Fed Warns US Pensions Reaching For Yield

Klement on Investing posted a nice comment on the sad state of public pension plans in the US:
I recently came across a fascinating paper from the Board of Governors of the Federal Reserve that investigated the financial situation of public pension plans in the US and their reaction to low interest rates. In three simple charts this paper clearly demonstrated the dire straits public pension funds are in.

Let’s first take a look at the official funding ratios (the ratio of assets relative to the present value of future liabilities) of the 100 or so public pension plans in their sample. In 2001, the average funding ratio was close to 100% but despite the strong returns of stock markets and declining yields in Treasuries that boosted returns on all fixed income asset classes the average funding ratio declined to about 70% in 2016.

Funding ratios of public pension funds in the US

Source: Lu et al. (2019).

And these are only the official funding ratios. Currently, generally accepted accounting standards require pension fund liabilities to be discounted by a discount rate derived from the asset mix of the pension fund. But why the risk and timing of liabilities should be related to the risk and timing of equity returns, for example, is anyone’s guess. It makes no sense to use equity returns as input into the discount rate for pension fund liabilities. Instead, as practically every pension fund expert in the world has pointed out by now, one should use discount rates that reflect the risk and timing uncertainty of the liabilities of the pension fund.

Unfortunately, this data is not publicly available but in the paper the researchers made an effort to estimate the true funding ratio of the public pension funds in their sample. And the results are disastrous. The average funding ratio is more likely around 40% than 70%!

Estimated actuarial funding ratios of public pension funds in the US

Source: Lu et al. (2019).

In most countries, private pension funds with such massive underfunding would be forced to either restructure or close. Yet, thanks to some simple accounting tricks, the true level of underfunding has been covered up for years.

One of the tricks to use to keep funding ratios high is to keep expected returns for risky assets, such as equities high because that will allow the pension fund to discount liabilities at a higher discount rate. And despite a ten-year equity bull market and steadily declining Treasury yields, the expected returns of public pension funds remain stable at 8% per year. Every investor knows that expected returns vary over time, depending on the valuation of the assets today. But public pension funds seem to be oblivious of that fact.

Of course, they aren’t. They know full well that their expected returns are too high at the moment and that their true funding ratios are lower than their published numbers. But they can’t say that out loud because that would start a discussion about who is going to fund the gap. And in the case of public pension funds the answer to that question is very simple: it’s taxpayers.

So instead of owning up to the problem, politicians and trustees of public pension plans rather kick the can down the road and assume unrealistically high expected returns and discount rates for liabilities. And to achieve these high returns, they have to reach for yield, particularly in the fixed income space where current yields are extremely low. No wonder, bank loans, high yield and private debt are so popular with pension funds these days…

Expected returns of public pension funds in the US

Source: Lu et al. (2019).
First, let me thank Drew Wells, an astute blog reader of mine, for bringing this comment from Klement on Investing to my attention. As a grumpy, middle-aged Canadian who is increasingly cynical on politics and markets, I enjoy reading thoughts on the markets from a "grumpy, middle-aged German" who rightly points out that the true liabilities of US public pension plans are grossly understated.

I read the paper from the Federal Reserve Bank of Boston, Reach for Yield by U.S. Public Pension Funds, and while I found it a bit dry and theoretical, it covers important issues.

Below, I list the five main results that emerge from the authors' theoretical analysis:
  1. The effect that the funding ratio has on risk-taking captures the effect of reach for yield. In our regression analysis we interpret the coefficient on the funding ratio as capturing reach-for-yield effects. 
  2. After controlling for funding ratios, interest rates may also affect risk-taking because they affect risk-premia. We interpret the coefficient on interest rates in our regression analysis as the risk-premium effect. 
  3. The reach-for-yield and risk-premium effects interact in theory. We allow for their interaction in our empirical specification. 
  4. How state finances affect risk-taking depends on whether states can shift risk to their debt holders. If they do shift risk, then state debt-to-income ratios that are large enough lead to higher risk-taking, especially for underfunded pension plans. If states don’t shift risk, then greater state debt is predicted to lead to lower risk. 
  5. The effect that state finances have on risk-taking also interacts with the risk premium channel of risk-free rates. If states can default on their debt, then for state debt-to-income ratios that are high enough, lower risk-free rates lead to higher risk-taking.
As stated above, in order to better measure the extent of underfunding, the authors revalued the funds' liabilities using discount rates that better reflect their risk.

They find that funds on average took more risk when risk-free rates and funding ratios were lower, which is consistent with both the funding ratio and the risk-premia channels. Moreover, consistent with risk-shifting, they also find more risk-taking for funds affiliated with state or municipal sponsors with weaker public finances. They estimate that up to one-third of the funds' total risk was related to underfunding and low interest rates at the end of their sample period.

