The Company That Insured Itself Against Technological Change
Late in 1980, the U.S. Securities and Exchange Commission (SEC) was still waiting for audited 1979 financial statements from a company most people today have never heard of. When the annual report (Form 10-K) finally arrived — on October 16, 1980, more than nine months past fiscal year-end — it reported a net loss of $433.3 million — about $2.0 billion in 2026 dollars — one of the largest corporate losses recorded to that date.
Eighteen months earlier, Itel Corporation of San Francisco had been the largest independent computer lessor in the world, with a $1.7 billion lease book — some $8 billion today — second only to IBM itself. The failure mechanism was not fraud and not demand. It was a residual-value assumption meeting a new generation of hardware. This is the sequence, from the contemporaneous record.
| Date | Event | Figure | 2026 $ |
|---|---|---|---|
| 1975 | Lloyd’s of London begins insuring Itel against early lessee returns — "in essence, Itel was insured against technological change." | ||
| 1978 | Record profit reported on a $1.7B lease book — the largest IBM-compatible lease portfolio after IBM itself. | $47.2M profit | ≈$243M |
| late 1978 | Speculation about IBM’s next generation halts customer purchases. The record profit fails its audit and is restated — explicitly for a residual-value writedown. | → $21.5M | ≈$111M |
| Jan 30, 1979 | IBM announces the 4300 series: ~3.2× the performance of the 370/138, up to 30% cheaper. Lessees begin canceling. | ||
| Q1 1979 | First-quarter loss — small. The repricing has happened; the recognition has not. | −$4.4M | ≈−$20M |
| H2 1979 | Cancellations and lease-ends convert January’s repricing into booked losses, one contract at a time. | ~−$370M | ≈−$1.7B |
| FY 1979 | Full-year net loss; workforce falls ~7,000 → 1,400; Data Products sold to National Semiconductor. | −$433.3M | ≈−$2.0B |
| Oct 16, 1980 | The audited 1979 10-K is finally filed — more than nine months past fiscal year-end. | $1.2B debt | ≈$4.9B |
| Jan 19, 1981 | Chapter 11 — at the time among the largest and most complex bankruptcies in U.S. history. Lloyd’s computer-leasing losses reach ~$340–400M market-wide. | ~$340–400M | ≈$1.5B |
| 1983 | Reorganization plan confirmed. |
Sources for every row are linked at the end of the note. 2026 dollars: Bureau of Labor Statistics (BLS) consumer price index for urban consumers (CPI-U), figure-year annual average deflated to May 2026 (index 335.1; 1978 ×5.14, 1979 ×4.62, 1980 ×4.07).
01 The Lessor
Itel is best understood as a finance company that lived inside the computer industry. Its core business bought IBM's own machines — System/360s, then 370s, mostly with borrowed money — and leased them to corporate users at rates below IBM's own rental terms: a bet that the machines would hold value longer than IBM's pricing implied. The leases were leveraged — lenders funded most of the purchase price against the machine and its lease payments, and the lease revenue largely serviced that debt — so Itel's own profit was concentrated in the tail of each deal: the residual value, what the machine would be worth when it came back. The accounting front-loads on the same assumption: much of the profit is recognized early, against the residual. Itel's residual assumptions were, by the standard of the day, aggressive, and the company's answer to the obvious objection was an insurance policy. From 1975, underwriters at Lloyd's of London insured Itel against lessees returning machines early. As it was put at the time: in essence, Itel was insured against technological change.
The model worked spectacularly while it worked. Itel grew into the largest non-manufacturer lessor of IBM-compatible equipment, shipped its own IBM-compatible processors, and reported a record profit for 1978: $47.2 million.
02 The Market Whispers Begin
IBM's next generation was announced on January 30, 1979 — but it was telegraphed well before that, and the market acted on the telegraphing. Through late 1978, speculation about the coming machines halted customer purchases across the industry. Itel's lessees were already behaving as if the 370 era was ending while Itel's books still carried 370 residuals at levels that supported a record year.
The record year did not survive its audit. The $47.2 million profit was restated to $21.5 million — a write-down taken explicitly against residual values, with the auditor declining to certify the original figure. The customers had repriced the machines before IBM confirmed why; the balance sheet caught up under audit.
