What Happened to Archegos Capital? The $10 Billion Collapse

Archegos Capital Management was the New York family office of Bill Hwang. It defaulted on 26 March 2021 after margin calls of more than $13 billion that it could not pay. Using total return swaps with about ten banks, it had built concentrated, leveraged positions in a handful of US media and Chinese internet stocks. When ViacomCBS sold new shares and its stock fell, the banks' fire sales cost them more than $10 billion.

The rise was as fast as the fall. According to the SEC, Archegos had about $1.5 billion of capital and about $10 billion of exposure in March 2020. A year later, at its peak, it had more than $36 billion of capital and about $160 billion of exposure. That is a roughly 24-fold rise in capital in twelve months. Within a week it was gone, and in July 2024 a jury convicted Hwang of fraud and market manipulation.

Who was Bill Hwang?

Hwang learned the trade at Julian Robertson's Tiger Management, covering Asian equities from the mid-1990s. When Tiger closed its outside funds in 2000, Robertson seeded former analysts, and Hwang launched Tiger Asia Management in 2001 with about $25 million, as reported.

In December 2012 the SEC charged Hwang and Tiger Asia with insider trading in Chinese bank stocks and with a scheme to manipulate Hong Kong-listed bank shares. The SEC settlement totaled about $44 million, and Tiger Asia Management pleaded guilty to wire fraud in a parallel criminal case. Hwang returned outside capital and turned the firm into a family office, renamed Archegos in early 2013, as reported.

As a family office managing only Hwang's money, Archegos had no outside investors, no SEC adviser registration and no Form PF filing. Within eight years it held more stock exposure than most hedge funds in the world.

How did Archegos build its positions?

Through total return swaps. In a swap, the bank buys the shares as its own hedge and pays the client the stock's total return in exchange for financing costs and margin. The client gets the economics of owning the stock without legal ownership.

That structure had three consequences, and together they made the collapse possible.

The banks competed for the business, and that competition lowered the terms. According to the Paul Weiss report to Credit Suisse's board, published in July 2021, Credit Suisse agreed in 2019 to cut Archegos's default swap margin to 7.5% after Archegos said rivals offered better terms. At 7.5% initial margin, a position is leveraged about 13 to 1, and a 7.5% fall in the stock uses up the margin. Credit Suisse's revenue from the client in 2020 was reportedly about $16–17 million. It bought a loss of about $5.5 billion.

Gains that were partly self-made

Prosecutors argued, and the jury accepted, that Archegos bought swaps "without any economic purpose other than to artificially and dramatically drive up the prices" of its holdings, in the SEC's description of the case. The trading lesson does not depend on the verdict. When one buyer is a large share of the demand for a stock, its mark-to-market profit is partly its own price impact. The gain reverses when the buyer becomes the seller.

What triggered the Archegos collapse?

On Monday 22 March 2021, ViacomCBS announced an offering of about $3 billion of common and mandatory convertible preferred stock, with the common priced at $85, as reported. Companies sell shares when their price is high, and ViacomCBS was near a record. Its board did not know that much of the recent buying had come from one family office.

The shares fell. During that week, the banks' margin calls on Archegos rose to "over $13 billion," according to the CFTC. Archegos could not pay.

The banks met on the evening of 25 March to discuss an orderly unwind and reached no binding agreement. The same lenders that had competed to lend to Archegos now competed to sell. Those that sold first, Morgan Stanley and Goldman Sachs, then Deutsche Bank and Wells Fargo, lost little or nothing. Credit Suisse and Nomura sold late and lost most: about $5.5 billion and about $2.9 billion. The damage to Credit Suisse's capital and reputation came two years before its 2023 rescue by UBS, which the finance book's profile of Credit Suisse covers.

Why was Archegos's leverage so dangerous?

On the SEC's figures, Archegos ran about 6.7 times exposure to capital in 2020 and about 4.4 times at the 2021 peak. That is high but not extreme by hedge fund standards. Three things made it far more dangerous than the ratio suggests.

