The bank merger environment & its effect on commercial & industrial real estate markets.

The advent of a substantial number of intrastate and interstate bank mergers and acquisitions has led to a large volume of research that has questioned the potential economic and political implications of these events (1,3,7,8,9,10,11,13,19,24). The vast majority of this research has focused on two issues: (1) the potential anti-competitive effects; or (2) the potential cost differentials that are likely to exist in a post-event environment (12,15,16,17,18,20,21,25,26,27). Most of this research has tested for the likelihood of significant differences in the level of interest rates paid on bank deposits, or the availability of total loan-able funds in a banking market before and after a merger or acquisition event. In general, this research has suggested that the likelihood of differentials in interest rates on loans or deposits would indicate a competitive advantage for a merger partner relative to its local counterparts. Any sustained differential would therefore suggest that bank mergers or acquisitions aff ect the competitiveness of the local post-event bank environment.

All of the studies have concluded that no "local effects" are evident in the data and therefore mergers and acquisitions do not create any anti-competitive elements. (1) Furthermore, it is argued that because banking products are generally homogenous and substitute sources of funding are readily available, future mergers or acquisitions are unlikely to create an anti-competitive environment (4).

However, when the focus of the research is shifted from the deposit-side of the balance sheet to the asset-side of the balance sheet, and the post-event effects in the discrete lending environment are tested, i.e. the commercial and industrial real estate markets, the agricultural production lending market, etc., rather than the availability of total loanable funds, the findings of "no local effects" may no longer be valid. Furthermore, the assumptions of homogeneity and substitutability do not appear to be supported since, for example, the risk, earnings, and maturities, etc., of a residential real estate loan are not comparable to a loan to a small business for an expansion.

The data and analysis in this study demonstrate that "local effects" do exist when the discrete lending categories are analyzed. The results of the analysis demonstrate many instances where significant concentrations and market dominance in post-acquisition environments exist (5,14,22,23).

In addition, Besanko points out that the lack of monopoly pricing elements, (in this case higher interest rates charged), is not necessarily indicative of the level of competition in the market. Instead, the existence of a lack of inter-firm competition may be evident in the operational characteristics of the market (4). In the market for commercial and industrial real estate lending, the lack of competition can lead to a situation where very few banks are setting virtually all of the policies and standards for a very large group of borrowers. For example, the parent organization's loan committee would likely set credit analysis procedures, credit scoring requirements, collateral requirements, repayment schedules, etc., for all operating units. Since extensive intrastate merger activity could result in a situation where a substantial number of previously independent banks are now governed by a single, more standardized lending policy, the potential is increased for commercial and industrial real estate borrow ers to be penalized or even excluded.

PURPOSE OF THE STUDY

The purpose of the study was to demonstrate the extent to which a discrete category of lending; i.e., commercial and industrial real estate lending, can become very concentrated in a very few banks in local banking markets as a result of inter- and intrastate bank mergers and acquisitions. The study results show that in some states these concentrations are so significant in the post-event environment, that there is virtually no competition among banks in the market for commercial and industrial real estate lending.

FRAMEWORK OF THE STUDY

The study period 1982 to 1999 was chosen because it encompassed a vast number of bank mergers and acquisitions and is consistent with the 1982 Justice Department Merger Guidelines. These revised guidelines provided for a more lenient regulatory environment with respect to approval of merger and acquisition activity. In addition, this time period allows the use of the most complete FDIC and Federal Reserve Bank data relative to bank merger and acquisition activity including the year-end FDIC Call and Income Reports and the Federal Reserve Bank Holding Company Acquisition and Merger Data Report.

Specifically, the Department of Justice has for many years published formal guidelines that identify structural changes resulting from mergers that are likely to cause the department to challenge a merger. Since 1982, the department has based its merger guidelines on the Herfindahl-Hirschman Index of Concentration (HHI). This measure, which is also used by the bank regulatory agencies, is calculated by squaring the market share of each firm competing in a defined geographic banking market and then summing the squares. The HHI can range from zero in a market having an infinite number of firms to 10,000 in a market having just one firm (with 100 percent market share).

The HHI is a particularly useful tool for bank merger analysis because it accounts for the presence of every competitor in a market and provides a measure of the structural effect of a merger of any firms in a market. In addition, the squaring of the market shares gives greater weight to firms that have large market shares. This weighting of the largest competitors in a market is consistent with the economic theories that predict weak competition in markets in which a few competitors hold a large combined market share (14).

This study used all commercial banks in the 50 United States over the period 1982 to 1999. Each bank's total assets, total loans, total deposits, and total commercial and industrial real estate loans were obtained from the FDIC year-end Call and Income Report data (8). A Herfindahl-Hirschman Index number was calculated for each of these balance sheet variables on a state-by-state basis for each study year (14,22,23). The HHI therefore provides a summary measure of market concentration that reflects the proportion of the total assets, deposits, or loans, etc., accounted for by each firm serving the market (25). The HHI is calculated in the following manner:

C = [summation over (N/i=1)] [A.sup.2.sub.i]

Where [A.sup.2.sub.i] represents the percentage of the market-area deposits or assets controlled by the i'th bank in the market. For presentation purposes, C is divided by 10,000 in order to demonstrate the percentage of the market controlled by the largest banks. The number of equivalent firms is then calculated by dividing one by the percentage of the market controlled by the largest banks. (2) The Justice Department defines bank markets where C exceeds 1800 as a highly concentrated market (14,22). This translates to a decimal of .1800 as a highly concentrated market with a numbers-equivalent threshold of 5.556 banking units (23).

DATA & ANALYSIS

Table 1 presents the total number of dollars of bank loans classified as commercial and industrial real estate loans by year for selected states and U.S. totals. Tables 2 through 7 present the HHI and the numbers-equivalent calculations for six representative states. (3) Each table, by state, contains the variables: year, the number of banks as of year-end, the HHI for total assets, the HHI for total loans, the HHI for total deposits, the HHI for commercial and industrial real estate, and the numbers-equivalent for the number of banking units based on the HHI for commercial and industrial real estate loans.

In order to assess the degree of concentration in a post-merger market environment for commercial and industrial real estate lending, an analysis of the HHI for commercial and industrial real estate loans and the numbers-equivalent of units on a state-by-state basis provided the most insight. For example, Tables 2 and 3 depict the post-merger commercial and industrial real estate lending environment of two states, Pennsylvania and Texas, with very large commercial and industrial bases. Note that in Pennsylvania, the number of banks declined from 349 to 193 and in Texas from 1601 to 753 over the study period. In both of these states as the number of operating banking units has fallen, the numbers-equivalent columns, column 8, in both Tables 2 and 3, indicate that the number of active bank lending participants in the commercial and industrial real estate market has also fallen, indicating an increased pattern of concentration in both of these markets. Yet the HHI figures and the numbers-equivalent figures indic ate the commercial and industrial real estate environment remained relatively broad-based, and dispersed across a large number of banks with no Pennsylvania bank controlling more than 9 percent and no Texas bank controlling more than 5 percent of the commercial and industrial real estate market within the state.