⛏️ RIOT: A Put Sold at 10:40 That Was Already Underwater by the Close

📅 2026-08-13 | 🤝 Multi-Leg Auction + Floor Block

---

🎯 The Quick Take

Two separate prints in Riot Platforms today, pointing the same direction — and one of them has already gone wrong inside the same session.

At 09:42:10 a desk bought 9,480 October-2 $24 calls for ≈$1.52M. At 10:40:41 another sold 10,468 August-28 $20.50 puts for ≈$1.54M. Bullish call buying further out, put selling nearer in.

The put is the story. It was sold with Riot at $20.30. The stock closed the session near $19.245 — down ≈5%. That put is now roughly $1.26 in the money against $1.47 collected, leaving about 21 cents, or 1.1%, of cushion. A position that looked comfortable at breakfast is nearly offside by lunch.

---

🏢 Company Overview

Riot Platforms mines bitcoin and is pivoting part of its power capacity toward AI and high-performance-computing tenancy — a transition that has re-rated several miners over the past year.

---

💰 The Trades, in Plain English

The call printed as a floor block; the put as a multi-leg auction — a worked order exposed for price improvement, not a lit sweep. Neither took liquidity, so both side labels are reported rather than tape-proven.

---

✅ RESOLVED — Both Legs Opened, Including the One We Could Not Baseline

Updated 2026-08-14 pre-market. Resolving OPRA snapshot timestamped August 14 (reflects the August 13 close, after this print); baseline is the August 13 snapshot (reflects the August 12 close, before this print).

The put leg is settled and needs no discussion: 28 contracts became 10,508, contract for contract.

The call leg resolved better than we expected it to. Yesterday we said we had no baseline and would not guess. The reason is now clear: OPRA published no open interest for that contract on any date we queried, going back to July 1 — and the neighbouring Oct-2 $24 call's sibling strike ($25) also appears for the very first time in the August 14 file. The October 2 weekly series simply had no open interest to close against before this print.

That makes the opening read solid rather than merely likely. You cannot close a position that does not exist, so the 9,480-lot call purchase must have created exposure. The line now stands at 12,729 — our print plus ≈3,249 contracts of other call buying at the same strike.

One honest caveat: because there was no prior published figure, this is a first-observation baseline rather than a before-and-after delta. It proves the position is new; it does not let us measure how much of the day's total belonged to the trade in this article.

---

🤓 What This Actually Means — Plain English

Selling a put means being paid to promise you will buy the stock at the strike. Here: $1.47 collected for the obligation to buy Riot at $20.50 through August 28. That works out to an effective purchase price of $19.03 — the strike minus the premium — which is the seller's breakeven.

Why it has already deteriorated. When the put was sold, Riot was $20.30, so the option was barely in the money and the $19.03 breakeven sat comfortably below. By the close the stock was $19.245 — only 21 cents above breakeven. The obligation is ≈$21.5M if assigned.

The honest framing: this is not yet a loss. The seller would still net a small profit if Riot expired here. But 1.1% of cushion on a stock with a beta of 3.85 and no scheduled catalyst is very thin — a single ordinary session can erase it. This is what "collecting premium" actually looks like when the underlying moves against you: the income is fixed, the risk isn't.

Buying the $24 call is the opposite shape — ≈18% out of the money when bought, 100% time value, needing $25.60 to break even, or roughly +33% from where the stock now sits.

---

📈 Technical Setup

One-Year Performance

Note a data conflict we could not resolve: our chart computes ≈+43% over the past year, while a sourced figure shows +68.18%. We flag it rather than pick one.

🔵🟠 Gamma-Based Support & Resistance

The stock now sits between the two strongest gamma levels, having fallen through the $20 line during the session — which is precisely why the short put's cushion evaporated.

🎯 Implied Move

A ±12.73% expected range by August 21 against 1.1% of cushion tells you the whole risk story. The market is pricing moves an order of magnitude larger than the buffer protecting this short put.

---

🎪 Catalysts — And a Pivot the Market Rejected

The story changed three days ago. On August 10 Riot's Q2 release disclosed a 20-year, 191 MW lease at Rockdale worth ≈$9.1B (≈$16.1B with extensions; $365–411M of average annual net operating income) with what the company called "a leading frontier AI lab."

