Archegos Capital Management and the hidden swap book

A family office, not a hedge fund

Archegos Capital Management was not a hedge fund in the way most people mean the term. It was a family office: a private vehicle managing money for one person, its founder Bill Hwang, and nobody else. That single fact took it outside many of the disclosure and leverage rules that apply to registered investment managers. Hwang had started the firm in 2013 with about US$500 million, after his earlier fund, Tiger Asia Management, closed following an insider-trading settlement. By 2020, Archegos was reported to be managing about US$10 billion.

Ten banks, one hidden book

The danger was not the size of that capital. It was how the firm built exposure on top of it. Instead of buying shares outright, Archegos entered total return swaps with at least ten different banks, one private, bilateral contract at a time, each referencing the same handful of stocks, including ViacomCBS and Discovery. Each bank could see only its own slice of the position; none of them could see the whole book. Between March 2020 and March 2021, the fund's equity grew from about US$1.5 billion to more than US$36 billion, while its gross exposure across all its swap counterparties grew from about US$10 billion to more than US$160 billion, according to the US Securities and Exchange Commission's later civil complaint.

The ViacomCBS offering that broke the position

On 22 March 2021, ViacomCBS attempted a US$3 billion share offering, led by Morgan Stanley and JPMorgan. It struggled to find buyers at the intended price, and the stock began to fall. Because Archegos held such a large, concentrated swap position referencing ViacomCBS, that fall triggered margin calls from several of its banks at once, and Archegos could not meet them. On 25 and 26 March, the banks began liquidating the shares they held as hedges against their swap exposure. ViacomCBS and Discovery both closed down more than 27% on 26 March alone, with Viacom down more than 50% and Discovery down 45% over the week.

The losses land on the banks

The losses fell on the banks, not on outside investors, because it was the banks that had ended up holding the underlying shares as hedges. Credit Suisse disclosed an initial loss estimate of US$4.7 billion on 6 April 2021, later revised to US$5.5 billion. Combined losses across the banks involved, including Nomura, UBS and Morgan Stanley, exceeded US$10 billion. Goldman Sachs, which exited its positions fastest, disclosed no material loss.

A total return swap is a contract, not a share purchase, and the distinction is the key to the whole case. Archegos, as the total return receiver, paid each bank a financing rate plus any fall in the referenced stock's price. In exchange, the bank paid Archegos any rise in the price, plus dividends. The bank typically hedged its own side of the deal by buying the actual shares, so the bank ended up holding real stock, while Archegos held only a claim on how that stock performed.

In the United States, a direct shareholder above certain size thresholds must file public ownership disclosures. A total return swap holder does not own the shares, so those disclosure trip-wires do not apply in the same way. Archegos's exposure was real, and at its peak enormous, but it was legally invisible to the market as a whole. It was also invisible to each individual bank, which saw only its own bilateral contract and had no way of knowing what Archegos was doing with the other nine or so banks at the same time.

This case maps directly onto three ideas from the course: notional exposure versus disclosed ownership (Lecture 1), how margin transmits credit risk into liquidity risk (Lecture 2), and how a swap actually works as a bilateral contract (Lecture 7).

Notional versus disclosed ownership. Because Archegos never owned the shares directly, its position sat outside the ownership-disclosure rules that apply to a direct shareholder crossing the same size thresholds. The exposure existed, and it was large, but no single party, not the market, not any individual bank, could see its true aggregate size. The course's basic distinction between notional and gross market value is stretched to an extreme here, with a real notional scattered across ten or more separate, mutually invisible contracts.

Margin as the transmission mechanism from credit risk to liquidity risk. Each bank required Archegos to post margin against its own swap position. When the referenced stocks fell, several banks called margin at once. Any single call looked manageable. Together, they were fatal, because the same underlying cash was being demanded by multiple counterparties, none of whom knew the others were making the same demand at the same time. A position that looked adequately collateralised to one bank was drastically under-collateralised in aggregate. This is the margin-liquidity spiral from Lecture 2: a moderate price move becomes an unmanageable cash problem the moment several lenders pull the same lever simultaneously.

Concentrated leverage. Total return swaps let Archegos gain stock-like exposure while posting only a fraction of the capital a cash purchase would require. The US Commodity Futures Trading Commission's complaint states Archegos told its banks that its largest position was about 35% of net asset value, when the true figure was about 70%. Combined with gross exposure running at several multiples of fund equity, a fall of a few tens of per cent in a small number of stocks, not a market-wide crash, was enough to erase the fund's entire equity within days.

Precision on who was long what matters here, since it is easy to get backwards. Archegos was economically long the referenced stocks through its swaps: it gained if they rose and lost if they fell. The banks, in hedging their own side of each swap, ended up net long the same stocks in their own trading books. When Archegos defaulted on its margin calls, the banks were left holding those hedges outright, with no counterparty left to offset them, and had to sell into a falling market. That forced selling drove the sharp falls in ViacomCBS and Discovery in the last week of March 2021.