The Salad Oil Swindle: How Seawater Broke Two Wall Street Brokers
An inspector for the American Express Field Warehousing Corporation would climb the ladder welded to the side of a storage tank in Bayonne, New Jersey, unbolt the hatch, and lower a weighted measuring tape into the dark. He would draw it back up wet with vegetable oil to roughly the depth he had been told to expect, write the number on his clipboard, and walk to the next tank. Nothing in that procedure could reveal the thing that mattered, which was that most of the column beneath his tape was water pumped in from Newark Bay with a few feet of soybean oil floating on top.
On the strength of those clipboard numbers, American Express issued warehouse receipts. On the strength of the receipts, fifty-odd banks and commodity houses lent Anthony De Angelis something close to $180 million. When the receipts turned out to describe oil that had never existed, two member firms of the New York Stock Exchange went under in a single week, the Exchange reached into its members' pockets for the first time in its history to pay the customers of a failed broker, and one of America's best-known financial names spent the next four years arguing about how much of a fraud it had to make good.
The Man From the Bronx
Anthony De Angelis, universally called Tino, was born in the Bronx in 1915 and left school young to work in a Washington Market butcher's shop. He was good at it. By 1949 he was president of the Adolf Gobel Company, a large New Jersey meat processor, and had learned the trick that would define the rest of his career: government buying programmes are generous, and their inspectors are outnumbered.
Gobel sold to the National School Lunch Act programme, and in 1952 it was caught overcharging the federal government by $31,000 for substandard food and shipping more than two million pounds of uninspected meat. The Yugoslav government sued over lard that did not match specification; the West German government complained about quality as well. Gobel went into bankruptcy. De Angelis, who had been the company's president through all of it, was not prosecuted, and three years later he incorporated the Allied Crude Vegetable Oil Refining Corporation on a stretch of waterfront in Bayonne.
Allied's business was legitimate in outline. Cottonseed and soybean oil were bulk exports, the Food for Peace programme subsidised shipments abroad, and a refiner who could buy cheaply and ship in volume had a real trade. By the early 1960s Allied was moving a meaningful share of American vegetable-oil exports, and De Angelis was a figure of some standing in the commodity business, a large man in a cheap suit who kept a plain office and was said never to forget a shipping rate.
A Warehouse That Belonged to Nobody
What made the fraud possible was an arrangement called field warehousing, a device for lending against inventory that is too bulky to move. A warehousing company leases a corner of the borrower's own premises, takes legal custody of the goods stored there, and issues negotiable receipts that the borrower pledges to a lender. Since the warehouseman stands between borrower and bank, the bank is spared the trouble of inspecting anything itself.
De Angelis's cousin, Michael De Angelis, approached the American Express Field Warehousing Corporation about opening a facility on Allied's Bayonne tank farm. American Express took the business. Field warehousing in that era was thin-margin work, and the standard economy was to hire custodians locally rather than staff each site with outsiders — which in practice meant hiring the borrower's own employees and asking them to report on their employer.
At Bayonne, American Express let De Angelis choose the custodians. He chose friends and relatives, and the tank readings they filed were whatever Allied needed them to be. On the occasions when American Express sent its own inspectors, the physics of the thing did the rest: oil floats, so a tank nine-tenths full of seawater gives a correct reading on a dipstick lowered from the top. Where an inspector wanted to see several tanks, Allied's men moved the same oil between them through the yard's pipework ahead of the tour.
By the autumn of 1963, warehouse receipts in circulation described roughly 900,000 short tons of vegetable oil — about 1.8 billion pounds — held in Bayonne on Allied's account. The tanks held something on the order of 55,000 tons. The receipts also described more oil than the Department of Agriculture believed existed in the entire United States, a discrepancy visible to anyone who cared to lay the two published numbers side by side. Nobody did, because each lender saw only its own receipts, and each receipt carried the name of American Express.
The Corner
Fraudulent collateral was the means; the end was a bet. De Angelis used the borrowed money to buy cottonseed and soybean oil futures, on the theory that a large enough long position would lift prices, which would raise the value of the collateral behind his loans, which would support more borrowing. Jerry Markham's survey of American commodity regulation treats the episode as a textbook instance of a squeeze financed entirely by paper that did not exist (Markham, 2002).
