On 9 September 2026, Applied Materials CEO Gary Dickerson sat on an earnings call and said something that is hard to say more emphatically:
"AI is the biggest technology inflection I've ever seen in my life."
He backed it with numbers rather than adjectives. The service business was growing over 20% year over year with margins up 180 basis points. Advanced packaging growth was expected to exceed 70% for the year. On the transcript alone, there is no ambiguity about how management sees the next several quarters.
Nine days later, on 18 September 2026, the options tape in AMAT looked like this:
| Side | Strike | Expiry | Vol/OI | Premium | IV |
|---|---|---|---|---|---|
| Put | $480 | 09/18/26 (1d) | 9.4× | $25.0M | 127% |
| Put | $500 | 09/18/26 (1d) | 8.9× | $37.5M | 156% |
| Put | $520 | 09/18/26 (1d) | 8.9× | $30.1M | 317% |
That is $92.6M of put premium in a single session, in a name whose CEO had just delivered one of the most bullish sentences you will read in a semiconductor transcript this year.
The obvious conclusion is that someone disagrees with Dickerson. The obvious conclusion is wrong, and understanding why is the most useful thing you can learn about combining these two data sources.
The expiry column is doing all the work
Look at the expiry. All three prints expire the same day they were written. A one-day option is not a view on whether AI is the biggest inflection of anyone's life. It cannot be. By the closing bell the position is either exercised or worthless, and nothing management said about 2027 capacity has any bearing on it.
The implied volatility column is the tell. An IV of 317% on a one-day option is not a market forecasting a 317% annualised move. It is the mechanical result of pricing an option with almost no time left, where a small absolute premium divides by a very small time value. Read those numbers as a directional signal and you will reach a confident conclusion about nothing at all.
What the prints probably do reflect is ordinary expiry-day business: hedges being rolled, positions being closed, market makers managing inventory into the close. That is real activity and worth seeing, but it is a different category of information from the one the transcript contains.
This is the first rule of putting transcripts next to flow: compare horizons before you compare direction. A transcript is a statement about quarters. An option carries a date. If those two horizons do not overlap, the comparison is noise dressed up as insight.
What a genuine disagreement would look like
A real contradiction between narrative and positioning has a specific shape. It looks like sustained put buying at strikes well below spot, in expiries that sit past the next reporting date, building over multiple sessions rather than spiking on one.
Those conditions matter individually:
Sustained, not single. One large print is one decision by one desk, and you have no idea whether it opened a position, closed one, or hedged something in another asset entirely. The same strike printing across several sessions is a much harder thing to explain away. OptionsBell's own analysis of this pattern, in sustained flow beats single prints, is worth reading before you build a screen around any single day's tape.
Past the catalyst, not before it. If the expiry sits before the next earnings date, the position cannot express a view on the quarter. It can only express a view on the weeks between now and then.
Away from spot, not at it. At-the-money flow is where hedging lives. Positioning that reaches for strikes some distance from spot is making a claim about magnitude, not just direction.
None of the AMAT prints above meet those conditions. They are expiry-day activity in a heavily traded name, which is exactly what you would expect to see and exactly what you should not build a thesis on.
The comparison that does work
The useful version of this exercise runs in the other direction. Rather than starting from a dramatic print and looking for a transcript that explains it, start from what management actually committed to and ask whether the option market is pricing that commitment at all.
Dickerson's call gave you three testable commitments:
- Service business growing over 20% year over year
- Service margins up 180 basis points
- Advanced packaging growth exceeding 70% for the year
Each of those has a date attached, implicitly or explicitly. The question for the options market is not "does anyone disagree" but "is there positioning in the expiries where these claims come due." If a company guides to a step change two quarters out and there is no unusual activity in any expiry that contains those two quarters, the market is not taking the other side. It is not taking any side.
That absence is itself information, and it is invisible if you only look at the transcript, and equally invisible if you only look at the flow.
Where each data source is authoritative
It helps to be precise about what each source can and cannot tell you.
The transcript is authoritative about intent and framing. It is the only place where you get management's own words about why results looked the way they did and what they expect next. It also records what analysts pushed on, which is often the better signal. The difference between prepared remarks and Q&A matters here: the scripted section is a company's chosen story, and the unscripted section is where that story meets resistance.
The options tape is authoritative about money and timing. It tells you what someone was willing to pay and when they needed to be right. It does not tell you why, and any write-up that claims to know why a specific print happened is guessing.
Used together, the transcript supplies the "why" candidates and the tape supplies the "when." Used separately, each one invites you to fill in the missing half from imagination.
A worked routine
Here is the sequence that actually produces something, using the AMAT case as the template.
Step one: pull the most recent call. For Applied Materials that is the 9 September 2026 transcript. Read the Q&A before the prepared remarks. Note every claim with a date attached.
Step two: pull the flow, filtered by expiry. Discard anything expiring inside a week unless you are specifically studying short-dated behaviour. What remains is positioning with enough time to be about the business. The live options flow feed shows the expiry and days-to-expiry on every row, which is the column that makes this filter possible.
Step three: line up the dates. Does any surviving expiry contain a date that management put a number on? That intersection is where the two data sets genuinely talk to each other.
Step four: check whether it persists. Come back the next session and the one after. The daily archive at OptionsBell keeps previous sessions addressable by date, which is what makes "did this repeat" a question you can answer rather than a feeling you have.
Most days, this routine produces nothing. That is the correct outcome and the reason it is worth running. A method that finds a signal every day is not finding signals.
The habit worth building
The instinct that gets people into trouble is treating a large premium number as a large conviction number. They are not the same thing. Premium scales with the price of the underlying, the size of the position, and the amount of time bought. A $37.5M one-day put in a high-priced semiconductor name and a $37.5M eighteen-month call in the same name are separated by an enormous difference in what the buyer needed to believe.
Transcripts are the correction for this. They give you the timeline the business is actually operating on, which is the only frame in which an expiry date means anything. A CEO talking about a multi-year inflection and a trader buying a contract that expires this afternoon are not having an argument. They are not even in the same conversation.
If you want to see both halves side by side, the earnings call transcripts are free to start reading here, and the options tape is free to watch at OptionsBell. The work is in lining up the dates, and that part nobody can do for you.