Concentration. With one position at about 70% of net assets, a 50% fall in that one stock costs about 35% of capital.

Illiquidity. Several positions were reportedly large multiples of the stocks' daily trading volume. They could not be sold without collapsing the price.

Fragmentation. With ten lenders, a bank holding a tenth of the exposure could believe it faced a modest client.

The book's chapter on sizing, leverage and survival works through a stylized version with invented round numbers. A trader with $100 of equity holds a long book at 4.4 times gross exposure, either spread over 20 names or with about 70% of equity in one name.

Stylized; invented numbers, not Archegos's positions; prices move in one step with no trading in between.

The leverage ratio gives the same answer for both books. The single-name loss does not. That is why the CFTC's allegation that Archegos told banks 35% when the truth was 70% mattered as much as the leverage.

Why a low margin made it worse

The same example shows what the 7.5% margin did. Take the concentrated book after its largest name halves. Its equity is $64.8 against $405 of remaining exposure. At a 20% margin, the lender calls for $16.2 at once, while the trader still has most of its equity and can sell into an orderly market. At 7.5% the lender calls for nothing. The call comes only after a second shock, when equity is $9.4 against a requirement of $26.2, a shortfall the trader cannot meet. A low margin does not lower the risk. It removes the early warning and moves the forced sale to the worst moment.

The study of leverage, margin, funding and liquidation calls this the "margin in the stress, not margin today" rule.

How does Archegos compare with other blow-ups?

The LTCM parallel is the closest. In both cases each lender sized its exposure on the slice it could see, and in both the lenders became competing sellers when the end came.

What happened to Bill Hwang?

On 27 April 2022 the SEC, the CFTC and federal prosecutors in Manhattan charged Hwang and others. Hwang was convicted on 10 of 11 counts on 10 July 2024 and sentenced to 18 years on 20 November 2024; prosecutors had sought 21. Archegos's CFO, Patrick Halligan, was sentenced to eight years in January 2025. The head trader and the chief risk officer pleaded guilty and testified for the government. Hwang's appeal was pending in late 2026, as reported.

The banks paid as well. In July 2023 the Federal Reserve fined UBS, as Credit Suisse's successor, $268.5 million, about $387 million together with the UK regulator's penalty. The Fed called Credit Suisse's practices "unsafe and unsound" despite "repeated warnings." The Paul Weiss report found persistent limit breaches that were discussed but not acted on.

The rules changed too. After the collapse the SEC adopted an anti-fraud rule for security-based swaps in 2023 and shortened the deadlines for ownership disclosures in October 2023.

Archegos in the four factors of the edge equation

Edge was unproven. Archegos ran a concentrated, discretionary stock book with no public record. Its gains came largely from leverage and, on the jury's view, from its own buying.

Survival was the deciding factor. Exposure of 4.4 to 6.7 times capital, one name at about 70% of net assets, illiquid positions and margin that ten banks could call at once meant a fall in one stock could end a fortune of more than $36 billion in days.

Execution failed because the exit at marked prices was impossible. The counterparties, not Archegos, executed the exit.

Discipline was absent: adding to winners in a rising market and, on the verdict, lying to lenders.

The rules and the weather mattered as well. Disclosure gaps for swaps and family offices let the leverage hide, and a routine corporate decision, ViacomCBS selling shares into its own rally, was the trigger. For more cases in which survival went to zero, see our list of the biggest trading losses in history, the chapter on blow-ups, and our overview of how trading firms make money.

The takeaway

Archegos was not brought down by an extreme leverage ratio. It was brought down by concentration, illiquidity and leverage split across lenders who could not see one another, which turned an ordinary share offering into the end of a $36 billion fortune within a week. The banks' low margins did not reduce the risk. They removed the warning until it was too late. The lesson for any trader is to measure the largest position and the time it would take to sell it, not just gross leverage. The full case is told in Chapter 16 of The Edge Equation, and the Archegos company profile holds the dated record.