Public Pensions
Consider this: you are saving for retirement and you have $100 in your pocket earning 7.25% interest. (We use that rate reported by the National Association of State Retirement Administrators, NASRA, as the current average, down from 8%.) . But now you are in a 4.75% earnings environment. How much money do you need in your pocket to achieve the same level of savings as 7.25% over your 30-year career? (Hint: about 90% more). Where will those assets come from?
Before 2000, it was possible for pension funds to achieve a 7.25% earnings rate. Public pensions were well-funded then and numerous elected officials increased benefits with agreement from plan managers and actuaries. Today, however, long term savers, whether pension plan managers, individuals with 401K plans, endowment fund officials or life insurers need to amass many more assets to produce the same level of earnings that was possible twenty years ago. What happened? Stated differently, 21st century monetary policy has helped to shift the role that earnings played for long-term investors to state and local taxpayers, employers, employees and individuals.
To be sure, factors other than setting aside enough assets are at work in pension and individual retirement funding. Greater longevity than what was expected when plans were first designed is an obvious and important factor. Public policy certainly contributed too. Elected officials increased benefits in the heat of the “dot-com” bubble (notably California but elsewhere, too), never expecting the market volatility that would come over the next two decades. Low rates, of course boost asset values and encouraged expansion in the home mortgage market. We need not repeat all the details of growth in subprime as well as “jumbo” mortgage lending leading up the Great Recession. But who would have expected a 20-year procession of economic defibrillators whenever markets and the economy had a swoon?
The following table illustrates earnings volatility among public pension plans followed by the Census Bureau since 1993. It’s a quirky chart since each bar totals 100% of the contributions to state and local pensions from employees, employers and earnings. In the second half of the 1990’s, asset earnings made a significant contribution to the total. When the dot-com bubble burst, and the terrorist attack on 9/11/01 on home soil shattered confidence, market losses contributed to asset decline in 2001-2002. Assets declined significantly again in 2008-2009 during the Great Recession.
Employer and employee contributions increased steadily between 2000-2018. Employee contributions grew at about 4.6% while government employer contributions (aka taxpayers) grew at an average annual rate of 8.3%. (Of note, these are aggregate national figures. Plus, employer/employee contributions are far smaller in scale relative to total assets.) Some plans did better, some worse; some contributed more, some took payment holidays. Including employer and employee contributions, total cash and investment holdings did not return to the pre-recession peak in 2007 until 2013 and have since grown (albeit with drops in 2012 and 2016).
The effect of market volatility since 2000 highlights the importance of compounding for pension asset growth. The median pension plan assumed an annual compound earnings rate of 7.25% or more, according to NASRA. During eight of the 18 years since 2000 in the chart below, actual interest earnings fell quite short of that target. If 7.25% had been the earnings rate through this period, pension plans would have amassed about $3.5 trillion more than they actually achieved (on a real dollar basis). Actual average annual growth in assets over the period 2000-2018 came to roughly 4.75%. That reflects growth, certainly, and healthy recovery of losses since the Great Recession, but nowhere near the target rate.
As many have observed, low rates have also pushed investors into increasingly risky assets – earning more in some years but facing volatility in others. A 4.75% rate is reasonable and incorporates the real world losses and gains over the the 18 years since 2000. The big question, to repeat, is where in in this policy environment will those extra assets come from?
Taking a step back to look at market and policy behavior over the last 30 years, the chart below shows four measures — on two different scales. Left vertical axis shows yields for 10-year and 30-year treasuries as well as the US federal funds target rate during this time. Right vertical axis shows the S&P 500. Shaded bars represent official recessions. As you can see in the chart, when the federal reserve ceased its tightening and then plateaued, recession followed soon thereafter. In each of the three recessions, the fed funds rate at plateau exceeded treasury yields (curve inversion). Whether there are sufficient guardrails today, or the recent federal reserve’s softened stance will mitigate a downturn, we cannot say. But we do know that another major drop in the equity markets will not only hurt pension funding levels, but will slow income and sales tax receipts that state and local governments use to pay for basic services.
As we were writing this we noticed the “Big Read” in this weekend’s Financial Times entitled “‘Their house is on fire’: the pension crisis sweeping the world.” (You can find the article here, but there may be paywall.) Authors Josephine Cumbo and Robin Wigglesworth note that the Netherlands is cutting pensions 10%. They comment:
This is not merely a danger to individuals…it could also have wider impact on the economy…by reducing consumption — the opposite of the intention of central banks when they cut rates.
FT, November 17, 2019
Public Higher Education
Not so coincidentally, expansionary monetary policies beginning in 2001 not only encouraged mortgage loans for housing, but student loan debt as well. Of note, where most other segments of employment were shedding jobs after the two major recessions during this period, public higher education did not. Headcount employment remained relatively stable.
However, state budgeteers typically cut aid to higher education when budgets get tight. Coupled with looser lending, higher tuition rates became the “moral hazard” of the last two decades, filling gaps left by state appropriation cuts. Tuition hikes were met with higher levels of student debt, which lenders were willing to supply. One offset we should mention is that higher education enrollment (and hence tuition revenues) tends to be counter-cyclical – enrollment surges along with higher unemployment and falls as the economy improves.
Student debt is, after all, a bet on future earnings of the borrower – or for the lucky minority, an option call on their parents’ pocketbooks when things don’t go as planned. Crowding out of other spending is not just at the state and local level – but at the consumer level as well. Today, student loan delinquencies greater than 90 days are around 10-12% — reminiscent of the sub-prime mortgage debacle in the 2008-2009 meltdown.
