Archegos Capital and the March 2021 Swap Collapse

Key Takeaways

• Archegos went from $1.6 billion of invested capital on $10.2 billion of gross exposure at the end of March 2020 to over $36 billion of capital on over $160 billion of exposure by March 22, 2021, according to the SEC's own complaint, a roughly 22x growth in invested capital in twelve months.
• The concentration was extreme by design. The SEC alleges Archegos's chief risk officer circulated a daily report specifically to keep Archegos's direct share ownership under the 5 percent threshold that triggers public disclosure, shifting to swaps instead whenever a position approached that line.
• The collapse was measured in single trading days, not weeks. Archegos's capital fell from $36.2 billion to $32.7 billion on March 23, then to $16.9 billion on March 24, a one-day loss of 48 percent, then to $9.2 billion on March 25, a further 46 percent loss.
• Credit Suisse's own position tied to Archegos reached $24 billion in March 2021, four times its next-largest hedge fund client and more than half the equity of Credit Suisse Group AG, according to Swiss regulator FINMA's enforcement findings.
• The banks' outcomes diverged completely. Credit Suisse's audited accounts record a CHF 4.8 billion net charge for 2021. UBS disclosed a $774 million pre-tax loss in the first quarter of 2021 alone. Goldman Sachs stated in its own quarterly filing that the episode caused it no loss and no material impact.
• Morgan Stanley disclosed a loss from the episode without ever naming Archegos in any 2021 SEC filing, a fact confirmed by the absence of the word from every filing that firm made that year, which this page found by searching SEC EDGAR's full text index directly.
• The SEC charged Bill Hwang and three Archegos executives with fraud and market manipulation in April 2022, and FINMA closed a two-year enforcement proceeding against Credit Suisse in July 2023 by ordering corrective measures from its legal successor, UBS.

What Was Archegos Capital Management?

Archegos Capital Management, LP was a family office headquartered in New York, run by Sung Kook "Bill" Hwang and exempt from registration as an investment adviser under a rule that excuses firms managing only a single family's money from the disclosure obligations that apply to hedge funds open to outside investors. That exemption was not incidental to the story. The SEC's complaint states that in 2012 Hwang pleaded guilty to criminal insider trading charges connected to his earlier fund, Tiger Asia Management, and settled related SEC civil charges for insider trading and attempted stock manipulation. The guilty plea closed Tiger Asia's hedge funds. Hwang returned investor capital and, by around 2013, converted what remained, his own money, into Archegos.

That history matters because it explains why nobody outside Hwang's own office was managing risk on his behalf, and why a firm that would ultimately control tens of billions of dollars of market exposure filed nothing with the SEC describing its holdings, strategy, or risk profile until after it collapsed. A registered hedge fund answers to outside limited partners, a board, and periodic SEC examination. A family office with one investor answers to nobody until its creditors start calling.

Archegos ran a long and short equity strategy, taking concentrated long positions in individual companies and hedging with short exposure to indexes and baskets, according to the SEC's complaint. Its long book was both leveraged and narrow: leverage ratios generally between 400 and 700 percent, and as high as 1,000 percent, meaning $100 of capital could carry $1,000 of exposure. It held long positions in roughly 100 issuers in total, but typically between a third and half of its entire gross exposure sat in its ten largest names. That is the shape of a portfolio built to compound quickly. It is also the shape of a portfolio that cannot absorb a bad week in its two or three biggest bets.

How Did Archegos Turn $1.6 Billion Into $160 Billion of Exposure?

The growth was fast enough that the SEC's complaint states it in three snapshots rather than a smooth curve. As of March 31, 2020, near the bottom of the COVID-19 market shock, Archegos had approximately $1.6 billion of invested capital on gross exposures of approximately $10.2 billion. By January 1, 2021, that had grown to approximately $7.7 billion of capital on approximately $54 billion of exposure. By March 22, 2021, the day its largest position began to unravel, Archegos had over $36 billion of invested capital on over $160 billion of gross exposure.

Archegos's reported growth in invested capital and gross exposure, from the SEC's complaint against Hwang and Archegos.

