In the aftermath of the 2008 financial crisis, the Dodd-Frank Act of 2010 set out to limit the authority of the Federal Reserve to rescue insolvent financial institutions. Since 1932, Section 13(3) of the Federal Reserve Act had given the agency the power to lend to “any individual partnership, or corporation” in “unusual and exigent circumstances.” The 2010 Act now compels the Fed to consult with the Secretary of the Treasury before implementing a new lending program. In providing liquidity to the financial system, central banks should require sufficient collateral to protect taxpayers and not aid a failing institution in a way that could induce moral hazard (Vives 2008).
This change seems to rule out pre-emptive action by the Fed to halt a panic by rescuing a ‘systemically important financial institution’ (SIFI) – an institution that is so intimately connected with the financial system that its failure could induce a general panic. If this modification of Section 13(3) were strictly enforced, the Fed would be limited to reactive policy tools to calm a panic after it had begun. The costs to the economy might then be higher. This prompts the following question: Under what circumstances might a pre-emptive rescue work well?
A forgotten rescue
Much research on financial crises focuses on policy failures, but it is important to identify successes as well – including those outside of the well-studied Anglo-American examples. In our study of French central banking (Hautcoeur et al. 2014), we have recovered the history of an incipient crisis in 1889, when the Banque de France formed a ‘lifeboat’ to rescue one of the largest French banks from failing and inducing a general banking panic. The environment in which this operation took place is notable for two reasons:
1. The French banking system was largely unregulated – there were no required capital or reserve ratios and there were no constraints on universal banking.
2. The government provided no guarantees to the banks – deposit insurance and ‘too big to fail’ were far in the future.
The actions of the Banque are striking as they violated Walter Bagehot’s strictures that define the contemporary debate on how a central bank should behave. In his classic, Lombard Street (1873), Bagehot ordained that a lender of last resort should respond to a financial crisis by lending freely at a high rate of interest on all good collateral, thereby preventing illiquid but not insolvent banks from failing (Bignon et al. 2012). The duty of the central bank was not to prevent financial shocks, but to neutralise their effects. The danger of moral hazard was too great if the central bank acted otherwise. Yet in 1889, the Banque – acting at the behest of the Ministry of Finance – reacted to a run on a major bank by providing liquidity to the whole market as Bagehot recommended, but also pre-emptively launched a ‘lifeboat’ to rescue the bank.
The crisis of 1889
The names of the top contemporary French banks – Société Générale, BNP Paribas, and Crédit Agricole (until recently, Crédit Lyonnais) – are familiar to most. But there is one name that is missing but which formed part of this echelon in the 19th century – the Comptoir d’Escompte (the Discount Bank), that engaged in commercial and investment banking, financed international trade, provided credit to the stock market, and assisted with government bond issues. In the tight financial networks of 19th-century France, where the private investment banks were heavily represented in the management and boards of directors of the leading banking corporations, the Comptoir was privy to most major deals. These interlocking relationships carried over to the Banque de France, where most of the time they served to inform the central bank of the strengths and weaknesses of financial institutions. However, the demise of the Comptoir was the result of the manipulation of these potential conflicts of interest by a small group for huge gains.
Working through a large industrial metals concern, the Société des Métaux (the Metals Company), this group began an attempt to corner the world’s copper market in late 1887. Copper was a target because industrial and agricultural demand for the metal was quickly rising. Had the speculation been confined to this group, which provided the initial capital, its failure would have caused sharp losses but not a banking panic. However, to purchase and warehouse current copper stocks and control future deliveries with forward contracts was beyond its means, and it needed funding from the big banks. Overly eager for profits, the management of the Discount Bank was ensnared. In addition to granting loans to buy copper, the bank signed guarantees for the Metals Company’s forward contracts – liabilities that were off of its balance sheet. This assistance was critical, and enabled the syndicate to double the world price of copper.
