Bears added to semiconductor shorts on Friday with a ferocity not seen since August, driving put volume in the SMH ETF to 180,000 contracts against 50,000 calls by midday and pushing $46 million of premium into puts versus $26 million into calls. The put-to-call open interest ratio climbed to 1.95, its highest level since the second week of August, while the same measure for the Invesco QQQ Trust extended its recent advance to 1.51. SpotGamma data indicate 129,000 of the SMH put contracts were bought rather than sold.
The SMH tape tells the story
The sheer scale of the put buying suggests institutional positioning rather than retail speculation. At 3.6-to-1 put-to-call volume, the flow dwarfs the typical hedging activity that accompanies a routine pullback. The open interest ratio confirms the positions are being held, not flipped. For a sector that has led the market higher for two years, a 1.95 put skew in the benchmark ETF is a statement of conviction, or at least a very expensive hedge.
Nvidia's $21 million January bet
Right after the open, a single buyer paid $21 million for 100,000 Nvidia 180-strike puts expiring January 15, the largest options trade in the name that day. At the current share price, the position requires a 22 percent decline by expiry to break even if it is a directional bet. The strike sits well below recent support levels, making it either a tail hedge for a large long position or a wager that the AI trade unravels faster than consensus expects.
Micron's deep-in-the-money puzzle
Micron drew the most unusual flow. Call volume ran 40 percent above average, yet SpotGamma flagged $270 million in premium tied to likely put buying. The bulk centered on deep in-the-money puts expiring June 2028 with strikes between 2,250 and 2,500, roughly 125 contracts transacted near the ask, implying purchases. Another 50 contracts at the 2,050 strike traded near the bid, implying sales. With the stock around $1,030, these are not lottery tickets.
What the structure really says
Taken at face value, the Micron legs form a net $14.5 million bearish spread with a delta near negative one, effectively a synthetic short. Traders typically choose deep in-the-money puts over borrowing shares when the borrow cost is prohibitive or when they want defined risk, the most they can lose is the premium paid. But Jason DeLorenzo of Volland cautions that midpoint analysis struggles with far-dated, low-volume spreads: "Spreads have difficulty being categorized by midpoint analysis as well, because dealers are willing to take a haircut on one leg while getting a better premium with the other." The structure may be a borrow workaround, a volatility play, or something else entirely. What is clear is that someone deployed nine-figure premium to bet against the memory cycle, and did it with a three-year horizon.