In layman’s terms, US public pension funds that are woefully underfunded take more, not less risk, especially if they come from states with weaker public finances (think Illinois which is quickly running out of pension time).

Pause to think what that means. Instead of hunkering down, preparing for the next downturn, these public pensions are dialing up the risk to "reach for yield" and if another crisis hits, they will be obliterated and the state or federal government or even the Fed might be required to step in to bail them out to bring them back to fully funded status.

That is the ultimate end game and I've been warning my readers for some time to prepare for the Mother of all US pension bailouts.

I'm not stating this to scare or anger you, it's pure math and logic. There's a limit to how long this underfunded charade can go on and if a major crisis hits underfunded US public pensions, there will be a point of no return (there's only so much you can issue in pension obligation bonds and there's a limit to how high you can raise real estate and personal income taxes to make up for your pension shortfall).

Now, realizing the end game might be near, many US public pensions are finally abandoning their 8% pipe dreams even if that means a higher contribution rate for public sector unions and state local governments.

In my humble opinion, conditional inflation protection is next, there will be no choice but to partially or fully remove indexing of pension benefits, especially at mature, chronically underfunded pensions.

Go back to read my notes from the CAIP Quebec & Atlantic conference where I noted the following from my discussion with Anne-Marie Fink, Portfolio Strategist, Alternative Program Management - State Street Global Advisors and the former CIO of Rhode Island's state pension fund:
I couldn't resist but ask her about Gina Raimondo and whether her critics are right or wrong. She told me they are wrong and Rhode Island implemented three important things to address their shortfall:
  1. Adopted a hybrid DB/ DC model where if returns are over 7.5%, you're ok in DB model
  2. Made ARC payments, made sure budget surpluses are used to pay pension payments
  3. And most importantly, adopted conditional inflation protection where COLAs are partially lowered if the return target is not met (based 1/2 on CPI and 1/2 on return target).
If other states go the way of Rhode Island, perhaps we can avoid the Mother of all US pension bailouts but I'm highly skeptical.

And even in Rhode Island, these are not wholesale changes to the governance of their state plan and my fear is that until they change the governance and adopt a Canadian model of governance (ie. no government interference), they will never address what is truly ailing their state pension.

I can say the same thing for many other states, even the ones that are doing well like Wisconsin, you need to radically transform their governance structure to make sure there is no government interference whatsoever.

That's just not going to happen in our lifetime (too many special interests have their hands in the state pension cookie jars).

And that unfortunately, sums up the true sad state of US public pensions.

Again, just remember this: pension deficits are path dependent, meaning the starting point matters, and my biggest fear is many chronically underfunded US public pensions are already in big trouble, and won't be able to recover if another crisis strikes and and bond yields plunge to new secular lows, exacerbating pension deficits.

One thing is for sure, negative rates are already placing pensions in uncharted territory and in Europe even the best-run pensions are starting to worry, and rightfully so.

Below, two years ago, Howard Marks, Oaktree Capital Group co-chairman and co-founder, discussed US pension plans with Bloomberg's Erik Schatzker at the Bloomberg Global Business Forum in New York, stating a US pension crisis looms (September 2017).

Next, Nisha Patel, portfolio manager at Eaton Vance, examines underfunded pension plans across the United States and how some states could benefit from recent Supreme Court decisions. She speaks with Bloomberg's Taylor Riggs in this week's "Muni Moment" on "Bloomberg Markets" (August 2018).

Lastly, earlier this week, General Electric Co. announced it is freezing pension benefits for about 20,000 US employees in a move that could help trim $5 billion to $8 billion from the company’s pension shortfall. Bloomberg Opinion columnist Brooke Sutherland explains the details on "Bloomberg Daybreak: Americas."

America's ongoing public and private pension crisis is deflationary and bond friendly. It's also one of the structural reasons why I'm convinced deflation is headed to the US, and that's not good for US or global pensions.

Update: Arun Muralidhar, co-founder of AlpaEngine Global Investment Solutions, shared this with me after reading this comment:
I was a discussant for this paper at a recent MIT conference and I argued that this is largely the Fed's own doing (along with weak regulation and poor theory). Pension funds holding lots of equity against bond like liabilities got a shock with the Tech Bubble bursting/GFC and funded status declined below 100%. The Fed pushing rates lower only worsened the funded status (because it negatively impacts liabilities by more than it benefited assets) leaving funds with no option but to increase risk....which sadly will not end well with the next equity correction. My paper " The F-Utility of Wealth" in JOIM lays out the challenges of theory in more detail.
I thank Arun for sharing his comments with my readers but he's touching on a more complex issue which I can't get into here, namely, Fed policy and pension deficits.