The dominant form of short-term funding was traditionally bank deposits. Deposits are essentially loans made to a bank by its depositors. Deposits are legally short-term, but practically rest in the hands of banks for substantial periods. Short-term funding becomes problematic, of course, when depositors systematically demand repayment: this is the old-style bank run. Post-Depression era deposit insurance has largely eliminated bank runs, at least in the United States, and so the ordinary insured bank deposit is (from the perspective of the bank) a trusty source of short-term funding.
The bank deposit no longer forms the core of short-term funding for many banks (particularly larger banks) -- and other systematically important financial institutions (hedge funds, money-market funds, broker-dealers) lack the ability to take insured deposits. The mix of short-term funding has changed -- and with this change, there are new vulnerabilities.
Tarullo points out that the liquidity freeze that occurred during the Crisis involved the unexpected drying up of short-term funding. Short-term investors had been willing to lend against mortgage pool securities up until the onset of the Crisis -- and then had a rapid change of heart. These breakdowns of investor willingness were the modern equivalent of the bank run -- in that expected funds suddenly became unavailable due to panic.
Contemporary short-term funding is often wholesale -- that is, it is provided in substantial packets by a few institutional players. It chiefly takes the form of securities transactions, such as repos, reverse repos and other forms of securities lending and borrowing, and so depends on the value and liquidity of the pledged securities. Tarullo observes that the basic structure of short-term wholesale funding has not changed in the years following the Crisis -- and the continuing risk of short-term funding runs has not been directly addressed in current reforms.
Let's be clear as to what happens when a short-term funding crisis occurs. First, a bank faces the obligation to make substantial disbursements to the fleeing investors. Whether it involves unanticipated account withdrawals or the refusal to roll-over other forms of funding, a challenged bank will have to expend liquid funds to satisfy the demands of the stampeding short-term fund providers. And, to the extent these demands exceed the then current amount of liquid funds, the stressed bank will have to create additional free funds either by borrowing (if possible) or by selling assets. Once a troubled bank begins selling assets, the downward spiral proceeds: assets dumped on the market erode asset prices, inflicting further damage to the bank's balance sheet. Moreover, a drought of short-term funding immediately shrinks a bank, affecting the amount of credit an institution can create. This has clear macroeconomic implications -- as the shrinkage of credit is directly transmitted to the real economy.
Tarullo notes that these risks can be managed by either capital or liquidity tools. A well-capitalized bank is less likely to face a short-term funding crisis in the first place, and if struck by contagion, can better withstand a certain degree of defunding. A bank with greater liquidity can meet unanticipated demands and live for a better day. But it seems implicit in Tarullo's assessment that the mix of minimum capital and liquidity requirements imposed by Basel III are not adequate to stem potential short-term wholesale funding runs. Far too many systematically important firms fall outside the coverage of Basel III -- and the Crisis demonstrated that a funding run against a shadow bank (such as a mutual fund) can be as destabilizing as a run on a regulated bank. Simply requiring maturity-matching is not adequate, Tarullo argues -- as a funding crisis may develop due to contagion involving a particular kind of pledged asset. Tarullo strongly signals that the Fed may use the authority it acquired in §165 of Dodd-Frank to increase the capital and liquidity requirements of large banking organizations to address short-term funding risks.