Pensions are complicated.  Necessarily so, because of the wide numbers of parties involved, and the contingencies involved (mortality, morbidity, asset returns, insolvency) over a long period of time.  Anyone who has had a cursory look at the math (or regulations) behind setting pension liabilities, contributions, etc., knows how tough the issues are, and why real experts need to handle them.

I’ve worked at the edge of the pension business for much of my career.  I have designed defined contribution plans, created stable value products, done asset allocation for defined benefit [DB] plans, terminal funding, and other incidentals.  That said, I am a life actuary [FSA], not a pension actuary [EA].

Tonight’s main issue revolves around a good article by the estimable Matthew Goldstein of Business Week.  Steve Waldman, filling in at Naked Capitalism, commented on the article as well.

Here’s my take: it is legal today for companies to shift their pension liabilities to life insurance companies in the Terminal Funding business.  All they have to do is send a description of the liabilities of the plan to the dozen or so companies that are in the business with adequate claims paying ability ratings, and the companies will send back an estimate of what they would require as a single premium payment to take on the liabilities.  Low bidder wins (and loses — he mis-bid).

So, why don’t plan sponsors take the life insurers up on this?  Easy.  The cost of buying the annuities from the insurers is more expensive than the amount of assets in the trust.  For those companies that are overfunded, they don’t care to terminate — it is a great benefit for their employees.

Terminal funding was most common in the late 80s, when companies could terminate DB plans, and any excess assets would revert to the company.  Then the law changed, and most excess assets would be taken by the Federal Government.  Another reason why overfunded plans do not terminate — the excess assets are valuable to the plan sponsor, but are trapped assets.  They are valuable because they give flexibility, and reduce future contributions.

Why is it more expensive to buy annuities from insurance companies than the assets on hand in the trust?

• The main reason is that the plan sponsor gets to assume the rates he will earn on plan assets (within reason).  That rate will almost always be higher than the rate that an insurance company can invest at after expenses.  Pension funding rules are significantly more liberal than life insurance reserving and risk-based capital rules.
• Insurers must mainly invest in bonds, whereas pension funds can invest in any asset class, subject to the prudent man rule.
• Insurers must keep surplus assets to keep the company sound through downturns.  Pension plans have no such requirement.
• Insurance companies have profit margins and overhead that pension plans do not.
• Often there are funky, hard-to-value benefits in the pension plan.  Subsidized early retirement is the simplest of those.  The insurance companies don’t have a good way of pricing them, so they toss out some guesses.  Often the winner is the one that ignored the cost of the odd ancillary benefits.

Now, for a proposal from the Treasury to be effective, they somehow have to wave their hands at the issues that I just put forth.  Even if they allow other regulated financial companies to take over pension plans, they have the following issues:

• Who is responsible for shortfalls?
• Does the company taking over the plan have to put in some subordinated capital to give them “skin in the game.”  (Essentially, the life insurers have to do that today.)
• How do profit incentives work?  Do they accrue inside the plan as a buffer against shortfalls, or do excess earnings (however defined) get immediately  or over time paid to the buyer of the pension liabilities?  (You can guess what the liability buyers want.)
• How do underfunded plans get transferred compared to adequately funded plans?  Hopefully the plan sponsors of the underfunded plans have to pony up to fund them at levels that are adequately funded, then they can transfer them.  It would be a sham to transfer underfunded plans to an entity that says that can fund the plans because they have an ultra-aggressive investment strategy.  The blow-up will leave behind even bigger deficits.

Call me a skeptic here, while I call the head of the PBGC a Pollyanna.  To Bradley Belt: If you think this will solve your underfunding/insolvency problems, think again.  Only through high risk investment strategies succeeding can all of the underfunding be invested away.  Ask this: how would you feel today if the plan sponsors of underfunded plans all adopted highly risky investment strategies?  You would worry.  Well, unless the liability buyers have skin in the game, you will worry just as much after the sale of liabilities.

Sometimes I think politicians/bureaucrats believe in magic.  Some little tweak, a loosening of regulations, and poof!  The problem goes away.  It is rarely that simple, particularly when you are dealing with the math and complexities of long term compound interest, which in my opinion are inexorable.  (Kind of the inverse of compound interest being Einstein’s eighth wonder of the world — it is a wonder when you are compounding assets looking forward, without liabilities to fund, but when your discounted liabilities are greater than your assets, my but that eighth wonder of the world fights you fiercely.)

Now, I’m not going to discuss this at length, because I am getting tired, but the Wall Street Journal had another pension article this week.  A good article, and I must say that I don’t get how the practices described are legal.  The anti-discrimination rules were put into place to deter this issue.  Why they are not enforced here is a mystery to me.  Regulated pension plans should not be able to invest in the debts of non-regulated pension plans.  To allow anything else, is to make a mockery of the regulations.  (Another reason why regulated and non-regulated financials should be separated.)  The Treasury has anti-abuse rules that they can invoke against such practices.  Why don’t they use them?

My guess is that the Bush administration doesn’t care about the issue.  Perhaps the next President will care more.  And, with respect to the sale of pension liabilities, my guess is that that gets left to the next President and Congress, who will not allow the practice as proposed.

PS — One last note: what would be fair, if pension liability buyouts are allowed, is to allow participants the option to roll their net assets into a rollover IRA.  Back in the 80s, many people got burned by less than creditworthy companies who bought their pension liabilities and went belly-up themselves.  It is a normal aspect of contract law that you can’t take a debt and transfer it to another party unilaterally, unless the creditor consents.  So it should be in pension liability transfers.