The brief: Astera's options are among the richest in the AI-hardware complex (30-day implied volatility about 88%), the stock is 43% below its June high and the author would be content to own it. The question is whether that premium is a gift or a fair price, which comes down to four things this page works through: what the stock actually does day to day, what the 3 November report has to deliver, whether the drawdown is demand or positioning, and what an owner is buying at $220–240. The answer is a put, but not the obvious one and not before Tuesday night.
Our view in one paragraph. The premium is not a gift; it is the price of a stock that moves five percent a day. Thirty-day implied volatility of 88% sits a few points over realized volatility of 81% (98% over the year), 26 of this year's 206 sessions moved 10% or more, and the stock has gone $183 → $100 → $483 → $250 → $390 → $260 since February on estimates that barely changed. The report on 3 November is priced at about ±13%; the last eight reactions averaged 15%, split four up and four down, and the stock fell after three of the last four even though every one of the last eight quarters was a beat-and-raise. So the November expiry's premium is fair to the history, not rich, and the pattern it pays for (a beat that is sold) is the one a put seller is on the wrong side of. The drawdown itself is sector and supply, not demand: every down day of 5% or more since September maps to a macro or sector event (the Amodei essay, the OpenAI revenue reset, CoreWeave's bonds, Lam's warning, Alphabet's cash flow) or to a valuation downgrade (Northland, TD Cowen), with no guidance cut and no reported socket loss, while insiders sold $622 million in twelve months and Amazon holds 5.6 million registered warrant shares. The business is real and narrow: revenue +104% year on year, the Scorpio scale-up switch becoming the largest product line a quarter early, Amazon's Trainium as the lead customer with a $6.5 billion purchase warrant, and one end customer above 70% of revenue, which is the single biggest number on this page. The thesis risk is not November, it is 2027: Trainium's third generation moves to a UALink switch Astera does not ship until 2027, Nvidia's NVLink Fusion money went to Marvell, Microsoft's and Meta's scale-up is Ethernet, and the industry sizes PCIe/UALink scale-up at about $3 billion by 2030 against Astera's own $20 billion claim. At $230 an owner pays 36× next year's consensus earnings for 56% growth, net cash and 40% operating margins, a PEG near 0.6 if the numbers hold; it is a good entry if the Q4 guide clears $660 million with gross margin at 72%, and a stop on the way to the Street's $190 if it does not. Our view: sell the December put, not the November one, and sell it on 4 November with the report known, at the strike the reaction sets, because the 80% volatility that will still be in the December and January expiries after the event is the premium that is actually rich. If it has to be before the print, the 18 Dec $230 put at about $11.80, at half size, under the shelf and the 200-day, is the one that survives a one-move fall.
True where Astera plays. On Amazon's Trainium3 the lead customer uses Astera's Scorpio X as the scale-up fabric (160 lanes in the first generation, 320 in the second), four PCIe 6 retimers per JBOG, and management says Scorpio X content "could be well over $1,000 per accelerator" with content per XPU "to scale to multiple thousands of dollars" as Taurus cables and Leo CXL add on; SemiAnalysis puts the Scorpio attach at about $1,000 per XPU, $32,000 per 32-XPU rack. PCIe 6 to 7 adds reach loss that needs more retiming, and Aries 7 (sampling in the fourth quarter) is the only PCIe 7 copper-plus-optical portfolio.
Not true where Nvidia plays. In an NVLink rack Nvidia sells the scale-up switch; Astera keeps retimer and cable content only, and Morgan Stanley counts the retimers per server falling from 8–14 to 1–2 in NVL-class racks. Microsoft's Maia 300 and Meta's MTIA 300 scale-up fabrics shown at OCP are Ethernet (Broadcom silicon), not PCIe. Aries, 90% of revenue in 2023, is 53% now: the retimer franchise is plateauing while Scorpio carries the growth.
