At Russell, we see our role as being to help our clients earn the rate of return they require at a level of risk they can survive. That’s a neat formulation: but when it comes to defined benefit pension plans, we need to stop and ask ourselves what precisely is the rate of return that they require. (See last week’s blog for a discussion of the return objective for non-profit organizations.)
Making things harder is the fact that one obvious candidate—the plan sponsor’s Expected Long Term Rate of Return on Assets (ELTRA)—is, for reasons I will get to in a moment, usually not a good proxy for the true required return at all.
But, before we worry about what the required return is not, let’s think about what it is. The impact of pension plan returns is felt in contribution requirements and in the surplus/shortfall position of the plan. Investment returns plus plan sponsor contributions together need to (a) match the growth in existing liabilities, plus (b) cover the value of new benefit accruals, plus (c) get rid of any shortfall in the plan’s funding over some reasonable time horizon. So if it were not for the “at a level of risk they can survive” clause in our description above, then the ideal return target (sometimes called the hurdle rate) would cover all of these without any additional contributions at all being required.
But unless the plan is frozen and well-funded, that hurdle rate may well not be feasible. You cannot necessarily ask the investments alone to do all of the work; some contributions may be necessary too. So, in practice, the actual target return from the investments is derived in conjunction with a funding (contribution) policy. And it is generally implicit. That is to say, a decision on asset allocation policy is made based on expected contributions (or surplus) rather than directly based on expected return. But the return target is there nonetheless—typically somewhere around 5-8% in our experience.
So what about ELTRA? The typical U.S. corporation currently sets its ELTRA in the range of 7–8%. That’s just the U.S. approach, though: International Accounting Standards in effect set ELTRA equal to the liability discount rate. As my colleague Jim Gannon has pointed out to me, if a corporation decided to switch from Financial Accounting Standards (FAS) to International Accounting Standards (IAS) (resulting in a reduction in ELTRA), that would not be a reason to reduce the pension plan’s return target from 7.5% to 4.8%.
And, since we have just described the return target in terms of contributions and the plan sponsor’s balance sheet, ELTRA, which forms part of neither calculation, is irrelevant anyway. But what if we wanted to define risk in terms of pension expense and the earnings statement (which do use ELTRA)? In that case, we run into a different problem: the current approach to pension expense, which dates from 1985, is a truly bizarre calculation. One of its oddities is that if you earn less than the ELTRA, the impact of doing so does not show up in the earnings statement for quite a long time (if at all) thanks to a series of corridors and other smoothing mechanisms. The upshot of the bizarre calculation is that it’s almost impossible to meaningfully measure the risk of different investment strategies in terms of their impact on earnings.
I am not denying that, for some corporations (such as utilities whose rate case includes pension expense) ELTRA can matter. But, even in those cases, ELTRA matters as an end in itself, not as a return target; so even when ELTRA is important to a corporation, performance relative to ELTRA is still not a useful measure of success.
So, convenient as the idea of using the earnings statement’s expected return as the plan’s return target might appear, it’s not the right answer.
One afterthought: I have noted in the past (most recently in 2011) that a new U.S. standard for the treatment of pension expense should be expected at some point. Let’s just say that whatever momentum this initiative once had, it seems to have lost. I’m sticking to my “change is a-coming” stance, but it’s sure taking its time.