Media reported the tenant as Anthropic on August 11, but Riot never named the counterparty and it remains unconfirmed — we are not stating it as fact.

The market's verdict is the interesting part. Despite a reported +21% pre-market pop and five price-target raises ($25–$40) on August 11, RIOT trades at $19.24 today — below its $19.40 pre-announcement close and ≈11% under its August 4 high. A transformational-sounding lease was announced and the stock is lower than before it.

Mining economics explain the scepticism. Bitcoin sits near $63,228, roughly 49.8% below its October 2025 high. Riot's Q2 all-in cost to mine was $90,631 per bitcoin — 126.5% of production value (cash cost $49,912). The treasury has been drawn down from 19,223 → 15,680 → 11,380 BTC.

Expiration alignment — and this is decisive:

So the put seller has no catalyst to sell into — which is the structural argument for the trade (nothing scheduled to cause a gap) and against it (nothing scheduled to cause a recovery either). With beta 3.85, it is effectively a naked short bitcoin tail.

And the call needs +33% to its $25.60 breakeven in 50 days with no confirmed company catalyst inside. For reference, the 52-week high is $30.32, so breakeven sits ≈16% below it — demanding, but not an impossible level. The realistic paths are a bitcoin rally on a dovish Fed, a second AI tenant against Riot's 1.76 GW of uncontracted approved power, or permanent investment-grade financing replacing the $573M bridge.

---

👥 Four Ways to Read This Trade

🎲 The YOLO trader

The $24 call is your instrument — $1.60 for ≈+33% of required upside in 50 days, on a 3.85-beta name levered to bitcoin. Just be clear-eyed: there is no confirmed company catalyst inside that window, so you are really buying a bitcoin call with a mining wrapper. And we could not verify whether the desk that bought it was opening or closing.

📈 The swing trader

The levels are unusually crisp. $20 is Very Strong gamma resistance and the stock just fell through it; $19 is support; $19.03 is the put seller's breakeven. That cluster within 5% is where the next move gets decided. The macro trigger inside the window is the September 15–16 FOMC, and bitcoin at −49.8% from its high is the dominant driver — not the AI lease.

💰 The premium collector

This is the clearest cautionary tale on today's board. A put sold at 10:40 with ≈6% of cushion had 1.1% left by the close of the same session. The income was fixed at $1.47; the risk was not. If you sell puts on high-beta names, the lesson is that cushion measured in percent means very little when the underlying routinely moves 5% in a day — the implied move to August 21 alone is ±12.73%. Sizing, not strike selection, is what keeps you solvent here.

🌱 The beginner

Watch what happened in three hours. Someone sold a promise to buy Riot at $20.50 and was paid $1.47 for it. That looked fine at $20.30. By the close the stock was $19.245 and the promise was worth more than they collected minus 21 cents. Nothing dramatic happened — no news, no crash, just an ordinary down day. That is the honest picture of what selling options exposes you to.

---

⚠️ Honest Limits

• ⚠️ The October call's open-versus-close status is genuinely unknown. The tape returned no open-interest history for that contract at all. No direction should be inferred from that leg.
• The put leg is a proven open (prior OI 28) — that part is solid.
• Both side labels are reported, not tape-proven — neither print took liquidity.
• The AI-lease counterparty is unconfirmed. Riot did not name it; media reporting is not the company's word, and we do not treat it as such.
• A return conflict is unresolved: our chart shows ≈+43% over a year against a sourced +68.18%.
• Research gaps, disclosed: the search budget was exhausted, so all sourcing was by direct retrieval; regulatory filings returned errors, so convertible and at-the-market facility details are unresolved; several news bodies were inaccessible and only headlines were available; one provider's quarterly balance sheet was internally inconsistent and was discarded rather than used.

Last updated: 2026-08-14 — next-day OPRA open interest resolved both legs as opens, including the October call we could not baseline (see the ✅ RESOLVED section).

This is market analysis and education, not investment advice. Uncovered short puts carry substantial risk of loss.