When companies value pension obligations and required contribution rates, they make assumptions about the expected future investment returns. (Accounting standards require market-based rates reflecting fixed interest returns, but that’s a separate point).
So what assumptions are pension funds making? The WSJ has an interesting article showing that the average US pension fund is assuming future returns of approximately 8%. To put that in perspective, yields on 30 year T-bonds in the US are about 3.9%, 10-year yields are below 3% and inflation is currently about nothing. This is a huge real return and suggests that many of these pension funds may be underfunded.
It’s also interesting to work out what Equity Risk Premiums these valuation assumptions imply. FinanceClippings makes some educated guesses at likely portfolio construction, and estimates assumed ERPs of nearly 8%. For reasons I’ve described before, an 8% ERP is madness.
My own calculations
FinanceClippings assumes a simple portfolio mix of 50% equities and 50% government bonds in this calculation, and assumes the average yield will be consistent with 30-year assumptions. I would differ slightly here. If we are looking at an overall portfolio, I would expect some investment grade corporate bonds and property in the mix too. These assets could be expected to earn 1% to 2% over risk-free over time (after adjusting for expected default loss on the corporate bonds). These return assumptions may seem low to some, but this is another area where it’s easy to overestimate the possible returns based on inappropriate periods of data.
On the other hand, it might be a little optimistic to assume 30-year yields as the average yield for the portfolio, depending on the remaining term of these pension funds. If we use a shorter duration, the risk-free yields approach 3% quite quickly.
Using a range of assumptions that make more sense to me, I see the pension funds assuming ERPs of between 6% and 6.5%. Lower than the 8% to be sure, but optimistic all the same. (The sharp-eyed reader may see in my calcs that I am assuming an additive ERP rather than a compound ERP. Small differences, doesn’t change the conclusions at all.)
Check the calcs and assumptions and play around with them yourself –> Pension ERP implied assumption calculator tool [XLS]
Whichever methods one applies here, it’s clear that too-high ERPs are being assumed. It’s irresponsible and it’s setting up tomorrow’s employees, shareholders, customers and citizens for problems that companies and valuators simply didn’t have the courage to deal with today.