The Run Doesn't Need a Bad Bank
In one class during my MBA at Melbourne Business School, we watched a scene from It's a Wonderful Life (1946). Most people remember the film for its ending. The scene I still think about is a bank run, and it's one of the clearest pictures I know of confidence failing.
George Bailey runs the Bailey Building and Loan, a small community lender in Bedford Falls. A missing deposit and a rumour that the bank is failing bring a crowd to the lobby, all asking for their savings. George's problem isn't that the loans are bad. Most of the money is tied up in mortgages to people in town, and the cash on hand can't cover everyone at once.
George tries to explain. The money isn't in a vault; it's in the houses of people who borrowed from the bank, and those loans can't be called in a day. The crowd doesn't believe him until neighbours start to change their minds. Some leave their savings where they are. Others hand over cash to cover the withdrawals. The run stops, not because the loans have suddenly become liquid, but because enough people decide to wait.
What struck me in class was that the bank was never insolvent. Its problem was the gap between what it owned and what it could pay today. The only thing that closed that gap was a community choosing to keep its money where it was. There was no lender of last resort. The buffer was confidence.
Private credit faces a version of the same problem. A modern fund's buffer is built in advance: cash reserves, undrawn facilities and gate terms agreed before anyone asks. The question is whether investors can see those buffers before the crowd forms.
The run doesn't need a bad bank
A sound institution can fail on a bad day with the same loans, staff and capital it had the week before. What it doesn't have is enough cash for everyone who wants their money at once. Economists Douglas Diamond and Philip Dybvig formalised this in 1983. If you believe others will withdraw, the institution may fail, so you withdraw first. The run is self-fulfilling. Nothing about the loans has changed, only the belief about what everyone else will do.
Private credit runs on the same structure. Many funds hold five-year loans to mid-market borrowers. Those loans can't be sold in a day without a discount, yet many funds offer monthly or quarterly liquidity. That works while investors are calm and don't ask at the same time.
Two problems that look the same from outside
A credit problem is when borrowers stop paying and property values fall far enough to wipe out the buffer, so losses land on investors. A confidence problem is when the loans keep performing, but too many investors want out at once and the manager has to sell or pause.
In a credit problem, the buffers are thicker than most people assume. A property-backed loan usually has the borrower's equity beneath it, so values have to fall a long way, and the borrower has to default, before the lender loses principal. Mortgage securities add further layers of protection, and losses have to work through several of them before the most senior investors are affected. That is why confidence often moves before credit does.
Confidence moves first
A run rarely starts with a defaulted loan. It starts with a headline about a fund restricting withdrawals, a peer's returns softening, a revised valuation, or an adviser pausing new allocations to "review the position." None of these says the loans are bad. Each one changes what investors believe the others will do.
Listed vehicles show how quickly sentiment can reprice a portfolio. Their share prices can fall well below the value of the loans they hold, even when those loans are still performing. The price moves on mood; the loans don't necessarily move with it.
Unlisted funds behave differently. Their unit prices are set from loan values rather than daily trading, so they look stable. But stable unit prices don't mean credit risk is absent. The listed market reprices in days. An unlisted fund reprices when someone asks for their money, which is the worst time to find out.
Why leaving early is rational
When redemptions arrive, the manager has to raise cash. Loans are sold through bilateral negotiation, often at a discount. If the manager sells the better loans first, the investors who stay hold a weaker book, which gives them a reason to leave too. The first to redeem gets the best price, so waiting is the costly choice.
Gates exist to stop this. They protect the investors who remain. But from the outside, a gate looks like distress, even when every loan is performing. An investor who was comfortable last week reads it as evidence of trouble. The mechanism designed to prevent a run becomes a reason to run.
Walter Bagehot made the point about banks in Lombard Street (1873): "Every banker knows that if he has to prove that he is worthy of credit, however good may be his security, he will not get it." Once investors start asking whether a fund can pay, the quality of the portfolio stops being the main question.
Where confidence turns into credit
The argument has a limit. Confidence shocks hit sound funds hardest, but some structures are fragile on their own terms. Land and construction loans are often valued on an "as-if complete" basis and can fall faster than established housing. Capitalising interest adds to the loan balance over time, so the loan-to-value ratio rises if the property isn't sold. A fund with few loans has less room to absorb one bad one. These are the places where a confidence shock can turn into a genuine credit problem.
Confidence can break sound funds, and it can also expose weak ones. The job is telling them apart before the gate goes up.
Bank debt: safe, not risk-free
Major banks are very unlikely to default, but their debt still trades. Subordinated bank debt and hybrids can fall in value when markets are stressed, as they did in the March 2020 sell-off, even when the issuer is sound. A bond can be repaid in full and still lose value on the way.
What to check before committing
• Does the liquidity promise match the asset life? Monthly redemptions against five-year loans need a clear source of cash behind them.
• How much of the book is capitalising interest or in construction? These can erode the buffer quietly while the unit price stays flat.
• Who else holds the fund? A concentrated investor base can produce sudden, large requests.
• How often are assets valued, and how? Infrequent or model-based marks can delay recognition of a problem.
• What are the gate terms? Read them now. Learning them during a run is the worst time.
Most Australian private credit managers may well be sound, and that's the point. Soundness is necessary but not sufficient. What protects a fund in a run is the structure it chose before the run began, and the confidence of investors who understand that structure. A bank can be solvent and still close on a Tuesday afternoon, because the question was never whether its loans were good. It was whether everyone else would wait.
Comments