Critique of MEPSIM By Ronald J. Ryan, CFA, and Russell D. Kamp
The paper has several misleading, if not erroneous assumptions and facts.
You do NOT want to fund the deficit at a discount rate of 7% for several reasons:
- 7% discount rate is not a market rate and undervalues liabilities and the deficit significantly.
- Using a market rate yield curve of AA corporates (i.e. like FAS 158/ASC 715) would calculate @ roughly a 4.00% discount rate. This difference of 3.00% equates to a PV difference of 30% to 45% higher liabilities such that a 40% deficit would now jump to a 58% deficit.
- You want to fund and defease Retired Lives not the deficit. This produces certainty and transparency that these liabilities are fully funded. The Butch Lewis Act ONLY contemplates the loan being used to defease Retired Lives.
- Funding a deficit does not provide any evidence of where the deficit is (short, long?) or how to allocate funds to where the deficit is. Can’t cash flow match a single deficit number, as you need to understand where in the liability term structure is within this deficit.
The loan should be at the interest rate the Treasury can finance (even a little higher for a profit margin). You don’t want to create more of a Treasury deficit. If the loan rate was 15 to 25 bps higher than the Treasury financing rate this new government agency (PRA) could operate at a profit. Fifteen basis points on $100 billion in loans produces $150 million a year in revenue, which is more than enough to operate this agency at a profit.
You want time to cure the problem. The Treasury does not issue 20-year bonds. Let the Treasury issue its normal 30-year bond auction. According to historical facts, the average return on the S&P 500 over a rolling 30-year period is over 8%. If you buy time (extend the investing horizon), the odds of success dramatically improve.
Since this is an asset / liability objective running 500 iterations of a Stochastic model without running liabilities is not very helpful.
They calculate that the loan is expensive costing $56b defined as PV of loan plan repayments? We don’t understand this math. Why won’t the loans be repaid?
It is true that these plans currently invest in risky assets. If through our work a lower ROA can be used than the plan’s asset allocation should respond and switch to a less risky profile.
NO, NO, NO… you never want to cut benefits. This is certainly a last resort. Our approach buys time, lowers the hurdle rate (ROA), and fully funds Retired Lives. There should be no benefit cuts for Retired Lives.
The PBGC is not a solution because it has a very low cap (guarantee) for the payment of benefits (max of $17,160 for 40 years of service) and it is not currently solvent given the forecast of future needs. You want a loan to fully fund your Retired Lives. The PBGC could never fully fund Retired Lives given its current funding.
A 6% average return seems conservative over a long-time horizon (30 years). Only when an Asset Exhaustion Test (AET) is run will we know the true economic ROA needed to meet future plan obligations.
We recommend that the loan proceeds are used 100% to defease Retired Lives with investment grade bonds. This leaves the current assets in tact. The output from the AET will help us determine the economic ROA needed to fully fund residual liabilities. The asset allocation here could be heavily skewed to non-bond assets instead of the 40% allocation in the MEPSIM forecasts. Remember the loan is new money and does not affect current assets and/or contributions.
The Butch Lewis Act – The ONLY Solution
The loan from the government should be 30-years at the Treasury financing rate with a small premium (15 to 25 bps) to pay for this new agency’s (PRA) operating budget.
The loan proceeds should be mandated to fully fund Retired Lives through a defeasance process so there is no risk of default here or lack of funding resources.
The amount of the loan must be based on what it costs to defease Retired Lives with Treasury STRIPS. If a cash flow matching strategy is used on the loan proceeds (mainly investment-grade corporate bonds) it should fully fund Retired Lives at a cost savings of roughly 8%+.
This cost savings could be held in escrow or transferred over to current assets to help fund residual liabilities. If transferred over to current assets these savings will reduce the economic ROA hurdle rate to fully fund residual liabilities.
Moreover, we have replaced the actuarial ROA rate (@ 7.50%) on Retired Lives with a Treasury loan rate (@ approximately 3.25%), which lowers the economic ROA needed to fully fund residual liabilities.
The loan program’s (BLA) success must be predicated on work done on an economic basis (economic truth) and not actuarial forecasts and assumptions.