SEANBURNS
MM / 06Market MisstepsPUBLICATION: April 2024EXPANDED: September 7, 2026

Total return swaps · Counterparty blindness · Forced liquidation

The Archegos Capital Collapse

Synthetic exposure let one portfolio look like several bilateral relationships until margin calls revealed the system as a single crowded trade.

CENTRAL QUESTION

How did bilateral risk controls fail to see aggregate leverage?

01 / THESIS

The argument

Archegos used total return swaps to obtain economic exposure to concentrated equities while multiple prime brokers each saw only part of the portfolio. The structure separated beneficial ownership, financing, and public visibility. At its peak, the SEC alleged approximately $160 billion of exposure against roughly $36 billion of capital.

The collapse was not just a bad stock selection amplified by leverage. It was a coordination failure across counterparties, a margin failure inside banks, and—according to the later criminal verdict—a fraud and market-manipulation scheme. Once falling prices triggered margin calls, each prime broker’s rational attempt to exit made the common collateral less valuable for every other broker.

Peak exposure alleged by SEC≈$160bn
Credit Suisse loss≈$5.5bn
Primary analytical variableHidden gross exposure

02 / CHRONOLOGY

The sequence

  1. MAR 2020

    The SEC places Archegos at roughly $1.5 billion in value and $10 billion in exposure.

  2. 2020—Q1 2021

    Synthetic positions expand across prime brokers; concentration, volatility, and margin exceptions accumulate.

  3. MAR 22—24, 2021

    ViacomCBS weakness and an equity offering pressure core positions; available cash and trading capacity are exhausted.

  4. MAR 25—26, 2021

    Archegos misses margin calls. Prime brokers liquidate blocks, with sharply different loss outcomes depending on exit speed.

  5. 2024

    A jury convicts Bill Hwang and Patrick Halligan; Hwang later receives an 18-year prison sentence.

03 / MECHANISM

How the failure compounds

01

Total return swap

The client receives the underlying security’s economic return while the dealer finances and typically hedges the exposure.

02

Bilateral opacity

Each dealer underwrites its own relationship. Without a reliable aggregate view, identical exposures can be financed several times.

03

Margin contagion

Declining collateral creates margin calls; failure to meet them forces sales; sales depress the same collateral supporting other loans.

04

Exit-order asymmetry

The first broker to liquidate can preserve capital while slower counterparties inherit a less liquid, lower-priced residual block.

04 / JUDGMENT

What survives the case

The decisive risk control is not a more precise VaR estimate on the visible slice. It is the ability to estimate the client’s total position, common counterparties, wrong-way risk, and liquidation cost under a coordinated unwind. Bilateral data must be treated as an incomplete observation of a system.

Archegos also demonstrates the difference between market risk and counterparty conduct risk. If exposure and liquidity representations are unreliable, conventional stress tests are calibrated on false inputs. The appropriate response is reduced financing capacity, independent verification, and enforceable margin—not a more elaborate model built on the same representations.

EVIDENTIARY LIMIT

Allegations from the 2022 SEC complaint are distinguished from facts established by the 2024 criminal verdict and sentence. Appeals or later proceedings may affect the legal record.

05 / SOURCE DOCKET

Follow the evidence.

VERSION 1.0 · EXPANDED SEPTEMBER 7, 2026 · MATERIAL CORRECTIONS WILL BE RECORDED ON THIS PAGE.