Verdict: content per accelerator is rising inside the custom-ASIC programmes that chose PCIe (Amazon above all) and falling inside Nvidia's and the Ethernet scale-up camps; the thesis is a bet on which camp grows faster.
The Amazon anchor is documented: a February 2026 warrant for 3.26 million shares at $142.82 that vests against up to $6.5 billion of purchases through 2033 (1.66 million vested by June), the lead customer taking Scorpio X to high-volume production in the third quarter, "more than ten customers engaged" and "ship to revenue for additional customers by the end of the year". Scorpio is expected to become the largest product family in Q3, a quarter early, and the Orion rack-scale reference design (160 accelerators a pod, builds by Accton and Hyve from Q1 2027) is the attempt to sell the same fabric to neoclouds and enterprises.
Nvidia is a partner, not an anchor. Astera was an NVLink Fusion launch partner and ships Aries in Hopper, Blackwell, MGX and HGX, but when Nvidia put $2 billion to work on Fusion scale-up networking in March it went to Marvell, which has also bought the XConn team (a CXL and UALink switch) and is sampling a 260-lane PCIe 6 switch of its own; Broadcom shows PCIe 6 switches and retimers and claims "5+ hyperscalers running Ethernet scale-up in production"; Credo holds the active-cable market Astera's Taurus is trying to enter; Microchip has a 160-lane Gen6 switch. 650 Group sizes 2030 scale-up switching at NVLink $25 billion, Ethernet $8 billion, PCIe/UALink $3 billion: "a minor but growing slice". Astera's UALink silicon arrives in 2027, the year Trainium's third generation reportedly moves to a UALink switch.
Verdict: a merchant leader in a segment that may stay small, with one anchor customer that extracts a 23% effective discount through warrants and a second that is building the competition; the Scorpio ramp is real, the moat is a 2027 question.
Every session of −5% or worse since September has a sector or macro name on it: 14 September −11.7% (Amodei's slowdown essay, chips −5.9%), 8 October −9.2% (the OpenAI revenue reset, SOX −4%), 12 October −5.3% (the OCP week), 14 October −7.7% (CoreWeave's bonds at 15%), 16 October −5.4% (no news), 22 October −11.1% (Lam's supply warning, SOX −6.7%), 29 October −4.8% (Alphabet's free cash flow, Marvell −12% on its Lumentum talks). The up days are the same in reverse: +12.4% and +6.7% on 21–22 September on a semis rally, +10.5% on 27 October on the Anthropic S-1 and Broadcom's financing. There was no guidance cut, no reported socket loss, and the Q3 guide of $540–560 million is +40% on the quarter.
What the market did change its mind about is the multiple and the supply. Northland (9 October, valuation: "2028 EPS of $5.00 achievable, not a slam dunk"), TD Cowen (20 October, target $375 to $300, Hold) and Zacks moved to the sidelines; the forward multiple fell from 74× (its one-year average) and 62× on 8 October to about 44× FY27 now. Insiders sold $622 million in twelve months with no purchase (Gajendra $241 million, chairman Alba $183 million, Mohan $157 million), $127 million of it in the last ninety days; Sutter Hill filed to sell in April; 5.6 million Amazon warrant shares are registered for resale. Beta is 3.7: in every 2026 sell-off the stock fell two to four times what SMH did.
Verdict: the fall is beta and supply on top of a multiple that had run ahead; the demand evidence is intact. For a put seller that is the good news (the strike can sit on a tested shelf) and the caveat (the same beta turns a 10% sector fall into a 30% one).
The beats are a fact: eight consecutive beat-and-raise quarters, the second quarter at $392 million (+27% on the quarter, +104% on the year) with non-GAAP EPS of $0.80 against $0.69 expected, and a third-quarter guide of $540–560 million at a 72% gross margin and $1.16–1.21 of EPS. Consensus sits at $550.8 million and $1.19, the whisper at $1.30. The bar that matters is the fourth-quarter guide: full-year consensus of $1.91 billion implies about $660 million (+20% on the quarter) and $1.43 of EPS.