None of that growth came from outside capital. Archegos was a single investor's money compounding through concentrated, leveraged bets on a small number of stocks that were, according to the SEC, rising largely because Archegos itself was buying them in size. The complaint alleges Hwang directed traders to "set the tone" through large trades before the market opened, bid up prices with incrementally higher limit orders through the day, and "mark the close" with large trades in the final thirty minutes, all intended to push up the share prices of the names Archegos already held. In a June 2020 text message quoted in the complaint, when an analyst asked Hwang whether a rise in ViacomCBS's share price was "a sign of strength," Hwang answered, "No. It is a sign of me buying," followed by a laughing emoji. A portfolio whose gains are substantially a function of its own buying does not need bad news to fall. It only needs to stop buying.

How Did Total Return Swaps Let Archegos Stay Invisible?

A total return swap is a contract between two parties referencing a security that neither is required to own outright. Under the terms Archegos used, if the referenced stock rose, its bank counterparty owed Archegos the gain; if it fell, Archegos owed the bank. Archegos never had to buy or register a single one of the underlying shares to get the full economic effect of owning them, and, critically, was under no disclosure obligation that applied to a direct owner of stock.

The SEC's complaint alleges this was not an incidental feature of the strategy but its deliberate design. Section 13(d) of the Exchange Act requires an investor who directly owns 5 percent or more of a company's outstanding shares to file a public disclosure. According to the complaint, Archegos's chief risk officer, Scott Becker, circulated a daily internal report that tracked Archegos's cash equity position in each issuer relative to its total shares outstanding, specifically to ensure Archegos never crossed that 5 percent line. Once a position approached the threshold through direct share purchases, Archegos would shift to acquiring the rest of its exposure synthetically, through swaps, which carried no equivalent trigger. The vast majority of Archegos's long exposure was therefore synthetic, which the complaint describes as a deliberate strategy to limit what market participants and counterparties could see of its aggregate holdings.

Under the swap terms, Archegos and its roughly dozen bank counterparties agreed to post variation margin against daily price moves: if a stock fell, the bank could call on Archegos for cash; if it rose, Archegos could call on the bank, though it was not obliged to, and often left that "excess margin" sitting with a bank as a buffer against future declines. Banks generally hedged their own swap exposure by buying the actual referenced shares through specialist trading desks, which meant that once Archegos's positions began to unwind, the shares hitting the market were not Archegos's, which it never owned, but its banks' own hedges, sold to close out contracts with a client who could no longer pay. How that margin call and forced-liquidation sequence works in a normal brokerage relationship, at much smaller scale, is covered in how margin calls and forced liquidation work, and the disclosure threshold Archegos was built to avoid is explained in Schedule 13D versus 13G.

The structural consequence is the one worth carrying forward: each of Archegos's roughly dozen counterparties saw only the swap exposure it personally extended. None of them, on its own, could see that the same client held an even larger position at the bank down the street, because nothing in securities-based swap reporting required Archegos's banks to compare notes on a shared client in real time.

How Concentrated Were Archegos's Positions by March 2021?

In March 2020, Archegos's largest holdings included Amazon and Microsoft, mega-cap stocks with enormous daily trading volume. By March 2021, the SEC's complaint shows those had been replaced by a much narrower and less liquid set: ViacomCBS, two share classes of Discovery, Baidu, Tencent Music Entertainment, GSX Techedu, Vipshop, iQIYI, Farfetch, and Shopify. By Archegos's own internal estimates of each company's public float, cited in the SEC's complaint, its cumulative cash and swap exposure by late March 2021 equated to over 70 percent of GSX Techedu's outstanding shares, over 60 percent of Discovery's Class A shares, over 50 percent of iQIYI, over 50 percent of ViacomCBS, over 45 percent of Tencent Music, and over 30 percent of Discovery's Class C shares.

Archegos did not just hold these positions; it traded them at a scale that itself moved prices. The complaint documents that from November 16, 2020 to January 4, 2021, Archegos's trading in Discovery Class A shares reached or exceeded roughly 25 percent of that stock's daily trading volume on 17 of 33 trading days. Within the narrower stretch of that same window, November 24, 2020 to January 4, 2021, its trading exceeded 30 percent of daily volume ten separate times, with a high of about 40 percent, and dropped below 20 percent on only 7 of 27 trading days. From January 6 to March 24, 2021, its trading in Discovery Class C shares reached or exceeded 25 percent of daily volume on 29 of 54 trading days, and from March 2 to March 23, 2021 its trading in Farfetch reached or exceeded 20 percent of daily volume on 14 of 16 trading days, including a session on March 18 where it surpassed 50 percent of the stock's entire volume for the day.

