Friday, April 1, 2011

To Market Value or not to Market Value

Author's Note: As usual, this blog does not necessarily represent the opinion of my employer, my profession, or any other organization of which I am a member. In fact, by the time I get done, it may not represent my opinion either, but if by the time you get done, I've caused you to think, to ponder, to smirk, and to laugh, then I have done my job well.

As some of you know, I've recently returned from actuarial March Madness sometimes known as the Enrolled Actuaries (EA) Meeting. If you are an Enrolled Actuary and you've never been, you should attend. It's not a boondoggle. It's an action-packed meeting (well, maybe not action-packed, but it is packed with more learning activities per hour of attendance than any meeting that I know of) with lots of expert speakers. At this meeting the last fairly large number of years, there have been three General Sessions (plenary sessions if you like the idea of showing your audience that you can use a fancy word when a simple one will do better) and eight tracks of Breakout Sessions, as well as a usually entertaining luncheon speaker on the first day. That's nearly 13 hours of continuing education in 2.5 days, but since actuaries get to count 50 minutes as an hour for these purposes, we think it's more like 15 hours.

Anyway, Session 002 (or GS-2, if you prefer) was a public pension plan debate. Even if you have been under a rock, you have likely heard that public pensions are in a state of crisis. Hell, everything is in a state of crisis these days. I know that's a fact because I've heard it on TV. The economy is in a state of crisis, the War on Terror is in a state of crisis, Japan (sadly, this is true) is in a state of crisis, the middle east is in a state of crisis, kids are in a state of crisis, adults are in a state of crisis. Well, many of them probably are, but after the EA Meeting, I don't know about public pension plans.

Why is that? Well, as is within the norm for a general session, we heard from three experts, a think tank person who says that public pension plans in the US are about $500 billion in the whole, a think tank person who says that US public pension plans are more than $3 trillion in the whole, and a person who represents the state pension plans in the US who I think told us that it doesn't matter.

I digress. What are think tanks and what do think tank people get paid for? I think that the think tank takes a position and then asks its think tank people to think of things to make people think that this particular think tank has correct thoughts. I can think. Maybe this is something for me to do after I have had enough of my current career.

Anyway, back to the subject at hand. The big difference between the $500 billion thinker and the $3 trillion dollar thinker, other than $2.5 trillion, is that one of them thinks that a concept called market value of liabilities is the only thing that matters and the other one thinks market value of liabilities doesn't matter at all.

Who's right? We'll come back to that later.

First, let's explore what Market Value of Liabilities (MVL) is. Back in the early days of ERISA, and the current days for public pension plans (for the most part), actuaries got to discount pension plan liabilities at a rate representative of the expected long-term rate of return on plan assets. So, if you thought your plan assets could return 10% over the long haul, then you discounted at 10%. Gradually, in the world of corporate plans, that opportunity disappeared. Prevailing wisdom, or at least opinion, among those who legislated and their advisers began to move those discount rates toward a risk-free rate. What's that? Well, for the sake of argument, we usually assume that US Treasuries are risk-free instruments, because if the US government starts to default on its obligations, not much else will matter. And, since nothing else is truly risk-free, even many MVL advocates suggest that discount rates can be tied to the yield on very high-rated (Aa or better from Moody's) corporate debt instruments.

That may all be fine. But, here is where I have a problem. For each purpose (funding, accounting), a plan, usually through its actuary does a valuation of assets and liabilities. And, for each valuation, there is a measurement date. And, we are required to value the assets and the liabilities as of the measurement date. On the liability side, that means that we choose actuarial assumptions, including a discount rate that are appropriate on the measurement date. If the measurement date is December 31, and on December 30, the duration-weighted yield on corporate Aa bonds is 5.75%, but on December 31, it is 5.50%, then we are to use 5.50%, even if it's back up to 5.75% on January 2. This is nuts. I realize that we have to pick something, but why one moment in time should get to ruin a plan/fund/company/city/state for a year is beyond me.

How bad can this example be? Suppose we have a plan that on December 30 of some fine year is fully funded with $1 billion each of assets and liabilities (using a 5.75% discount rate). Let's assume that December 31 is a bad day for pensions. Discount rates decrease by 25 basis points and the value of the funds in the plan's trust fall by 1%. This has happened before, by the way. Then, assuming that the plan's liabilities have a duration of 12 (very reasonable by the way), at the close of business on December 31, liabilities will be approximately $1.03 billion and assets will be about $990 million. Stated differently, our funded status is now barely more than 96% and our plan is $40 million in the hole. So, which one is correct? Are we fully funded or 4% away from that? Are we fully funded or $40 million to the bad?

I think that the answer is neither one, at least not for purposes of reporting to the world at large, and not for purposes of doing calculations that affect the next year. MVL is a nice concept. Choosing assumptions that you think are long-term appropriate is a nice concept. Neither one is perfect; in fact, both have significant flaws. But, just like in every thing else being debated, the vocal ones are those toward the fringes.

In the case of public pension funds in the US, I think they are more than $500 billion underfunded, but I don't think they are nearly $3 trillion underfunded. Assets fluctuate. Discount rates are not static. Assumptions used for many public pension funds are way too aggressive.

Market value accounting is a good guide to choosing reasonable assumptions, just as is a prudent longer-term view. Did you hear that? Both are good guides, but nobody seems to want us to consider both and pick something appropriate. Perhaps it's because they think we've shown we can't. But, there is a bit too much hysteria right now.

It's time for a happy medium. I think we should call it reality value.

No comments:

Post a Comment