The reactions are the other fact. Close to close after the last eight reports: +37.7%, −10.9%, +7.6%, +28.7%, +1.5%, −21.4% (the February print, when the CFO left and the Amazon warrant's margin effect was guided), −0.8%, −12.0% (August, after a +12.7% run-up into the print). Average absolute move 15%, four up and four down, three of the last four down on beats. The options price about ±13% (the 6 November expiry at 122% implied volatility against 95% for the 20 November monthly). That is fair, not generous, and the direction has been the wrong one for a put seller on three of the last four occasions: a beat is already in the price, and what moves the stock is the Q4 guide against $660 million and the gross-margin line against 72%.
Verdict: the premium in the November expiry is correctly priced for a stock that moves 15% on its reports and has sold off on the last three beats; the mispricing, if any, is in the December and January expiries, which still carry 80–84% volatility for weeks after the event.
What confirms it: a Q4 guide of $690 million or more with gross margin at 72%; a close above $309 (the 50-day) on volume; Trainium3's second generation confirmed on Scorpio X at re:Invent.
What confirms it: a close below $245 (the shelf gone, with $190 the next cluster); a Q4 guide under $660 million; gross margin guided under 71%; a Trainium scale-up switch named that is not Scorpio.
| Date | Event | Why it matters for the thesis |
|---|---|---|
| 29 Oct (after close) | Amazon Q3 (AWS growth, capex, Trainium) | The lead customer's proxy, five days before Astera's own report. |
| 3 Nov (after close; call 4:30pm ET) | Astera Q3 2026 results and Q4 guide; AMD Q3 the same evening | Q3 vs $550.8M / $1.19; the Q4 guide vs ~$660M; gross margin vs 72%; Scorpio as the largest line; the second and third Scorpio X customers. |
| ~5–6 Nov | Q3 10-Q | Customer percentages, vested warrant shares, China ship-to. |
| ~wk of 2 Nov | Possible Marvell–Lumentum announcement | Optical PCIe partner moving to a competitor. |
| 17 Nov (after close) | Nvidia Q3 FY27 | Rubin rack detail: retimer content, NVLink Fusion partners. |
| wk of 16 Nov; 19–20 Nov | Anthropic pricing and first trade | The AI-hardware beta's supply phase begins. |
| 20 Nov | November monthly expiry | The report's volatility premium is gone by here. |
| 30 Nov–4 Dec | AWS re:Invent | Trainium roadmap: PCIe (Scorpio) or UALink/Ethernet for the next generation. |
| ~1–2 Dec | Marvell and Credo report | The competitors' read on PCIe switches and AECs. |
| ~4 Dec / 18 Dec | S&P 500 quarterly announcement / effective; December expiry, quad witching | Astera is a candidate at $49B; the December put's expiry. |
| 9 Dec | Broadcom Q4 FY26; FOMC decision the same week (a hike 85% priced) | Ethernet scale-up claims; rates. |
| 24 Dec | Target date | Every structure on this page is valued on this day. |
Daily closes. Hover for values; the table shows the same data.
| Total return | 1 week | 1 month | 3 months | 6 months | YTD | 1 year |
|---|
Astera is up this year against for the semiconductor ETF, for Marvell and for Credo; over one month it is against , which is the beta in one row: the stock does two to four times what the sector does in both directions (3.7 against the market, 1.6 against SMH on a year of daily data). The year's path, $183 to $100 to $483 to $250 to $390 to $260, happened on estimates that moved far less than the price. The relative-strength read is the opposite of a momentum argument: this is a name the market re-prices from 50× to 120× and back on the same numbers, which is exactly why its options cost what they cost.
Blue candles closed up, red closed down. Orange dashed: dealer-gamma levels (call wall, put wall, gamma flip); swing support and resistance are listed in the key-levels table. Hover for the crosshair: the price is tagged on the price axis and the date on the time axis. Right panel: net dealer gamma per strike on the same price axis (blue where dealers are long gamma, red where short).
