The measurement of integration is of considerable policy relevance. For example, the European Central Bank mission statement reads: "We in the Eurosystem have as our primary objective the maintenance of price stability for the common good. Acting also as a leading financial authority, we aim to safeguard financial stability and promote European financial integration" (italics added). Most observers have concluded that while money markets and government bond markets are rapidly integrating following the introduction of the common currency in the euro area, there is little evidence that a similar integration process is taking place for retail banking. Data on cross-border retail bank flows, cross-border bank mergers and the law of one price reveal no evidence of integration in retail banking. This paper shows that the previous tests of bank integration are weak, because they are not based on an equilibrium concept. They do not consider the possibility that due to violations of the Modligliani-Miller assumptions, demand for financial services may differ across countries. For example, if corporate income taxes differ, based on the trade-off theory of capital structure, firms would have different target capital ratios. This in turn implies that they demand different amounts of debt and exhibit different risk characteristics. Hence, we would observe a violation of the law of one price even if markets are perfectly integrated. Hence, this paper argues that the previous tests are neither necessary nor sufficient statistics for bank integration. The paper proposes a new test of integration based on convergence in banks’ profitability. Banks profitability should converge in integrated markets as contestability or entry will result in high profits being competed away. Further, a functioning market for corporate control will ensure that weak banks will either disappear or be taken over by stronger banks. Hence, the new test identifies the role of an active market for corporate control and of competition in banking integration. Unlike previous tests of integration, therefore, the test proposed in this paper emphasises the link between integration and efficiency: integration is only to the extent desirable as it results in a more efficient provision of banking services. A monopolist, while obviously "integrated", would not provide banking services efficiently. The results, using RoA or RoE as a measure of bank profitability, suggest that European listed banks profitability appears to converge to a common level. There is weak evidence that competition eliminates high profits for these banks, and underperforming banks tend to show improved profitability. Hence, both contestability and the market for corporate control appear to be operational for listed European banks. Unlisted European banks, savings banks and cooperative banks differ markedly. Their profits show no tendency to revert to a common target rate of profitability. Overall, the banking market in Europe appears far from being integrated. This is true not only across Europe, but also within individual European countries: Even within countries, listed banks do not appear to be integrated with unlisted commercial banks, savings banks and cooperative banks. We also estimated the model for U.S. banks as a benchmark. In the U.S. both listed and unlisted commercial banks profits converge to the same target, and high profit banks see their profits driven down quickly. However, there is no evidence of a market for corporate control driving out weak unlisted banks even in the U.S.. Still, using our measure of integration, the U.S. banking market appears significantly more integrated than the banking market in the E.U.. The evidence presented in the paper is consistent with earlier evidence on cross-border bank mergers and contagion among listed banks, both of which appears to have increased since the introduction of the common currency. However, unlisted commercial banks, savings and cooperative banks constitute about 50% of total assets of the banking systems of the major European countries studied here and the retail market share may be even larger. The governance of these banks is not subject to the same mechanisms as the governance of listed banks. The evidence shows that they neither respond to competitive pressures as much as listed commercial banks, nor do these banks face pressure to remedy underperformance through a threat of takeover. These rigidities remain in place, and as far as we can see, would be unaffected by the introduction of the common currency.

Gropp, Reint and Anil Kashyap (2008), A New Metric for Banking Integration in Europe, ZEW Discussion Paper No. 08-102, Mannheim. Download