Notes from
Douglas Diamond's Talk at Yale Economics Reunion

I began by recalling some of the many at Yale who changed my view of economics. I then talked about the Asian Debt Crisis.

• The severity and unpredictability of the Asian Debt Crisis has been interpreted as a financial panic.

• Velasco and Chiang
• Radelet and Sachs
• Krugman

• Changes in fundamentals, such as poor macro policies, are too small to explain its timing and severity.
• The basis for the panic is the liquidity gap between the illiquid assets of Asian corporate and bank borrowers and their short-term liabilities.

• If too many lenders demand their money back early, liquidation losses wipe out the long-term value of continuing to finance the borrowers. This creates a self-fulfilling prophecy of panic.

• The model used in these studies is Diamond-Dybvig [1983].
• In Diamond-Dybvig, the illiquid assets represent the aggregation of firms and banks: they are physically illiquid real assets. A run then requires inefficient physical liquidation
• Some policy conclusions they get and some criticism of the basic story depend on details of this reduced-form shortcut in the model.
• A more detailed model that accounts for the role of banks, and which has nearly the same reduced form, has different policy conclusions, and fits the facts better.
• The detailed model puts more emphasis on banks and on ex-ante benefits of financial fragility (the threat of a panic).

Questionable Policy Implications

• Ex-ante: Polices to discourage short-term borrowing
• Ex-post: only option is suspension of convertability = capital exit controls
• In series of papers with Raghu Rajan, we develop a model of illiquid financial assets due to bank specific loan collection skills

1. Liquidity risk, liquidity creation and financial fragility: A theory of banking ( w.p. 1997)

and

2. A theory of Bank Capital (w.p. 1999)

• An incumbent bank can collect more from a borrower than any other person can.
• Market value of loans is less than full value because the bank can't commit to collect the loan for anyone else in the future
• For the bank itself to raise more than the low market value of its loans, it mustcommit itself to collect them without renegotiating
• The way for the bank to borrow more than market value of its loans and create liquidity, is for the bank to issue some demand deposits as a commitment device

• If bank will not (or can not) pay full value to depositors, a run occurs
• A run drives the banker's rent to zero.
• Deposits are a commitment device because they create a collective action problem: a run occurs even if not in the collective interest of depositors, due to first come -- first served.
• The threat of a run allows banks to raise more from outside depositors who have little ability to collect loans on their own.
• Thus there is a trade-off between liquidity creation and stability

• More deposits imply more liquidity creation, but runs might occur after negative shocks (even without a panic)
• More deposits make panics more likely
• Requiring firms and bank to use long-term funding will greatly reduce firms access to credit

In developing economies with weak investor protection, local institutions such as banks have much greater ability to enforce repayment from borrowers

• Local banks have better information and their local clout gives them stronger weapons of persuasion
• Foreign investors (including foreign banks) who want to diversify internationally therefore have to rely initially on local banks to intermediate their investments.
• Foreign deposits in local banks must be short-term, to commit the local bank to collect on their behalf
• If local firms borrow directly, it also must be short-term, because foreign outside investors have little bargaining power other than that from pulling out funds.

Effects on financial structure after a run

• Once a run begins, the failure of banking institutions takes specific lending skills out of the system and prevents the system from recovering easily from the crisis.
• Foreign investors will not come in and buy cheap financial assets after the devaluation. With the local banks' collection expertise taken out of the system, these financial assets are illiquid and no longer credit worthy.
• DFI after a run??
• Rapid closure of failing banks may be the wrong thing-- recapitalization on a large scale may be better (but will be unprofitable)
• Banks are at the center of the problem. Economies where the specific collection skills of banks are needed must have fragile capital structures to attack outside capital
• Foreign borrowing that is not through a local bank in such an economy must have a bank-like short-term debt structure.