Trend: a five-fold run, a 43% drawdown, and a shelf that has held four times. Astera went from $97.89 on 30 March to $499.48 on 30 June, five times in three months, then lost half of it: $244.61 on 29 July. The recovery to the 25 September and 6 October highs ($377.87, $401.74) came with the sector and left with it: the October slide took the stock from $401 to $249.25 on 26 October, through the shelf, on 22 October's −11% (the Lam Research day, 6.3 million shares, three times average). The bounce has been as violent as the fall: $257 to $297 in three sessions (27–28 October), then −2% on Thursday after Alphabet's report. The four lows since July sit at $244.61, $266.31, $262.45 and $251.63, with the 26 October low at $249.25: a $245–252 shelf tested four times and undercut once, the builder's support cluster at $244.6 and $251.6.
Levels. Below: the $245–252 shelf, then the 200-day average at $255 rising into it (the two will meet in November), then nothing until $190 (the builder's lowest cluster) and the March–April gap. Above: the 50-day at $309 and the 20-day at $311, the $310–322 swing highs of late August and early September, $342–354, and the September–October highs at $378–402. RSI is ; the stock is from the 50-day and above the 200-day, which it has not closed below since April.
Volume. The heaviest sessions of the last year are reaction days, not trend days: the November 2024 report (+37.7%, 24.8 million shares), the June 2026 spike (+11.3%, 22.6 million), the August 2025 report (+28.7%), the February 2026 report (−21.4%, 16.4 million), and two non-earnings air pockets (27 January 2025 −28.0%, 14 October 2025 −19.0%). The last two weeks ran at 2–6 million a day against a 20-day average of ; the bounce from $249 came on rising volume (3.0, 4.5, 3.3 million), the first time since the 6 October high that up days carried the volume.
Technical verdict: a stock that has held a $245–252 shelf four times in three months, undercut it once on sector capitulation volume and recovered in three sessions, with the 200-day average rising to meet the shelf in November. That is the right place for a put strike: under the shelf and the 200-day, at $240 or below, not inside the $250–300 range where the stock spends its days moving 5%. For the thesis of a retest of $350 the stock first has to reclaim the 50-day at $309 and the $310–322 highs; until then this is a range between a tested floor and a falling average.
| Level | Price | vs spot |
|---|
Support and resistance are clustered swing lows and highs from the last eight months of daily bars; the gamma levels come from CBOE open interest.
What the profile says for Astera. This is a small, tight options book (open interest about contracts, net of gamma per 1% move), so the levels describe where the positioning sits rather than what it can enforce. The book is marginally positive around the price: the largest strikes are the $275 and $267.5 puts (net +$4 million and +$2 million as dealers' longs), the $280 put wall and the $300 call wall bracket the price within 6%, and the flip sits at $257, just above the shelf. Below $257 the book turns short and the moves extend, which is what the 22 October session looked like. For a put seller the structure says two things: the stock trades in a hedging-neutral zone between $257 and $300 where the 5% daily swings are the stock's own, and a strike at $240 or below sits under the level where the dealers would start selling with the stock. The sign convention (dealers long customers' sold calls, short customers' bought puts) is an assumption; open interest is end-of-day (see the footnote under the chart).
How to read this. Market makers who sell options hedge them with stock; how much stock they must buy or sell as the price moves is their gamma. We estimate it from CBOE open interest for every listed strike and expiry, recompute each option's gamma with Black-Scholes from its own implied volatility, and sign it by the usual convention: dealers are long the calls customers sell and short the puts customers buy. Positive net gamma means dealers sell as the stock rises and buy as it falls, dampening moves; negative means they chase the move and amplify it. The sign convention is an assumption, and open interest is end-of-day, so the profile changes after every expiry.
Realized volatility from daily closes; implied volatility is the one-year series from the put screen's Interactive Brokers data to 14 October, with the CBOE chain's 30-day value for 29 October.
