Pension plans and interest rates (revisited)

All investors follow the movements of interest rates closely, but none more so than pension plans, who feel the impact of those movements in both their assets and their liabilities.

Pension plans and interest rates

The value of a pension promise depends on interest rates: if rates increase then the value of pension liabilities falls, and vice versa. Corporate pension plans can offset that exposure through liability-driven investing (LDI), which involves replicating the liability’s sensitivity to interest rates in an asset portfolio—but most plans only do so partially: typical interest rate hedge ratios are around 30-60%. The net result is that pension plans in aggregate have a massive bet on rising interest rates.

The bad news is that the existence of a term premium means that the odds are probably stacked against that bet. I wrote about that a couple of years ago, and have just published an update of that work, adding another 23 months of data and including analysis of the ten-year Treasury yield (the original piece looked only at the two-year yield.)

The extra data do not change the conclusions about the behavior of the two-year yield. That yield has been rising pretty much throughout the whole of the period since the analysis was first published—but that increase has been gradual, with the result that a position based on an increase would not have paid off. This highlights a key finding of the work: that for an investor who has taken a position on rising interest rates, it is not the rise (or fall) in rates that determines whether they make a profit: rather, it is the rise (or fall) relative to the change that the market had priced in.

Evidence of a term premium in the ten-year yield

The results for the ten-year Treasury yield are summarized in the chart above. This compares:

  • the (twelve-month) change in the yield on a ten-year Treasury bond that was priced in to the market each day from the start of 1990 through the end of August 2014 (the “forward curve breakeven change”), with
  • the actual change in yield that subsequently occurred.

Throughout the great majority of the period covered, the breakeven change was an increase. Only 10% of the time did the market act as if it expected the ten-year yield to fall. Hence roughly 90% of the data points lie to the right of the y-axis in the chart above. Since the ten-year yield actually fell more often than it rose over this period, the persistence of the pricing seems to imply a term premium. It is, unfortunately, not possible to break out exactly how much of the forward curve pricing is due to true market expectations and how much is a term premium at any point in time. But it does seem reasonable to infer from these results that the break-even rate of change priced in to the forward curve was most likely higher than the market’s true consensus expectation for much or all of the period studied.

The outcome: about 26% of the data points (1,615 out of 6,172) lie above the orange breakeven line – so a position based on the ten-year yield rising would have resulted in a gain 26% of the time, and a loss 74% of the time.

The net impact on pension plans

So there are two issues at play here:

  1. Most pension plans have significantly greater exposure to interest rates in their liabilities than in their assets. This means that, in terms of the funded status of the plan, most pension plans are positioned for an increase in the Treasury yield.
  2. However, the existence of a term premium would mean that positions based on an increase in the Treasury yield result in a loss more often than a gain. That’s true even when the consensus market expectation is an increase – because market pricing already incorporates that expectation.

As of the end of August, 2015, the forward curve breakeven change in the ten-year Treasury yield was an increase of 0.26% over the next twelve months. To the extent that includes a term premium, the odds may continue to be stacked against pension plans.