The price behavior this produced is visible in a single stock's history. ViacomCBS traded around $40 in early January 2021, around $55 by late January, around $70 by early March, above $80 on March 10, above $94 two days later on March 12, and closed at $100.34 on March 22, an increase of roughly 150 percent in under three months that the SEC's complaint states was not supported by any publicly available information. A stock that rises 150 percent on the trading of a single concentrated holder, rather than on anything the company disclosed, is not describing a market's judgment about the business. It is describing that holder's own buying power reflected back at it.

How Did Archegos Keep Its Banks Extending More Credit?

Concentration on this scale should have run into a wall of risk-limit questions from every counterparty, and the SEC's complaint alleges it did, and that Archegos's executives answered those questions with false information rather than accurate ones. Hwang directed his CFO and chief risk officer not to give counterparties the full picture, while simultaneously pushing them to secure more trading capacity and better margin terms.

The complaint's specific example is instructive. Becker is alleged to have told multiple counterparties, repeatedly, that Archegos's single largest position never exceeded about 35 percent of its net asset value, a figure the complaint says was false and that Becker had been taught to repeat by his predecessor regardless of the truth. On March 8, 2021, in a call requesting an additional $2 billion of trading capacity from one bank, Becker allegedly told that bank's credit risk team both that its largest position there was different from, and smaller than, positions held at other banks, and that Archegos's true largest position was still around 35 percent of capital. Both statements were false: the same concentrated names sat at the top of Archegos's book at every counterparty, and the true concentration was substantially higher. When that bank then asked how quickly Archegos could unwind its book in a distress scenario, Becker, after conferring with Hwang's head trader and CFO, told the bank thirty days at 10 to 15 percent of average daily volume, a figure the complaint states Archegos's own internal analysis showed was more than twice too fast, with its largest positions alone requiring months to liquidate at that pace. The bank approved the increased capacity based on those representations, and Archegos immediately used it to buy more of the same concentrated names.

This is the part of the story that a chart of Archegos's exposure cannot show: the growth in the table above was not merely undetected risk. According to the SEC, it was risk that specific individuals were told, in specific phone calls and emails, was smaller and more liquid than it actually was, at the exact moments banks were deciding whether to extend more of it.

What Triggered the Collapse in the Week of March 22, 2021?

The proximate trigger was ordinary corporate finance, not a market shock. After the close on Monday, March 22, 2021, ViacomCBS, then Archegos's single largest long position, announced a $3 billion secondary offering of its shares. Announcing a large new share issuance typically pressures a stock's price because it increases supply, and ViacomCBS fell roughly 10 percent the next trading day, March 23.

Archegos's response, according to the SEC's complaint, was to buy more rather than reduce exposure. On March 23, in an effort to prop up the stock prices and head off further margin calls, Archegos put on approximately $2.6 billion of additional trading positions, mostly further swap purchases in its most concentrated existing holdings. It did not work. By the end of that day Archegos's capital had fallen from $36.2 billion to $32.7 billion, and it faced $2.5 billion of margin calls due the following morning.

Wednesday, March 24 compounded the damage from two directions at once. ViacomCBS priced its secondary offering at $85 per share, well below the $100.34 closing price from two days before, which pressured not only ViacomCBS but the correlated Discovery shares Archegos also held in size. On the same day, U.S. regulators adopted interim amendments implementing the Holding Foreign Companies Accountable Act, which pressured the American depositary receipts of the China-based issuers, including Baidu, GSX, Vipshop, and iQIYI, that made up the rest of Archegos's concentrated book. Two of Archegos's three clusters of exposure were under pressure simultaneously, for unrelated reasons, on the same trading day.

How Fast Did Archegos's Capital Actually Disappear?

What makes this collapse unusual even among leveraged blowups is how precisely it can be dated, because the SEC's complaint reconstructs Archegos's capital at the end of each of the four trading days that ended it.

Archegos's reported end-of-day capital during the collapse, from the SEC's complaint. Percentage changes are calculated from the prior day's closing figure.

