Underfunded pensions..not such a slam dunk

I picked up the link to the following report on twitter: THE FUNDING OF STATE AND LOCAL PENSIONS: 2013-2017

The report stated that the funding status of state and local pension funds was at some 72% of liabilities and would likely rise further assuming a return to historical returns.   The report did look at current funding liabilities under a range of equity return scenarios none of which were as low as many pundits (Hussman, GMO and myself) would model.   The lowest return figure of 4% (all assets combined) would yield a liability with a net present value of $3.8trn or circa 22% of current GDP or 32% of personal consumption expenditure ….this translates into my vague assumption that if future GDP rates are based on current consumption, then future return assumptions would need to incorporate pension fund liability shortfalls.  

The report also discussed the fact that many financial economists recommend using the risk free rate of return to model liabilities.   I am not a big fan of liability driven investment (that is matching nominal fixed liabilities with bonds, real rising liabilities with equities and index linked liabilities with index linked risk free assets) for very long duration liabilities.  It skews economic consumption towards the future (unless future pension objectives are reduced) and skews the demand for and hence price, risk and return of assets.   

I am however a big fan of adjusting return assumptions for cyclical market and economic risks (extended market and economic cycles are the easiest to model) and adjusting asset allocation for the size and timing of liabilities over the short to medium term (5 to 10 years).  Under a variable assumption regime for funding liabilities, liabilities would not be as heavily impacted by short term changes in asset valuations (rising markets would see falling returns and falling markets would see rising returns) and would be anchored by more realistic and risk focussed longer term return assumptions.  

Portfolio asset allocation should more or less reflect economy wide relationships: in other the words the portfolio balance needs to reflect asset/liability relationships which are themselves mirrored by consumption/saving, production and investment decisions.

Please note the following twitter conversation which spurred these comments (my thoughts are long held beliefs):


As I also noted, variable return assumption generation is not part of today’s orthodoxy.   Most equity return assumptions are based on an estimate of future nominal growth rates to which dividends are added.  This ignores valuation issues and the impact of the evolution of the return profile (short term risks are higher) on the ability of portfolios to meet liabilities.   So if we were to look at both valuation risks and structural risks to GDP growth going forward, we may well have at the lower bound much larger liabilities within pension funds that are just not going to be matched.

Leave a Reply