lower volatility and lower volatility assets; or b) encourages unsophisticated public sector actuaries and investment consultants to take inappropriate investment risks on the grounds that GASB accounting standards do not contemplate that an expected return can only be reasonable where it has been designed to achieve reasonable funding policy objectives.

    Reasonable funding policy objectives, include liquidity needs as well as the need to fund benefit cash flows in over the next five to ten years.

    This is especially the case in underfunded plans with large short term benefit payment obligations. Moreover, a reasonable funding policy will not permit the government plan sponsor to hide behind deterministic return assumptions (especially long-term based) when the critical funding policy need is to protect against the real and very material risk — even with the framework of normal Gaussian volatility without any black swan investment results — of running out of plan assets in 5 to 10 years.

    My position has been that the only way to read the intent of the GASB ED — given that the intent is not to encourage the burning of low volatility and lower return assets — is that more lower volatility and lower returning assets need to be included in the expected return when a reasonable funding policy calls for such an investment approach.

    I agree that the language in the ED referencing “long term” expected returns is absurd when, for example, the plan only has short term pension obligations, or any material short term obligations. My point to you is that it would be helpful if you not belittle GASB or say that its ED demands the absurd, but rather that, to say the opposite.

    If you say that (because the members of GASB are intelligent are reasonable people) that the ED language must mean something other than encouraging investment policies that are not ordered to achieving reasonable funding policy objectives (e.g., burning money) then perhaps the language will be changed.

    That result would seem to be more important than insisting that GASB use your entire framework or nothing.

    There are other areas of common ground that your (Robert’s) analogies might be very useful; for example, 1) pushing GASB toward disclosing its fundamental benefit cash flow assumptions; 2) disclosing the individual asset class earnings (arithmetic), volatility, and correlation assumptions on which the actuarial earnings assumption is based; and 3) showing a simple funded ratio based on the market value of existing (not expected future contributions) plan assets against the present value of accrued benefits calculated off the benefit cash flows.

    If we aan investmentespecially when the

  • Jeffrey Brown says:

    I think we found a point that we can agree on – I told GASB at least 6 times during my discussion that the single most important thing they could do would be to mandate disclosure of year-by-year cash flows. That would allow a far more robust analysis and understanding of the true fiscal impact. I believe this point resonated with them, as it is clearer how to do it, less controversial, and would make much of the discount rate debate a less central side show.

    I’ve enjoyed our exchange. I guess time will tell what they choose to do …