Systemic Risk and the Great Depression in France during the 1930's – SYSRI-30

This project is based on an interdisciplinary approach within the framework of the digital social sciences: it puts together historians, economists and IT persons.
IT persons develop tools in order to collect in the more efficient ways data from historical sources. Thanks to an approach based on OCR techniques and expertise on historical sources, software was developed to extract from yearbooks of French professions the names and the addresses of French bankers and banks. On the other side, data entry masks have been developed to speed up the entry into the Equipex DFIH relational database of handwritten balance sheet data on French banks and their failures.
The empirical and theoretical literature shows that variables of different order can act upon banking stability. First, the balance sheet structure: to sketch the argument, high liquidity and capital ratios can make a bank resilient, but they can also hamper its lending capability. Second, the governance rules of a bank (already in the Equipex DFIH database) may prevent it from excessive risk taking, but may also decrease its financing role in the economy. Panel econometrics will be mobilized to understand this dimension. A specific analysis of the central bank’s behavior in time of crisis will be run with quantitative and qualitative approach. The networks of administrators among firms and banks may explain the relative weakness of a bank. Network analysis and panel econometrics will be mobilized to understand these dimensions.
Qualitative analysis of archives and literature will enlighten the institutional framework. In spite of the lack of banking regulations in France, other rules matter (tax system, bankruptcy law, jurisprudence discount window access rules). The representations of finance following financial and banking crisis will be analyzed with a qualitative approach.

Following the construction of an original database consisting of the balance sheets of nearly 500 banks from 1910 to 1939, some important results emerge and call into question the traditional history of the banking crisis of the 1930s in France. The common opinion is that France escaped the financial crisis of the 1930s, unlike Germany and the United States for example. The first analyzes of SYSRI show, on the contrary, that while the main deposit banks maintained their activity during the crisis, the rest of the banking system experienced a major shock, leading to a drastic reduction of banking loans to the economy. The French economy was thus strongly and permanently affected by a banking crisis that began in 1929-1931.

The current main objectives of the research project are twofold. First, we are studying the determinants of banking failures and try to assess whether interbank exposures, liquidity ratios or capital ratios were associated with weaknesses of individual banks. Second, we are studying the role of the central bank during the banking crisis. The main question is to understand why the central bank did not rescue part of the banking system.

The main production at this stage was the construction of the database of bank balance sheets. The first results of the project were presented at international conferences at the Federal Reserve Bank of St Louis, the Business History Conference in Baltimore, and at the World Economic History Congress in Boston. Two main software packages have also been developed within the framework of this project. One allows manual entry of bank and banker balance sheets over the period and inserts them into the database in DFI-H compatible form. This software allows the integration of a specific accounting plan for SYSRI-30. The other software, currently in the test phase, allows automated data entry on banks and bankers during the interwar period in the view of establish a demography of banking operators. It is based on an optical character recognition (OCR) technique. Automatic verification tests are currently very successful. A notable success of the project was the organization of the international conference Building up Historical and Financial Databases held in PSE in April 2016. This led to the formation of a consortium for the submission of the EURHISFIRM project within the framework of the program INFRADEV - H2020. This submission was successful and the first tranche of European funding was obtained. The European project, supported by the coordinator of this ANR and by the institution that administers it, sees among its protagonists, the other members of the steering committee of this ANR, as well as others of its members. The three project members’ publications focus on accounting and the role of public debts in the interwar period.

This project aims at studying the casual dynamics of banking crisis during the Great Depression in France, their impact on the performances of French economy at national and regional levels, and the systemic risk management by the monetary authorities, with a concrete interdisciplinary approach based on quantitative and qualitative exploitation of primary and secondary sources.
Shortcomings in the management of banks’ risks undermine the well-being of citizens, weaken their strength to future shocks, and arouse mistrust vis-à-vis both decision makers and scientists. Banking crises thus affect society’s capability to innovate and to be open, tempted by both turning in on itself and refusing the challenges raised by rapid global changes.
By gathering together historians, economists, sociologists, accounting and IT experts, the project proposes for the first time in Europe a comprehensive country-level study of the banking crisis during the Great Depression, topic till now dominated by the literature on the United States. It is well known that the United States’ responses to the current financial crisis were based on the study of the 1930s banking crisis. The lack of methodical studies of banks’ systemic risk in 1930s European countries may lead European decision makers to base policy responses on the results of research focused on the United States, without properly taking into account both US and European specificities. To study French banking crisis, the project constructs original datasets that merged with the data already stored into the Equipex DFI-H provide researchers with databases comparable to the US ones and allow conducting for the first time large-scale quantitative studies. On the other hand, qualitative analysis of primary and secondary sources sheds light on the actual working of banking system and power balances among stakeholders surrounding authorities’ decisions.
This ANR project profits from the data of the Equipex DFI-H, but also from its resources already in place and its innovative process of sources’ digitization. It cooperates with the TGIR Progedo and Huma-Num to disseminate collected data and make scanned primary sources public. The project targets various audiences to disseminate its scientific results. First, it informs decision makers involved in the debate on systemic risk management, particularly by presentations at seminars and conferences organized by central banks members of the Eurosytem thanks to institutional links to and affiliation of project’s members with the Banque de France. Second, it disseminates the results through presentations at international conferences and publications on leading journals as well. Third, it disseminates the results among society by setting a website with tailored documentation. Fourth it contributes to high education training with material adapted to university students, while it prepares teaching material for high-school senior students.
The principal investigator has acquired significant managerial skills within the Equipex DFI-H. His research interests focus on Italian and French financial systems in the nineteenth and twentieth centuries. His approach, based on extensive use of primary sources (he co-wrote the inventory of the archives of the Paris and Lyon Stock Exchanges), is centered on studying the actual working of financial institutions and the interactions among them through quantitative methods and tight assessments of statistical series. After extensively working on financial markets, he has recently started research on banking crisis. He regularly adopts interdisciplinary methods: he is qualified by the Conseil National des Universités in economics, business administration and contemporary history, while his more frequent co-author is a social scientist.