The recent financial crisis and the ensuing economic slowdown have heightened the importance of better understanding the interconnectedness between the industrial and banking sectors. While several recent studies undertake this endeavor, the transmission mechanism in these studies is almost always from the banking sector to the industrial sector. In contrast, in our paper, The Effect of Industrial-Sector Transparency on Bank Risk-taking and Banking System Fragility, which was recently made publicly available on SSRN, my co-author (Sudarshan Jayaraman) and I provide evidence of the chain of causality working in the reverse direction, i.e., from the industrial sector to the banking sector. In particular, we document the important role that industrial-sector transparency plays in the efficient functioning of the banking sector. We argue that greater transparency in the industrial sector facilitates firms’ access to financing from capital markets and thus diminishes their reliance on banks. As a result, we expect banks to face increased competition in their product markets and to offset their lost rents by: (i) taking on more risk, (ii) reducing their cost structures, and (iii) increasing the intensity of intermediation.
Using a large panel of banks from 37 countries, we find strong support in favor of our predictions. We find that banks in countries with greater borrower transparency do indeed take on more risk, improve cost efficiency, and undertake greater intermediation. Moving from the median level of transparency (which corresponds to Japan in our sample) to the upper quartile (Norway) increases bank risk-taking by 13%, cost efficiency by 5%, and bank intermediation by 6%. We find that the increases in risk-taking emanate through both lending and non-lending activities, but that influence of the latter dominates. The overall effect of these actions by banks is to lower the likelihood of a banking crisis in countries with greater industrial-sector transparency. Moving from the lower quartile of transparency (Israel) to the upper quartile (Norway) reduces the likelihood of a banking crisis from 55% to 40%. These results speak to the beneficial role of bank diversification in reducing bank fragility.
We perform several sensitivity checks to ensure that our results are not confounded by endogeneity concerns stemming from reverse causality, correlated omitted variables, etc. In particular, we use the mandatory adoption of a common set of accounting standards (called International Financial Reporting Standards or IFRS) by several countries in 2005 as a shock to industrial-sector transparency and find within-country increases in bank risk-taking, cost efficiency, and bank intermediation.
Our study is novel in several aspects. First, it documents the important role that industrial-sector transparency plays in the efficient functioning of the banking sector. While a long stream of research documents the role of firm transparency on firm outcomes and similarly of bank transparency on banking outcomes, ours is the first study to document the interconnectedness between transparency in the industrial sector and economic outcomes in the banking sector. An important implication of our study is that one of the ways in which the banking sector can be made more efficient and stable is by improving transparency in the industrial sector—a channel that has not been recognized as yet.
Second, our study broadens the economic consequences of adopting a common set of accounting standards. We show that the adoption of common accounting standards by countries promotes the development of their banking sectors—an important driver of economic growth. Finally, our study contributes to the bank competition risk-taking literature—an area of keen interest to regulators. In contrast to most studies that employ cross-sectional associations in this literature, we provide evidence of causality by documenting that shocks to borrower transparency lead to increases in bank risk-taking, cost efficiency, and bank development.
The full paper is available for download here.