Why do banks acquire non-banks?

For example, SouthTrust Bank acquired Carolina Financial Corp. (investment securities) and Citizens Fidelity Bank acquired Cowger & Miller Mortgage (a mortgage company) in early 1992.

As we revise this version of manuscript as of 2009, we are witnessing continuing turbulences in financial sector, where fire sale types of acquisitions are ubiquitous. JP Morgan Chase bought Bear Sterns at a bargaining price after Bearn Sterns went down under the wind of sub-prime mortgages crisis, started in 2007. Lehman Brothers Holdings were purchased piecemeal by Barclays Capital and Nomura Securities. On 22 September 2008, Barclays Capital completed its acquisition of Lehman Brothers’ North American Investment Banking and Capital Markets businesses. As a part of the transaction, Lehman Brothers indices have become part of Barclays Capital. Japan’s Nomura announced they will buy most of Lehman Brothers’ business in Europe in an attempt to catapult itself into the major league of Western investment banks. The federal government bailed out Fannie Mae and Freddie Mae as well as AIG with cash infusion of $85 billion. JP Morgan acquired major assets of Washington Mutual after WaMu sets a record as the largest bank failure in US history. Wells Fargo acquired the troubled Wachovia.

See Berger et al. (1998); Berger et al. (1999); and Knapp, Gart, and Becher (2005) for surveys of literature on bank consolidation activities.

See Puri (1999); Bhargava and Fraser (1998); Cornett et al. (2002); Gande et al. (1999) for studies of Section 20 affiliates.

Merger strategy among banks to reduce costs is known as mergers strategy to gain the economies of scale. Merger strategy to increase revenue from activities other than traditional banking is known as mergers to explore the economies of scope. Therefore, the terms revenue-enhancement and cost savings mergers refer to the economies of scope and scale.

Ang et al. (2002) show that the bank’s CEO is not the only top executive that is affected by bank performance. We add all top-five executives’ salaries, bonuses, and stock options (incentives) to compute their annual pay. We use only observations that have the pay of at least five top executives. When firms report more than five top executives, we pick the five executives who have the highest total pay. The univariate test between the full sample of 1,009 and a subsample of 598 is reported in Table 7 (and available upon request).

We exclude foreign banks (SIC 6082) as targets and acquirers and bank holding company (SIC 6712) as targets.

We conduct a sensitivity analysis by excluding Citigroup and Chase Manhattan only and excluding the top ten of the non-bank acquisitions. We find that all our results in this study remain robust with these exclusions.

Gleason et al. (2006) also utilize the two-digits SIC codes to distinguish non-banks from banks.

The bank decision to acquire non-bank instead of other bank is conditional upon its decision to acquire or not to acquire. Therefore, we also conduct a conditional logistic regression estimation to test the robustness of our results. We find that the results from conditional logistic regression estimation are qualitatively the same as logistic regression reported on Table 3.

The inverse Mills ratio (sometimes also called the correction for “sample selection bias”) is used in regression analysis to take account of a possible endogeneity bias due to omitted variables that influence the choice variable. If a dependent variable is censored, i.e., not all observations have an observable outcome, it causes a concentration of observations at zero values. With a censored dependent variable, there is a violation of the Gauss-Markov assumption of a zero correlation between independent variables and the error term. Heckman (1979) proposed a two-stage estimation procedure using the inverse Mills ratio to take account of the endogeneity bias due to omitted variables. In the first stage, a regression for observing banks acquiring non-banks is modeled with a probit (logit) model. The estimated parameters from the probit (or logit) are used to calculate the inverse Mills ratio, which represents the unobservable factors that affect banks’ decisions to acquire non-banks beyond the bidders financials and deal characteristics, and is then included as an additional explanatory variable in the second-stage OLS estimation.

The estimated slope coefficients for year dummies are not reported in column (3) to conserve space. They are available from the authors upon request.

The Execucomp database covers top executives compensation for mostly large firms since 1992. Due to the data limitation from Execucomp, our sample is significantly reduced from 1,009 to 517 completed acquisitions. We conduct a sensitivity test by using only the CEOs’ salary, bonus, and incentive compensation and find that the results are qualitatively similar to the results of the compensation of the total top-five executives. We also conducted a univariate statistics between the full sample and the subsample that contains the executives compensation. The results of univariate analysis are presented in Table 7.