Soon, deliveries of copper began to exceed demand, leading to the growth of a vast stockpile. Covering nearly two years of the world’s copper supply, the guaranteed forward contracts left the Discount Bank in a highly leveraged position. After discounting some copper warrants, the senior management of the Banque de France became concerned, and summoned the head of the Discount Bank to present detailed records. Revealing it to be deeply insolvent, the CEO committed suicide on 5 March 1889 – an event that turned a slow drain of deposits into a run.
The design of the ‘lifeboat’
The Banque initially hoped other banks would privately come to the aid of the Discount Bank. Then on 7 March, the Minister of Finance called a midnight meeting with the leading banks and formed a rescue plan to be managed by the Banque. Presented with this plan the next morning – with only an hour before the banks opened – the governing body of the Banque, the Council of Regents, debated whether to acquiesce. A supermajority – required to suspend the Banque’s rules – voted to proceed, but the debate was intense, highlighting the concerns about rescuing an insolvent bank.
The Banque agreed to provide an emergency loan of 100 million francs against collateral of all of the bank’s assets, regardless of quality. However, it required a guarantee syndicate of banks that would absorb the first 20 million francs of losses. Although ability to pay – measured by bank capital – partly determined which banks would contribute, responsibility for the Discount Bank’s collapse was a key factor in assigning shares in the syndicate. Conflicts of interest measured using network analysis of the interlocking directorships or subscription lists to the Metals Company’s 1888 capital increase econometrically identify this factor in our study. The Banque provided additional liquidity as Bagehot recommended, but it was secondary and did not concern the Regents’ debate – even as a second loan of 40 million francs with a guarantee syndicate of 20 million was made to finally squelch the panic. Ultimately, these guarantees were not called upon, but that was because losses were recouped by large fines – most notably on the Discount Bank’s board of directors, who were left ruined or bankrupt. The void left by the Discount Bank was then filled by a new Comptoir National d’Escompte, recapitalised with private funds.1
The story of the Crisis of 1889 leaves two questions. The first is why didn’t Bagehot discuss the potential problem of failing SIFIs? The answer may be that in 1873, when he published Lombard Street, the merger waves that swamped British banking in the last quarter of the 19th century had yet to produce a SIFI, and insolvent banks were not seen as so interconnected as to threaten the solvency of the many more financial institutions. By the time that Barings failed in 1890, the Bank of England believed it necessary to intervene with a lifeboat à la française. Certainly, the Rothschilds believed this to be the case, and saw Barings as a far worse threat to the financial system than the failure of Overend, Gurney and Co. in the previous crisis of 1866. The strict use of Bagehot’s rule may thus require some caution.
The second question is whether lessons from the 19th century are applicable in the 21st. The 1889 crisis – like so many others before 1914 – had its origins in the exploitation of conflicts of interest by a single institution or a small group of institutions. The widespread insolvencies of the US Savings and Loan Crisis of the 1980s or the financial crisis of 2008 are different, and pose a larger problem because of systemic risk-taking incentives created by deposit insurance or ‘too big to fail’. Nevertheless, the response of the French authorities is instructive because they did not hesitate to impose harsh penalties on those responsible, which may have helped to contain moral hazard – for the next 25 years, there were no more banking crises in France.
Bagehot, W (1873), Lombard Street: A Description of the Money Market, London: H S King.
Bignon, V, M Flandreau, and S Ugolini (2012), “Bagehot for Beginners: the Making of Lender-of-Last Resort Operations in the Mid-Nineteenth Century”, Economic History Review, 65(2): 580–608.
Hautcoeur, P-C, A Riva, and E N White (2014), “Floating a Lifeboat: the Banque de France and the Crisis of 1889”, Journal of Monetary Economics, forthcoming.
Vives, Xavier (2008), “Bagehot, central banking, and the financial crisis”, VoxEU.org, 31 March.
1 The Comptoir National d’Escompte survived until 1966, when it merged with the Banque Nationale pour le Commerce et l’Industrie to become the Banque Nationale de Paris. In 2000, The Banque de Paris et des Pays-Bas was absorbed by the Banque Nationale de Paris and became BNP Paribas.