10.11 Case studies

10.11.1Negative WTI, 20 April 2020#

Context The COVID-19 lockdowns cut oil demand faster than producers could cut supply. Inventories rose toward tank capacity, above all at Cushing, the delivery point for NYMEX WTI.

Timeline (sourced)

• As of 17 Apr 2020, about 60 mb of Cushing's roughly 76 mb working storage capacity was reported full (JPT / S&P Global Platts coverage).
• 20 Apr 2020: the May 2020 WTI contract settled at −$37.63/bbl, down $55.90 on the day, the first negative settlement in the contract's history. The June contract settled at $20.43 (S&P Global Platts, 20 Apr 2020).
• 21 Apr 2020: the May contract's last trading day; it settled back above zero.
• MCX's April crude contract expired on 20 Apr. MCX first fixed an interim settlement of ₹1/bbl, then settled at −₹2,884/bbl, the NYMEX settlement converted at the RBI reference rate of 76.6335 ($-37.63 \times 76.6335 = -2{,}883.7$) (Business Standard and ETV Bharat, 21–22 Apr 2020).
• Reported losses on Indian long positions: about ₹418 cr (one report, with about 11,000 contracts open at the close) to over ₹450 cr across more than 100 brokers (another). Brokers threatened litigation. MCX later added provisions for trading at negative prices. The Bombay High Court later upheld the settlement (reported in a broker's market note and by Rawlaw; secondary, date not verified).

Mechanism Proposition 10.2 in its purest form. A long WTI futures position at expiry is an obligation to take barrels at Cushing. Holders without tank space (financial longs, retail-linked products) had to sell to the few participants who could take delivery. With storage nearly full, $u$ (the marginal cost of storage at Cushing) became enormous, so the full-carry upper bound was high, and there was no lower bound for anyone without storage. The price fell to the level at which those with storage were paid to take the oil. The June contract, a month further from the storage constraint, stayed positive: the May–June spread was a measure of how close Cushing was to full. Sources disagree on whether USO's roll added to the pressure; USO had largely rolled out of May before the 20th, according to one account.

The Indian twist: MCX's contract is cash-settled against NYMEX, so Indian traders who could never take delivery inherited the physical squeeze through the settlement formula. Several broker systems could not handle a negative price.

10.11.2The Hunt brothers and silver, 1979–80#

Context Nelson Bunker Hunt and William Herbert Hunt accumulated physical silver and futures through 1979, with associates (including the Hunt-related International Metals Investment Company).

Timeline (sourced) Sources. The primary record is the regulators' own: the SEC staff report The Silver Crisis of 1980 (Division of Market Regulation, October 1982), the CFTC's investigation and its later manipulation case against the Hunts, and Fed Chairman Paul Volcker's statement to the House Subcommittee on Commerce, Consumer and Monetary Affairs of 30 Apr 1980 (FRASER, https://fraser.stlouisfed.org/title/statements-speeches-paul-a-volcker-451/statement-subcommittee-commerce-consumer-monetary-affairs-committee-government-operations-house-representatives-8216/fulltext), which describes the brokerage emergency of 26 March and the bank rescue but gives no prices. The price points below are as commonly reported in secondary accounts (bullion-dealer histories and press); the SEC report was not available to re-check them.

• Silver traded below $10/oz before August 1979 and reached about $34.45 by 31 Dec 1979.
• Holdings controlled by Nelson Bunker Hunt, William Herbert Hunt and International Metals Investment Company rose from about 123 Moz at the end of July 1979 to about 195 Moz by 1 Jan 1980, about half of it in futures (as quoted from the regulators' reports in secondary accounts). An October 1979 CFTC staff report worried that the Hunts held "perhaps as much as half" of the silver traded on the Chicago and New York exchanges (same).
• 7 Jan 1980: COMEX imposed position limits ("Silver Rule 7"); later amendments tightened them, and margins were raised.
• January 1980 peak: reported as $48.70 (close, 18 Jan) and as about $50 intraday. Treat the peak as about $49–50.
• 21 Jan 1980: COMEX restricted silver trading to liquidation orders only; the Chicago Board of Trade did the same.
• Silver fell from about $29.75 on 10 Mar 1980 to close at $10.80 on 27 Mar 1980, "Silver Thursday", when the Hunts failed to meet margin calls (a one-day fall reported as $21.62 to $10.80, −50%). The unsecured debit in Hunt accounts at the broker Bache was reported as about $122 m at 27 March prices.
• A small group of banks then arranged new loans to Placid Oil, a Hunt company, secured on its assets (Volcker's statement of 30 Apr 1980 describes the arrangement while it was still incomplete); the amount is usually reported as about $1.1 bn from a consortium of 13 banks (secondary).
• 1989: in a settlement with the CFTC, Nelson Bunker Hunt and William Herbert Hunt were each fined $10 m and permanently barred from trading on US commodity markets (as reported; check the CFTC order).

Mechanism A corner: accumulate a large share of the deliverable supply and stand for delivery, so that shorts must buy back from you at your price. In storage-theory terms, removing metal from the visible, deliverable stock moves the market to the steep left of the supply-of-storage curve: the convenience yield of a deliverable ounce explodes and the spot price with it. The corner was broken not by arbitrage but by rule changes (position limits, higher margins, liquidation-only trading) that made it impossible to add longs, and by the high price pulling scrap silver into the market. Once forced selling began, leverage converted a price fall into insolvency.

10.11.3The LME nickel squeeze, March 2022#

Context Russia, a major nickel supplier, invaded Ukraine on 24 Feb 2022. Tsingshan, a large Chinese nickel producer, reportedly held a large short position (partly through over-the-counter positions with banks) as a hedge.