The bet had a specific catalyst. Through 1963 traders expected the Soviet Union, having lost much of its harvest, to buy American grain and oilseed on a scale that would clear the market. De Angelis positioned for it heavily. On 15 November the United States Senate suspended work on the wheat-export negotiations with Moscow, and the trade concluded that the vegetable-oil orders behind the rumour were not coming.
Source: Contemporary New York Produce Exchange and Chicago Board of Trade settlement quotations
Soybean oil settled at 9.875 cents a pound on Friday 15 November, at 9 cents on Monday, and at 7.75 cents on Tuesday 19 November. A fall of just over two cents sounds like nothing. Against a position built with borrowed money on thin margin it was fatal, and it arrived at exactly the moment when De Angelis needed the collateral to be worth more, not less. Allied Crude Vegetable Oil Refining filed for bankruptcy that Tuesday. The New York Produce Exchange, where most cottonseed oil traded, halted dealings in the contract and never really recovered the business; within a few years it had ceased to function as a commodity market at all.
Two Brokers and Twenty Thousand Customers
Allied's collapse landed first on the firms that had carried its futures account. Ira Haupt & Co. was the largest of them, an old and respectable Stock Exchange member with about 20,000 customers, most of whom had never heard of vegetable oil. Haupt had allowed Allied to trade on thin margin and had gone further than that: it had borrowed from banks against Allied's warehouse receipts and passed the proceeds back to its client.
When the exchanges called for roughly $14.1 million of additional margin on the Allied positions, Haupt could not produce it, and the receipts it had pledged to Chase Manhattan, First National City, Morgan Guaranty and Manufacturers Hanover were worth a fraction of their face. A federal court later described the mechanism without ornament: Haupt's dealings with Allied had caused "the ratio of Ira Haupt & Co.'s assets to its liabilities to fall below the minimum requirements of the New York Stock Exchange and of the Securities and Exchange Commission." The Exchange suspended the firm on 20 November 1963, along with J.R. Williston & Beane, which had carried a smaller piece of the same account. By 21 November the size of the hole was, in the court's phrase, "in the neighborhood of twenty million dollars."
| Date (1963) | Event |
|---|---|
| 15 November | Senate halts Soviet wheat talks; soybean oil settles at 9.875 cents |
| 19 November | Allied Crude Vegetable Oil files for bankruptcy; oil settles at 7.75 cents |
| 20 November | NYSE suspends Ira Haupt & Co. and J.R. Williston & Beane |
| 22 November | Kennedy is shot; NYSE halts trading at 2:07 p.m.; Dow closes at 711.49 |
| 25 November | Exchange and creditor banks sign the Haupt rescue agreement |
| 26 November | Market reopens; Dow gains 32.03 points, then its largest one-day rise |
Two days after the suspensions, on the afternoon of Friday 22 November, President Kennedy was shot in Dallas. The Exchange stopped trading at 2:07 p.m. with the Dow Jones Industrial Average at 711.49, down 21.16 points on the day, and stayed shut through Monday for the funeral. Whatever else that closure did, it bought the Exchange's governors a weekend they would not otherwise have had.
The Weekend the Exchange Decided to Pay
Nothing obliged the New York Stock Exchange to protect the customers of a bankrupt member. A brokerage failure in 1963 meant that customers joined the queue of general creditors and waited years for cents on the dollar. The Exchange's governors looked at 20,000 such customers, at a market about to reopen after a presidential assassination, and concluded that the queue was the greater danger.
They met through the weekend. On Monday 25 November, five days after Haupt closed its doors, the Exchange, Haupt's creditor banks and the firm's general partners signed an agreement whose terms the Second Circuit later set out plainly: "the Exchange advanced $9,500,000 from which to pay Haupt's customer creditors; the Banks agreed to defer their claims against Haupt to the extent of two dollars for every dollar advanced by the Exchange." The Exchange raised its share by assessing its own members, and the commitment eventually reported ran to about $12 million. Charles Seligson was appointed trustee in the Haupt bankruptcy and spent the next several years unwinding what was left.
