The Italian Banking Structure in the 1990s: Testing the Multimarket Contact Hypothesis

Abstract

The multimarket contact hypothesis holds that more contacts between firms competing in the same markets may induce more collusion. This paper tests the hypothesis for the Italian banking market, analysing the behaviour of the largest Italian banks from 1990 to 1996. Market rivalry is gauged by changes in loan market shares and interest rates in each Italian province. We estimate the effects of increasing multimarket contacts, concentration indicators, banks' costs and loan demand on variations in market shares and interest rates. No support is found for the multimarket contact hypothesis. Geographical overlap in banking is positively correlated with changes in market shares, confirming the thesis of an overall increase in competition within the Italian banking system. Greater multimarket links also seem to correspond to lower lending rates. (J.E.L.:G21, C33, L40.)

Multimarket Contact and Profitability in Banking: Evidence from Italy

“…To our knowledge, the only empirical test of the linked oligopoly theory for the Italian banking industry is performed by De Bonis and Ferrando (2000), who estimate the effects of increasing multimarket contacts, concentration indicators, banks' costs and loan growth on variations in market shares and interest rates. They do not find any support for the multimarket contact theory; quite to contrary, geographical overlap in banking is positively correlated with changes in market shares and lower lending rates, confirming the hypothesis of an overall increase in competition within the Italian banking system in the last years.…”

“…Actually, a multimarket bank needs also to meet the same rival in more than one market. 19 For instance, Heggestad and Rhoades (1978) consider only the U.S. banks which are "dominant" according to various criteria (such as the position in the size ranking or the combined market shares), Whalen (1996) focuses on the top three banks in terms of deposits for each U.S. state, and De Bonis and Ferrando (2000) concentrate on the first 55 Italian banks. 20 Following the classification of the Central Bank of Italy, our macro-regions are: North-West (including Piemonte, Valle D'Aosta, Lombardia, Liguria); North-East (Veneto, Trentino Alto Adige, Friuli Venezia Giulia, Emilia Romagna); Center (Toscana, Umbria, Marche, Lazio); South (Abruzzi, Molise, Campania, Basilicata, Puglia, Calabria); Islands (Sicilia, Sardegna).…”

“…20 Following the classification of the Central Bank of Italy, our macro-regions are: North-West (including Piemonte, Valle D'Aosta, Lombardia, Liguria); North-East (Veneto, Trentino Alto Adige, Friuli Venezia Giulia, Emilia Romagna); Center (Toscana, Umbria, Marche, Lazio); South (Abruzzi, Molise, Campania, Basilicata, Puglia, Calabria); Islands (Sicilia, Sardegna). De Bonis and Ferrando (2000) use a similar criterion: particularly, because of data availability, they focus on the first 15 banks for each macro-area. 21 The values of the third quartile for total assets (in real values) and branches are 688.7 millions euro and 19 units, respectively.…”

Structure and performance in the Portuguese banking industry in the nineties

“…On the other hand, Neven and Roller (1999) have found significant increases of competition over time in the mortgage market, leading the authors to conclude that the behavior of banks in seven European countries (France, Denmark, Germany, Spain, UK, Belgium and Netherlands) has become less collusive over time. Cerasi et al (2001) claim that deregulation has increased the degree of competition within the European retail banking industry, and both De Bonis and Ferrando (2000) and De Bandt and Davis (2000) conclude that there has been an increase in competition within the Italian banking system. Nevertheless, Gual (1999) concludes that concentration will diminish with European market integration and enlargement, but one would expect consolidation only at the level of some domestic markets.…”

Bank Competition and Outreach: Evidence from Turkey

“…17 The theory of multimarket contact is suitable for the banking industry as banks usually meet one another in several geographical markets via subsidiaries or branches. At the bank level, we measure multimarket contact (MMC) by considering the number of geographical contacts between banks (Coccorese and Pellecchia 2009;De Bonis and Ferrando 2000). Let D ij be equal to 1 if bank i operates in province j, and 0 otherwise, for i = 1, ..., n, and j = 1, ..., m. We construct a symmetric (n × n) matrix A = (a ik ), where its generic element a ik = S m j=1 D ij D kj represents the number of markets in which bank i meets bank k, while the diagonal element a ii measures the number of markets serviced by bank i.…”