Capital Market Integration and Branch Banking

[B]ranch networks allow lenders to mitigate contracting frictions, and play an important role in financial integration.

Bank branch networks help integrate U.S. lending markets in segments where arm's - length financing is costly or even infeasible, according to Exporting Liquidity: Branch Banking and Financial Integration (NBER Working Paper No. 19403) by Erik Gilje, Elena Loutskina, and Philip Strahan. By studying deposit windfalls from the purchase of drilling rights following the discovery of oil and gas shale resources in some areas, the authors demonstrate that banks benefiting from the windfalls extend their origination of new loans into outlying, non-boom areas. This increase in loans only happens in counties where banks have branches, and it is most evident in the market for mortgages that are hard to securitize. The estimates suggest that retained mortgages increase somewhat more than 2 percent for every 1 percent increase in deposits. The authors do not find any effect of deposits on mortgages that are sold off to national capital markets.

The authors explain that "[t]hese results suggest that banks export deposits to non-booming markets and increase credit supply rather than merely retain more loans. ... [B]ranch networks allow lenders to mitigate contracting frictions, and play an important role in financial integration."

National capital markets have transformed the banking system over the past three decades. In 1980, only 12 percent of all home loans were securitized. By 2011, that share was up to 52 percent. This rise could have made branch banking less important for integrating lending markets. Yet the number of branches per bank also increased, from five per bank in 1990 to 14 per bank in 2011. These similar trends suggest that securitization and branch networks act as complementary, rather than substitute, ways to integrate local lending markets.

The authors use the unexpected, large wealth windfalls stemming from shale booms to test this thesis. They study the 327 banks that received deposits in the seven states with major shale discoveries during the 2003-10 period: Arkansas, Louisiana, North Dakota, Oklahoma, Pennsylvania, Texas, and West Virginia. Average mortgage lending grew 11.7 percent per year for the banks that benefited from shale discoveries over this period, compared with 11.2 percent for banks that did not participate in the shale booms.

By discarding the lending activity in the 124 counties that experienced booms and focusing instead on the mortgage loans that the 327 banks made in the 515 counties that did not participate in shale booms, the study avoids potential distortions to credit demand that arise directly from the shale booms. The authors find that the banks exposed to these booms increased lending in non-boom counties more than non-exposed banks, but only when the exposed banks had branches located near borrowers. As a result, shale-boom banks saw retained mortgage growth average 9.1 percent per year, substantially greater than the 7.7 percent per year for non-boom banks. New lending expanded the most in home equity lines of credit, the category of mortgages that is hardest to securitize, and the least in mortgage re-financings, the easiest to securitize.

The authors conclude that by allowing capital to flow more easily across local markets, deregulation of bank branching fostered a denser branch network that improved capital mobility and thus investment allocation efficiency.

--Laurent Belsie