Fifty-seven US companies report, and we cover the twenty most widely held: Dell, Palo Alto Networks, Hewlett Packard Enterprise, Snowflake, Ciena, MongoDB, Medtronic, NetApp, UiPath, Samsara, Guidewire, Five Below, GitLab, Netskope, Victoria’s Secret, DocuSign, Planet Labs, Argan, NIO and Sibanye Stillwater.

More than half of it lands on Tuesday evening, when Dell and Palo Alto both report after the market closes.

Last week Okta beat what analysts expected it to earn, and rose 28.6%. Rubrik beat too, comfortably, and fell 13.1%. Beating the estimate is not what moves a stock.

In this issue:

• Who reports on which day, and how big a move is expected from each.

• Why beating the estimate did not save seven stocks last week.

• The five reports that matter most, and the number that decides each one.

• A buy, hold or sell view on all twenty with a confidence level, five free and fifteen for paid subscribers.

Do you own something that reports this week?

Most investors find out what the market expected only after the stock has already moved. Earnings Intel arrives the Sunday before, free: for each big report, the number the company is expected to print, the size of the move the options market is pricing, and how the stock has reacted to its own past reports.

Subscribe now

If you hold one of them, two minutes is enough: find it in the day map below, then set the move it is priced for against what it usually does on a report. That gap is the size of night to plan for. If you hold none of them, this is the reading you would want done before a name goes on a watchlist. First, the week.

Who reports when, and what is priced for each day

Find the day your stock reports and how big a move is priced for it, so the night is not a surprise.

Money at stake is a company’s market value multiplied by the size of move expected from its report: paper value that could swing that night, not cash changing hands.

The priced move is how big a swing the options market expects, up or down. It says nothing about which way. If a stock you hold is priced for a 10% move, expect a night that size in either direction.

The typical name this week is priced for a 9.7% move, and twelve of the twenty are priced above their own history, four at more than double it.

Monday 31 August. One company reports, none of our twenty.

Tuesday 1 September. Twelve report and seven are ours: Medtronic, NIO and Sibanye Stillwater before the market opens; Dell, Palo Alto Networks, MongoDB and GitLab after it closes. Job openings and the manufacturing survey both land at 10:00 ET. Dell and Palo Alto alone carry $53.1 billion of the day’s $64.3 billion.

Wednesday 2 September. Twenty-two report, six of ours, all after the close: HPE, Snowflake, NetApp, Netskope, Five Below and Argan. The August private-payrolls estimate lands at 8:15 ET, before any of them.

Thursday 3 September. Twenty-two report, seven of ours: Ciena and Victoria’s Secret before the open, then UiPath, Samsara, Guidewire, Planet Labs and DocuSign after the close. The services survey lands at 10:00 ET.

Friday 4 September. None of ours report. The jobs report lands at 8:30 ET, and Thursday’s after-close reporters take their first session against it.

Nine software companies report across those three nights, and their shares have historically moved together on earnings days about 61% of the time. So how Tuesday’s names are received is a fair early signal for Thursday’s.

A profit beat is not enough: which of this week’s twenty it has actually moved

Your stock can beat expectations and still fall: seven did last week. This shows which of the twenty that has happened to before, and how our own views scored.

Eighteen of the nineteen that reported beat on profit, and eleven of the twenty finished higher. Both of the biggest moves happened at the open: Okta gapped 23.6% and Elastic 24.1% before a share changed hands.

Five of the twenty carried a buy or a sell view. Three were right and two were wrong: Zoom, a buy at high confidence, fell 7.0% after beating on both profit and sales, and Workday, a sell, rose 5.8%. Owning the buy and holding cash on the four sells came to −1.4% for the week; simply holding all twenty through their reports returned 4.4%. One week is evidence rather than a record, and this one went against us.

The history splits this week’s twenty. At five, the beat and the move after it pointed the same way in at least four of the last five reports: Dell, Hewlett Packard Enterprise, Snowflake, MongoDB and Five Below. At seven, the headline number has rarely decided the reaction: Palo Alto Networks, NetApp, Medtronic, GitLab, NIO, Planet Labs and DocuSign. The other six are mixed, and Netskope and Sibanye Stillwater have too little history to say. Five reports is a small sample: a pattern worth knowing, never a measured edge.

For three names the profit line is the wrong one to watch. Planet Labs has cleared the revenue estimate five times in five and the profit estimate once; Argan and NIO do the reverse.

The five reports that matter most, and what decides each one

If you hold one of these five, the number to look for and what has usually happened after.

1. Dell Technologies (DELL): $30 billion of market value moving on one evening

Tuesday after the close · priced ±9.3% against a usual 8.9% · expected profit $4.91 a share on $44.9 billion of revenue · beat profit in 4 of the last 5, rose after 3 of 5

A 9.3% move on $322.5 billion of market value puts $30.0 billion of paper value on one evening. The stock is up 48.7% since its last report, 8.3% below its high.

That last report is why. In May, Dell reported $4.86 a share against a $2.96 estimate and rose 32.8% the next day. Its recent reports have averaged a 27.3% move against 5.6% for the older ones, so anyone sizing this off the old history is sizing it wrong.

Guidance is the target management sets for itself, and May’s was high: $44 billion to $45 billion of quarterly revenue, $15.5 billion of it AI servers. Its AI-server backlog, orders taken but not yet delivered, stood at $51.3 billion.

What decides the reaction: AI-server revenue against that $15.5 billion target, and profit margin on everything else. The risk: memory and processor supply, which management named as the thing holding it back.

2. Palo Alto Networks (PANW): five straight beats, and it fell after four of them

Tuesday after the close · priced ±8.1% against a usual 6.8% · expected profit $0.98 a share on $3.35 billion of revenue · beat profit in 5 of the last 5, rose after 1 of 5

Palo Alto has beaten the profit estimate five reports running and fallen after four of them. Its usual reaction is a 6.8% drop, and its one gain, 3.1% in August 2025, was the smallest move of the five. It arrives 4.0% below its high, with thirty analysts agreeing closely on the quarter. In June management set the bar at $8.9 billion to $8.95 billion of newer-product subscription revenue, up 59% to 60%.

What decides the reaction: how much of that subscription growth is its own rather than bought in with CyberArk and Chronosphere, and the guidance for next quarter. The risk: five beats have produced four drops, so the guidance has to carry the day.

3. Hewlett Packard Enterprise (HPE): priced for 9.7% against a usual 1.9%

Wednesday after the close · priced ±9.7% against a usual 1.9% · expected profit $0.93 a share on $11.95 billion of revenue · beat profit in 5 of the last 5, rose after 4 of 5

This is the widest gap on our list after NIO. Its median reaction is 1.9%, and yet the options market expects 9.7% this time, 5.1 times that. June is why: HPE reported $0.79 against $0.535 and rose 19.5%, after four reports that moved it three times under 2% and once down 3.3%. It is up 15.7% since that day.

In June management raised the year to $3.35 to $3.45 a share. Dell reports the night before in the same business, so by Wednesday’s close the market will already have a read on server demand and parts supply.

What decides the reaction: AI-systems orders and undelivered backlog against that raised target, and whether networking orders are turning into revenue. The risk: memory and networking parts are in short supply, which management called the thing holding it back, and a priced move that assumes June was the new normal.

4. Snowflake (SNOW): up 88% into a night the market is treating as routine

Wednesday after the close · priced ±10.3% against a usual 13.4% · expected profit $0.45 a share on $1.48 billion of revenue · beat profit in 5 of the last 5, rose after 4 of 5

Snowflake is priced for a quieter night than it usually delivers, 10.3% against a usual 13.4%. It has beaten the profit estimate every time and risen after four of five, the last a 36.5% jump in May. The exception was December, when a beat still produced an 11.4% drop.

Analysts agree on the sales and not on the profit. Their revenue estimates sit within 4% of each other while their profit estimates are 29% apart, one of the widest disagreements on our list. The argument is about margin, meaning how much of each dollar of sales it keeps, and in May the company set a target of 75% on its core product for the year. The worry analysts raised on that call: AI work carries thinner margins than the rest.

What decides the reaction: that 75% margin, and whether existing customers are spending more than last year. The risk: Snowflake deliberately sets a low bar, so a beat is expected rather than impressive after an 88% run.

5. Ciena (CIEN): it keeps beating and the stock keeps falling

Thursday before the open · priced ±10.3% against a usual 12.9% · expected profit $1.72 a share on $1.63 billion of revenue · beat profit in 4 of the last 5, rose after 2 of 5

In June it reported $1.64 against $1.46 and fell 13.7%. In March it beat again and fell 12.9%. Its usual reaction across five reports is a 12.9% drop. It reports Thursday morning from $399.85, still 35.5% below where it stood before June, after rebounding 21.0% this month.

In June management raised the year to $6.3 billion of revenue, with orders taken but not yet delivered set to finish above $7.7 billion. The priced move, 0.8 times history, says the market expects a quieter night than the last two.

What decides the reaction: how fast that order book turns into revenue, and the margin target for the final quarter. The risk: supply shortages that management says limit how fast it can deliver, and two reports running where a beat still sold off.

The other 15: when each reports, and what is priced

The same facts for the other fifteen: when each reports, what is priced, and how that compares with its own history.

Seven have no option expiring near their report, so their priced move covers several weeks and is marked as such. NIO’s profit estimate is reported in a different currency, so treat it as low-confidence. GitLab and DocuSign trade thin options; their figures carry a wider margin of error. Four carry heavy short interest, money betting on a fall, as a share of the stock available to trade: UiPath 28.7%, Victoria’s Secret 14.4%, GitLab 11.0%, Planet Labs 9.4%.

Buy, hold or sell: the view on all 20, with a confidence level

What we think of each of the twenty, and how sure we are.

The views below are general information for a general audience, not personal advice; the note at the end of the issue applies to them.

Each view is built from two inputs: positioning in the options market, which way the outstanding bets lean and whether traders pay more to protect against a fall or to catch a rise; and confirmed analyst rating changes over the last ninety days, counted rather than named. Most weeks most names read hold at low confidence, and this week fourteen of the twenty do.