By close on March 24, with margin calls of $10.7 billion anticipated the next day and its cash reserves essentially gone, Archegos was, according to the complaint, still telling counterparties it was solvent. When one bank's credit officer asked why Archegos was withdrawing excess margin held there, given the day's losses, Becker allegedly told him Archegos still had $9 billion in cash on hand and that the withdrawal was routine rebalancing, not distress. The complaint states that representation was false and that Becker knew it; Archegos had virtually no cash left. The bank wired $248 million anyway. Hours later, on a call with a different counterparty asking for Archegos's current capital, Halligan is alleged to have instructed Becker over a Bloomberg chat to claim Archegos still had about $20 billion of assets under management, a figure both knew was materially inflated.

Thursday, March 25 brought the final leg down, to $9.2 billion, and after the close Archegos worked with one remaining counterparty to execute a block trade at a discount simply to raise cash against outstanding margin calls, a sale that itself added to the losses. That evening Archegos convened a group call with its largest counterparties to try to coordinate an orderly wind-down rather than a chaotic one. The talks, the complaint states, persisted through Friday and into the weekend, and ultimately failed.

What Happened When the Banks Started Selling?

Once the group call among counterparties broke down without an agreement, each bank was left holding its own hedge shares against a client who had defaulted and could not be relied on to coordinate an exit with anyone else. The SEC's complaint states plainly that Archegos's counterparties ultimately delivered default notices and unwound its positions, including their own hedge positions, and that the resulting failure led to billions of dollars of credit losses spread across those counterparties.

The mechanism matters more than any single day's headline. A bank holding shares as a hedge against a swap has no reason to sell them in an orderly, patient way once the swap counterparty has defaulted; its job at that point is simply to close out its own risk. When several large banks holding the same concentrated names decide independently to do that within the same short window, the resulting selling pressure is large relative to those stocks' normal trading volume for exactly the reason the exposure had grown so large in the first place: Archegos, and by extension its hedging banks, had become an outsized share of the market in each of these names. That is the same dynamic in reverse as the buying that had inflated ViacomCBS to $100.34 two trading days before. FINMA's later review of Credit Suisse's own hedge book, discussed below, confirms that the shares a bank had bought to hedge Archegos's swaps were themselves concentrated in the same few names Archegos held, which is precisely why FINMA found that hedge could not do its job once those names collapsed together rather than independently.

Which Banks Lost the Most, and Which Avoided Losses Entirely?

The same collapse produced almost the full range of possible outcomes across Archegos's roughly dozen counterparties, from the largest single-client loss in Credit Suisse's modern history to a bank that says it lost nothing at all, and the divergence traces directly to how each bank managed its own risk limits, not to how large Archegos's position happened to be with each of them.

Disclosed or stated outcomes by counterparty, from each bank's own SEC filings or, for Credit Suisse, its regulator's enforcement findings. Figures are not directly comparable across banks: some are pre-tax, some net of tax, and reporting currencies differ.

Goldman Sachs's own words are the clearest data point on this page for how differently the same event can land on two banks that were, by the SEC's account, misled by the same client using the same false statements. Its Form 10-Q for the first quarter of 2021 states: "the events related to Archegos Capital, a client with highly concentrated and leveraged positions, did not result in a loss to us or otherwise have a material impact on the net revenues of Global Markets. As part of our risk management process, we identified the risk early and took prompt action consistent with the terms of our contract with the client." Its provision for credit losses that quarter was, in fact, a net benefit of $20 million.

Morgan Stanley's disclosure is notable for what it does not say. This page searched SEC EDGAR's full-text index for the word "Archegos" across every filing Morgan Stanley made in 2021 and found none. Multiple financial news outlets reported at the time that Morgan Stanley disclosed a loss on its earnings call without identifying the client by name, in contrast to UBS and Credit Suisse, both of which named Archegos explicitly and repeatedly in their own SEC filings. This page does not restate the specific dollar figure widely reported for Morgan Stanley's loss, because it could not independently verify that figure against a Morgan Stanley SEC filing or press release this session; the pattern of naming, or not naming, the client is independently verifiable and is reported here instead.

Other banks, including Nomura and Wells Fargo, also disclosed exposure to the episode in their first-quarter 2021 results. This page does not state a dollar figure for those banks because it could not verify one against a primary filing or the bank's own press release this session.

