Archive for January 22nd, 2013
One of the main points of contention about PERA has been the expected rate of return. Up to 2001, PERA used an expected rate of return of 8.75%, but lowered that to 8.5% from 2002-08, and again to 8% in 2009, where it now stands.
So how has PERA done since 2001, when it stood at a 100% funding level? The chart below shows the annual return over those 11 years for which they’ve reported, plus an estimated return based on a similar portfolio from CalPERS, the California pension plan. CalPERS has a .99 correlation with PERA, which is about as close to metaphysical certainty as anything human gets. In 2012, CalPERS reported a 13.26% return, which would project a 13.9% return from PERA, so just for grins, I’ve plugged that into the 2012 spot.
The red line is the running CAGR from 2001, or Cumulative Average Growth Rate for 2001 through the current year.
As you can see, over time the red line becomes less volatile, as more years are factored into the average.
The green and purple lines are where it gets interesting. Those lines are the return that PERA would have to see over the remainder of the 30-year window from 2001-2030, in order to meet the 8.75% and 8% targets. So, for instance, after 2007, in order to have a CAGR of 8.75% from 2001-2030, PERA would have to have returns of 9.3% from 2008-2030.
Note that using the 8% return from 2001 basically gives hindsight credit to PERA, since in 2001 they were projecting 8.75%, not 8%. Let’s blow up that part:
As you can see, PERA started out with poor 2001 and 2002 returns, which put it in an immediate hole. Even good returns from 2003-2007 only got the required returns down to 9.3% and 8.3%, respectively. Then 2008 happened. By now, we’re 12 years into that 30-year window, and even a good year like 2012 will only push the required return down a little bit, about 0.3%. For instance, it would take 7 years of 15% returns to get the required return for the 8.75% line down to below 8.75%, in effect, to catch up to the 8.75% projection, and it would take 5 more years to catch up to the 8% projection. It’s highly unlikely that even 5 years will pass without a normal, cyclical recession, and attendant lower returns.
The average return is 6.28%, the CAGR is lower, just over 5%. In fact, any set of returns with an average of 8%, but which are not actually 8% each year, will compound to less than 8%, making it that much harder to make up ground. Bear that in mind when someone talks about how easy it is to make up for a couple of bad years.
All of this argues for using a Monte Carlo simulation to determine solvency, rather than a simple rate of return. Such a model is a little more expensive to run, but will give a far more accurate picture of the health, or lack thereof, of any defined benefit plan.