The Herstatt Bank Failure: How One Afternoon in Cologne Created Settlement Risk (1974)

At 15:30 Central European Time on Wednesday, 26 June 1974, officials from the Bundesaufsichtsamt für das Kreditwesen walked into Bankhaus I. D. Herstatt KGaA at Unter Sachsenhausen 4 in Cologne and informed its partners that the bank's licence had been revoked. The doors of a 247-year-old private house, founded by a silk merchant in 1727, were closed to depositors and counterparties. Within hours, correspondent banks from Chase Manhattan on Park Avenue to Hill Samuel on Wood Street discovered that the Deutsche marks they had wired to Cologne that morning had been received — and that the matching US-dollar payments due in New York that afternoon would never arrive. The Cologne bank was not large. It ranked 52nd among West German institutions and had a balance sheet of roughly DEM 2 billion. But the manner of its collapse introduced a word into the language of international finance that has survived every subsequent reform: Herstatt risk.

A private house on the Rhine

Bankhaus Herstatt had been bought in 1955 by Iwan David Herstatt, a Cologne-born banker with family roots in the city's Protestant Huguenot community, and by Hans Gerling, the insurance magnate whose Gerling-Konzern held the majority stake. Under Herstatt's stewardship the bank grew quickly, specialising in merchant banking for Rhineland industry and, from the late 1960s onward, in foreign-exchange trading. By 1973 the FX book dwarfed the rest of the business. A senior trader named Daniel "Danny" Dattel ran a team on the third floor whose daily turnover in Deutsche marks, dollars, Swiss francs and sterling ran into the hundreds of millions — a sum entirely disproportionate to the bank's DEM 85 million of capital.

The context matters. The Nixon Shock that ended the gold standard in August 1971 had pulled the last pillar out from under the Bretton Woods system of pegged currencies. By March 1973 the major industrial countries had abandoned fixed parities entirely, and exchange rates began to float. What had been a staid interbank plumbing job became, almost overnight, a speculative casino. German commercial banks, newly exposed to dollar volatility, generated huge demand for FX dealing. Small and mid-sized Landesbanken and private houses rushed in. The Bundesbank, whose own Deutsche-mark interventions had financed the end of Bretton Woods, watched the expansion with unease (Galati, 2002).

Dattel bet that the dollar, having already fallen against the mark from DEM 3.66 to DEM 2.40 between 1971 and 1973, would continue to weaken. Through the first months of 1974 he doubled down on long-mark, short-dollar positions. When the dollar rallied sharply in May and June, the book was caught.

The spring of 1974

By late April the Bundesaufsichtsamt, the federal banking supervisor in Berlin, had begun receiving signals from several Landesbanken that Herstatt's FX counterparty exposures looked wrong for a bank of its size. The supervisor dispatched examiners to Cologne in May. They found open positions whose mark-to-market loss, using spot rates on the day of the visit, already ran near DEM 470 million — more than five times the bank's stated capital (Schenk, 2014). Dattel and his superiors had been hiding the losses by rolling forward contracts at off-market rates and booking notional "profits" on paper trades with affiliates.

Iwan Herstatt was summoned to Berlin. Gerling, the majority owner, initially offered to inject fresh capital and obtained a short forbearance. Through June the bank continued to trade, hoping the mark would weaken again. It did not. On 25 June, the supervisor informed Herstatt that the licence would be pulled the following day unless the full DEM 470 million was plugged. Gerling declined to proceed.

A one-day timeline

The closure was scheduled for 15:30 Frankfurt time on Wednesday, 26 June — after the West German interbank day had effectively ended but during the New York morning. The choice of hour was intentional from the Bundesbank's perspective: it minimised domestic panic. It also, unintentionally, maximised foreign-exchange havoc.

The scene at CHIPS in lower Manhattan was chaotic. CHIPS was a net-end-of-day system: during the day it tallied dollar payments between member banks and then settled the net at close of business. When the Herstatt news hit at roughly 10:30 EDT, member banks began revoking payment orders to try to avoid paying dollars for marks they knew they would not receive. The Clearing House, facing a cascade of reversals, imposed a 72-hour delay on provisional settlement, which Hildebrand (2007) has described as the single most disruptive operational moment in CHIPS's history.

Who took the hit

The loss distribution became clearer over the following weeks as counterparties filed claims with the Cologne liquidator. The US Comptroller of the Currency and the Bank of England compiled unofficial tallies that circulated between supervisors long before they were published. Arthur Burns, chairman of the Federal Reserve, personally telephoned Helmut Schmidt, then West German chancellor, to press for an orderly process — a call that Mourlon-Druol (2015) argues was the first concrete instance of Fed-Bundesbank crisis coordination in the floating-rate era.