That refusal is worth pausing on. The auditor had no price index that contradicted the marks — none existed — and no announcement to point to. Every input was gray: the residuals rested on dealer quotes and recent transactions, trade sentiment on used 370s was divided even months after the 4300 arrived, and the client was reporting the best year in its history. Certifying was the path of least resistance. Refusing meant striking more than half from a record profit — a swing of roughly $130 million in 2026 dollars — on judgment alone. It was also the only recognition in the entire episode that arrived by choice rather than by force: every markdown after it was extracted by events, one contract at a time.
03 January 30, 1979
IBM announced the 4300 series — the 4331 and 4341 — on January 30, 1979: roughly 3.2 times the performance of the 370/138 at up to 30% less cost. The announcement repriced every 360/370 lease book in the industry overnight. Itel's customers began canceling leases and returning machines — to a lessor whose balance sheet said those machines were worth far more than anyone would now pay to use them.
Itel's first quarter of 1979 showed a loss of only $4.4 million. The full year lost $433.3 million — roughly $370 million of it landing in the second half. Nothing about the 4300 got worse between March and September; losses reached the books as individual contracts produced events — cancellations, machines returned into a market that had moved, leases expiring against residuals booked years earlier. The market repriced the machines in one day; the books caught up one contract at a time.
04 The Unwind
The arc is legible in the headlines alone. By midsummer, The New York Times had the company "at the brink again" (July 29, 1979). Headcount went from roughly 7,000 to 1,400. The Data Products division was sold to National Semiconductor in October 1979. The Lloyd's policies — the insurance against technological change — turned into the largest computer-leasing insurance loss on record, some $340–400 million across the market, and the claims paid slowly: by August 1980, Itel had collected $8.4 million of the $21.5 million then claimed. The audited 1979 statements took until October 16, 1980 to file, with the company roughly $1.2 billion in debt and negotiating a restructuring as it filed; in December the chief executive was "painting a bleak picture". Chapter 11 followed on January 19, 1981 — at the time among the largest and most complex bankruptcies in American history. By February the Times' framing was already posthumous: "trying to salvage the little that's left." The reorganization took until 1983.
05 What Nobody Could Know
There was a great deal nobody knew in 1979, and no way to know it. Residual values were set at origination and rested on judgment thereafter — management's estimates, tested once a year by the auditor against dealer quotes and recent transactions, because no published price series for used IBM mainframes existed. People disagreed in good faith right to the end: in mid-1979 — months after the 4300 announcement, weeks before roughly $370 million of second-half losses landed — some dealers were predicting a used-370 shortage. Nobody could check anyone's number. It was an unforgiving time to be aggressive with structures whose entire profit lived in the residual.
06 The Same Trade, Structured Differently
Itel's exposure was built into the design of its leases: the economics rested on what the machines would be worth at lease-end, profit was recognized up front against those residual positions, and the residual risk was insured rather than eliminated. The equivalent financings of 2026 — credit secured by GPU fleets — are structured far less aggressively on exactly this point. Of the 65 GPU-collateralized instruments on the CCIR credit ledger, every one that discloses its amortization mechanics (17) schedules the collateral's value to zero over the life of the loan, and none references a market residual. Equipment-finance markets have rarely stayed at that conservative end for long — competition for deals historically pushes structures toward partial amortization and residual dependence. So far, this one has not taken that step. The 1979 book carried the machines' future value as an asset; the 2026 books carry it at zero.
07 Sources
- The New York Times, July 29, 1979 — Itel Corporation Is at the Brink Again
- The New York Times, October 17, 1980 — Delayed Itel Report Puts '79 Loss at $433.3 Million
- The New York Times, December 19, 1980 — Itel Chief Paints a Bleak Picture
- The New York Times, February 15, 1981 — Itel in Bankruptcy: Trying to Salvage the Little That's Left
- United Press International (UPI), October 16, 1980 — Itel files its delayed 1979 report
- UPI, January 19, 1981 — Itel files under Chapter 11
- The Washington Post, October 17, 1980 — Troubled Itel Lost $443 Million in '79 (headline figure differs from the 10-K's $433.3M)
- The Christian Science Monitor, December 30, 1980 — Itel struggles to stay on its feet in technological storm
- Background: Encyclopedia.com — Itel Corporation; FundingUniverse — History of Itel Corporation; Fortune, October 8, 1979 (print).