Thursday, December 3, 2015

Managing the Risk in Active Pension Liabilities

If you work with defined benefit (DB) pension plans at all, or if you happen to be among those responsible for them at a company that still sponsors an ongoing DB plan, you know that de-risking has been all the rage in recent years. And, this is not to say that this is not without good reason. With the inherent volatility in cash flow and accounting expense that has been added to the DB world by what some (including me) would consider poorly conceived funding and accounting rules, risk mitigation is often critically important.

So, let's consider what have been the trends in de-risking. There have been a mass of so-called lump sum windows in which former employees have been offered one-time opportunity to take a single sum distribution of the value of their remaining benefits thereby removing those liabilities from the plan's 'balance sheet' and thus eliminating the company's future liabilities with respect to them. And, there has been lots of liability driven investing (LDI). That is, companies look to invest assets in fixed income instruments of durations similar to the liabilities in the plan. By doing this, liabilities in the plan tend to move roughly in lockstep with the underlying assets intended to support those liabilities.

All this is great, especially for well-funded plans, but how about all the rest of those plans? And, these strategies tend to look at the liabilities that currently exist. How about the liabilities that will exist in the future (we used to consider those in actuarial valuations, but the law changes made them irrelevant in the annual valuations)? What do I mean by that? Consider an example.

The XYZ Pension Plan currently has its funding target (liability for accrued benefits). We can split that out among three groups of participants, two of which are composed of former employees. Suppose they are broken out as follows:

Retired participants: 100,000,000
Terminated Vested participants (former employees who are not yet in pay status, but are entitled to a benefit in the future): 30,000,000
Active participants: 70,000,000

for a total funding target of $200,000,000.

What do we know about the liabilities that compose this funding target? We know that the amount attributable to current retirees will generally decrease because for each year that a retiree lives, there are 12 months fewer of benefit payments that the retiree will receive in the future. This might tend to imply that this group represents a diminishing risk.

We know that the terminated vested group will become retirees. Generally, these are former employees with smaller benefits (they didn't work for the company until a retirement age). In the usual profile, they represent a far smaller liability than do retirees or actives. For a given vested term, the liability with respect to their benefit will generally increase as they approach retirement and then decrease. Lump sum windows were designed to target this group and have been quite successful. But, because their accrued benefits, and therefore attendant liabilities are smaller, the effect of having done this has not been as significant as it would be with a group of active employees.

The active employees currently accruing benefits are the most important group to de-risk. Why is this? There are several reasons:

  • In our example, and typically, their accrued benefits represent a meaningful liability to the plan and plan sponsor.
  • With each year that an active employee remains with the company, they are one year closer to retirement meaning that the discounting period until retirement has shortened.
  • In an active plan, these employees continue to accrue benefits (more on this to follow).
Yes, active employees continue to accrue benefits. Suppose we consider a concept that used to be paramount to actuarial valuations of DB plans -- the actuarial present value of projected benefits under the plan (PVB). This is the actuarially determined amount that if it sat in a pool of assets today would be sufficient to eventually fund the benefits of that individual. 

For retirees and vested terms, the PVB and the accrued liability or funding target or accumulated benefit obligation (ABO in accounting parlance) are, save differences in actuarial assumptions, the same as each other and the same as the PVB.

For active participants, they are not. In fact, for a mid-career person, the PVB may be more than twice as large as the liabilities we are consider. And, if we are de-risking, isn't it really the eventual benefit liability that needs to be de-risked? 

If you agree, you're correct, and if you're a plan sponsor, we need to talk. 

If you disagree, you're likely not correct, and if you're a plan sponsor, we need to talk.

There is a really easy approach to this that many companies have taken and that is to freeze benefit accruals under the plan. But, not everyone wants to do that. Some companies actually philosophically like the concept of DB plans. But, the funding and accounting rules have made it more difficult to get comfortable with sponsoring them.

Suppose I told you that you could mitigate the risk in the active PVB. And, suppose I told you that you could do that without freezing accruals. 

You'd be interested, wouldn't you?

Let's talk. You won't regret it. Here is one way.

No comments:

Post a Comment