On Risk-Based Liquidity and Systemic Risk

I can’t help but think after the financial crisis that we have drawn some wrong conclusions about systemic risk. Systemic risk is when the financial system as a whole threatens to fail, such that short-term obligations can’t be paid out in full.  It is not a situation where only big entities fail — the critical factor is whether it creates a run on liquidity across the system as a whole.

Why does a bank fail?  It can’t pay in full when there was a demand for liquidity in the short run.  Typically, there is an asset-liability mismatch, with a lot of payments payable now, and assets that cannot be easily liquidated for what their stated value reported to the regulators.

Imagine the largest bank failing, and no one else.  Yes, it would be a mess for the FDIC to clean up, but it could be done.   Stockholders and preferred stockholders get wiped out. Bondholders, junior bondholders, and large depositors take a haircut.  Future deposit insurance premiums might have to rise, but there would be enough time to do that, with banks adjusting their prices so that they could afford it.

But banks don’t fail one at a time, except perhaps in good times with a really incompetently managed bank.  Why do some banks tend to fail at the same time?

  • They own many of the same debt securities, or same types of loans where the underlying asset values are falling.
  • They own securities of other banks, or other deposit-taking institutions.
  • Generalized panic.

What can stop a bank from failing?  Adequate short-term cash flow from assets.  Why don’t banks make sure that they always have more cash coming in than going out?  That would be a lower profitability way of running a bank.  It is almost always more profitable to borrow short and lend long, and make money on the natural term spread that exists — but that creates the very conditions that makes some banks run out of liquidity in a panic.

You will hear the banks say, “We are solvent, we just aren’t liquid.” That statement is always hogwash.  That means that the bank did not adequately plan to have enough liquidity under all circumstances.

Thus, planning to avoid systemic risk across an economy as a whole should focus on looking for the entities that make a lot of promises where payment can be demanded in the short run with no adjustments for market conditions versus assets available to make payments.  Typically, that means banks and things like banks that take deposits, including money market funds.  What does it not include?

  • Life insurers, unless they write a lot of unusual annuities that can get called for immediate payment, as happened to General American and ARM Financial in 1999.  The liability structure of life insurance companies is so long that there can never be a run on the bank.  That doesn’t mean they can’t go insolvent, but it does mean they won’t be part of a systemic panic.
  • Property & Casualty and Health insurers do not have liabilities that can run from them.  They can write bad business and lose money in the short-run, but that doesn’t lead to systemic panic.
  • Investment companies do not have liabilities that can run from them, aside from money-market funds.  Since the liabilities are denominated in the same terms as the assets managed, there can’t be a “run on the bank.”  Even if assets are illiquid, the rules for valuing illiquid assets for liquidation are flexible enough that an investment firm can lower the net asset value of the payouts, while liquidating other assets in the short run.
  • Even any large corporation that has financed itself with too much short-term debt is not a threat to systemic panic.  The failure would be unique when it could not roll over its debts.  Further, it would take some effort to actually do that, because the rating agencies and lenders would have to allow a non-financial firm to take obvious risks that non-financial firms don’t take.

What might it include?

  • Money market funds are different because of the potential to “break the buck.”
  • Any financial institution that relies on a repurchase [repo] market for financing is subject to systemic risk because of the borrow short to finance a long-dated asset mismatch inherent in the market.
  • Watch any entity that has to be able to post additional margin in order maintain leveraged asset finance.

How then to Avoid Systemic Risk?

  • Regulate banks, money market funds and other depositary financials tightly.
  • Don’t let them invest in one another.
  • Make sure that they have more than enough liquid assets to meet any conceivable liquidity withdrawal scenario.
  • Regulate repurchase markets tightly.
  • Raise the amount of money that has to be deposited for margin agreements, until those are no longer a threat.
  • Perhaps break up banks by ending interstate branching.  State regulation is good regulation.

But aside from that, there is nothing to do.  There are no systemic risks from investment companies or those that manage them, because there can’t be a self-reinforcing “run on the bank.”  Insurance companies are similar, and their solvency is regulated far better than any bank.

Thus, there shouldn’t be any lists of systematically important financial institutions that contain investment managers or insurance companies.  Bigness is not enough to create a systemic threat.  Even GE Capital could have failed, and it would not have had significant effects on the solvency of other financials.

I think it is incumbent on those that would call such enterprises systemically important to show one historical example of where such enterprises ever played a significant role in a financial crisis like the ones that happened in the 1870s, 1900s, 1930s, or 2000s.  They won’t be able to do it, and it should tell them that they are wasting effort, and should focus on the short-tailed liabilities of financial companies.