Why Did Credit Suisse Lose Four Times More Than UBS?

The gap between Credit Suisse's loss and every other bank's is not a rounding difference; it is close to an order of magnitude, and Switzerland's financial regulator, FINMA, spent over two years investigating why. FINMA opened enforcement proceedings against Credit Suisse on April 22, 2021, and closed them on July 24, 2023, publishing findings specific enough to function as a case study in how a single relationship manager's business can outrun a bank's own risk controls.

The headline figure from FINMA's findings is the size of the position itself: Credit Suisse's own exposure tied to Archegos reached USD 24 billion in March 2021, a value FINMA states was four times the position of Credit Suisse's next-largest hedge fund client and more than half the equity of Credit Suisse Group AG. A single family office's business had grown, inside one bank, to a scale that rivaled that bank's entire capital base, and FINMA found the bank was not able to adequately manage the risks associated with it.

FINMA's report describes four separate, compounding failures rather than one. First, Credit Suisse's executive board members were never informed of the size or risk of the Archegos relationship; there was no standing requirement that they be told about significant risky relationships on their own initiative. Second, the bank's own risk monitoring regularly flagged that internal limits on the Archegos exposure had been breached, and staff responded not by reducing the exposure but by repeatedly raising the limits to make the breaches disappear on paper, while the underlying risk of loss kept growing. Third, the shares Credit Suisse bought to hedge its swap exposure to Archegos were themselves concentrated in the same handful of names, rather than diversified, so when those names fell together in the final week, the hedge could not do the one job a hedge exists to do. Fourth, in the two weeks before the collapse, when Archegos's positions still carried a high paper value, Archegos exercised its contractual right to call back USD 2.4 billion of excess margin from Credit Suisse, and FINMA found no evidence the bank seriously examined whether it could delay or avoid that payment before wiring it out, draining liquidity from the bank at the worst possible moment.

Credit Suisse's own interim disclosures show how the loss built: a provision for credit losses of CHF 4,430 million in the first quarter of 2021, plus CHF 594 million of further losses in the second quarter as it finished closing out the remaining positions, made up of CHF 493 million of trading losses from market movements during the close-out, a CHF 70 million credit loss provision, and CHF 31 million of severance and professional-services costs. Its later audited 2021 results state the full-year net charge at CHF 4.8 billion, reflecting year-end adjustments to those interim figures. The bank also clawed back approximately USD 70 million of previously granted compensation from individuals involved, and cut its Investment Bank's risk-weighted assets by USD 20.4 billion and leverage exposure by USD 41.5 billion in the second quarter of 2021 as it shrank its prime services business.

FINMA closed its proceedings by ordering corrective measures against Credit Suisse's legal successor, UBS, which absorbed Credit Suisse in the emergency rescue covered in the Credit Suisse-UBS rescue: group-wide limits on the bank's own positions relative to any single client, and compensation criteria that require control functions to assess risk before bonuses are set, rather than after. FINMA also opened a separate enforcement proceeding against a former Credit Suisse manager. The Archegos charge was one of the scandals cited when Credit Suisse's own 2022 annual report later recorded a CHF 4.8 billion charge for the matter, a loss that sat alongside Greensill and other failures as part of the confidence collapse that eventually forced the bank's rescue two years later.

What Did the SEC and FINMA Actually Do About It?

The response to Archegos ran on two separate tracks in two countries, addressing two different failures, and it is worth keeping them apart.

The SEC pursued the fraud, not the risk management. On April 27, 2022, the SEC filed a civil complaint in federal court in Manhattan against Hwang, CFO Patrick Halligan, head trader William Tomita, and chief risk officer Scott Becker, alleging violations of the antifraud and market-manipulation provisions of the federal securities laws. The SEC's own press release states that the U.S. Attorney's Office for the Southern District of New York announced parallel criminal charges the same day, and that the Commodity Futures Trading Commission brought its own parallel civil charges. The SEC's complaint seeks permanent injunctions, disgorgement of gains, civil penalties, and bars against the individuals serving as officers or directors of public companies. "We allege that Hwang and Archegos propped up a $36 billion house of cards by engaging in a constant cycle of manipulative trading, lying to banks to obtain additional capacity, and then using that capacity to engage in still more manipulative trading," SEC Enforcement Division Director Gurbir Grewal said in the agency's own announcement.

