Block Trading in a Fire Sale: The Protection That Only Works Alone

Morgan Stanley sold $5 billion of stock in one night and came away largely unscathed. Goldman Sachs sold more than twice that the very next morning and did the same. Credit Suisse, negotiating the identical kind of trade on the identical stocks a little over a week later, lost $5.5 billion. Same tool. Same handful of company names. The entire gap between walking away fine and losing billions came down to who reached the negotiating table first.

Block Trading explained why the tool all three of those banks reached for exists in the first place: a large seller negotiates one price for an entire position privately, so their own order never has to fight its way through the public order book and move the price against itself on the way through. That protection works exactly as designed for a seller acting alone.

On the night of March 25, 2021, five banks needed to sell the same handful of stocks at the same time, and none of them was acting alone.

What block trading protects against, and what it doesn't

A block trade's whole value proposition is narrow and specific: it keeps your selling from telling the market how much you're selling, so the price doesn't run away from you one print at a time the way it would on the open book. Nothing about that design does anything for a second problem — someone else selling the identical stock, through a different desk, an hour later, for a completely unrelated reason of their own. Your privately negotiated price already reflected the size and urgency of your order. It said nothing about theirs.

Prime Brokerage and Total Return Swaps both covered why Archegos Capital Management ended up owing margin calls at Credit Suisse, Nomura, Morgan Stanley, UBS and Goldman Sachs simultaneously, on positions in the same small set of stocks — ViacomCBS and Discovery among them. Each of those banks, as the total-return payer on its own swaps with Archegos, actually held the underlying shares on its own balance sheet as a hedge. When Archegos couldn't meet the calls, each bank had to sell the shares it was individually holding — and every one of them reached for the same tool, on the same handful of names, inside the same general window of days.

Five banks, one exit door

ViacomCBS's own $3 billion public stock offering, priced at $85 a share on March 23, 2021, is what actually started the clock — investor demand came in tepid, the stock slid on the news, and Archegos's swap positions across all five banks began failing their margin calls almost immediately. By the close of trading on March 26, ViacomCBS shares had fallen to $48.23 — a decline of roughly 44% from that offering price in three trading days.

Speed inside that window mattered more than almost anything else. Morgan Stanley, dealing with a comparatively small circle of buyers, sold roughly $5 billion of Archegos-linked stock on the night of March 25 — quietly enough that the market didn't learn about it until well afterward. Goldman Sachs moved the next morning, March 26, executing roughly $10.5 billion in block sales that were, unlike Morgan Stanley's, widely and immediately reported. Both banks came out of the episode largely intact. Credit Suisse and Nomura, still working through their own positions after Morgan Stanley and Goldman had already sold, absorbed losses of roughly $5.5 billion and $2.9 billion respectively. Credit Suisse didn't finish unloading its own block of Archegos-linked stock — the same underlying names — until more than a week later, into a market that Morgan Stanley's and Goldman's own block sales had already pushed lower in the meantime.

None of the five banks did anything a block-trading desk doesn't do every ordinary day. Each one negotiated a single price, for its own full position, away from the public book — exactly the mechanic Block Trading describes. The tool worked, individually, for every bank that used it. It just couldn't do anything about the fact that four other banks were using the identical tool, on the identical stock, in the same collapsing window — because protecting a seller from that was never what the tool was built to do.

The rule that bends under duress, and the one that can't

Here's the part that decided how large a discount a forced seller could actually be made to accept. A US block trade's price sits close to the prevailing market price as a matter of convention and best-execution practice — brokers have every commercial reason to negotiate near the tape — but no US rule caps that discount at a fixed percentage the way Block Trading describes SEBI enforcing in India: every block deal on the NSE or BSE must sit within ±3% of a defined reference price, full stop, with no exception for a seller who has no choice about whether to sell. A US desk facing a genuinely desperate counterparty — a bank that must move billions of dollars of stock today, not eventually — has no equivalent floor stopping the negotiated price from settling wherever the buyer's own appetite for risk says it should. That's the gap a 44% decline in three days moved through.

