What Interest Rate Trends Tell Us about Pension Plans in the Age of COVID
State of Pensions 2020 Series
At the end of the Great Recession in 2009, statewide retirement systems in America reviewed their funded status and found a $1.16 trillion shortfall. Teachers, police, firefighters, and other public workers had been collectively promised $3.1 trillion in retirement benefits, but states had less than $2 trillion in assets to pay them after two years of steep losses from the financial crisis.
In the decade that followed financial markets experienced historic growth and a bull market ran from April 2009 all the way until the COVID-19 pandemic set in during the spring of 2020. State pension fund assets took advantage of that market and grew by $1.6 trillion over this period — but promised benefits grew faster and increased by $1.8 trillion.
As a result, going into the COVID Recession, statewide pension funds in America have more in unfunded liabilities ($1.35 trillion as of 2019) than they did after the end of the last recession, and things are only going to get worse from here.
One of the most fundamental reasons for this phenomenon is the decline in interest rates over the past few decades and the failure of state legislatures and state pension boards to adequately adapt and update their assumed rates of investment return. We documented this in State of Pensions 2020, and laid out the following charts.
- Interest rates have been in steady decline.
Lower interest rates have meant the returns pension funds have been able to earn on fixed-income investments like bonds have been much lower than previous points in history.
At the same time, the cost of making financial guarantees generally has grown over time as interest rates have declined. Whether looking at financial guarantees like pensions or life insurance, it is simply more expensive to make long-term promises today than ever before in history. In fact, the cost of guaranteeing payments just 10 years in the future is nearly 10 times more expensive today than in the 1980s.
2. States took several years before making meaningful reductions to their investment assumptions.
The nearby chart shows the delta between the highest and lowest assumed rates of return used by statewide retirement systems, going back to 2001. The average assumption was around 8% until the financial crisis of 2008, after which states gradually started to lower their returns. But it was slow going. In 2016 the average assumed rate of return was still 7.5%, only since then dropping more rapidly as pension boards were encouraged by their actuaries and outside analysts to adopt more realistic assumptions.
3. State pension fund assumed rates of return have been particularly slow compared to declining interest rates
The slow pattern of assumed return reduction relative to interest rates has tacitly meant pension funds are taking on two kinds of risk. One form of risk is associated with underperformance — with assumed returns so much higher than risk free rates of return, pension funds have to take on more high risk, high reward types of investments to try and hit their goals. These alternative investments, such as private equity, hedge fund strategies, and others can produce strong returns, but also carry their own form of performance risk.
Trying to hit a 7.2% annualized assumed rate of return (the average as of 2019) with bonds yielding less than 3% (at the end of 2019) has meant piling investment risk on top of the risk of unfunded liabilities increasing if those investments do not perform.
4. State assumed returns are still between 60 and 200 basis points higher than they should be.
There is no objectively “correct” assumed rate of return. Actuaries and economists have debated in academic literature over whether expected returns on investments should be used as a proxy for “discounting” the value of liabilities. But, at a minimum, there is a consensus that the assumed rate of return shouldn’t be much higher than the 50th percentile return on a projection of a given pension fund’s asset allocation.
At the end of 2019, one of the nation’s largest actuarial consultants Milliman estimated that the 50th percentile long-term return for the 100 largest plans in the country was 6.6%. Using that standard, the current 7.2% average assumed rate of return is roughly 60 basis points higher than it should be.
By another measure, the current average is roughly 200 basis points higher than it should be. If statewide pension plans had anchored their investment assumptions to interest rates as of 2001, then by 2019 the average assumed rate of return would be closer to 5.1% than 7.2%.
Exactly what probability a state might want for its assumed rate of return could be related to the state’s underlying economy and capacity to absorb underperformance (e.g. California can take more risks than Kentucky) and related to the risk appetite of its state legislature.
Looking to the Future
Even before the pandemic hit, states were using assumed rates of return at the most optimistic end of realistic expectations about the future. Now with years of COVID-19 putting downward pressure on the economy, states will need to lower their expectations even further than they should have. Using the post-Great Recession behavior of states as a guide, we should expect states to continue lowering their assumed rates of return, albeit slowly. The speed at which this change is made will likely influence how much risk persists within public plans.
This article is part of a series based on Equable Institute’s “State of Pensions 2020” report.