Solid: implied volatility of the at-the-money option for each listed expiry, on a calendar axis. Dashed: the forward volatility the market prices for the period between one expiry and the next; the spike is the earnings report. From CBOE's delayed chain.
Last three months of closes, then the range the market prices to each expiry: today's price plus and minus the at-the-money straddle (shaded, the expected absolute move) and ± one standard deviation from at-the-money IV (dotted). Green dashed: the report's target.
Out-of-the-money puts left of spot, calls right of spot, for the first monthly expiry after a month, the expiry that carries the earnings report, and January. A smile that rises to the right means upside calls are bid: the market pays up for a squeeze.
The premium is high because the stock moves, not only because of the report. Thirty-day implied volatility is against realized over the last month: the market charges a few points over what the stock has actually delivered, which is normal with a report inside the window, and nowhere near the 142% of the March crash (the rank is in a 54–142% range, the th percentile of the year). A seller of a $240 put is being paid for 5%-a-day behaviour and a ±13% report, and gets both.
The term structure dates the report. At-the-money IV is 122% for 6 November, the first expiry after the report, 104% for 13 November, 95% for the 20 November monthly, 88% for 27 November, 84% for 18 December and 79% for January. The 6 November straddle prices ±14.6% for eight days, of which the base week is worth about ±6%: the report itself is priced at roughly ±13%. The realized reactions in the bars are larger: +37.7% (Nov 2024), −10.9% (Feb 2025), +28.7% (Aug 2025), −21.4% (Feb 2026), and the research brief lists the rest. The point for a seller: the report premium is fair to the history, not generous, and the direction has been both ways on beats; the premium that is actually rich is the 84–95% in the December and January expiries, which price a 5%-a-day stock for weeks after the event.
The smile leans to calls. The 25-delta skew is negative in every expiry (−3 to −8 volatility points: calls trade over puts), which is what a stock that doubled in a quarter and halved in the next looks like: the market pays for the squeeze, not the crash. Out-of-the-money puts carry no extra fear premium, so the edge in selling them is the level (87% against 81%) and the time, not the skew.
The cone. The one-standard-deviation move to 24 December is ±, a third of the price; the 20 November straddle prices ±19% and the January straddle ±29%. On that distribution $240 is 0.5 standard deviations below spot for November and $230 is 0.4 for December: these strikes are not far in volatility terms, which is why they pay what they pay. The market gives $350 by Christmas a chance.
| Structure | How it is built | What you want | Risk | Best used when |
|---|---|---|---|---|
| Long call | Buy a call | A big move up, fast | The premium (100% if it expires below the strike) | You expect a move larger than the market prices; implied volatility is low |
| Bull call spread (debit) | Buy a call, sell a higher call | The stock at or above the upper strike at expiry | The net debit | You have a price target and a date; you want the highest return per dollar to that target |
| Cash-secured short put (credit) | Sell a put, hold the cash to buy the shares | The stock to stay above the strike; you keep the credit | Owning the shares at the strike in a sell-off (large, stock-like) | You would be happy to own the stock lower; volatility is high |
| Bull put spread (credit) | Sell a put, buy a lower put | The stock above the short strike at expiry; keep the credit | Spread width minus credit (defined) | You are bullish-to-neutral and want to be paid to wait with a known worst case |
| Risk reversal | Sell a put, buy a call (often for zero cost) | A large rise; behaves like stock above the call strike | Stock-like below the put strike; margin required | You want stock exposure without paying premium and accept the downside |
| Put-financed call spread | Sell a put, buy a call spread | The stock at the top of the spread; the put pays for it | Stock-like below the put strike | You want a call spread for free and accept assignment risk |
| Put spread + call spread | Sell a put spread, buy a call spread | The stock above the call spread's top strike | Both spreads' widths minus net credit (defined both ways) | You want a leveraged payout at the target and a known worst case, with part of the cost paid by the put spread |
Returns on the short-put structures below are shown on a cash-secured basis (capital = strike minus credit). Brokers typically require about 20% of that as margin; for a stock this volatile expect more.