NIO (NIO), Tuesday before the open: Buy, high confidence. The firmest view of the twenty. Positioning leans up, upside costs 14% more than downside, the second-widest gap of the week, with no analyst cuts. Against that: the stock sits 1.4% above its 52-week low, profit estimates have been cut 12.5% in two weeks, and the 7.0% priced move is 9.6 times its 0.73% median reaction. The buy rests on positioning alone; everything else here points the other way.

Argan (AGX), Wednesday after the close: Hold, low confidence. More downside protection outstanding than upside positioning, no analyst cuts. Down 33.3% since a June report that beat by 40% and moved the stock under 1%. No option expires near the report date, so the 12.2% priced move covers several weeks rather than the night itself.

Sibanye Stillwater (SBSW), Tuesday before the open: Hold, low confidence. Bets lean up, no analyst cuts, and with no option expiring near the report the cost of protection cannot be compared. Two reports of usable history: one up 4.4%, one down 7.5%.

Planet Labs (PL), Thursday after the close: Hold, low confidence. Positioning leans up, upside and downside cost the same, no analyst cuts. The 12.4% priced move is 0.35 times a 35.0% median reaction, the cheapest against history of the twenty.

DocuSign (DOCU), Thursday after the close: Hold, low confidence. Upside costs 21% more than downside, the widest gap of the twenty, while outstanding positioning is balanced and there are no analyst cuts. Five beats in five, gains after two; average drop 11.3% against an average gain of 3.8%.

Fifteen of the twenty sit behind the cut, Dell and Palo Alto Networks among them, the two biggest reports of the week. Five of the fifteen carry medium confidence rather than low.

Below the line: the five the data leans against, what a move that size would be worth on 100 shares, and the view on the other ten.

The other fifteen views, starting with the five the data leans against

Read more

Estée Lauder rose 16.3% on Wednesday, the biggest gain on our list.

Fabrinet beat on profit and sales on Monday evening and lost 19.4% the next day.

That was last Sunday’s list, and the full scorecard is below.

This week the market re-judges NVIDIA, CrowdStrike, Intuit, Synopsys and Workday, along with PDD, three of Canada’s big banks and twelve other widely held names.

If you own any of them, this is the week their price gets reset, and the reset usually happens overnight.

Eighty-one companies report in all, and 25 of them land on Wednesday.

Put each one’s market value against the move the market is pricing for it and the week comes to $380.5B on paper, as of Thursday’s close; NVIDIA is two thirds of that on its own.

For each one you get the number the company is expected to print, the size of the move the options market has already priced, and how the stock has reacted to its own past reports.

Do you own something that reports this week?

Most investors find out what the market expected only after the stock has already moved. Earnings Intel arrives the Sunday before, free: for each big report, the number the company is expected to print, the size of the move the options market is pricing, and how the stock has reacted to its own past prints.

Subscribe for free now and don’t miss the next report.

Subscribe now

Two things to know about the numbers: the priced move is one plain percentage, the average-sized move the options market is braced for, up or down, and it predicts size, never direction.

We set that beside how the stock moved on its last five reports and beside what analysts expect, on the adjusted basis companies guide on, so you can see where the market is paying for more than usual and where a beat has failed to move the stock before.

If you hold none of these, the same three numbers are the homework you would want done before any of them earns a place on a watchlist.

First, the grades.

Last Sunday’s list: what the market priced, what each stock did

Eleven of the twenty finished higher over their report, seven finished lower, and two, Home Depot and Analog Devices, moved less than 1% either way. The tables below mark each one buyer won, seller won or neutral: buyer won means anyone who held the stock through the report came out ahead. The options market had priced a median move of 6.26% across the twenty; the median actual move was 4.33%. Six moved more than priced and fourteen moved less.

The biggest moves: Estée Lauder rose 16.3% on a 22.2% profit beat, priced for ±9.71%. Fabrinet fell 19.4% despite beating on both profit and sales, priced for ±11.59%. Walmart beat profit by 9.2% and fell 9.2%, double its priced ±4.64%, on the question we raised about where its growth was coming from. Baidu missed on profit and fell 12.7%.

Seventeen of the twenty beat on profit. Five of the seven the seller won had beaten. A beat on its own decided very little last week, which is why this week’s featured names lead with the track record rather than the estimate.

The week by day: one quiet Monday, then everything lands on Wednesday

Eighteen members of the S&P 500 report, 9.3% of the index by value, and 13 of our 20 are technology companies, most of them software names that report in a chain from Tuesday to Thursday.

Monday. PDD before the open, a stock that has fallen after three of its last five reports, and little else.

Tuesday. Bank of Nova Scotia and Gold Fields before the open. Consumer confidence at 10am ET. Intuit and Zoom after the close.

Wednesday. The July inflation reading (core PCE), personal income and durable goods orders at 8:30am ET. After the close, 25 companies, including NVIDIA, Synopsys, CrowdStrike, Agilent, Trip.com, Everpure, HP, Veeva, Okta and Nutanix: ten of our twenty and $309.7B of the week’s $380.5B.

Thursday. TD and Royal Bank before the open. Workday, Rubrik and Elastic after the close, by which time seven software peers have already reported.

Friday. Four small reports. The week is over by Thursday night.

The 5 highest-stakes reports this week: what is expected, what is priced, and what decides each one

1. NVIDIA (NVDA)

Wednesday after the close · ±4.78% priced, against a usual move of 3.15% · $2.09 a share expected on $91.96B · beat 5 of 5, and the stock fell after 4 of them

NVIDIA has beaten the profit estimate in each of its last five reports, and the stock closed lower after four of them. That is the whole problem with Wednesday. The company keeps clearing the bar, and the bar is no longer what the stock trades on. In February it beat by 5.2% and fell 5.5%.

The reported quarter is already modelled almost to the cent: twenty-six analysts sit inside a seven-cent range, and management guided to $91.0B plus or minus 2%. So the number that moves the stock is the one for a quarter that has not happened yet. Options are priced for a move of about ±4.78% either way, roughly $1,037 on 100 shares at Thursday’s $216.85, and on a $5.32T company that is $254.1B of market value on paper, more than any other name this week.

It goes in up 2.3% over the past month, 8.3% below its 52-week high, and slightly below where it stood before the May report. Results land about 4:20pm ET, the call at 5pm.

What decides it: the revenue guide for the October quarter, how data-centre growth splits between the largest cloud buyers and everyone else, whether gross margin holds the 75% management promised, and whether the call puts a date on Vera Rubin shipments. The risk is the obvious one: a beat alone has not been enough four times running.

The full NVIDIA deep dive, with three scenarios and what each has done to the stock before, is here:

2. Synopsys (SNPS)

Wednesday after the close · ±7.37% priced, against a usual move of 5.16% · $3.67 a share expected on $2.44B · beat 4 of 5, and the stock fell after 4 of them

Across its last five reports, the average gain on the days Synopsys rose was 0.3%. The average drop on the days it fell was 12.8%, including a 35.8% collapse last September. Nothing about the coming quarter says a fall is likely, but the shape of this stock’s history is lopsided in a way almost nothing else on the list is.

The price already carries a bad year. It reports 35.4% below its 52-week high and about a quarter below where it stood before the May report, having gained 5.4% in the past month. Management raised the full-year revenue guide last time, to $9.625B to $9.705B, on the view that the chip-design licensing business had bottomed. Wednesday is the first full check on that.

Twelve analysts cover the quarter, the fewest of the five, and a move of ±7.37% is about $2,933 on 100 shares.

What decides it: whether licensing shows the improvement the raised guide assumed, and what the Ansys engineering-software business contributes. The risk is that the downside here has historically been the far bigger half.

3. Intuit (INTU)

Tuesday after the close · ±6.97% priced, against a usual move of 5.03% · $3.58 a share expected on $4.27B · beat 5 of 5, and the stock rose after 3 of them

Intuit’s last five reactions averaged 11.9%, against 5.7% for the five before. The market has been treating each report as more consequential than the last, and the most recent one was a 20.0% fall on a small profit beat. When this stock has gone wrong it has gone wrong badly: an average drop of 12.5% against an average gain of 5.3%.

Since that fall it has climbed 27% in a month while two analysts cut their ratings, an unusual pair of facts to carry into a print. It still sits 48.7% below its 52-week high.

This is the fiscal year-end report, which means Tuesday evening is about next year more than this quarter. The quarter itself is expected inside management’s own $3.56 to $3.62 range, so the guide is the event. A move of ±6.97% is about $2,522 on 100 shares.

What decides it: the full-year guide, TurboTax Live growth against the 38% management put on it, and the cheap end of do-it-yourself filing after this year’s pricing changes. The analysts who cut were pointing at that last one.

4. CrowdStrike (CRWD)

Wednesday after the close · ±5.9% priced, against a usual move of 4.15% · $0.29 a share expected on $1.44B · the stock has moved with the profit surprise in 4 of its last 5 reports

Most stocks on this list have stopped moving with their profit surprise. CrowdStrike still does, which makes Wednesday simpler to read than the rest: in four of the last five reports the stock went the way the surprise pointed. Its reactions have also been the tightest of the five, worst of the last five at −5.8% and best at +4.6%.

One complication. The July inflation reading lands the same morning, so a large enough macro surprise can move the stock before the report has anything to do with it.

It goes in up 1.0% over the month, 14.8% over three months, and 16.3% below its 52-week high. Twenty-nine analysts cover it, more than any other name here, and a move of ±5.9% is about $1,123 on 100 shares.

What decides it: new annual recurring revenue against the $284M to $286M management guided, and whether the AI security products convert the pipeline described in June. The sceptical question on the last call was whether that spending is new money or budget moved from somewhere else.

5. Workday (WDAY)

Thursday after the close · ±7.28% priced, against a usual move of 5.16% · $2.61 a share expected on $2.64B · beat 5 of 5, and the stock fell after 3 of them

Workday goes into Thursday evening having run 49% in a month and 62% since its May report. It has also beaten the profit estimate in all five of its last reports and fallen after three of them, so a beat is not the thing to watch here either. When it has fallen it has fallen hard: an average drop of 7.7% against an average gain of 3.7%.

