Banking supervision long before Basel

By Clemens Jobst (University of Vienna and CEPR) & Kilian Rieder (Oesterreichische Nationalbank and CEPR)

This blog post is based upon the authors’ article forthcoming in the Economic History Review.

—

Lending of last resort can invite imprudent behaviour by banks. Knowing that the central bank will step in whenever they run into liquidity problems, banks may reduce their holdings of cash. If most banks behave this way, aggregate liquidity declines and the financial system becomes more prone to systemic shocks and financial crises—the very scenario the central bank endeavoured to prevent in the first place.

Today, minimum liquidity standards like Basel III attempt to address these externalities. Yet, the lender of last resort is not an invention of the twenty-first century. European central banks (or note-issuing banks, as they were known in earlier periods) orchestrated liquidity support at least since the mid-eighteenth century. At the same time, most European countries neither introduced a banking code, nor conferred an official banking supervision role on central banks until the interwar period. This raises an important question. How did European central banks manage moral hazard when they still lacked formal regulatory powers?

Our forthcoming paper in the Economic History Review argues that central banks attempted to set up a system of “supervision without regulation” for this very purpose. This system created incentives for financial intermediaries to self-manage their balance sheets in a prudent manner.

We look at the case of the Oesterreichisch-ungarische Bank (OeUB), the central bank of the Austro-Hungarian Empire, to study how this “informal” supervisory mechanism worked in practice. We show that OeUB maintained a system of credit limits which provided the backbone of its supervisory activities.

All OeUB counterparties received an individual limit defining the maximum exposure the bank was ready to accept in discount lending. Since counterparties had an interest in receiving higher credit lines to streamline their day-to-day liquidity management, the OeUB offered a deal: higher credit limits in return for more prudent liquidity and capital management.

Figure 1. Postcard showing the branch office of the Austro-Hungarian Bank in Lemberg (now L’viv/L’vov, Ukraine). The panic of 1912 mostly affected branch offices in Galicia, close to the Russian border.

Our interpretation is based on a wide range of archival sources, notably the local lists of credit limits that provincial branch offices had to send to the Viennese headquarter for final approval. We hand-collected 3,400 individual decisions on credit limits from these lists, covering the period between 1909 and 1913. As Figure 2 shows, the documents give the name of the counterparty, their occupation, a qualitative statement on their (financial) management and, finally, the limit accorded by the OeUB.

A first descriptive statistical analysis of the qualitative statements reveals that leverage- and liquidity-related considerations played a more important role for the credit limits of financial intermediaries than for those of non-financial institutions—an intuitive finding, given the special risks associated with maturity transformation. In a next step, we matched credit limits with hard information on banks’ balance sheets. Controlling for time and counterparty fixed effects, we confirm econometrically that liquidity and leverage ratios represented significant determinants of the size of credit limits.

Figure 2. Entry for the Moravian Commercial and Industrial Bank in Ostrava from 1913. The branch suggested to increase the limit (“Zensoren-Antrag”) from 1.2 million to 1.5 million crowns. The directors in Vienna, refused, citing “a lack of liquidity, high rediscounts, and a mismatch between own and borrowed funds.”

But what happened in a crisis? Enforcing individual credit limits in the face of an aggregate liquidity shock is a ticket to panic and bank runs. To understand how credit limits could be squared with ample central bank liquidity provision in crisis times, we look at a banking panic in late 1912. The 1912 crisis was caused by fears about an imminent war with Russia. These fears led to a sudden increase in the public’s demand for cash and heavy withdrawals from banks situated close to the Russian border.

The reading of the Bank’s protocols and actions during the episode reveal that the Bank was ready to relax limits as long as the shock was perceived to be unrelated to individual banks’ behaviour in the run-up the crisis, or if the shock propagated like a “contagious disease”, as the Bank’s secretary general put it in a meeting on 19 December 1912. Under these circumstances, so the Bank’s management opined, commercial ‘banks had no other choice but to obtain funds for further repayments from the bank of issue, as a last resort’, (in letzter Linie, as the line reads in the German original).

Yet, the Bank did not provide liquidity unconditionally. While it intervened to avoid defaults that could fuel the panic, there was a crucial flipside to the extraordinary accommodation: once the panic had abated, the OeUB used its informal regulatory powers to impose a strict resolution and deleveraging regime on counterparties whose large-scale liquidity needs during the crisis it deemed to be ‘homemade’, rather than exogenously driven. This ‘carrots and sticks’ approach fostered time consistency. It meant that the Bank did not implicitly endorse imprudent behavior in the run-up to the crisis by indiscriminately relaxing credit limits during the panic.

Our article adds to recent research using micro data to understand historical central bank policies. While we concentrate on Austria, some form of credit limits existed at several other major banks of issue, including the Bank of England, the German Reichsbank, the Banca d’Italia and the United States Federal Reserve System.

Future research should explore when and why systems of credit limits were adopted by other central banks in the past. A comparative perspective would allow to shed more light on the parallels and differences in the design and purposes of credit limits across Europe and North America. It may also reveal to what extent the OeUB’s system was representative of similar mechanisms used by central banks elsewhere.

To contact the authors:

Email: kilian_rieder@oenb.at
Email: clemens.jobst@univie.ac.at