Publicly Traded Versus Privately Held Commercial Banks: Sensitivity to Growth Opportunities

In 1984, 2.7% of commercial banks were publicly traded and these banks held 68% of total industry assets; by 2012, these figures increased to 5.9% and 88%, respectively.

There were 5288 commercial banks or bank holding companies in the U.S. in 2012, of which 312 were publicly traded. In comparison, there were approximately 1.2 million large business enterprises in the U.S. in 2012, of which about 4100 were publicly traded. Data from the Federal Reserve, the Federal Deposit Insurance Corporation, the U.S. Census Bureau, and the World Bank.

From the call reports dataset, we exclude all non-commercial bank financial institutions (RSSD9048 ≠ 200). From the Y9-C dataset, we exclude any holding company that does not have a commercial bank subsidiary. If a holding company contains multiple levels of affiliated banks and bank holding companies, we aggregate quarterly financial data to the top holder level. We exclude all observations with missing total assets, total loans, or equity information, and we exclude banking companies at which foreign ownership equals or exceeds 25%.

Banks headquartered in rural areas (i.e., cities or towns never included in a metropolitan statistical area, or MSA, during our sample period) face growth opportunities that are both qualitatively and quantitatively different from those faced by banks in urbanized areas. Banks less than 10 years old exhibit periods of unsustainable asset growth as they lever-up their equity capital for the first time (DeYoung and Hasan 1998). Subchapter S corporations face severe, self-imposed limits on equity capital formation that will limit their rates of growth. U.S. banks were prohibited from organizing as S corporations prior to 1997. S corporation status shows up initially in the call reports in the third quarter of 1997 and in the Y-9C reports in the first quarter of 1998. We classify any bank identified as an S corporation as of these dates as an S corporation for the entire 1984–1998 period.

We use the earliest of the following three dates to determine when a bank became a public corporation: The IPO date reported in the COMPUSTAT database, the first date with non-missing stock price in the CRSP US Stock Database, and the start date of the RSSD ID-PERMCO link in the New York Fed CRSP-FRB link file. A handful of banks delisted during our sample period and we use the latest of the following three dates to identify these events: The delisting date reported in the CRSP US Stock Database, the last date with non-missing stock price in the CRSP US Stock Database, and the end date of the RSSD ID-PERMCO link in the New York Fed CRSP-FRB link file.

The first of these three results is calculated as 0.585*2.93 = 1.71, where 2.93 is the standard deviation of OPPS from Table 1. Unless indicated, this approach is used to calculate the economic magnitudes of other coefficient estimates reported in the text.

The geographic locations of the four cities (New York, San Francisco, Boston, Chicago) used to calculate lnDISTANCE were established based on transportation routes and economic activity at least a century prior to our 1986–2012 data period. While lnDISTANCE contains little time series variation (headquarters cities occasionally change after an acquisition), neither does the variable PUBLIC for which it is instrumenting. The value of PUBLIC changes, always from 0 to 1, just 647 times for the 106,181 bank-quarter observations in our full data sample.

Compounded annual GDP growth across the 50 states averaged 5.42% with a standard deviation of 0.74% between 1986 and 2012. California (5.5%), Massachusetts (5.0%), New York (4.9%) and Illinois (4.8%) were the 22nd, 32nd, 35th and 39th fasting growing U.S. states during this time period. Data from the Bureau of Economic Analysis website and authors’ calculations.