Two things make the move potentially sharper than the ±7.28% priced, about $1,436 on 100 shares. About 12.7% of its freely traded shares are held short, roughly five days of normal trading volume, and the twenty-three analysts covering it disagree more about the number than on any other name in this five.

It also reports last. By Thursday evening Intuit, Zoom, Synopsys, CrowdStrike, Veeva, Okta and Nutanix have all printed, so the market will have decided something about business software before Workday says a word.

What decides it: new contract value, whether the AI-agent products approaching $500M of recurring revenue show up in the guide, and whether that guide pushes estimates higher. After a run like this, a guide that merely meets expectations gives the stock nothing to work with.

Read-through: the software chain runs Tuesday to Thursday

Ten of our twenty are software and internet names reporting in sequence: Intuit and Zoom on Tuesday evening, then Synopsys, CrowdStrike, Veeva, Okta and Nutanix on Wednesday, then Workday, Rubrik and Elastic on Thursday.

By our measure their earnings moves have been similar in size 67% of the time over the last five quarters, the strongest chain this week. If Tuesday’s two prints move the group, the later names arrive into a market that has already repriced the theme, and Workday, the last of them, carries the most run-up into it.

The banks are a shorter chain: Bank of Nova Scotia on Tuesday morning, then TD and Royal Bank on Thursday morning.

The other 15 on the list, by the day they report

NVIDIA, Synopsys, Intuit, CrowdStrike and Workday are above. These are the other fifteen, by day, with the priced move, how it compares with each stock’s usual reaction, and the track record. Four carry a caveat: Trip.com, TD, Royal Bank and Bank of Nova Scotia have no option expiring near their report, so their figure covers about three weeks of ordinary trading and is not the earnings move. For PDD, Gold Fields, Trip.com and the three Canadian banks the estimate feed is in local currency, so no profit estimate is shown.

What the data leans toward: a buy, hold or sell view on all 20

The views below are general information for a general audience, not personal advice; the note at the end of the issue applies to them.

Everything above describes what is priced and what has happened before. This section adds a lean: for each name, a buy, hold or sell view with a confidence level, built from how the options market is positioned going in and from recent changes in analyst ratings. It is a starting view, not a forecast, and this week most names read “hold, low confidence”, because the signals behind them are weak or disagree. Where the view is firmer, it says so.

Five of the twenty in full:

Zoom (ZM), Tuesday after the close: Buy, high confidence. Options positioning leans to the upside and no analyst has cut the stock. Priced ±7.20% against a usual 9.85%. Rose after 3 of its last 5 reports, the last one by 9.2%.

Elastic (ESTC), Thursday after the close: Hold, medium confidence. Positioning leans up, one analyst cut in 90 days, so the signals split. Priced ±15.56%, the largest on the list, against a usual 12.31%. Fell after 4 of 5, all of them profit beats.

Royal Bank of Canada (RY), Thursday before the open: Hold, low confidence. No readable lean. Fell after 3 of 5; its beats have narrowed each quarter.

Bank of Nova Scotia (BNS), Tuesday before the open: Hold, low confidence. No readable lean. Rose after 4 of 5 on a usual move of 1.03%.

Nutanix (NTNX), Wednesday after the close: Hold, low confidence. Positioning leans up, no analyst data. Priced ±10.62% against a usual 4.81%. Beat profit 5 of 5, fell after 3 of 5.

The fifteen highest-attention names, including the five the data leans against going in.

The five the data leans against. Where the view is sell, or where the stock’s own history says the downside has been the bigger half, with what the priced move means in dollars for a holder of 100 shares at Thursday’s price.

If you hold these five, pay attention this week:

Read more

In four of those five, the stock closed lower the next day.

It reports again on Wednesday 26 August, after the US market closes.

This piece covers what the options market is pricing for Wednesday, what the company itself promised back in May, what has happened since that promise was made, the three line items that have decided the reaction in past quarters, and what happens when.

The rest of this week’s reports are in today’s Sunday Earnings Intel Report. This one is about the report everything else this week gets measured against.

Do you own any stock that reports earnings this week?

Most investors find out what the market expected only after the stock has already moved.

Earnings Intel arrives the Sunday before, free: for each big report, you get the expected numbers, the move the options market is pricing, and how each stock has reacted to its own past reports.

Subscribe for free below so that you don’t miss the next Earnings Intel report.

Subscribe now

What the market is paying for, and what this stock has actually done

The options market is pricing a move of about 4.8% in either direction on the first session after the report, based on Friday’s data.

On a $216.85 share price, that is roughly $10.37 a share, or about $1,037 on a holding of 100 shares. Counted against NVIDIA’s market value before the report, it is a swing of about $254 billion.

The way to read that number: options prices carry an estimate of how far a stock is expected to travel by a given date. Strip out the ordinary day-to-day movement and what remains is the market’s estimate of the earnings reaction itself. It carries no view on which way.

Set that against what NVIDIA has actually done. The median move on its last five reports was 3.15%, and the average was 2.8%. A 4.8% move would be larger than four of those five reactions, and smaller than one.

Here is the record:

Two things in that table are worth holding onto. Five beats out of five is a clean record on the numbers themselves. One positive session out of five is a separate fact about how the market has received them.

The stock has not gone nowhere over that period. The closing price before each of those five reports rose from $134.81 to $223.47, a gain of about 66% across the span. The gains arrived between the reports rather than on them.

Going into Wednesday, NVIDIA trades at $216.85, about 8.3% below its 52-week high of $236.54 and well above the 52-week low of $164.07. It is down about 3% since the last report in May, up 2.3% over the past month and up 0.7% over three months.

What management promised in May, and where the estimates sit against it

On the May call, management gave specific numbers for the quarter that has now finished.

Revenue of $91.0 billion, give or take 2%. Non-GAAP gross margin of 75.0%, give or take half a percentage point. Operating expenses of about $8.3 billion. A full-year gross margin in the mid-70s.

One detail in that guidance matters more than the headline: it assumed no data centre compute revenue from China at all. Anything China contributes to the quarter arrives on top of a promise that excluded it.

Published estimates now cluster at about $91.96 billion of revenue and $2.085 of adjusted earnings per share, drawn from 26 analysts, with individual estimates spread from $90.55 billion to $93.33 billion.

That consensus sits above the midpoint the company guided to. There is a reason for the gap. Three months ago management guided the April quarter to $78.0 billion give or take 2%, and reported $81.615 billion, comfortably above the midpoint. Analysts appear to be modelling a repeat of that behaviour, which means clearing the company’s own guidance and clearing the published estimate are two different bars, and the second one is higher.

Estimates have barely moved in the run-up. Earnings estimates are up 0.07% over the past week and 0.17% over two weeks. Revenue estimates are up 0.05% over the week. Nobody has been revising toward anything in the final stretch.

The forward number is where agreement breaks down. Public estimates for the October quarter range from about $101.88 billion at one provider to about $103.46 billion at another, and the two use different analyst populations and update at different times. The honest way to hold that is as a range of roughly $102 billion to $103.5 billion rather than a single figure.

What has happened since that call

Fourteen events between the May call and this week bear on whether the promise holds. Most of them come from other companies.

The demand side kept confirming itself. Meta reported $31.08 billion of capital spending in a single quarter and raised its 2026 guidance to a range of $130 billion to $145 billion. Microsoft said AI demand still exceeded its supply and expects to deploy Vera Rubin. Amazon said its AI and chip businesses each passed a $25 billion annualised run rate. TSMC’s monthly revenue rose through May, June and July.

The competition kept confirming itself too. Broadcom reported $10.8 billion of AI chip revenue in a quarter and guided the next one to about $16 billion. AMD’s data centre revenue reached $6.718 billion, up 107% year on year. Microsoft promoted the economics of its own Maia 200 chips in the same breath as its NVIDIA deployment. Amazon’s Trainium commitments are scaling.

Two NVIDIA announcements sit slightly apart from both groups. On 10 August the company announced agreements with large financial institutions aimed at mobilising more than $500 billion of third-party capital for AI infrastructure. On 17 August it announced credit support and a $1.5 billion investment tied to an AI campus whose large capacity phases begin in 2028. Both expand the amount of buildout that can be financed. Both also make the capital behind future demand a more visible part of the story than it used to be.

Read together, that evidence gets to an estimate rather than a certainty. The research behind this piece puts the quarter at roughly $92 billion to $94 billion of revenue, with data centre at roughly $85.5 billion to $87.5 billion and gross margin near the 75% guide, and rates its own confidence in that as medium. The reason it is not higher: TSMC’s revenue is a whole-industry proxy rather than an NVIDIA shipment count, and July’s figure spans NVIDIA’s quarter end.

The three lines that have decided the reaction

Data centre revenue. This is most of the company. In the April quarter, data centre brought in $75.246 billion of $81.615 billion total. Within that, hyperscale customers accounted for $37.869 billion and the rest of the AI and enterprise business for $37.377 billion, growing 12% and 31% respectively from the previous quarter. Networking, at $14.8 billion, grew 35%. The consensus for data centre revenue this quarter is about $85.67 billion.

The October guide. Management gives next-quarter guidance in written commentary published immediately after the results. Against a public bar spread across $102 billion to $103.5 billion, where that number lands relative to the range matters more than where it lands relative to any single provider’s figure.

Gross margin. The promise was 75.0%, give or take half a point, with the full year in the mid-70s. The April quarter came in at 75.0% exactly. The pressure points named in the research are Vera Rubin transition costs, a 25% tariff tied to H200 imports, product mix and supply-chain economics.

Behind those three sits a commitment with a date attached: management said Vera Rubin production shipments would begin in this October quarter, which makes Wednesday the first point at which that promise becomes testable.