FINMA pursued the bank's own controls. Its two-year proceeding against Credit Suisse, described above, ended not with a fine but with an order for organizational change, aimed at the specific gaps its investigation found: no client, anywhere in the group, should again be allowed to grow to a scale unknown to the executive board, and bonus decisions should weigh risk before, not after, the fact.

Neither track produced a new capital or liquidity rule of the kind that followed a bank run like Silicon Valley Bank's 2023 failure. Archegos was not a run on deposits; it was a single client's fraud interacting with several banks' individually inadequate risk controls, and the regulatory response, enforcement against the people responsible and corrective orders against the bank that failed most badly, reflects that different shape of problem.

Were the Warning Signs Visible Before the Collapse?

Unlike a case where the fragility sits in a public filing anyone can read, Archegos's fragility was, by design, not publicly visible at all. That asymmetry is the honest answer to the hindsight question here.

Visible in principle, invisible in practice. The share price behavior of ViacomCBS, Discovery, and the China-based ADRs Archegos held was extraordinary and, in ViacomCBS's case, ran far ahead of anything in the public record about the company. A sufficiently skeptical market participant could have observed that a roughly 150 percent rise in ViacomCBS's share price in under three months, with no supporting news, looked like a stock being pushed up by concentrated buying rather than an improving business, without knowing who was doing the buying or how leveraged that buyer was. That is a real signal, but it is a signal about the stock, not about Archegos specifically, since Archegos's own identity and position size were the things being actively concealed.

Invisible even to sophisticated counterparties. The people best positioned to see the danger, the risk officers at Archegos's own banks, are alleged by the SEC to have been given false answers when they asked the right questions. A counterparty that asked how concentrated Archegos's book was, or how quickly it could unwind, received a number the SEC says Archegos's own internal analysis showed was wrong by a factor of roughly two. Asking the right question is not protective if the answer you receive is a lie designed specifically to defeat that question.

Invisible structurally, regardless of anyone's diligence. Even a counterparty that asked every right question and received an honest answer would still only have known its own slice of Archegos's total exposure. No aggregation mechanism existed, and none was legally required, to show any one bank, or any regulator, the sum of Archegos's position across all dozen of its counterparties in real time. That gap, more than any individual's diligence failure, is why this case belongs in conversations about market structure rather than only about one man's fraud.

Could a Fund Like Archegos Build the Same Position Today?

Two separate questions hide inside this one, and they have different answers.

Could the specific fraud recur, the deliberate lying to counterparties the SEC describes? That is now the subject of a completed enforcement action naming the individuals involved by name, which raises the personal cost of repeating it and gives every prime brokerage risk team a documented playbook of exactly which questions to ask more skeptically. Credit Suisse's successor, UBS, is now operating under FINMA's specific corrective order on client position limits and risk-adjusted compensation, and every other major prime broker had its own internal post-mortem to absorb the same lessons without needing a regulator to order it.

Could the structural gap recur, the fact that no single counterparty or regulator sees a client's aggregate cross-dealer exposure in real time? That is a harder question, because nothing described in the sources for this page indicates that gap has been closed. Each bank still primarily manages the risk it extends directly to a client, and the swap-based route around direct-ownership disclosure thresholds that the SEC alleges Archegos used deliberately is a feature of how those instruments work generally, not a bug specific to Archegos. A determined, sufficiently leveraged client willing to lie convincingly to several banks at once starts from a genuine structural advantage: it is the one party in the relationship who can see the whole picture, precisely because it is the only party present in all of the relationships at once.

Common Myths About the Archegos Collapse

"It was a hedge fund." Archegos was legally a family office managing only Bill Hwang's own money, specifically exempt from the investment adviser registration and disclosure obligations that apply to hedge funds managing outside capital. That exemption, not a regulatory oversight, is why it filed nothing publicly describing its holdings before it collapsed.

"The banks didn't know who Archegos was." They knew exactly who their own counterparty was; several of Archegos's largest banks had lending relationships with the firm for years. What they did not know, according to the SEC, was the true size of Archegos's position, because Archegos's own executives are alleged to have repeatedly misstated it to them, and because no bank could see what Archegos held at any other bank.

