Internal organization of banks and cross-border transmission of shocks

The internal functions of global banks could be decisive factors in the transmission of shocks both across a country's regions and internationally. However, there is still little knowledge of how the internal organization of these financial conglomerates is connected with their cross-border lending decisions. A major obstacle for such an analysis is the lack of information about the degree to which a parent bank affects the decisions of its foreign subsidiaries. Few studies have focused on confidential information about the activities of the internal capital markets in banking conglomerates (see, for instance, Cetorelli and Goldberg (2012a) and Cetorelli and Goldberg (2012b)), but such information is usually either not available to the general scientific community, or available only for a small number of countries, such as the U.S., which makes the results difficult to apply elsewhere. In the proposed project, we introduce a new measure of bank integration, based on the organizational culture within a global bank, reflected by the strength of the language in its publicly available financial reports. After establishing the validity of this approach for our purposes, we will investigate which social and bank-specific characteristics determine the degree of integration within global banks and whether that degree of integration affects the transmission of solvency and liquidity shocks from parents to their subsidiaries.

We base our method on the General Inquirer Approach developed by Philip Stone and his collaborators (Stone et al. (1966)) at the Harvard Laboratory of Social Relations. The General Inquirer is a computer software that calculates the frequency of appearance of a predefined set of words in a given document. In particular, we use the "Power" category of the Lasswell value dictionary to gauge markers for the prevalence of a language of power, authority and control in 267 annual financial reports of 105 global banks for the years 1997, 2005 and 2012, totaling at 22.4 million words. Then, we calculate our measure of bank integration, the Power Ratio, as the ratio of the number of authority-related words to the total words in the particular document. Since we consider the language of authority to be an indicator of the intrinsic corporate culture within a bank, which is stable across time, we pool all documents for each bank to derive static measures of bank integration, arriving at a cross-section of 105 Power Ratio values. Subsequently, we analyze whether bank integration is determined by individual bank characteristics or by country-related social and economic factors. Our hypothesis is that the degree of centralization of the society from which a bank originates determines how centralized it is in its internal operations. Thereafter, we will focus on the main part of our analysis: whether the degree of bank integration, as measure by the Power Ratio, affects the transmission of parent shocks to domestic and foreign subsidiaries.