Three ways Wednesday goes

Both bars cleared. Revenue at or above roughly $94 billion with data centre clearly above the mid-$80 billions, gross margin at or above about 75%, an October guide above the top of the $102 to $103.5 billion range, and Rubin shipment timing reaffirmed with named customer deployments. That combination answers the questions raised on the last call rather than deferring them.

The quarter beats, the forward line holds. Revenue modestly above the $92 billion cluster, earnings above $2.09, and an October guide sitting inside the public range rather than above it. The research describes this outcome as carrying less information than the word “beat” suggests, because a beat of this shape has already happened four times in the record above.

Something in the forward set breaks. Revenue lands below the estimate cluster, or data centre mix disappoints, or margin guidance weakens, or Rubin timing slips. The mixed version, where the reported quarter beats and the forward evidence disappoints, is the one that has no clean read.

For a holder of 100 shares, a 4.8% move is about $1,037, and the arithmetic is the same size in either direction.

What the model says. It declines to take a side. The positioning read comes back as hold at low confidence, with none of the three pressure indicators triggered and no clear signal from any of them. The research puts it plainly: no high-confidence directional edge, because the headline quarter is so heavily modelled that the published estimates already carry most of what can be publicly known. Where public evidence is thinner, it argues, is the composition of the quarter, the Rubin conversion and the financing behind customer spending.

What would make all of this read wrong

Five quarters is a short record. Every pattern above rests on it, and a five-observation sample supports a description of what has happened, not a rule about what happens next.

Consensus is not one number. The figures printed here come from a 26-analyst pipeline set; other providers show different numbers on different timestamps, and the gap widens on the October quarter.

Several inputs look more solid than they are. The traded earnings whisper of $2.15 sits alongside a revenue whisper that appears stale, so it counts as weak evidence about appetite for another beat and nothing more. Reported H200 shipments to China are not recognised revenue, since licensing, delivery, acceptance and tariff economics are separate steps. Daily short-sale volume from FINRA is not short interest. The April quarter’s reported net income was flattered by large unrealised gains on equity investments, which is why the operating figures are the useful ones here.

The $1 trillion of Blackwell and Rubin revenue visibility that management has cited is a demand estimate rather than a contracted order book, and public filings do not allow anyone outside the company to reconcile it. The April quarter also disclosed roughly $145 billion across inventory, purchase commitments and supply-related prepayments, which is a large bet on demand arriving on schedule.

The demand evidence and the competition evidence in the timeline above can both be true at once, and treating either as proof against the other would be too simple.

This piece rests on 33 sources, 20 of them primary company or regulatory documents, and the research behind it is point-in-time through 21 August. Estimate revisions, macro releases, China policy decisions and company updates after that date are not in it.

Wednesday, in order

US income and spending data lands before the market opens that morning, which can move long-term interest rates before NVIDIA reports anything.

Results are published at about 4:20pm New York time, roughly ten minutes after the close. The chief financial officer’s written commentary goes up on investor.nvidia.com immediately afterwards, and that document carries the October guidance. The conference call starts at 5pm New York time.

Four questions are worth listening for on that call. How much of the October revenue step comes from Blackwell against Vera Rubin, and what milestone confirms Rubin started on time. Whether the October guidance includes any China data centre revenue, and on what licensing and tariff assumptions. How much of the $1 trillion visibility figure is firm purchase commitments rather than customer forecasts. And where, if anywhere, management is seeing measurable change in workload share as customers scale their own chips.

Federal Reserve communications from Jackson Hole run from 27 to 29 August, the two days immediately after, which is worth knowing before attributing everything the stock does that week to the report.

The reported quarter ended on 26 July 2026 and the company calls it the second quarter of its 2027 fiscal year, which is why the fiscal labels in press coverage will not always agree.

The rest of the week’s reports, and what is priced into each of them, are in today’s Sunday Earnings Intel Report.

Do you own any stock that reports earnings this week?

Most investors find out what the market expected only after the stock has already moved.

Earnings Intel arrives the Sunday before, free: for each big report, you get the expected numbers, the move the options market is pricing, and how each stock has reacted to its own past reports.

Subscribe for free below so that you don’t miss the next Earnings Intel report.

Subscribe now

This publication provides research and educational content for investors rather than investment advice. We present valuation data, options-market information, and model estimates along with their assumptions; however, we do not offer buy or sell recommendations, price forecasts, or specific portfolio advice. All figures are timestamped as indicated and may be updated. Always conduct your own due diligence or seek advice from a licensed professional before acting on any information here.

Most investors find out what the market expected only after the stock has already moved.

Earnings Intel arrives the Sunday before, free: for each big report, you get the expected numbers, the move the options market is pricing, and how each stock has reacted to its own past reports.

Subscribe for free below.

Subscribe now

Last week, Cisco beat on both lines and fell 8.4%. Tapestry beat and fell 16.5%. Nebius jumped 34% in a day.

If Walmart, Home Depot, Target or Alibaba sits in your portfolio, this is the week the market re-judges it.

Earnings day is when a position you’ve barely thought about for months moves like that, and by then it’s too late to go looking for context.

That’s what Earnings Intel is for.

You’ll know which of the week’s 66 reports actually touch the stocks you own and which are noise.

You’ll know what the options market has already priced into each big name, so when the move comes you can tell a real surprise from a stock just doing what was expected of it.

For each of the five highest-stakes reports, you’ll know the specific number and the specific pattern that has decided the reaction in past quarters, the things worth checking before the market opens on the day.

And if you don’t own any of these names, the setups below are exactly the homework you’d want done before deciding whether any of them deserve a place on your watchlist.

First, the grade on last week’s list: what we flagged, what the pricing expected, and where both were wrong.

The week’s headline: $86B of earnings risk lands on Thursday

Sixty-six US companies in our coverage report this week, carrying a combined $160.9B in expected market-cap swing. Thursday holds $85.6B of it, 53% of the week, and almost all of that lands before the bell: Walmart, Alibaba, Deere, NetEase and Futu print Thursday morning, with Ross following after the close. If you own big retail, Thursday is your week.

Last week, graded: what options priced vs what stocks actually did (Aug 10–14)

The pricing came closer to honest last week. Across the 15 names with a clean straddle price, the median implied move was ±13.85% and the median realized move was 11.85%, a two-point gap where earlier weeks ran to multiples. Option buyers won 8 of those 15 and sellers won 7, the first card where the buyers edged it.

The other five names had no clean straddle price, so we graded them against how much they normally move: one moved more than normal, four stayed within it.

The beats came through again, 15 of 20 on EPS and 15 of 20 on revenue, and the reactions still went their own way:

• Nebius, priced for ±16.0%, moved +34.1% on an 83% EPS beat. The week’s biggest overshoot.

• Rocket Lab, priced for ±14.2%, closed 0.04% lower on an EPS miss. The flattest print of the week, and the biggest overpricing.

• Cisco beat on both lines and fell 8.4%.

• Tapestry beat on both lines and fell 16.5%.

The attention-rank correlation with actual move size printed at +0.00 this week, after −0.21 in the last issue. Our ranking measures where the money and the option flow are concentrated, and so far it has told us nothing about where the surprises land. We publish this number every week regardless of what it says.

Carry that into the rest of this issue: even in a week where the median pricing was nearly right, the single names ranged from dead flat to +34%. Now, the week ahead.

Do you own any stock that reports earnings this week?

Most investors find out what the market expected only after the stock has already moved.

Earnings Intel arrives the Sunday before, free: for each big report, you get the expected numbers, the move the options market is pricing, and how each stock has reacted to its own past reports.

Subscribe for free below.

Subscribe now

The day-by-day risk map: when each big report lands (Aug 17–21)

Three things the chart can’t say on its own:

• Tuesday morning stacks Home Depot against Housing Starts and Building Permits, all before or at the open, with BHP’s full-year numbers already out overnight. A housing-data surprise and an HD print in the same session will be hard to untangle in real time. Keysight and Toll Brothers also report Tuesday, but after the close, so their moves land Wednesday.

• Wednesday afternoon brings the FOMC minutes (2pm ET), hours after ADI, TJX, Target, Lowe’s, Estée Lauder and Viking all report in the morning. Anything rate-sensitive that prints Wednesday trades into that release.

• Monday and Friday are near-empty. Monday’s only name of size is Fabrinet, and it reports after the close, so nothing on Monday’s list can move during Monday’s session. Friday belongs to BJ’s alone.

The week is also unusually concentrated in one theme: this is retail’s quarter-end. Home Depot, Lowe’s, TJX, Target, Walmart, Ross and BJ’s all report within four sessions. Only 12 S&P 500 members report this week, 3.1% of index market cap, so the index itself has limited earnings exposure. The single-stock risk is where the action is.

Which brings us to the names carrying it.

The 5 highest-stakes reports this week, and what’s already priced into each

Sixty-six companies report, but five setups stand out, either because of sheer size or because the pricing tells a story. As always, what follows is a description of what’s priced in and what to watch. It is not a view on where any of these stocks go next.

Home Depot (HD) kicks off the retail run on Tuesday before the open, and the pricing here is the odd part.

HD is normally a sleepy earnings stock: its median post-earnings move over recent quarters is about 2%. This time, options are pricing ±3.7%, close to double that. Somebody expects this print to matter.

Two reasons they might be right. The stock is up 13% since its May report, so there’s a run to defend. And HD’s recent record with expectations is shakier than its reputation: it has beaten the EPS consensus in only 2 of its last 5 quarters, the weakest beat record among this week’s large caps, and its ugliest recent day (−6.0%, last November) came on a small miss. Consensus this time sits at $4.73 on $47.3B of revenue, from 21 analysts on the EPS line. One more wrinkle: Housing Starts and Building Permits hit at 8:30am the same morning. If housing data and the HD print disagree, the first hour of trading will be genuinely hard to read.

Walmart (WMT), Thursday before the open, is the biggest single event of the week at $42.9B of expected swing, and it comes in dragging a strange pattern behind it.

Walmart keeps beating and the stock keeps falling. Four EPS beats in the last five quarters; four negative next-day closes in those five. In May it beat on both lines and dropped 7.3% anyway, and it’s down 11.9% since.

