A fire truck.

A fire truck. Credit: iStock

There’s a bank run happening in Dallas. The funny thing is, it’s not happening at a bank.

The city’s pension for firefighters and police is disastrously underfunded. Start with the normal problem that virtually all public pension plans are facing (overpromising benefits, underinvesting in the fund). Add in an attempt to make good on promises by investing in “non-traditional” assets like timber and real estate, which didn’t pan out as the investors had hoped. Add in a deferred retirement program which allows people to accumulate assets in the fund long after their retirement date — and to withdraw that money in a lump sum whenever they want. And what you get is a big mess.

Alarmed by the state of its pension fund, city officials have started talking about fixes. Folks with money in the fund have become justifiably suspicious that those fixes might include cuts. They are hastening to withdraw their money. At the moment, those withdrawals actually help the fund, because it’s required to pay out a fixed rate of interest that is below the current investment returns. But if the withdrawals continue, the fund will soon be in a parlous state. Naturally, this makes people want to withdraw their money, in order to get it out before the thing collapses.

As an example of what can happen to a pension fund, this is interesting, but not particularly representative. Pension funds do not have to allow lump-sum distributions, though many do. But it is nonetheless useful to look at what’s happening in Dallas, because it illustrates why bank runs happen, and why they’re so hard to stop once they’ve started.

The ultimate source of a bank run — the original sin of the financial system, if you will — is the conflict between what we want as borrowers and what we want as savers. As savers, we want a vehicle that will guarantee that we will not lose money and that we can get that money out before we need it. As borrowers, we want a predictable, fixed payment, preferably for a long time at a low interest rate. In economic terms, savers have a very high preference for liquidity (the ability to easily turn your investment back into cash), while borrowers have a very high preference for illiquidity.

That translates into the price of investments: the more illiquid it is, the more the investors are willing to pay for the privilege in the form of higher interest rates or other investment returns. This is why 30-year mortgages have higher interest rates than 15-year mortgages — and 30-year mortgages nonetheless dominate the market. The same sort of logic applies to assets, like stocks, that are generally relatively liquid, but are also volatile — meaning that their value can move around a lot over short periods, so that while you can usually liquidate, you might have to do so at a loss.

The good news is that while people want to have the option to liquidate their investments, they don’t necessarily use that option very often. So if you pool a lot of savers together, and a lot of borrowers, you can make a sort of simulated asset that is both liquid and not volatile, even though it’s made out of assets that are actually fairly illiquid and/or volatile. This has real benefits to savers, who get a better return on their money than they would by lending it out short term or leaving it in a stack of bills; and to investors, who get lower rates than they would if they had to persuade an individual to give them a 30-year mortgage.

This is fundamentally what banks do, and in a way, pensions do it too. While a pension is not as liquid as a savings account, it guarantees low volatility and regular withdrawals, while investing the money in assets that have variable returns and are potentially hard to sell without taking a loss. Because at any given time, people are going to be accessing only a small amount of the overall assets in the funds, they can do this without going broke.

That is, most of the time. The problem is that, like some sort of arcane magic, the simulation works only as long as everyone believes in it. If people stop depositing money in the system, and start taking it out instead, the underlying volatility and illiquidity reassert themselves, and you get a run. The paradox of a bank run is that it doesn’t matter whether the bank’s assets are basically healthy or not; if enough people try to leave, the bank will be insolvent, even if its underlying assets are throwing off more than enough cash to keep operating normally.

There are ways to mitigate this problem, but it never really goes away. The government can guarantee the assets, as we do with deposit insurance — but then people start worrying about the government’s soundness, as we saw in many places in Europe during the eurozone crisis. (The Dallas government has just had its bond rating downgraded by Standard & Poor’s, thanks in part to worries about the police pension fund.) You can stop people from making withdrawals (as many banks did during the Great Depression), but then your liquid, low-volatility asset has turned into something riskier, and people will demand higher returns to put their money in it, which means the underlying financial model no longer works the way it did.

In other words, as I’ve written before, risk can neither be created nor destroyed, only transferred or transformed. And the process of transforming or transferring risk creates its own, new risks: that when the illusion cracks, people will start rushing for the exits, leaving the slowest and most credulous holding an empty bag.

McArdle is a Bloomberg View columnist.


Unlimited Digital AccessOnly 25¢for 5 months