"Every bank lost roughly the same amount." The range ran from Credit Suisse's CHF 4.8 billion audited net charge to Goldman Sachs's own statement, in a public filing, that the episode caused it no loss and no material impact. The difference traces to each bank's own risk management, documented in detail for Credit Suisse by its regulator, not to the size of Archegos's exposure, which was large at multiple banks simultaneously.

"This was a stock market crash." The S&P 500 was largely unaffected by the week that destroyed Archegos; this was a handful of individual, relatively less-liquid stocks falling sharply because their single largest concentrated holder was forced to liquidate, and because that holder's hedging banks then sold their own hedge positions in the same names. It is a counterparty and concentration story, not a broad-market story.

"Regulators fixed the problem afterward." The SEC's case addressed the individuals who allegedly committed fraud, and FINMA's order addressed Credit Suisse's specific internal controls. Neither closed the structural gap that let the fraud work in the first place, which is that no counterparty or regulator aggregates a client's total cross-dealer swap exposure in real time.

What a Reader Can Actually Carry Forward

The temptation with a case this dramatic is to file it away as "don't use excessive leverage," which is true but not the useful lesson, since almost nobody reading this page runs a family office with a dozen prime brokers. The more transferable ideas sit one level below that headline.

What generalizes

• Synthetic exposure hides economic reality, not just accounting reality. A total return swap gives the same economic outcome as owning the stock, without triggering the disclosure rules built around actual ownership. Any instrument that separates legal ownership from economic exposure, whether a swap, a contract for difference, or a structured note, can be used to build a position that looks smaller to outside observers than it actually is. Reading a company's Schedule 13D and 13G filings tells you who owns the stock directly; it does not tell you who has bet on it synthetically through a bank.
• Counterparty risk is a portfolio-level question, not a relationship-level one. Every one of Archegos's dozen banks could truthfully say its own individual exposure to the firm looked manageable in isolation. None of them could see, or was structurally positioned to see, that the sum across all of them was a bet the client could not survive losing. Any institution extending credit or margin to a client that also borrows elsewhere is exposed to a risk it cannot fully measure on its own.
• A hedge that is concentrated in the same names as the risk it hedges is not really a hedge. FINMA's finding on Credit Suisse's own hedging book, shares bought in the same handful of stocks Archegos already dominated, is the clearest lesson here: diversification is not a nice-to-have layered on top of a hedge, it is the property that makes a hedge actually work when the underlying risk materializes.
• Rapid, unexplained price appreciation in a name with one dominant holder is a signal worth investigating, even without knowing who that holder is. ViacomCBS's rise had no supporting news. The absence of a fundamental explanation was itself informative, whether or not a given investor could have identified Archegos as the cause.

What does not generalize

• The specific fraud. Repeated, allegedly deliberate lies to multiple bank counterparties about position size and liquidation timelines is not a passive risk that materializes on its own; it required active, sustained deception the SEC now treats as a completed enforcement case with named defendants.
• The four-day collapse timeline. That speed reflected swap contracts with daily variation margin calls on a book with leverage as high as 1,000 percent. A less leveraged, less synthetic portfolio does not unwind on the same clock.
• Credit Suisse's specific failures. Not every bank that lent to Archegos had an executive board kept in the dark, or a hedge concentrated in the same names as the risk, or a habit of resolving limit breaches by raising the limit. FINMA's findings describe one bank's control failures in particular, not an industry-wide pattern; Goldman Sachs's own filing is the direct counterexample.

The one question worth asking of any counterparty relationship after reading this case is not "how large is my exposure to this client," but "how would I know if this client's exposure to everyone else had also grown dangerously large, and what would I do differently if I did know?" For Archegos's banks, according to the SEC and FINMA, the honest answer in early 2021 was that most of them would not have known, and at least one of them, when it suspected something was wrong, chose to raise the limit rather than ask harder questions.

