Pension Math

Over $10 Trillion Is Managed In a Very Strange Way

Know someone who might like Capital Gains? Use the referral program to gain access to my database of book reviews (1), an invite to the Capital Gains Discord (2), stickers (10), and a mug (25). Scroll to the bottom of the email version of this edition or subscribe to get your referral link!

A defined-benefit pension is, in one sense, a very clever financial product—if it didn't exist, the first fintech to build it would be worth a ton of money. Here's the problem statement: people tend to save a bit less than they should for retirement, and not everyone is equipped to make good investment decisions on their own. And those people can be retired for a very long time; some might die early, but some stick around until their 90s (and more now than ever).

If everyone tried to save enough to ensure that they had a good income past 65, society would, in the aggregate, save way too much. And some of that money would be drained by bad luck, scams, or avoidable mistakes.

Enter the defined-benefit pension: if every worker at a given company sets aside some portion of their paycheck, and the money gets pooled and invested, then the aggregate portfolio, if managed properly, will have high actuarial confidence in being able to provide a set income for beneficiaries. That’s why retirees with a defined-benefit pension get a fixed amount of cash in the mail each month. But arranging this is hard, which is why these pensions have gotten rarer outside of public-sector jobs. Still, government and private-sector defined-benefit pensions manage about $10 trillion.

So, what does proper management look like? Since the money is pooled, it makes sense to hire professional managers. These managers aren't necessarily setting out to beat the market or anything—they’re just trying to figure out the highest expected return they can get for the appropriate level of risk. And since this pension money is locked up for a period that's both long and predictable, the pension fund is able to invest in assets that will be illiquid for long periods, at least some of which should be accorded a higher return for that exact reason.1

To give you some context here, academic finance really only got going in the 1970s, and it then took a couple of decades for it to become widely adopted. So historically pension fund asset allocation was more a matter of taste than rigor. Which isn't really a commentary on the quality of people doing the work; there simply wasn't enough data out there to be especially precise. So pensions tended to aim for conservatism, and used broad asset allocations. For example, in 1980, General Motors changed their pension allocation from 50/50 to 70/30 stocks and bonds. In a modern context, this is both a simplistic allocation—no alternative assets? No breakdown by type for either category?—and a massive swing. Today, a five-point swing in asset allocation at a large pension produces multiple articles in the relevant trade magazines.

The metric that ultimately governs pension behavior is its target return. A pension is a pool of assets with some associated liabilities; a "100% funded" pension fund is, phrased another way, an investment vehicle with infinite leverage: every single dollar of assets is matched by a dollar of liabilities, and there's no equity cushion at all. But pension funds' liabilities are extremely predictable, based on when people retire and how long large numbers of them live on average. So this leverage doesn't have the same effects that it would in other contexts. A 90%-funded pension isn’t a disaster (yet), even though in a sense it’s an entity whose debts are larger than its assets. In fact, pensions can and do get underfunded, where if their assets produce the return the pension targeted, it still won’t meet its obligations. This is not necessarily a crisis, because pensions have time to make up the gap; they're expected to have some volatility from year to year, so a well-managed pension will spend some fraction of its existence underfunded.

Hold the pension's payout constant, and then raise the target return, and you get two effects:

  1. It makes the pension better-funded: the quickest "solution" to being underfunded is just to raise expected returns so the fund requires less money today. This reduces the amount of money beneficiaries need to pay into the system over time; if a pension expects 7% annual returns instead of 6%, it needs a smaller starting base in order to get a given annual return.

  2. It increases the amount of risk the pension manager needs to take. If your pension targets a 4% return, you can get most of that with treasury bonds. If it's aiming for 8%, you're going to need equities at a minimum, and almost certainly some other asset class like private equity or venture.2

Choosing an expected rate of return means balancing different interests. It also means reacting to how the pension has performed recently. The paradox of pensions is that underperformance is simultaneously empirical evidence that the fund is being too aggressive (because equities introduce volatility that is hardly there for a fund invested totally in bonds), and a reason for pension fund managers to want to aim for higher returns to make up the gap without annoying anyone by demanding more money.

