We were surprised to see two articles implying a connection between municipal bankruptcy, underfunded pensions and other post-employment benefits (aka OPEB).  The first, from the Wall Street Journal (WSJ), channeled Meredith Whitney’s 2010 prediction that there would be 50-100 defaults stating that her views were correct, just too early.  The second, published by ZeroHedge, suggested that New York City faces imminent bankruptcy due to the low funding level of the city’s retiree health plan. 

We find both predictions misleading for investors and it is worth delving into why.  Funding levels alone do not tell the whole story. Patricia Healy of Cumberland Advisors also penned commentary on the WSJ article and we have also had discussions with other analysts. 

The key reference in the WSJ article is a study by Pew Trusts that shows a decline in aggregate public pension funding over the last three years.  Before associating this change with bankruptcy (insolvency for states) the important question to ask is: what caused the decline?

  • Is funding low because officials took recurring “payment holidays” skipping required contributions to the plan?
  • Or is the funded ratio low because the state has reduced its discount rate assumption (lower discount = higher liability).  Many have called for lower discount rates to more realistically reflect the interest rate and earnings environment since 2000.  Michael Cembalest, Chairman of Market and Investment Strategy at JPMorgan Asset management has a chart showing discount rate changes over the last three years.  Notable among those that lowered their assumptions are Colorado, Indiana, Minnesota, Kentucky and New Jersey. 
  •  The bottom line here is that the persistently low interest rate environment requires far greater accumulation of assets to meet liabilities — no easy task for pension, OPEB or individuals with 401(k) accounts.  
  • OPEB sponsors typically use a low discount since most governments have not pre-funded in the same way as pensions.  New York City used 3.01% for FY2018, compared with typical discounts between 6-8% for pension obligations. According to Cembalest, “the average OPEB discount rate used by the states is 4.5% with a standard deviation of 1.4%.” 
  • Pre-funding of OPEB is a good idea, but OPEB is a more flexible benefit than pensions and pre-funding is even arguable for this benefit.  First, the lower public sector retirement age (particularly for police and fire) leaves a gap before Medicare kicks in.  Many governments pay for healthcare during the gap.  Others cover the cost of retiree’s healthcare premiums.  Over time, as longevity has increased, many governments have raised the age of retirement shortening the gap.  Others have modified healthcare benefits; yet others have reduced costs by implementing wellness programs.  
  • To reduce OPEB costs, many governments have setup “Section 115” irrevocable trusts; there are an estimated 150 trusts in California alone.  These give employers greater control over investments and help smooth out the required payments overtime. 
  • (We caution that GASB allows allows for a a higher discount rate when a trust is established.  Like pensions, to the extent there are insufficient resources to meet the liability, a blended rate is supposed to be used.  However, we observe some governments using the same 7.5% rates across the board.  For example, aee Michigan Public School Employees Retirement System, MPSERS.)

But what about bankruptcy?

To see if a government is headed for bankruptcy, it is more appropriate to look at the total cost of fixed obligations as a percent of the government’s budget.  These include debt service, pension and OPEB according to what the government should be contributing rather than what they actually contribute.  Contributions are set to amortize the unfunded amounts on a level dollar basis over 30 years.  (Level dollar payments avoid an escalating cost over time.  without going into greater detail, you can find more discussion here (SM7; page10).)

The fixed cost is measured against “own source revenues”.  This is important, too, particularly at the state level since Medicaid is a significant share of state budgets but a portion is a pass-through from the federal government.  Cembalest uses the described methodology in his annual study “The Arc and the Covenant”; the Center for Retirement Research at Boston College (CRR) does as well (see their report) and we believe Build America Mutual employs this methodology when evaluating the fiscal condition of governments they are considering for bond insurance.

The Results:

Cembalest and CRR show a wide spectrum of these fixed cost burdens – that, for sure,  defies a simple aggregation.  Using 15% of revenues as a reasonable cut off, ten states were above that level in 2017 in the Cembalest report, with Illinois leading the pack at nearly 50% of budget.  The others include New Jersey, Kentucky, Hawaii, Massachusetts, Connecticut, Maryland, Delaware, Pennsylvania and West Virginia.  The rest fall below the 15%.  We note that West Virginia, Maine, Rhode Island, Louisiana and Michigan are paying close to what they should contribute based on this methodology. 

Also, if you compare the Cembalest and CRR approaches with Pew, you arrive at different baskets of states in the “worst” condition vs. “best” condition.  For example, Cembalest includes Minnesota and New Hampshire in the best category with Colorado, Rhode Island, South Carolina, Vermont, Louisiana, Alabama in the second-best category.  On the other hand, Pew includes Colorado, Minnesota, New Hampshire and South Carolina in their “worst” categorization.  If you are worried about bankruptcy, we believe that the Cembalest and CRR approaches more accurately reflect fiscal condition.   

Finally, as we’ve already said in different ways, “macro” averaging of all state funded levels is of little use to investors looking to purchase a particular bond from a particular state.  CRR published a report in October highlighting the growing divide among states’ funding of pension benefits.  They found the top third of states had average funding levels of 90% whereas the bottom third averaged 55% funding.  Most states are fine and are tackling the cost of their programs in a variety of ways while there are a few standout “bad actors”, the worst of which may have approached a level of no return without major fiscal crisis.  (See our prior piece “Is Illinois Insolvent” from 2010).