The Pass-Through Test: From Withdrawal to Failure

The subject now is not the definition of a run or the prediction of one variable at a time, but the connection among trigger, bank condition, institutional response, and outcome. A useful judgment about a reported run should answer four questions in that order: What caused the withdrawals? How strong was the bank before they began? What could the bank and its surrounding institutions do? What actually happened to the bank and the economy around it? This pass-through test turns a headline into a conditional assessment without pretending that the evidence can reveal every depositor’s motive.

The same run can carry different information

A withdrawal wave is an observation, not yet an explanation. It tells us that depositors wanted to change their position, but not whether they were responding to losses at the bank, fear about another institution, a broad economic signal, or confusion. The first step is therefore to classify the trigger without treating the classification as certain.

If the trigger is a false rumor or a run at another bank, it may primarily be noise or contagion from the depositor’s point of view. If it follows a rise in nonfinancial business failures, it may carry information about borrowers and the value of bank assets. If it follows weak stock-market returns or a broader banking panic, it may combine both: depositors are reacting to public news while the same economic deterioration may be weakening banks in the background.

This distinction matters because the historical evidence assigns different failure implications to different signals. A run following another local bank’s run was associated with a 14-percentage-point lower pass-through to failure than other runs. A one-percentage-point rise in the state business-failure rate—roughly a one-standard-deviation increase—was associated with about a seven-percentage-point increase in pass-through. The first pattern is consistent with healthy banks being pulled into a fearful environment. The second is consistent with a deteriorating economy damaging bank assets.

Neither estimate is a universal conversion rate. The signal variables are proxies, and a common shock can produce both business failures and bank withdrawals. But the contrast tells us what a good diagnosis must ask: is the adverse signal mainly changing beliefs about the target bank, or is it also changing the target bank’s ability to repay?

The bank’s starting condition is the central filter

The second question is what the bank looked like before the run. The historical study uses capitalization, leverage, liquid assets, deposits relative to assets, noncore funding, and related balance-sheet measures as proxies for fundamentals. These indicators do not reveal every hidden loss, but they describe how much room the bank has to absorb pressure and how exposed it may be to a loss of funding.

That starting condition changes the meaning of the same withdrawal wave. A strong bank may experience a run because depositors cannot distinguish it from a troubled neighbor. A weak bank may experience a run because the public signal has finally made its existing problems salient. In both cases, depositors withdraw. In only one case does the bank begin with little margin for error.

The historical estimates quantify the difference.