Lessons from 1763: the central bank as lender of last resort

It is a real pleasure to be here with you today. Although I must admit I am standing in for our governor, Olaf Sleijpen, who unfortunately could not join us.

But perhaps that is not entirely a disadvantage. Because coming from a supervisory background, I can offer a slightly different perspective. One lesson that has stayed with me over the years is that, in times of stress, the strength of the system largely depends on how well central banks and supervisors work together across borders.

And with that in mind, let me take you back in time, with a story. A story about Amsterdam. A story about a crisis.

Picture this. We are in Amsterdam. It is 29 July, 1763. Leendert Pieter de Neufville, one of the most prominent bankers of his time, is sitting behind his desk in his richly decorated office. He has his hands in his hair. Or more likely, in his wig. He has just gone bankrupt. A bankruptcy that will send shock waves through large parts of Europe.

What had happened?

The banking house of De Neufville was a big player in Amsterdam, one of the major financial centres in Europe. And the times had been good, very good. The Seven Years’ War had opened up new opportunities. Capital flowed from regions like Holland to capital-hungry regions such as Prussia, what we might call today an emerging market.

But there was a tension. Dutch lenders were typically conservative and risk-averse. They were only prepared to lend short term, to reputable names. This problem was overcome by a financial innovation, the acceptance loan. This basically transformed the short-term bills of exchange into the long-term loans needed in Berlin. In this construction, the trusted merchant bankers in Hamburg were crucial intermediaries.

Credit expanded rapidly, as we can see from the increasing size of bank balance sheets during that time. All was well. Until it wasn’t.

In 1763, the war ended. Grain prices collapsed by more than 75% in just a few months. And at exactly the wrong moment, De Neufville was deeply involved in a large, highly leveraged speculative deal involving grain from the departing Russian army in Poland. The deal failed. His capital evaporated.

The collapse of De Neufville triggered a classic credit crunch. Other banks in Amsterdam immediately failed, including some of the long-established banking houses. Two weeks later, there was a first wave of bank failures in Hamburg, in turn triggering a new wave in Amsterdam. There was no coordinated public intervention. No effective lender of last resort. Ultimately, Prussia was hit hardest, plunging into a deep and long-lasting recession.

I must admit, this story was not ready knowledge for me. Much of what I just told you is derived from a fascinating 2003 article by Hyun Song Shin and - a name that may sound familiar to you - Isabel Schnabel.

The crisis of 1763 was one of the first cross-border financial crises. Its ingredients are all too familiar. Leverage, maturity mismatch, complex products, interconnectedness. And then: a liquidity squeeze, fire sales, collapsing asset prices, and bank failures.

Only one key ingredient was missing, an ingredient that could have saved the day: central banks. Institutions capable of stabilising the system. Of providing liquidity, especially in moments when markets fail. In other words: lenders of last resort.

Emergency liquidity assistance, or ELA, would have certainly helped in that crisis, especially if coordinated cross-border, as in the Game without Frontiers you are simulating today.

But, and this is an important ‘but’, having the tools is not enough. The events of 2023 reminded us that banks must also be ready to use them.

In both the case of Silicon Valley Bank and Credit Suisse, operational readiness was an issue. The Swiss National Bank had to provide ELA to Credit Suisse based on a government guarantee, because Credit Suisse was unable to post sufficient collateral in time.

That is a powerful lesson. Collateral prepositioning matters. The question, of course, is how far we should go.

A strict prepositioning requirement could increase resilience, but at the cost of flexibility. This could result in the central bank shifting from being a lender of last resort to a lender of first resort.

Similarly, including prepositioned collateral in the LCR, as has been suggested in in recent policy debates, would fundamentally alter the nature of that metric and potentially weaken banks’ self-reliance.

In the UK, the PRA has proposed that, provided they are operationally prepared, banks should be allowed to assume central bank facility drawings for internal stress testing purposes. This is closer to what we are doing in the SSM, where supervisors are allowed to consider borrowing capacity with central banks in internal supervisory analyses.

And importantly, we are not starting from zero. Following the 2023 turmoil, we conducted a stock-taking exercise in the BCBS on operational readiness and prepositioning. The exercise revealed that banks in various countries are already prepositioning collateral voluntarily.

After the events in 2023, US banks have prepositioned 1 trillion dollars as collateral. And as you are well aware, banks in the EU also pledge collateral for contingency reasons. The question is whether this is sufficient or that more needs to be done. More prepositioning will further improve resilience and would reduce the likelihood of ELA, but this may come at higher costs for banks.

It is very important that both banks and central banks prepare for the provision of ELA. Today you are in the middle of a crisis management exercise, which is extremely useful and I am very interested in the lessons learned.

Before I conclude, I would also like to make a few comments about the use by banks of the central bank’s regular refinancing operations.

Banks may sometimes have the impression that using central bank operations in normal situations may trigger questions from their supervisor. This is absolutely not the case, and I think it is important to correct this misperception to avoid stigma in the market. It is true that most supervisors are alert when banks under their supervision make use of central bank operations. Especially when this might signal that banks have no access to market funding. But this doesn’t mean that we don’t want banks to use the regular operations: banks can certainly use them in their day-to-day liquidity management, as long as they do not rely on them as their sole source of funding.

I think supervisors and central banks should clearly communicate that there is no issue with using the regular operations. Therefore, the blog by Claudia Buch and Isabel Schnabel on Managing liquidity in a changing environment was a very good idea. And I think we should continue to repeat its message from time to time. Using the ECB’s standard refinancing operations is a routine and integral component of banks’ day-to-day liquidity management, and constitutes an important part of a diversified funding mix.

Let me end where I began. The story of 1763 reminds us of the consequences of having fragile and interconnected financial systems without crisis coordination or a lender of last resort. Today, we are in a very different place. You, as members of the Market Operations Committee, fulfil an important role in our modern financial system.

Through your work in this Committee, you help ensure that, when the next moment of stress comes, we are not reacting, but ready.

So I would like to thank you for that commitment, and I look forward to hearing the lessons from your discussions today.

Thank you very much.