We show a series of charts to illustrate. The first, which we copy in its entirety from a report written in 2012 by John Quinterno for Demos, shows this trend. It illustrates the decline in state appropriations for higher education following three recessions, from 1990-2010 (real dollars).
As a percent of total state General Fund expenditures, Quinterno shows General Fund spending for higher education as a proportion of the total declining from 14.1% in 1990-1991 to 11.5% in 2010-2011, citing figures from the National Association of State Budget Officers, NASBO. NASBO’s latest report shows that proportion has declined further to 10.1% in fiscal year 2018.
Shown somewhat differently, and with updated figures, the following chart from the State Higher Education Executive Officers, SHEEO, illustrates how tuition growth filled the budget gap in public higher education, as state appropriations declined from 2001 and again from 2008 recession periods.
Filling the Gap: Tuition growth fueled by student loans helped to fill the gap left by declining state appropriations. Growth of international students, who pay higher tuition than local residents, also helped fill the gap. We show another chart from Quinterno’s report, illustrating the growth in student loans from 1999-2011. It is important to note that the Federal Reserve Bank of New York, FRBNY, revised its numbers in 2011 when it realized that it had undercounted the level of student loans and re-adjusted its figures. Given our focus on the implications of expansionary monetary policy we wanted to span the recession in 2001-2002, so we show the original figures, and then follow with the FRBNY’s corrected, updated chart. Note the significant rise in loans between 2004-2005 and steady growth since that time. The amount of student loan debt surpassed all other categories of household debt (except mortgages, which is not shown below) beginning in 2010 and has continued to grow.
In addition to the counter-cyclical character of higher education, the leading age of millennial generation reached college age in 1999. When looking at dollar growth of student loans (as in FRBNY figures) rather than dollars per FTE, it is important to unpack the various drivers. FT Alphaville’s Cardon Garcia (citing a chart from Goldman which does the unpacking) shows the composition of student loan debt growth among enrollment, higher tuition, and a greater proportion of students that borrowed. You can find that chart here.
Student debt hit $1.5 trillion in Q3 2019 according to the FRBNY’s latest report. Delinquencies >90 days in student loans have also has hit a high of 10.9% in the same report although the transition into delinquency has fallen from the last quarter to 9.3%.
There is also some weakening of international growth in attending US higher education institutions. This is less a measure of disinterest but rather a reaction to the slowdown in visa approvals as well as anti-immigrant sentiment under the current federal administration. This is not great news for higher education admissions officers. Educationdive.com reported that fewer international students are heading to the United States for the 3rd straight year. Dozens of college officials (including Harvard, Yale and Princeton) have written to the Department of Homeland Security urging that visa approvals be sped up so they don’t lose their international students. While the top ranked schools are somewhat buffered from this trend, smaller, less selective schools are seeing big losses, according to Rahul Choudaha, who was quoted in the article. A Chinese education official recently warned the country’s students about the risks of studying in the U.S. Iowa’s public universities have seen a 44% decrease in Chinese students since 2015 and Arizona State University has lost 20% of its Chinese student population since 2016.
Inequality Accelerator? Needless to say, pushback against high tuition rates and resistance to student debt, falling state support and loss of higher paying international students are significant financial stressors for higher education. Plus, the same kind of earnings volatility faced by pensions affects endowments. In September’s New York Times Education Issue, Paul Tough highlighted the difficulty for admissions officers who would like to diversity their student body but cannot do this easily, given their tuition dependence.
Elite schools say they’re looking for academic excellence and diversity. But their thirst for tuition revenue means that wealth trumps all.
Paul Tough, “What College Admissions Offices Really Want”, September 10, 2019
To be fair, the Tax Cuts and Jobs Act preserved the charitable tax deduction, while significantly capping the state and local tax deduction. Charitable contributions have grown since the act was passed, including major donations greater than $100 million. (See this report from the Council for Advancement and Support of Education.)
Finally, the connection between the student loan burden and low homeownership among young adults has now been well documented and we link just a few resources here and here. Once considered a great equalizer and ticket to future earnings and mobility, the four-year college degree has lost its luster. Aside from a slower housing recovery post-recession, slower wealth accumulation and weaker consumption for millennials and later cohorts will continue feed through the economy.
Municipal Bonds
To end on a more positive note we reflect on how the extraordinarily low interest rate environment has helped state and local governments to address infrastructure needs. The level of borrowing has surged in 2019 well beyond many market participants’ expectations, given the absence of the federal government’s infrastructure promises, the loss of the state and local tax deduction and elimination of advanced refunding.
Total borrowing as of the end of October, which was $330 billion, inched within $8.6 billion of last year’s 12-month post-tax reform total, according to the Bond Buyer. “New money” borrowing for infrastructure projects were a healthy $215 billion, for the 10-month period.
Of interest, with rates at rockbottom lows, refunding bonds for the 10-months through October were $73.8 billion, $14.4 billion higher than last year’s 12-month total. One contributor has been a new advance refunding construct: taxable refunding of tax-exempt debt, since spreads between taxable and tax exempt bonds are so tight and short-term rates elevated. Taxable borrowing as a whole was $16.3 billion higher for the 10-month period in 2019 vs. 12 months in 2018. A portion of taxable borrowing in 2019 were pension obligation bonds, particularly in California, in an effort for issuers to close the gap on their unfunded liabilities.
We expect these trends of state and local borrowing to continue through 2020, given the soft stance of the Federal Reserve.