The consensus asks for $0.74 on $186.7B, though the last five quarters suggest the beat is the easy part. What the beat is made of has decided the reaction. What stands out is how routine the options market is treating this: the ±4.6% priced is almost exactly Walmart’s normal earnings-day move, despite the streak.

Alibaba (BABA), also Thursday morning, is the week’s biggest argument.

Analyst EPS estimates for this quarter are scattered across a 53% spread, the widest disagreement on our entire list. Last quarter EPS landed 91% below consensus as the company’s spending ramp swallowed profits, and the spread now is a real fight over how deep and how long that investment cycle runs. When nobody agrees on the number, nobody agrees on what a beat even is.

Here’s the twist: despite all that disagreement, options are pricing ±6.5%, a touch below BABA’s usual 7.6% earnings-day move, with a mild lean toward the call side. The stock still sits 36% below its 52-week high. In recent quarters the reaction has been set by what the company says about cloud and AI capex, which usually lands well after the headline EPS line has been read. One caveat we owe you: parts of the estimate feed for this name mix reporting currencies, so hold the precise consensus figures a little loosely. We’ve flagged it in our data.

Analog Devices (ADI), Wednesday before the open, is the least dramatic setup on the list, which is why it’s here.

Five straight EPS beats, estimates sitting in a 3% band, options pricing an ordinary ±4.9% against a 4.6% usual move. Consensus: $3.34 on $3.91B. With a beat essentially assumed, the information in this report lives in the guide. ADI sells analog chips into industrial, auto and comms, which makes it the closest thing this week has to a reading on the broad economy. Worth remembering that a routine beat guarantees nothing: last May the company beat by 7% and closed down 3.9%.

And then there’s Target (TGT), Wednesday morning, the most stretched setup on the list.

The stock is up 21% since May and sits at 97% of its 52-week range, within 2% of the high. The odd part is how calm the positioning looks: short interest fell 17% into the print and the options skew is flat, so the hedging pressure you might expect at the high simply isn’t showing up in the data.

Consensus wants $2.26 on $26.1B. Target’s recent earnings days argue for humility: the median next-day move over the last five prints is −3.9%, including a 3.9% drop in May on a 16% EPS beat. Wednesday tests whether a beat is enough at the top of the range. Over the last five quarters, a beat on its own has not been.

Read-through: Home Depot on Tuesday sets up Lowe’s on Wednesday

There’s a reason we spent so long on Home Depot: it doesn’t just report for itself. HD and Lowe’s have posted next-day moves of similar size in 84% of comparable quarters, the strongest pairing in this week’s list. HD prints Tuesday, Lowe’s prints Wednesday, so by Wednesday’s open the market will have already marked LOW off whatever HD said. Options there are pricing ±3.2%, close to double Lowe’s usual move, partly because Wednesday is also FOMC-minutes day.

The discount-store sequence, Target (Wed) → Walmart (Thu) → BJ’s (Fri), looks like a chain, but move similarity across the three is a weak 50%. We’d read each on its own setup.

The other 15 names on this week’s list, one line each

Home Depot, Walmart, Alibaba, Analog Devices and Target got the detail above. Here are the other fifteen reports on this week’s list, grouped by the day they land, with what the options market is pricing for each.

Wildcards outside the top 20, flagged for unusual pricing only: Mercury Systems (Tue after close, ±17.2% priced vs 10.5% median history) and Klarna (Tue, ±17.7% priced, which is actually below its 20.3% median; one of the few names where options are pricing less drama than the stock usually delivers).

The chart below draws the eight most unusually priced names from the whole top 20, ranked by how far the priced move sits above each stock’s own history. This week none of the five featured names makes the cut; the strangeness lives further down the list.

Next Sunday: this week’s list gets graded

The frozen list above becomes next week’s report card: 20 names, graded against exactly these numbers, including whatever we got wrong.

This publication provides research and educational content for investors rather than investment advice. We present valuation data, options-market information, and model estimates along with their assumptions; however, we do not offer buy or sell recommendations, price forecasts, or specific portfolio advice. All figures are timestamped as indicated and may be updated. Always conduct your own due diligence or seek advice from a licensed professional before acting on any information here.

The week in 60 seconds

Earnings season has slowed from high to moderate activity. This week, 249 companies with revenue over $1 billion will report earnings, a decrease from 943, reducing the expected total earnings swing from $1.53 trillion to $234.6 billion. Only ten S&P 500 companies are scheduled to report, representing 1.8% of the index’s market cap, down from 14.2% last week.

This week focuses on key macroeconomic events. July CPI data will be released on Wednesday at 8:30am, followed by PPI on Thursday, with retail sales and Michigan sentiment reports scheduled for Friday. Wednesday is particularly busy, with the inflation report in the morning and Cisco's earnings announced after market close, expected to cause a $75.0 billion fluctuation. Investors will scrutinize Cisco's after-hours earnings in relation to the CPI figures released earlier that day.

Concentration has significantly risen. Last week, our 20 stocks contributed to 23% of the expected weekly movement, while this week, they represent 65%. Despite fewer reports, the remaining ones encompass most activity, primarily focused on AI infrastructure. The top fourteen features include seven key companies: Lumentum, Cisco, Applied Materials, Coherent, CoreWeave, Nebius, and Cerebras.

Method summary: The attention score integrates the expected dollar swing—calculated as market cap times implied move—with factors like liquidity, options-market focus, and how rare this quarter's pricing is compared to the company's history. It indicates where investments are heading rather than surprises from the company. The report card below illustrates what that focus translated to last week.

Report card #1: options priced 10.8%, stocks moved 3.3%

As promised last Sunday, each issue now begins with a public grade of the previous week’s 20, including the parts that embarrass us.

From August 3–7, expectations for a volatile week in the options market were high, with fireworks anticipated. However, only three major movements actually took place. The median implied move across 20 options was 10.8%, but the actual median move was just 3.3%. In 15 of 18 cases with specified event-driven expiries, the actual moves were smaller than what options pricing had forecasted.

The sequence of events, arranged by their significance, is as follows:

• Three stocks significantly exceeded expectations, with some experiencing substantial gains. Palantir soared by +29.5%, well above the anticipated 11.4%, marking the largest weekly overshoot, supported by a 19% EPS beat and 175 million shares traded. Airbnb rose by +17.4% compared to 7.4%, and Shopify gained +17.0% after an initial rise of +21.7%. The growth in short interest, which increased by 25% before earnings—something we had previously noted—likely contributed to this upward trend.

• Beats were common, but they didn’t indicate direction. 80% of the 20 EPS beats and 85% exceeded revenue expectations. CVS beat EPS estimates by 38%, the biggest surprise, yet it closed 5.1% lower. Astera Labs surpassed both earnings and revenue forecasts but fell 12%, its second consecutive double-digit decline despite beating estimates. AppLovin exactly met its EPS estimate and moved 0.45%, compared to a 14% shift, making it the week’s biggest shortfall.

• The CVS puts we identified were partly accurate. We saw that the options indicated downside risk for a company that hadn’t missed earnings in five quarters. The stock opened with a 10.3% gap down, then recovered 5.8% during the day before ending down 5.1%. While the concern was justified, the panic subsided by lunchtime.

• The booking anomaly was explained through historical behavior. Options were priced for a 6.6% move—six times the usual market response—but only moved 0.8%. When a stock's history indicates stability, yet the options market signals volatility, last week’s historical trend was prioritized.

• Gap-and-fade was the predominant pattern this week. Six stocks opened with gaps exceeding 10% (CAT +11.1%, ANET +10.3%, SHOP +21.7%, CMI -10.8%, CVS -10.3%, SNDK -13.1%), but they all closed closer to their prior levels. Cummins, in particular, regained 9.9% during the trading day. Each night, the assessments consistently outstripped the daytime evaluations.

• Caterpillar extended its winning streak to six consecutive wins, outperforming EPS estimates by 31% and posting a 5.6% increase.

Once again, our attention rank displayed a -0.21 correlation with actual move size, marking the second week in a row of a negative relationship. For instance, Astera Labs, ranked first, moved 12%, while Palantir, ranked twelfth, moved 29%. The score shows where money is at risk but doesn't predict where surprises may occur. We will keep publishing this weekly correlation, regardless of what it suggests about us.

A quick note on grading: since trust outweighs tidiness, the grading archive froze its top 20 rankings on August 2, the day after our data cutoff. Four companies we highlighted—AMD, Uber, Datadog, and Vistra—slipped just outside the frozen list, while Airbnb, ON Semiconductor, McKesson, and Cummins entered. Going forward, our grading will be based solely on the exact snapshot we publish.

SpaceX after the print

Last Sunday, we acknowledged that SpaceX’s initial report highlighted a limitation in our approach: a market cap of $1.52 trillion with a 16.6% implied move, implying roughly $250 billion in expected fluctuation. It didn't make it into our top 20 rankings because these rankings rely on historical data, which SpaceX lacked. Here’s how the market responded to a debut without an anchor.

The stock fell about 7 to 8% the day after Tuesday’s report but rebounded most of its losses by the end of the week, closing near its pre-report level on Friday. Two main points are noteworthy. First, even SpaceX followed this pattern: options pricing at 16.6% reflected less than half of that stock movement—similar to what occurred in seventeen out of twenty cases graded. Second, the dollar impact remained significant: a 7 to 8% swing in a $1.52 trillion company equates to roughly $110 billion changing hands over four sessions, exceeding the total movement of any single graded stock that week, including Palantir’s 29% one-day move. The market’s initial response to a no-history report lasted about three days before reversing completely.

A transparency note: SPCX was not on the frozen grading list, so the figures above are estimates rather than pipeline-graded data. It will be included in the ranking and report card once it has accumulated sufficient history for our method.

The Five

This is a comprehensive analysis of the top five most important names for this week. The fair-value model remains under development; "setup" indicates the current price, positioning, and the company’s guidance as they appear now.