Timeline (sourced)

• LME three-month nickel closed at $29,494/t on 4 Mar 2022 (WisdomTree) and at $48,078/t on 7 Mar (S&P Global; WisdomTree gives $48,211), up about 63% in one session.
• 8 Mar 2022: the price hit an all-time high of $101,365/t in early trading, about 250% above two sessions earlier. The LME suspended nickel trading from 08:15 UK time and cancelled all trades executed on or after 00:00 UK time that day (S&P Global; LME statement as reported).
• The value of cancelled trades is reported inconsistently: "over $4 bn" in one report; the claimants' lawyers later told the High Court that $12 bn of trades had been cancelled. Treat the figure as uncertain.
• Trading resumed on 16 Mar 2022, with further disorder (daily limit moves in the following days).
• Elliott and Jane Street sought judicial review. The Administrative Court dismissed the claims on 29 Nov 2023 (Stephenson Harwood; City AM). One source reports that the Court of Appeal upheld the decision in October 2024; verify before relying on it.

Mechanism A short squeeze in a market with thin visible inventory and opaque positioning. Short holders facing margin calls had to buy; prompt metal was scarce; there was no price at which new supply could arrive within hours. In storage-theory terms, prompt convenience yield went vertical, as in the Hunt case but from the short side. The exchange judged that letting the price stand would bankrupt members and threaten its clearing house, and chose to cancel trades. Arbitrage could not cap the price because no one could deliver nickel to LME warehouses within the day.

10.11.4Gold in 2026: the boom, the Warsh crash, and the March fall during a war#

Context Gold rose through 2025 on central-bank buying and ETF inflows (2025 demand a record 5,002 t; WGC). By January 2026 the move had become parabolic, with silver rising faster.

Timeline (sourced; Jan–Mar figures largely secondary)

• Gold all-time high about $5,608/oz (spot; some sources give $5,626) around 28–29 Jan 2026. Silver all-time high about $121.64 (spot). Both are secondary spot quotes; the LBMA benchmark history is licence-restricted.
• 30 Jan 2026: President Trump named Kevin Warsh as Fed Chair. The dollar rose; gold fell 7–12% intraday, silver about 15% in a day and about 30% over 48 hours (several aggregators; exact figures vary). Yahoo's continuous front-month futures closes (checked for this course) show GC=F falling from $5,318.4 (29 Jan) to $4,745.1 (30 Jan), −10.8%, and SI=F from $114.04 to $78.29, −31.3%, in a single close-to-close session (continuous series can contain roll effects, so treat this as indicative). CME raised COMEX silver margins from 11% to 15% (secondary).
• 28 Feb 2026: US–Israeli strikes on Iran begin. 2 Mar: gold spikes to about $5,419 (secondary).
• March 2026: gold falls 11.8%, its worst month since 2013 or 2008 depending on the source (Reuters via MarketScreener; ABC, 27 Mar 2026). About $4,367 by 23 Mar (secondary).
• 16 Sep 2026: Fed hikes 25 bp to 3.75–4.00%, the first hike since July 2023 (FOMC statement, 16 Sep 2026; target range in FRED DFEDTARL/DFEDTARU). The 10-year real yield was 2.68% that day and reached its 2026 high of 2.95% on 5 Oct (FRED DFII10).
• 23 Oct 2026: gold below $4,000; about $3,876–3,881 on 30 Oct. About $4,098 on 6 Nov, roughly 27% below the January high (secondary spot quotes).
• Meanwhile, central banks bought 321 t in Q3 2026 and ETFs added 211 t (WGC, 3 Nov 2026).

Mechanism In GRAM terms (10.17), the dominant driver changed. In 2025 the residual (official buying, event premium) dominated. On 30 Jan a rates-expectation shock (a chair perceived as hawkish) hit a crowded, leveraged, momentum-driven position: margin hikes and options hedging (dealer gamma; secondary reports) amplified the fall. In March the war raised oil, oil raised inflation expectations, and the market priced a more hawkish Fed: real yields and the dollar rose. By Proposition 10.8 the opportunity-cost channel pushed gold down, and it outweighed the safe-haven channel. Liquidation to meet margin calls elsewhere added selling. Rupee gold fell far less because of the weak rupee and, from 13 May, the duty hike (10.5.4).

10.11.5The 2026 Hormuz shock and oil#

Context and timeline The dated timeline of the 2026 war, the oil price path and the policy responses is in M12 · 12.8 Case 6, the course's reference timeline; M11 Case 3 takes the rates angle and M02 · 2.10.4 the return-distribution angle. This case takes the oil-market angle: the term structure and the transmission to Indian fuel retailers. The numbers it needs: the EIA Brent spot price went from $71.32 on 27 Feb to $127.61 on 2 Apr and a 2026 high of $138.21 on 7 Apr (EIA RBRTED, via FRED DCOILBRENTEU), while the highest front-month futures settlement was about $118 (31 Mar; exchange settlement as reported, secondary), and dated Brent was reported at $141.37 on 2 Apr (Reuters; secondary).

Mechanism A supply shock in Kilian's sense: oil up, growth expectations down. Physical prices outran futures (10.8.4), the signature of a convenience-yield spike. The IEA release covered only about 20 days of lost supply, so it barely moved the market along the storage curve. In India the chain of 10.8.5 ran in full, adding a gold-duty hike and a basket reweighting. For the OMCs the shock arrived as marketing under-recoveries that lasted long after the futures peak (10.8.5: still about ₹5/l on petrol and ₹23/l on diesel in September, as reported), because pump prices lagged and excise cuts absorbed only part of it. The 8.7% one-day drop in Brent futures after the US midterms (M12 Case 7) shows how much of the price was a geopolitical risk premium that news could withdraw.