Working Paper: NBER ID: w23113
Authors: Matthew S. Jaremski
Abstract: Operating in individual cities, U.S. clearinghouses were the closest thing to a central bank before 1914, but they only assisted banks that chose to join the association. Using an annual bank-level database for seven states between 1880 and 1910, this paper shows that after the entry of a clearinghouse member banks were less likely and non-member banks in the same city were more likely to close. The results are driven by the fact that the presence of clearinghouses led all banks to become more exposed to systemic liquidity risk, yet only provided liquidity to member banks during panics.
Keywords: No keywords provided
JEL Codes: G21; G32; N21
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Creation of a clearinghouse (D40) | Decreased probability of closure for member banks (G21) |
| Creation of a clearinghouse (D40) | Increased probability of closure for nonmember banks (G21) |
| Clearinghouse membership (G23) | Protective effect due to access to emergency liquidity (E44) |
| Clearinghouse entry (Y20) | Increased systemic liquidity risk for nonmember banks (F65) |
| Differential closure risk between member and nonmember banks (G21) | Interbank liquidity risk created by clearinghouse network (F65) |