LITE · Tuesday after close

Key figures: consensus EPS is $2.97 (within a narrow $2.89–$3.06 range among 13 analysts), with revenue totaling $987.7 million.

Setup: Our top attention score this week declined by 15.7% since May's high. Although earnings per share (EPS) beat estimates and fell 5.1%, it has increased 18.5% over the past month but remains 23% below its $1,086 peak. The company has reported five straight EPS beats, yet revenue has only exceeded forecasts in two of the last five quarters, indicating recent results are driven more by earnings than revenue. Barclays upgraded the stock to Overweight on July 20, with a target 29% above the current price. Insiders sold $35.7 million worth of shares during this period, with no insider purchases.

Implied move: 11% on the clean term-stripped read and 15% on the raw straddle, versus a typical 5.1%. Both basis are priced 2 to 3 times higher than usual, leaning towards calls.

Red flags are a 40% beat rate on the revenue line and a stock decline following its last beat.

Key points to watch are revenue staying within the $984–997M analyst range, commentary on AI transceiver demand, and margin trends.

CSCO · Wednesday after close

Key figures include an EPS of $1.17, which is within a three-cent range of estimates, and revenue totaling $16.82 billion. The most significant dollar movement expected this week is $40.4 billion.

Setup: Cisco's profile has grown more intriguing. Its latest two reports indicated earnings changes of +13.4% and -12.3%, representing the most significant fluctuations in our data, primarily driven by EPS beats of 2.9% and 2.0%. Despite five consecutive modest beats, market reactions haven’t mirrored this pattern. Since May, the stock has risen 18.7%, but it still trades 7% below its peak, with short interest also rising by 20.5% before the most recent report.

The implied move stands at 8.1%, which is 1.7 times higher than the median of 4.9%. This indicates that the options market has revised its perception of what a Cisco quarterly report implies.

Red flags: The small-beat, large-move approach can go either way; the report comes in hours after CPI, so Wednesday’s macro data affects the mood before the call starts.

Observe whether guidance, instead of the quarter, is driving the move again; compare the beat size to the last five, all under 5%; and assess short-covering pressure if the figure appears softer.

AMAT · Thursday after close

Key figures: EPS of $3.39 surpassing the company’s forecast of $3.36 ±$0.20; revenue reaching $9.0 billion against the expected $8.95 billion ±$500 million. Consensus closely matches the guidance, making the primary challenge to outperform market expectations, which are subtly incorporated into the forecast.

Since May's report, the stock has increased by 19.7%, outperforming the previous gain of 6.7% (which was a decline of 0.9%). It is now 29% below its recent peak. Despite recording five straight earnings beats, the market responded positively only twice. A year earlier, it also beat estimates but fell 14.1%. Management's projections for 2026 include over 30% growth in equipment and more than 50% in advanced packaging. CEO Gary Dickerson sold $50 million worth of stock at prices between $700 and $735 in late June; the stock currently trades around $527. Insider sales total $170 million, with no insider purchases reported.

Implied move: headline at 10.1%, but the closest expiry is the day after the print date, showing a cleaner event-day reading of 7.5%, 1.4 times the historical average. The skew heavily favors calls.

Red flags include the echo from a year ago, China's pressure on gross margin, and the absorption of ICAPS demand, which was the analysts’ main concern during the last call.

Monitor key points such as gross margin, the 2027 WFE language, advanced packaging in relation to the 50% claim, and Applied Global Services versus its target in the mid-teens.

RKLB · Monday after close

Key figures show a consensus revenue of $231.6M, which is slightly above the midpoint of the company’s $225–$240M guidance range; EPS is around -$0.06.

Setup: The May print experienced a +34.2% jump, making it the biggest one-day response in our data, but the stock has since erased all those gains, declining 3.7% and halving from its peak. In early July, Peter Beck sold $96.5 million worth of stock over two days, part of $363 million in insider sales during that period, with no reported purchases. Despite this, retail investors haven’t shown concern: there were 2,813 mentions tracked, showing the second-highest bullish tone of the week, and the options skew is +15.7 toward calls, signaling a strong bullish sentiment.

Implied move: 13.9% (excluding the term, 11.2%), which is 2.8 times higher than the median of 4.9%.

Red flags comprise Neutron’s delayed launch window in late 2026, which is both the primary thesis and a major concern for analysts; Space Systems' margin structure; and the expected negative free cash flow through 2026, as per management’s guidance.

Concentrate on any date-related language in Neutron, the gross margin in comparison to the 38–40% target, and the backlog conversion rate versus the 36% goal over twelve months.

CAVA · Tuesday after close

This week's highlight is the 'unusualness' star, combined with the messiest estimate image.

Key figures indicate a headline EPS consensus of $0.18, but the detailed analyst average is $0.14. Estimates from individual analysts range from $0.06 to $0.16, making it the second-largest disagreement this week. Revenue projections vary between $311 million and $379 million. Overall, there is no clear consensus on expected performance for this quarter.

Setup: Shares have decreased by 20.3% since May’s peak, surpassing EPS by 18% and revenue by 22%, and have increased by 3.1% during this period; however, they are still 37% below their highest point. Over the past three months, upgrades from Morgan Stanley, UBS, and Argus have improved the stock’s outlook, with a consensus target about 50% higher than current levels, despite investor Artal’s $271 million block sale in June.

Implied move: 14.2% on the clean read versus a 3.1% median, marking a 4.6x increase—the highest for the week. Still, caution is needed: the option chain backing this has mediocre quality with wide spreads, so interpret the level informally and focus on the overall trend.

Red flags include last August’s echo, which exceeded EPS estimates by 23% but then declined by 16.6%, along with the significant variation in estimates.

Monitor the exact EPS figure the print matches, compare revenue to the $311–379M range, and check if the upgrade cycle continues throughout the quarter.

Priced unusually, and priced backwards

After a week where most options were heavily overpriced, leading to investor losses, the market has now corrected its prices. Currently, only 4 out of 20 carry moves are at or above twice their historical levels, a decrease from 12 last week. CAVA trades at 4.6x, QXO at 2.4x, Rocket Lab at 2.8x, and Lumentum also appears on the list. It's important to note a chain-quality concern for QXO due to its wide spreads, which raise questions about its valuation.

The stranger story is located at the bottom of the chart, featuring two names priced below their historical values.

• CoreWeave (Tue AMC) trades at 0.8x. The stock has consistently declined after each of its five earnings reports as a public company, with a median decrease of 16.3% and a maximum drop of 20.8%. Options are priced at 13% ahead of this report, which is lower than its median. Since May, the stock has decreased by 34% and is down 44% from its peak, amid ongoing share sales by the CEO. Despite market expectations about this quarter, recent trading activity is more subdued than on any previous earnings day.

• Cerebras (Wed AMC) is trading at 0.7x and has a history of sharp reactions: it outperformed EPS by 76% in June but then fell by 19.6%. The 17.5% implied move is lower than its only previous earnings response, which occurred on a 27% free float. Management expects Q2 core revenue to nearly reach $194M, an 88% increase, though margins are temporarily expected to decrease by 10–15 points due to data-center rentals.

Further down the tape, CEVA is valued at 19.6%, much higher than the typical 1.8%. While this is the most extreme ratio this week, it involves a low-quality chain, so we view it as noise unless there's evidence to the contrary.

The AI hardware gauntlet

The week’s read-through chain follows two primary paths. The networking spine includes Lumentum on Tuesday night and Cisco on Wednesday night, together showing a 95.8% move-magnitude similarity across over five quarters. The AI compute chain involves CoreWeave on Tuesday night, Nebius on Wednesday morning, Cerebras on Wednesday night, and Applied Materials on Thursday night. Holding two of these stocks provides six data points for the week, with CPI data released between them.

The chain highlights three key points: CoreWeave and Nebius are almost complete opposites. CoreWeave has not received any positive feedback, while Nebius has four positive responses out of five, with a median increase of +4.2%. Last quarter, Nebius raised its 2026 capex guidance to $20–25B, emphasizing funding as the main focus during the call. Additionally, Applied concludes the week as both the chain’s verdict and its largest company, with the CEO’s June selling price remaining 30% above current market value.

Monday belongs to retail

The week kicks off with two key names in our coverage appearing on the same evening: Rocket Lab and AST SpaceMobile. AST SpaceMobile leads in mentions this week with 5,101 and also has the most optimistic outlook. Despite this positive sentiment, the company's fundamentals tell a different story: ASTS has failed to beat an EPS estimate in five attempts, missed last quarter’s revenue target by 62%, and has considerable analyst disagreement on revenue projections, ranging from $25.6M to $41M. Nonetheless, two upgrades were announced in the past month. Options are priced at 15.1%, favoring calls. For investors, satellites are crucial; management targets about 45 Bluebirds in orbit by year’s end.

Monday morning, Barrick is in focus as short interest rises by 32.5% ahead of the report, showing put skew, and the stock has beat earnings three times in a row. Ferguson, which Goldman downgraded on July 7, was upgraded by Zelman on July 17. Ferguson also displays put skew and is trading 7% below its recent high.

Further down the tape

Tapestry (Thu BMO) leads this week’s trading with five straight EPS and revenue beats. Although it is 1.2% below its 52-week high, its median next-day change is -9.6%, with a notable 12.3% drop after a 28% earnings surprise last quarter. It also has the strongest put skew of the week. Meanwhile, Nu Holdings (Thu AMC) has experienced three downgrades in 90 days, including a downgrade by Bank of America to Underperform, with only 6% of analysts anticipating a positive move. Data notes indicate that JD and NetEase have estimate discrepancies, and JD’s weak option chain suggests its 2.9% implied move may not be dependable.

The initial reports include Pershing Square Inc., which announced on Thursday morning that it is entering its first full quarter since its April listing, with a 6.7% implied move. Quantum firm Quantinuum is set to report on Tuesday night with an implied move of 17.9%, while geothermal IPO Fervo Energy will report on Wednesday morning at 13.4%. Regarding squeeze activity, Immatics reports on Wednesday with a 13.4-day-to-cover ratio against a 32.4% implied move, indicating the most volatile activity on the tape. A noteworthy insider event this week is that five EagleRock directors bought $10.5 million worth of stock in May, ahead of the company’s first earnings report scheduled for Monday after the market closes. This is significant because the company has no prior reporting history or insider sales, making such insider purchases before a debut rare and worth noting each time it happens.

