The Ledger and the List
Moshe Bejski knew what it meant to be kept off a record. Born in Działoszyce, Poland, in 1921, he survived the Plaszów labour camp under Amon Göth and was saved only because his name appeared on a list compiled by a German industrialist named Oskar Schindler. That list recorded approximately 1,200 names. Each one represented a life that might otherwise have vanished without a trace.
Bejski emigrated to Palestine in 1945, studied law at the Sorbonne, and rose to the Israeli Supreme Court. He spent twenty five years chairing Yad Vashem’s Righteous Among the Nations Commission, honouring the non Jews who had saved Jewish lives. He understood lists. He understood what happens when obligations are recorded, and what happens when they are not.
Which is why, in January 1985, the Israeli government asked him to investigate the worst banking crisis in the country’s history.
On 6 October 1983, the Tel Aviv Stock Exchange closed. It would remain shut for eighteen days. Israel’s four largest banks had spent years buying their own shares through intermediaries and lending clients the money to buy more, creating a loop that drove bank stock prices up by three hundred per cent in real terms. The true leverage was hidden. The obligations were circular. By the time the loop broke, the banks were insolvent. Seven institutions, constituting more than ninety per cent of the banking sector, were nationalized. Bejski’s 560 page report, delivered in April 1986, concluded that the banks had manipulated share prices “through a series of actions designed to affect share prices and returns.” Sixteen of Israel’s most senior banking and government finance officials were censured. Bejski recommended that four be barred from the industry for life.
Bejski’s principle was simple. If an obligation exists, it must be recorded. If it is not recorded, someone will eventually pay for its absence. The cost of concealment is never zero. It is merely deferred.
That principle is now being tested on a scale he could not have imagined. Last week’s Financial Times reported that technology companies including Meta, Nvidia, and Broadcom have issued up to three hundred billion dollars in guarantees over the past twelve months to finance artificial intelligence infrastructure. The guarantees are structured through special purpose vehicles. The tech company does not borrow the money itself. A separate entity does. The tech company merely promises that if the data centers or chips backing the debt lose value, it will cover part of the shortfall. Because the promise is contingent rather than certain, it does not appear on the balance sheet as ordinary debt. It appears, if it appears at all, in the footnotes.
[https://www.ft.com/content/7f11afae-c4e3-4054-a65b-873f3647f563?syn-25a6b1a6=1]
The numbers buried in those footnotes are staggering. Morgan Stanley estimates that total off balance sheet commitments across the major technology firms now exceed three trillion dollars. Meta alone carries an estimated four hundred and twenty billion dollars off balance sheet, nearly three times its disclosed debt. Google’s purchase commitments surged from seventy two billion dollars across all of 2025 to seven hundred billion in a single recent quarter. Amazon and Google have already seen their free cash flow turn negative. Oracle’s insurance against default has risen fivefold.
Andrew Fastow, the chief financial officer of Enron, used hundreds of special purpose entities with playful names like JEDI and Raptor to hide billions of dollars in debt. When the structure collapsed in 2001, it destroyed the seventh largest company in America and the accounting firm that had approved every page. In 2008, banks used structured investment vehicles to keep mortgage exposure invisible. When property values fell, the guarantees were called, and the cost returned to the institutions that thought they had transferred it.
Today’s structures are legal and disclosed in ways Enron’s were not. But the pattern is consistent across decades and continents. Complex structures push obligations into vehicles whose purpose is to separate the appearance of risk from the reality of risk. Regulators tighten the rules after each crisis. New structures emerge that comply with the letter of the tightened rules while reproducing the same incentive. The guarantor enjoys the benefit of debt funded expansion without the visible cost of debt. The hidden exposure grows until something forces it back onto the books.
Israel learned this in 1983. The banks had not broken every rule. They had been granted exemptions from insider trading laws precisely so they could trade their own shares. The regulators knew. The lesson of the Bejski Commission was not that the banks were uniquely dishonest. It was that a system designed to tolerate concealment will produce concealment, and that the cost will fall on those least equipped to see it coming.
The trading floor of the Tel Aviv Stock Exchange sat empty for eighteen days in October 1983. The man who had survived because his name appeared on a list spent the rest of his career ensuring that nothing was left off the record. Three trillion dollars in the footnotes suggest the lesson was not learned. The room is full again. The ledger is not.