DiNapoli announced today that he’s approved a recommendation by the State Retirement System Actuary to reduce, from 6.8 percent to 5.9 percent, the assumed rate of return (RoR) on investments by the $268 billion Common Retirement Fund, which underwrites the New York State and Local Employee Retirement System (NYSLERS) and Police and Fire Retirement System (PFRS), of which the comptroller is the sole trustee.
To be sure, even at 5.9 percent, the RoR that the pension fund literally counts on to pay constitutionally guaranteed benefits will remain considerably higher than the yields from commensurate low-risk U.S. Treasury or high-quality corporate bonds, which currently range from 2.3 percent to 3.3 percent. Nonetheless, in isolation, cutting the RoR assumption is an unequivocally good and prudent thing for the comptroller to do.
Assuming lower earnings also tends to result in higher required contributions by employers—which is why politically sensitive public pension fund administrators across the country have tended to set their RoRs at much higher levels than those required for private corporate plans. To guard against volatility in investment returns, which has been especially pronounced over the past 25 years, DiNapoli and other pension fund administrators also resort to “asset smoothing” — i.e., counting average market returns over several years—as a basis for estimating the assets available to pay retirement benefits. In New York’s case, the smoothing period is five years.
For the last 20 years, state and local pension plans’ assumed rates of return have been far too optimistic. The distributions of average (geometric mean) assumed investment returns and actual returns from 2001 to 2020 demonstrate this. The figure below shows the distribution of the average assumed investment return rate versus actual investment returns for 200 of the largest state and local pension plans in the United States. The median assumed rate of return over the last 20 years was 7.7 percent per year, the median actual rate of investment return for these public pension plans was 5.7 percent.
This two percent difference helps to explain the nearly 30 percent drop in the average pension plan funded ratio over the same period. In recent years, many pension plans lowered their assumed rates of return.
State pension systems dropped the rate of return they assume for their investment portfolios again, continuing a two-decade long trend that public-finance experts say is necessary, even as it presents some challenges for the entities that participate in such plans.
A dramatic, recent example of this dilemma occurred in a Massachusetts district court proceeding, when an employer agreed to a $59.17 million settlement in a proposed ERISA class action accusing it of using outdated mortality rates to calculate pensions. Cruz v. Raytheon Co., Mass. Dist. case number 1:19-CV-11425-PBS, Feb. 16, 2021.
The employer had argued in its motion to dismiss that the retirees failed to make the case that the plan violated ERISA by unreasonably using a mortality table created in 1971 and a 7% interest rate to calculate retirees’ alternative annuity benefits it said would be “actuarially equivalent” to the plan’s benefits. The employer argued that its conversion factors for determining the alternative annuity benefits were reasonable and that the retirees were attempting to force their own arbitrary actuarial assumptions. The employer further asserted that under ERISA, employers sponsoring pension plans have wide discretion in determining which actuarial assumptions or conversion factors can be used, requiring only that the single life annuity (SLA) normal form of benefit is equivalent by actuarial standards.
Author(s): Jeffrey D. Mamorsky, Richard A. Sirus, Greenberg Traurig, LLP
Overall, the typical pension fund is now expected to return approximately 6% annually over the next 20 years, compared with 6.4% pre-pandemic (Figure 4). This change in outlook is consistent with the reduction in the median long-term return found by a recent survey of pension investment consultants13 and aligns with revisions for public pension return expectations published by S&P Global.14
The $116 billion North Carolina Retirement Systems has lowered its assumed rate of investment return for the third time in four years, cutting it by 50 basis points (bps) to 6.5% from 7% annually.
The target return had already been reduced to 7.2% from 7.25% in 2017 and again in 2018 to 7%. Prior to then, the rates had been left unchanged for nearly six decades even though the two main state pension funds—the Teachers’ and State Employees’ Retirement System and the Local Government Employees’ Retirement System—have, on average, underperformed their assumed rates of return over the past 20 years. In fact, the new target rate of 6.5% is still higher than the fund’s estimated 20-year return of 6.28%.
One of the most important assumptions built into pensions is the guess at how much investments will grow over time. If investments provide high returns, then lawmakers don’t need to set aside as much money today to pay for pensions to be paid out in the future. If investments do not return as much as assumed, a gap develops between what has been promised and what has been saved. The bills cap the rate at which administrators can assume their investments will grow, allowing them to be no more risky than current policy allows. They also let administrators use less risky assumptions if they think it is prudent.
This is a smart approach. Much of the current pension debt exists because administrators overestimated investment returns. Taxpayers now owe more to the pensioners than they do to the lenders and bondholders who willingly lent the state money. Putting a cap on the assumptions administrators make can prevent future pension debt.
Author(s): James M. Hohman
Publication Date: 1 March 2021
Publication Site: Mackinac Center for Public Policy