What would change the week

On Wednesday at 8:30am, the CPI re-prices everything prior to any company comments that day, with the PPI following on Thursday morning before Applied Materials releases its report in the evening. Friday is relatively quiet, with only four reporters and $780M in potential moves, marking the end of the week’s main earnings event with AMAT on Thursday night. Friday's focus shifts to retail sales and Michigan sentiment. If inflation data remains stable, the top 20 trades for the week will be driven mainly by these figures; if not, significant price changes are unlikely on this page.

This publication provides research and educational content for investors rather than investment advice. We present valuation data, options-market information, and model estimates along with their assumptions; however, we do not offer buy or sell recommendations, price forecasts, or specific portfolio advice. All figures are timestamped as indicated and may be updated. Always conduct your own due diligence or seek advice from a licensed professional before acting on any information here.

100% written by a human analyst

Next week in 60 seconds

Tuesday stands out as the key day of the week. Of the expected $1.53 trillion market-cap change across 943 US companies with over $1 billion in reports this week, $687.6 billion happens on Tuesday, August 4. This accounts for 45% of the weekly total in a single session, with Caterpillar and Merck releasing reports before the market open, followed by AMD, Arista, Western Digital, AppLovin, Astera Labs, and SpaceX’s first report after the close.

Wednesday has the highest count with 312 reporters and an expected swing of $423.8B. Monday is dominated by Palantir and Booking, while Friday is mostly empty, aside from Vistra and the July jobs report.

A total of 136 S&P 500 companies report, with 14.2% of the index’s market cap re-pricing within five days. The macroeconomic calendar accompanies this, featuring ISM Manufacturing on Monday, JOLTS data on Tuesday, ISM Services on Wednesday, and payrolls on Friday at 8:30 am.

Our selection process for 20 out of 943 involves assigning each name an attention score based on expected dollar swings (market cap times implied move), trading liquidity, options market activity, and how its current quarter’s pricing compares to its historical patterns. This score indicates the potential money at risk but doesn’t predict surprises. Therefore, starting next Sunday, each issue begins with a report card on the previous week’s 20: what we identified, what actually moved, and where the pricing was off. We openly evaluate our performance, even when it’s embarrassing.

The Five

AMD · Tuesday after close

Key figures include a consensus EPS of $1.61 and revenue of $11.3 billion, based on August 1 estimates. The company forecasted approximately $11.2 billion in revenue, representing a 46% increase year over year, with a gross margin close to 56%.

Setup: The stock has increased by 96.7% since the February 3 report, which exceeded EPS expectations by 16%. Despite this, the stock declined by 17.3% the following day. AMD has exceeded estimates for four consecutive quarters, with a median next-day reaction of -2.3%. Over the past month, the stock has retreated by 8%.

Implied move: the headline figure is 9.7%, with a complication. The closest option expiry occurs one day after the print, and removing background volatility adjusts the event-day estimate to about 6.6%. In any case, the expected dollar change is $52 billion, making it the largest single-name figure of the week.

Red flags include management's own warning about gross-margin dilution as MI450 ramps up in the second half, the reduction in China data-center GPU demand, and rising memory costs affecting client and gaming sectors.

What to monitor: gross margin compared to the ~56% guidance, server CPU growth relative to the “over 70% year over year” statement, and any dates related to Helios shipments.

PLTR · Monday after close

Key figures include a consensus EPS of $0.35 and revenue of $1.81 billion, both exceeding the company’s guidance range of $1.797–$1.801 billion. Since the estimates already account for a beat against guidance, this positivity is already factored in.

Setup: The stock is down 41% from its 52-week high and has declined 16.7% since the February report, despite exceeding expectations that quarter. Retail interest remains strong: last week, there were 1,144 tracked social mentions with a notably bullish tone among our top 20. Analyst sentiment is improving, with three upgrades in the past 90 days.

Implied move: 11.4%, representing a $36–40B fluctuation. Interestingly, this is just 1.4 times Palantir’s past median of 7.9%. The options market views a double-digit move in Palantir as common.

Red flags include estimates exceeding guidance, defense revenue vulnerable to ongoing budget uncertainties, and a 150% net dollar retention rate that must keep surpassing expectations.

Key points to observe: US commercial growth relative to the “at least 120%” full-year guidance, the Rule of 40 score (145 in Q1 compared to the 129 projected for the full year), and the RPO growth (134% in the last quarter).

SHOP · Wednesday before open

Key figures: consensus EPS at $0.39 and revenue at $3.39 billion. Guidance indicates revenue increasing in the high-20% range, but gross profit is only expected to grow in the mid-20s, which raises the main question for the quarter. This gap suggests that growth is becoming more costly to achieve, and during the last call, analysts focused on AI token expenses.

Setup: 36% below its high. Short interest increased by 24.7% before the Jul 15 settlement. The streak of beating estimates is interrupted: last quarter, EPS missed by 9%, causing the stock to drop 6.7%, marking the third negative response in five reports. The only exception was a 22% gain.

Implied move: 12.0%, representing an $18.3 billion swing, which is 1.8 times the historical average.

Red flags: Two July upgrades (Jefferies and Stifel, both to Buy) increased expectations just before the report.

Key indicators to monitor include the gross profit versus revenue growth spread, payments penetration, and free cash flow margin, all compared to the mid-teens guidance.

CVS · Wednesday before open

This week's top attention score, along with its most unusual setup.

Key figures include a consensus EPS of $1.86 and revenue of $100.1 billion. In May, the full-year guidance was increased to an EPS range of $7.30–7.50 and at least $405 billion in revenue.

Setup: Since February, the stock has increased by 37.8%, remaining 5.6% below its 52-week high. It has reported five consecutive EPS beats, each with an average surprise of 22%. Despite these strong results, the median next-day market response is -0.2%. The stock outperforms expectations, but the market remains indifferent.

Implied move: 6.4%, well above the 2.0% historical median, over three times the usual price, with a strong preference for put options. In other words, after a 38% rise, the options market is costly for downside protection on a company that has not missed earnings in five quarters. The market's concerns are unlikely focused on EPS.

Red flags include the Medicare Advantage margin trajectory, which targets a 3% margin by 2028 and remains under stress testing; legislation affecting state PBMs, especially Tennessee; and director Larry Robbins selling approximately $297 million of stock in May.

Focus on the medical benefit ratio compared to the 90.5% ±50bps guide, monitor Aetna's cost trend, and track PBM client retention.

CAT · Tuesday before open

Key figures include a consensus EPS of $6.22 and revenue of $19.3 billion, with estimates showing greater variation than usual: EPS predictions range from $5.96 to $6.69 among 15 analysts.

Setup: Caterpillar has a consistent winning streak, with five consecutive days of positive next-day reactions, median +3.4%, peak +11.6%, including a +9.9% jump the day after the April report. However, the market seems skeptical about a sixth gain: the stock has fallen 15.4% in the past month and is 24% below its high.

Implied move: 6.8% ($27.4B), twice the historical average.

Red flags include tariff costs continuing to be a key variability, as management repeatedly re-quantifies them; weakness in Resource Industries' margins; and the fact that the same quarter last year was its only EPS miss in the past five reports.

Focus on the tariff-mitigation calculations, the backlog in power-generation (notably Caterpillar’s data-center segment with large engines approaching nearly three times the 2024 levels), and the comparison between dealer inventory and end-user demand.

Priced unusually

Twelve out of our 20 options show implied moves that are at least twice their own historical median, with two exceeding three times. This essentially provides a list of the quarters the market views as notably different this time.

Booking (Monday after close) stands out as an anomaly this week: 5.4% implied versus a 0.9% median, making it 6.1 times higher. BKNG has been one of the most stable mega-cap stocks, with little movement after four of its last five earnings reports. Currently, the options market is anticipating its largest earnings move in years, with skew favoring calls over puts. The key questions from the last quarter are how extensively Middle East disruptions affected bookings, and whether AI travel agents pose a genuine threat to disintermediation or are just a topic of discussion.

A quick overview of the remaining members of the over-2x club:

• VST (Friday before open), 3.1x. The key name on Friday. Vistra has missed EPS in four of its last five quarters, with analyst estimates ranging from $1.38 to $2.08. It is priced for a 7.9% return, compared to a 2.5% median.

• MRK (Tuesday before market open), trading at 2.8x, shows a strong call skew, which is uncommon for large pharmaceutical companies. Focus on Keytruda's growth outside of timing and Gardasil’s recovery in China as key indicators.

• SNDK (Wednesday after close), trading at 2.8x. More details provided below; it also has the biggest raw implied move for the week.

• CEG moves on Thursday before the open at 2.6x and then passes the theme to VST on Friday. These two power producers exhibit a 96% similarity in move magnitude, making them the most closely correlated pair on the tape.

• APO (Tuesday before opening), 2.4x. The key figure to verify is Q2 origination compared to management’s estimate of a "possibly near record $97 billion.”

• NVO reports results on Wednesday before opening, with 2.3x consensus EPS of $0.77 on $10.88B revenue. The stock has dropped 20% since the last report, which included a -14.6% day, and management guides to sales declining 4–12% this year. Despite this, the options skew favors calls, indicating someone is positioning for a potential relief rally.

The anti-anomaly: AppLovin’s 13.3% implied move is only 1.1 times its median. For APP, a 13% earnings day is typical on Tuesday.

The software gauntlet

Seven internet and software companies are reported over four days, each previewing the next: Palantir on Monday night, AppLovin on Tuesday night, Shopify and Uber on Wednesday morning, Cloudflare and DoorDash on Wednesday night, and Datadog on Thursday morning. Owning two of these provides you with seven data points this week.