On the CFA exam, there will sometimes be a question like this: a pension fund recently lost money on its investments and is now only 90% funded. Should the fund manager:

a) Reduce the risk of the portfolio and aim for a lower expected return.

b) Do nothing.

c) Increase the risk in the portfolio to make up the gap.

The correct answer from the perspective of someone who likes it when the checks to retirees never bounce is a), and that’s the answer the CFA Institute wants you to pick. But c) is the realistic answer that describes how pensions actually behave. To make a) happen, you need two different psychologically difficult things: first, the pension managers have to admit to themselves that they made a mistake, and to subsequently lock in that loss. Second, lower expected returns mean higher required contributions to get the same payout, so pension beneficiaries all have to be told "Your take-home pay is being docked a bit because, unbeknownst to you, the person entrusted with your retirement decided to take a flier on international stocks/currency derivatives/structured products/crypto." Since many pension recipients are unionized, and since they typically make less money than the people running the pension fund, the politics of this are challenging.

For people who just want to have a safe retirement and not worry too much about it, this is terrible news; it means their pension manager is incentivized to run the Gambler's Fallacy strategy at all times, doubling down whenever they've lost money in the hope that they can make it back. Interestingly enough, though, there's a macroeconomic upside. Pensions will underperform in two scenarios: 1) a long period of slow growth and low rates, such that their fixed income assets are a drag on performance; or 2) a big market drop. If their response to this is to pile into risk assets even more, this has two effects:

  1. In the slow-growth scenario, it subsidizes new technologies. It also encourages private equity firms to do some brutal but necessary reallocation of talent, by buying underperforming firms and "restructuring" them (read: firing people). This is unpleasant for the employees involved, but for the economy as a whole it's bad for talent to be locked up in companies that aren't using it effectively. In some situations, that's exactly what causes slow growth, and it tends to be self-reinforcing over time.

  2. In a market crash scenario, the riskiest assets tend to perform the worst. If pensions lose money and want to make it back quickly, they'll allocate more to those risky assets. So the pensions end up being a countercyclical buyer that stabilizes the market.

This is by no means the goal of designing pensions the way they currently work. It's entirely possible to have a pension system where expected returns are set by looking at interest rate benchmarks—since pension funds have long-duration liabilities, they should own lots of long-duration debt to hedge, but doing this also means giving up some expected returns. It's an important thing to understand because pensions are such a huge asset class, and because these incentives are unique to them. Everyone wants to make pro-growth investments at times of low growth, or to buy when there's blood in the streets. But only pension funds are, quite by accident, structured in a way that more or less forces them to.

Read More in The Diff

The Diff has alluded to pension problems in a few different places.

A Word From Our Sponsors

Every day Refind picks 5 articles that make you smarter, tailored to your interests. Loved by 200,000+ curious minds.

1. As an individual investor, you should demand a higher expected return for the illiquid version of something for which a liquid market exists: a loan is worth less money than a bond to an equivalently creditworthy borrower; a single property needs a higher return than a publicly traded REIT; if you're buying shares of a unicorn on the private market, you expect better returns than investing in a similar company that's already public. In practice, these return gaps are often low, in part because the entities whose equity is publicly traded are the competing bidders. If every office building traded at, say, a 10% discount (to where it would be valued if it were owned by a REIT), any REIT that levered up to buy additional properties would automatically create value by making them liquid. So the real takeaway for individual investors is usually "Don't buy the illiquid version unless you're highly confident that this specific asset is materially undervalued. And if you're good at spotting such opportunities, maybe you ought to be doing it as a full-time job.

2. One way to understand the rise of pod shops, and the fact that their absolute returns have come down over time, is to think of them as an asset class that outperforms pension funds' hurdle rates and doesn't correlate that well with anything else the pension invests in. That's what they're selling—and, critically, if they target a 10% return instead of a 20% return, pensions have to allocate twice as much to them in order to get the same contribution to overall returns, and that allocation leads to higher fees. 

Share Capital Gains

Subscribed readers can participate in our referral program! If you're not already subscribed, click the button below and we'll email you your link; if you are already subscribed, you can find your referral link in the email version of this edition.

Help us learn more about you!

We're building a media kit and need some anonymous aggregated subscriber data. If you work in finance and have time for a 3-question survey, that would be very helpful:

Join the discussion!

Join the conversation

or to participate.