Markets reopened on Tuesday 26 November. The Dow rose 32.03 points to 743.52, a gain of 4.5 per cent and at that date the largest one-day advance in the average's history. Haupt's customers got their securities. No run on any other brokerage materialised, which was the entire point of the exercise.
What It Cost
Losses across the lenders came to more than $180 million, a sum worth roughly two billion dollars today. Bunge Corporation and Continental Grain took losses; so did banks in New York, Britain and Israel that had discounted Allied paper without ever seeing Bayonne.
American Express carried the heaviest single burden, and the question its board faced was whether it had to. The warehousing receipts had been issued by a subsidiary, and a determined legal defence might have limited the parent's exposure or pushed the subsidiary into bankruptcy. The company chose to pay, on the reasoning that a firm whose entire business was travellers' cheques and charge cards could not be seen to repudiate its own paper. American Express settled claims for about $60 million, funded out of reserves, insurance recoveries and new debt.
Its shareholders were less sanguine. The stock fell from around $65 in October 1963 to about $37 by January 1964, a decline of more than forty per cent in ninety days. Warren Buffett, then running a partnership out of Omaha, spent the following two years buying it, ultimately committing about $13 million — some forty per cent of the partnership's capital — on the judgement that the charge-card and travellers'-cheque franchises were untouched by anything that had happened in New Jersey. Roger Lowenstein's account treats it as the trade in which Buffett first backed a qualitative reading of a business against a quantitative reading of its balance sheet (Lowenstein, 1995). He sold most of the position by mid-1967 at around $92.50.
De Angelis pleaded guilty to four counts and was sentenced to twenty years. He served seven and was released in 1972. Norman C. Miller of The Wall Street Journal, whose reporting on the scheme won the 1964 Pulitzer Prize for national reporting and became a book the following year, found no evidence that anyone at American Express had been in on it; what he found was a chain of people each of whom had relied on somebody else to have looked (Miller, 1965). Time's summary of De Angelis, published as the receivers were counting the tanks, ran under the headline "The Man Who Fooled Everybody."
What the Fraud Left Behind
Field warehousing never recovered. Lenders had treated a warehouse receipt as a document that made physical verification unnecessary, and Bayonne demonstrated that a receipt is only ever as good as the independence of the person who counted the goods — a lesson that would have to be relearned by the banks that financed the Sumitomo copper positions of the 1990s and by everyone who accepted a custodian's statement at face value in the Madoff fraud. The specific structural weakness at Allied was that the custodians drew their wages from the party whose inventory they were certifying, which is the same weakness that recurs whenever an auditor, a rating agency or a warehouseman is paid by the subject of the report.
Commodity regulators drew a narrower conclusion about position limits and margin on oilseed contracts, and the squeeze De Angelis attempted looks in retrospect like an underfunded rehearsal for the Hunt brothers' silver corner of 1980: the same arithmetic, the same dependence on borrowed money, the same discovery that a corner collapses the moment the financing does.
The durable consequence was on the securities side, and it was not one the Exchange had intended to set. By paying Haupt's customers, the New York Stock Exchange established for the first time that a member firm's insolvency was the Exchange's problem rather than the customer's misfortune. That precedent was tested to destruction seven years later, when the paperwork crisis of 1967 to 1970 buried more than a hundred firms and made plain that a voluntary fund assembled by governors over a weekend could not absorb a failure wave. Congress replaced it with a statutory scheme, and Richard Nixon signed the Securities Investor Protection Act on 30 December 1970.
Alice Schroeder's account of the Buffett purchase notes that the men who ran American Express spent that winter arguing about a liability they had not incurred and could probably have escaped (Schroeder, 2008). They paid it because the alternative was to explain to the holders of several billion dollars of travellers' cheques why the company's signature on a piece of paper meant less than they had assumed. That was the real exposure, and no tank at Bayonne had anything to do with it.
Related
Historical records Learn more about our methodology.