References

Every figure on this page was verified against the following sources, each retrieved on 26 August 2026:

• SEC: Complaint, SEC v. Sung Kook (Bill) Hwang, Patrick Halligan, William Tomita, Scott Becker, and Archegos Capital Management, LP: Archegos's invested capital and gross exposure at each snapshot date, the family office exemption and Hwang's 2012 guilty plea, the leverage ratios and issuer concentration, the Top 10 Holdings share table, the float-percentage figures, the trading-volume-dominance figures for Discovery Class A, Discovery Class C, and Farfetch, the ViacomCBS price history and the $100.34 close on March 22, the misrepresentation allegations against Becker and Halligan, and the complete March 22-25, 2021 capital collapse chronology including the $2.6 billion of additional positions, the $2.5 billion and $10.7 billion margin calls, the $248 million wire, and the block trade with CP6.
• SEC: SEC Charges Archegos and its Founder with Massive Market Manipulation Scheme: the April 27, 2022 filing date, the individuals and entity charged, the parallel criminal charges from the U.S. Attorney's Office for the Southern District of New York and civil charges from the CFTC, and the quoted statement from SEC Enforcement Division Director Gurbir Grewal.
• FINMA: Credit Suisse: FINMA Opens Proceedings in Archegos Case and Confirms Ongoing Proceedings in Greensill Case: the April 22, 2021 opening of enforcement proceedings against Credit Suisse.
• FINMA: Archegos: FINMA Concludes Proceedings Against Credit Suisse: the July 24, 2023 closure of proceedings, the USD 24 billion Credit Suisse position size and its comparison to the next-largest client and to Group equity, the finding that the executive board was not informed, the finding on repeated limit-breach overrides, the finding on concentrated rather than diversified hedging, the USD 2.4 billion payout in the two weeks before the collapse, and the corrective measures ordered against UBS as Credit Suisse's legal successor.
• Credit Suisse Group AG: Form 6-K, Second Quarter 2021 Results, Exhibit 99.1: the CHF 4,430 million first-quarter 2021 provision for credit losses, the CHF 594 million of further second-quarter losses and its CHF 493 million trading loss, CHF 70 million credit provision, and CHF 31 million operating-expense components, the approximately USD 70 million compensation clawback, and the USD 20.4 billion risk-weighted-asset and USD 41.5 billion leverage-exposure reductions in the Investment Bank.
• Credit Suisse Group AG: Annual Report on Form 20-F for the Year Ended December 31, 2022: the audited full-year 2021 net charge of CHF 4.8 billion for the Archegos matter.
• UBS Group AG: Form 6-K, First Quarter 2021 Earnings Call Transcript: the $774 million pre-tax loss and $434 million net loss UBS disclosed for the first quarter of 2021.
• Goldman Sachs Group, Inc.: Form 10-Q for the Quarterly Period Ended March 31, 2021: the statement that Archegos did not result in a loss or material impact, and the $20 million net benefit in the provision for credit losses that quarter.
• SEC EDGAR full-text search system, searched directly for this page: the confirmed absence of the word "Archegos" in any Morgan Stanley SEC filing dated in 2021.

Figures deliberately not stated. This page gives no dollar figure for Nomura's, Morgan Stanley's, or Wells Fargo's individual losses from Archegos, no per-share or intraday price data for ViacomCBS, Discovery, or the other concentrated holdings beyond the specific closing levels stated in the SEC's complaint, no total industry-wide loss figure beyond FINMA's own characterization of Credit Suisse's loss as the largest "of over USD 5 billion," and no detail of Bill Hwang's criminal trial, verdict, or sentence, because no primary or institutional source verified for this page this session supplied them. Widely reported figures exist in the financial press for several of these, particularly Nomura's and Morgan Stanley's quarterly losses and the outcome of Hwang's criminal trial, but this page states only what its own source list can support; a figure repeated frequently in secondary coverage is not the same thing as a figure this page has verified.

Method note: CHF figures are stated as Credit Suisse itself reported them, in Swiss francs, without an independent currency conversion by Swoopr Investment; FINMA's own characterization of the loss as "over USD 5 billion" is quoted directly rather than derived from a Swoopr-calculated exchange rate. Percentage changes in Archegos's daily capital are calculated by Swoopr Investment from the dollar figures stated in the SEC's complaint. UBS's $774 million and $434 million figures are for the first quarter of 2021 only; UBS indicated on the same earnings call that it expected a further, smaller loss in the second quarter as it exited its remaining position, but this page does not state that second-quarter figure because it was not independently verified against a UBS filing this session.

This is educational content about a historical episode. It is not investment advice, it is not a forecast, and nothing here should be read as a claim about the current risk management practices of any bank named on this page.

Frequently Asked Questions