The total US counterparty loss — roughly $328 million in 1974 dollars, or around $2 billion in current money — was modest against global FX turnover but enormous against the paid-in capital of the exposed institutions. Chase, as the largest single claimant, wrote to Herstatt's liquidator on 2 July 1974: "Our claim arises not from lending the borrower funds, but from having paid Deutsche marks this morning for dollars we will never receive" (Chase letter, quoted in Schenk, 2014).

The concept is named

The failure reframed a risk that had always existed but never been labelled. In a spot FX trade, one party pays one currency in one centre at one time; the counterparty pays another currency in another centre at another time. The gap between the two payments is the window of settlement risk. Before 26 June 1974 that window was discussed in technical papers as "temporal mismatch in foreign-currency settlement". Afterward it was simply called Herstatt risk.

Galati (2002), in the Bank for International Settlements quarterly review, put the shape of the exposure plainly: "The maximum loss a bank can suffer from a Herstatt-type event is the full principal of the currency it has already paid, not the usual small credit spread on an uncollateralised loan." In an FX market whose daily turnover by 1989 had reached $620 billion and by 2022 exceeded $7.5 trillion, that principal exposure aggregated across the banking system became a potential systemic hole.

The Basel Committee, born in a hurry

Four months after Herstatt, the governors of the G10 central banks met in Basel under the presidency of George Blunden of the Bank of England. In December 1974 they chartered the Committee on Banking Regulations and Supervisory Practices, housed at the Bank for International Settlements. Its first chairman was Blunden himself; its first deliverable, in 1975, was the Concordat — a short text allocating supervisory responsibility between home and host authorities for cross-border banks. The Concordat's central principle, that no internationally active bank should escape supervision in any jurisdiction where it operates, was a direct answer to the Herstatt case, where the German supervisor had no formal window into Herstatt's New York-side exposures (Goodhart, 2011).

From that narrow beginning the Basel Committee grew into the most influential standard-setter in global banking. Basel I in 1988 imposed the 8 percent capital ratio and the risk-weighted-asset framework. Basel II in 2004 refined risk weights and introduced the three-pillar structure. Basel III, assembled in the wake of the 2008 financial crisis, added the leverage ratio, the liquidity coverage ratio, and the net stable funding ratio. Herstatt is rarely mentioned in any of those documents. But the committee exists because of it.

Twenty-eight years to a fix

Basel I and II addressed capital. They did not address settlement. Through the 1980s and 1990s FX settlement risk quietly grew with the market. The New York Fed's 1996 Allsopp Report, commissioned to measure the exposure, found that for a typical large bank the daily peak Herstatt exposure ran into the tens of billions — in many cases larger than the bank's own capital. The report's author, Patricia Allsopp, concluded that national supervision alone could not shrink the number; only a structural change to how FX trades settled would.

The industry built one. Between 1997 and 2002, a consortium that eventually numbered seventy of the world's largest FX banks constructed CLS Bank — the Continuous Linked Settlement Bank — as a special-purpose US Edge Act institution, supervised by the Federal Reserve. Live settlement began on 9 September 2002 in seven currencies. CLS operates on a payment-versus-payment model: both legs of an FX trade settle simultaneously across CLS's own books, so neither side pays unless the other pays. By 2024 CLS was settling approximately $6 trillion of FX obligations daily across 18 currencies, and the BIS estimated that CLS had eliminated roughly half of the global FX settlement exposure outright.

Plenty remains unsettled on CLS — trades in non-CLS currencies, trades among counterparties who are not CLS members, and trades whose settlement falls outside CLS's operational window. The BIS Triennial Survey of 2022 found that FX settlement exposure still ran above $2 trillion on a typical day, much of it in emerging-market currencies. Herstatt risk has not disappeared; it has been compressed.

Legacy

Iwan Herstatt was charged with false accounting and fraud in 1976. He was fined DEM 60,000 and given a suspended sentence; Danny Dattel served four and a half months. Gerling-Konzern eventually paid out roughly DEM 100 million in settlements to counterparties. The Cologne bank's 247-year corporate history ended in liquidation in 1984, long after the name had detached itself from the building and become a technical term.

Structural echoes of the episode are audible in every FX-related reform since. The post-2008 push for central clearing of derivatives, explored alongside reforms that followed the collapse of Barings Bank and the LIBOR rigging revealed in the LIBOR scandal of 2008–2012, rests on the same logic CLS embodied a decade earlier: when a market's plumbing is fragile, the fix is not more rules but a different plumbing.

The irony of the Cologne afternoon is that the bank itself was ordinary. A mid-sized private house, an overambitious trader, a bad FX bet, an orderly Bundesbank closure at 15:30 — nothing in that chain was novel in 1974. What was novel was the realisation that a 247-year-old bank in Cologne, closed before tea-time, could leave the world's most important market with a hole at its centre for the next twenty-eight years. The phrase that failure gave the language has outlasted the bank, the building, and most of the people who wired the marks that morning.