The brief is to sell a put on a stock the author is content to own, in a chain where everything is expensive. Three expiries are priced: the 20 November monthly (22 days, carries the report, 95% implied), 18 December (50 days, carries the report plus six weeks of 84% volatility after it) and 15 January (78 days, 79%). The target of $350 (the late-September swing high, +24%) and the 24 December date are there so every structure is valued on the same day; for a short put the question is where the strike sits against the $245–252 shelf, the 200-day at $255, the report's ±13% and the stock's 5% days, and what the premium pays per month of that.
Scenarios on 24 December: Bear $198 (−30%, a bad report through the shelf and the 200-day to the spring gap; implied volatility +8 points) · Flat $283 (volatility −12 points after the report) · Base $317 (+12%, back to the 50-day) · Bull $350 (+24%, the target; volatility −15 points) · Overshoot $400 (+41%, the 6 October high).
| Trade | Net | Max loss | Break-even | Delta | PoP |
|---|
Net is the mid-price debit (paid) or credit (received) per share; one contract is 100 shares. Max loss is the capital at risk per share (for a short put, the strike minus the credit). PoP is the lognormal probability, at today's implied volatility, of finishing above break-even at expiry. Scenario cells show profit or loss as a percentage of the capital at risk on 24 December (20 November trades have expired by then, so their cells are the result at expiry); −100% means a total loss. Highlighted rows are the recommendation.
| Trade | Capital at risk | P&L | Return |
|---|
Pick up to four trades. Values on the target date use Black-Scholes with each contract's own implied volatility; trades expiring before then show their value at expiry.
The recommendation is about timing as much as strike. The November expiry sells you the report at a fair price for a pattern (beats that are sold) a put seller does not want; the December and January expiries sell weeks of 80% volatility after the event on a stock whose realized volatility is 81%. So: not the November put; the December put, preferably on 4 November with the reaction known, at a strike under the shelf and the 200-day; half size, because a 3.7-beta stock is already a leveraged version of the semiconductor exposure the author holds.
| Strike | Bid | Ask | Mid | Break-even | IV | Delta | Theta/day | Volume |
|---|
Option quotes, Greeks, open interest, term structure, smile and dealer gamma: the CBOE delayed chain at about 12:00 New York time on 29 Oct 2026 (fetched ; IB Gateway was not running; the far strikes are $0.50–1.50 wide, work the mid). Daily prices: Yahoo Finance to 29 Oct (intraday bar); one-year implied-volatility range: Interactive Brokers via the put screen of 14 Oct. Fundamentals: Astera Labs Q2 2026 8-K and 10-Q (4 Aug 2026), the Q2 call transcript (Motley Fool, 11 Aug 2026), the FY2025 10-K, the Amazon warrant 8-K (5 Feb 2026) and S-3ASR (20 Mar 2026), the Q3 call notice (6 Oct 2026), Form 4 filings via SEC EDGAR and Massive; consensus and targets: StockAnalysis, MarketBeat, TIKR, WhisperNumber; implied moves: Bloomberg via Investing.com (27 Oct), earnings-watcher (28 Oct), MarketChameleon; competition: Broadcom OCP 2026 materials, Dell'Oro OCP note (19 Oct 2026), 650 Group (20 Jan 2026), SemiAnalysis Trainium3 deep dive (4 Dec 2025), Futurum (9 and 14 Oct 2026), DCD on the Nvidia–Marvell investment (31 Mar 2026), Investing.com on Credo/MaxLinear (23 Oct 2026); news of 4 Sep–29 Oct 2026: CNBC, Reuters, Bloomberg via 24/7 Wall St, Investing.com, MarketBeat, Seeking Alpha, TradingKey, TechTimes. Earnings-day moves: computed from the daily bars (vendor tables that quote report-day moves were not used). Two research briefs compiled on 29 Oct 2026 with web search; figures that drive the verdict were checked against the primary release or the built data.