That gap didn't go unnoticed by the people it would have protected. Weeks after the collapse, in April 2021, foreign portfolio investors began lobbying SEBI directly to loosen India's own block-deal band — citing Archegos by name as the reason they might one day need the room a hard 1% ceiling didn't give them. That was the actual band in force at the time: a single 1% window around the reference price, tighter even than the ±3% Block Trading describes today. SEBI's eventual answer, the same 2025 framework that essay cites, widened the band to 3% for most stocks — a real concession, arrived at four years and a full consultation process later, and still a fixed percentage ceiling rather than the open-ended discretion a US desk was working with the whole time. The lobbying moved the band. It never removed it.

India's own version of a forced, pledge-driven sale shows what happens when that discretion isn't available at all. Under SEBI's Regulation 31, a listed company's promoters must disclose to the exchanges whenever they pledge shares as loan collateral — and disclose again the moment a lender invokes that pledge. In January 2019, media reports that lenders were growing nervous about loans backed by pledged shares in Subhash Chandra's Essel Group sent Zee Entertainment's stock down roughly 30-35% in a single session on January 25 — before a single pledged share had actually been sold, purely on the disclosed size of the pledge and the visible risk that lenders might invoke it. When Bank of Baroda and Credit Suisse — the same Credit Suisse absorbing Archegos losses in New York, sitting as an entirely unrelated lender in Mumbai — did go on to sell part of their pledged Zee shares to recover the underlying debt days later, they sold into the open market rather than through a privately negotiated block window — a sale of that size, at whatever discount actually clearing the position required, would have had nowhere to hide inside a rule that caps the negotiated price at 3% off a public reference point. Essel Group ultimately reached a standstill agreement with most of its lenders, trading a time-bound asset-sale-and-repayment plan for a promise not to invoke the remaining pledges at all.

Run the two failures side by side and the trade-off is exact and symmetric. Archegos's exposure was invisible until the moment it broke — Rule 10B-1 covered why a swap-based position carried no disclosure obligation at all — so the market found out only when Morgan Stanley and Goldman's block trades started printing, by which point the two fastest banks had already gotten most of their stock out the door. Zee's pledge was disclosed from the moment it was created, so the market priced in the danger of a forced sale before any bank actually needed to execute one — and when a bank did sell, India's own price-band rule made sure it couldn't do so quietly or steeply in one print, pushing the sale into the open market's slower, publicly visible price discovery instead. Hidden exposure with a fast, contained exit. Disclosed exposure with a slow, public one. Neither design stops a forced seller from losing money. Each one decides, in advance, whether the market finds out before the damage or during it.

Why this matters for a Business Analyst

Go back to what a block trade actually promises: it protects a seller from their own order moving the price against them. A risk system that checks "will this trade's own size move the market against the desk executing it" and stops there is checking the thing block trading was built to solve — and staying completely silent about a genuinely different risk sitting right next to it: how many other sellers, at other desks, at other banks, are holding economically identical exposure to the same underlying stock, for reasons that have nothing to do with this desk's own trade.

That second question is exactly the one Prime Brokerage already showed no single bank could answer about Archegos's financing. It resurfaces here on the other side of the same collapse, at the point of actually selling the collateral rather than the point of lending against it — a reminder that a control built around one desk's own execution quality can be functioning perfectly, every single time it's checked, while the market-wide concentration around it builds into exactly the kind of five-banks-at-once exit that no single desk's risk model was ever scoped to see coming.

Lighthouse Insight

Go back to the chocolate shop from Block Trading — the private negotiation for the whole box, agreed before the crowd finds out.

That negotiation still works exactly as intended when you're the only person trying to sell a box that morning. It stops working the moment four other people, each holding an identical box for their own unrelated reasons, walk up to four other counters in the same shop within the same hour. Every one of those five negotiations will still produce a single, contained price for its own box — the tool never breaks. What breaks is the assumption underneath it: that the danger worth protecting against was ever just your own order. Morgan Stanley and Goldman got out first and priced their boxes closer to what the shop still believed chocolate was worth. Credit Suisse and Nomura negotiated the same fair deal, for the same tool, one price point later than everyone else — and paid for a difference the tool itself had no way of seeing coming.