Three components within the chain:

• Uber has declined after each of its last four reports, even including one with a +350% EPS surprise. Since mid-2025, there has been no positive market reaction. Options are pricing in a 7.0% move, roughly twice its typical volatility, leading into Wednesday morning.

• DoorDash has the largest analyst disagreement this week, with EPS estimates ranging from $0.27 to $0.73—almost triple the low. Whichever figure is reported, half the models on the street are significantly inaccurate.

• Cloudflare and Datadog are both trading above their consensus price targets—by 10% and 2%, respectively—after gains of 55% and 135% since their February reports. In July, Datadog experienced two Buy-to-Hold downgrades based on that calculation but still enters Thursday with a 14.5% implied move, the second-highest among our 20 stocks.

Tuesday’s hardware pile, and a first

Sandisk, after market close on Wednesday, displays the largest implied move of the week: 19% for a $1,215 stock, equating to an expected swing of about ±$230 per share. The company has reported five consecutive earnings beats, with surprise sizes of 60%, 71%, 38%, and an extraordinary 522%, even amid a NAND shortage that the company itself identified as structural. Despite this, the stock has declined 30% over the past month and is 48% below its peak. Reports from Arista and Western Digital scheduled for Tuesday night offer insight, as these three hardware stocks show an 82% similarity in their move magnitudes.

Astera Labs (reporting Tuesday after close) has an implied volatility of 15.9%. Last quarter, it exceeded EPS estimates by 12% and revenue by 8%, yet its stock declined by 21.4%. At this point, beating estimates is no longer the main concern.

SpaceX (reporting after Tuesday's close) releases its first earnings as a public company. With a market cap of $1.52 trillion and an implied move of 16.6%, the expected swing is approximately $250 billion for a single stock with no prior reporting history—more than all Thursday's reports combined. It isn’t in our top 20 rankings because the method relies on historical data the company hasn't established yet. This is a known limitation of our approach; nonetheless, keep an eye on it.

Further down the tape

Wildcard implied moves are highlighted in small caps: Digital Turbine at 26.9% and Fluence Energy at 25.7%, making option pricing highly volatile. GPGI reports Thursday with 25.6 days to cover short interest, matching a 25% implied move—our most volatile find. Additionally, insiders purchased shares in three companies—Fortune Brands' Ed Garden, who bought $28.7M worth with no sales; Cascade Investment added $202M in Republic Services in May; and Robert Duggan invested $26M in Pulse Biosciences—an uncommon and noteworthy pattern.

What would change the week

The final print of the week isn’t an earnings report. Payroll data is released on Friday at 8:30 am, after four days and about $1.5 trillion of anticipated earnings swings have already been settled. A strong or weak jobs report can quickly influence Thursday’s winners before their transcripts are even analyzed. Additionally, Exxon, Linde, and Eaton announced earnings after market close on Friday, making Monday’s opening bell the first real market assessment of the week.

This publication provides research and educational content for investors rather than investment advice. We present valuation data, options-market information, and model estimates along with their assumptions; however, we do not offer buy or sell recommendations, price forecasts, or specific portfolio advice. All figures are timestamped as indicated and may be updated. Always conduct your own due diligence or seek advice from a licensed professional before acting on any information here.

One analyst says the turnaround has begun.

Another says the brand is losing relevance.

Bulls point to improving revenue.

Bears point to declining margins.

Options traders expect a large move, but no one explains whether that expected move makes sense.

Investors are surrounded by conflicting opinions about earnings.

That is the problem Earnings Intel was built to address.

Earnings Intel is a research publication focused on studying major upcoming earnings events from three perspectives:

1. What the company has been reporting

2. What related companies are revealing

3. What the options market is pricing

The objective is to understand the possible outcomes, identify where expectations may be wrong, and show investors what information is most likely to move the stock.

Every week, we reduce a large earnings calendar into a small number of events worth paying attention to.

Most companies are excluded, and this is intentional.

Subscribe now

How We Analyze Earnings

We begin with the options market.

Before earnings, options prices indicate how large a move traders collectively expect after the announcement.

For example:

Stock price:             $100
Options-implied move:    ±8%
Expected lower boundary: $92
Expected upper boundary: $108

Options are pricing a move of approximately 8% in either direction.

We compare that figure with the stock’s actual movement after previous earnings reports:

Previous earnings moves:

Last quarter:       −11.2%
Two quarters ago:   +4.8%
Three quarters ago: +9.5%
Four quarters ago:  −6.1%

Average absolute move: 7.9%
Current implied move:  8.0%

This immediately gives investors useful context.

• Is the market expecting an unusually large move?

• Is the current straddle cheap or expensive compared with recent history?

• Has the expected move increased sharply over the past week?

• Are traders paying much more for downside protection than upside exposure?

The expected move is the starting point because it establishes the hurdle.

A stock falling 5% may sound dramatic. But if options priced a 10% move, the result was relatively ordinary.

However, a stock falling 12% when options priced 6% is a very different event.

What this tells you: how much uncertainty is already reflected in the market price before the company reports.

Every earnings report revolves around a small number of unresolved questions.

For Nike, those questions might include:

• Is the turnaround gaining momentum?

• Is growth returning to Nike’s direct and digital channels?

• Is China stabilizing?

• Are inventories becoming healthier?

• Are margins recovering?

• Is growth coming from strong consumer demand—or from discounting and liquidation?

• Can one successful category compensate for weakness elsewhere?

For another company, the questions may concern AI spending, subscriber growth, credit losses, advertising demand, cloud margins, store traffic, or production capacity.

We identify the variables that matter most to the current investment debate.

This prevents the analysis from becoming a list of financial statistics with no clear meaning.

Revenue matters. EPS matters. But the number that changes the stock is often something else. It may be:

• forward guidance;

• gross margin;

• a regional business;

• customer concentration;

• inventory;

• pricing;

• capital expenditure;

• management’s explanation of why growth changed.

What this tells you: which parts of the earnings report are most likely to determine the stock reaction.

Before Nike reports, Adidas, On, Puma, Lululemon, retailers, and other related businesses may already have disclosed important information about:

• consumer demand;

• pricing;

• promotions;

• inventory;

• tariffs;

• product categories;

• geographic markets;

• retailer behavior;

• market-share changes.

The important part is interpreting those results correctly.

Strong competitor earnings are not automatically positive for the company we are studying.

If Adidas grows rapidly in a market where Nike is declining, that may mean athletic-footwear demand is healthy—but Nike is losing share.

That produces two simultaneous conclusions:

Industry demand: Positive
Nike-specific read-through: Negative

Weak competitor earnings can also be informative.

If several brands report rising promotions and excess inventory, that may signal a broader margin problem.

We therefore separate competitor evidence into different categories:

• Industry demand

• Market-share transfer

• Pricing pressure

• Inventory pressure

• Geographic exposure

• Channel strength

• Company-specific execution

This is one of the most important parts of Earnings Intel.

It helps distinguish between:

“The entire industry is weak.”

and:

“The industry is functioning, but this company may be falling behind.”

Those are very different investment conclusions.

What this tells you: whether recent competitor results support or contradict the story currently priced into the target company.

Once we understand the company, its competitors, and the expected move, we look for contradictions.

Examples:

• Competitors are growing, but the company’s consensus estimates have barely changed.

• The company’s direct sales are declining, but analysts remain focused on headline revenue.

• The options-implied move is close to its historical average, even though current risks are unusually concentrated.

• Investors expect a turnaround, but management has not yet demonstrated improvement in the most important business segment.

• The stock has already declined heavily, but long-term analyst forecasts still assume a rapid recovery.

These contradictions form the earnings setup.

We ask:

• What result would confirm the bull case?

• What result would confirm the bear case?

• What could happen even if the company beats EPS?

• Which risk appears understood?

• Which risk may still be underpriced?

• What would cause the stock to move outside the options-implied range?

The objective is not to force every company into a bullish or bearish conclusion.

Sometimes the correct conclusion is:

The market is pricing the event reasonably, and neither side has a clear advantage.

Those situations are useful too.

Knowing when there is no clear edge can be as valuable as identifying an unusual setup.

What this tells you: where expectations, fundamentals, and market pricing may be misaligned.

The final report is designed to be read quickly. It contains:

The expected move

Options-implied move: ±9.2%

Lower boundary: $48.30
Upper boundary: $58.10

Historical context

Last four earnings moves:

−10.5%
+6.4%
+15.2%
−5.5%

The unresolved questions

1. Is China stabilizing?
2. Is Digital demand improving?
3. Are wholesale gains sustainable?
4. Will management extend the turnaround timeline?

The evidence

Positive:
- Strong performance-category growth
- Improving wholesale relationships
- Lower investor expectations

Negative:
- Direct and Digital weakness
- Competitors gaining share
- Margin and inventory pressure

The scenarios

Bull case:
A convincing improvement in guidance and operating trends.

Base case:
Results close to consensus, with an uneven recovery.

Bear case:
Headline numbers hold up, but guidance reveals a longer downturn.

The Earnings Intel conclusion

Directional bias: Bearish asymmetry
Primary risk: Forward guidance
Most important metric: China and Digital
Options assessment: Large move priced, but downside tail may remain underpriced

The report is not intended to bury readers in research.

It is intended to compress the research into the few variables that matter.

The Weekly Earnings Intel Schedule

We publish two types of research.

Every Sunday, we publish a focused overview of the coming earnings week.

It includes:

1. The biggest earnings events

The large companies and heavily followed stocks likely to dominate market attention.

2. The largest expected moves

Companies where options traders are pricing the greatest percentage movement.

3. Historical earnings movement

For each selected company, we show:

• the current options-implied move;

• the previous four post-earnings moves;

• the average historical move;

• whether the current expectation appears high or low by comparison.
  

The purpose is to answer:

Where is the market expecting the most action this week?

Before selected major earnings events, we publish a deeper company report.

These reports examine:

• the central investor uncertainty;

• what the company must prove;

• what competitors have already reported;

• what options traders expect;

• the most important bull and bear arguments;

• possible outcomes inside and outside the expected range.

We will not publish a deep report on every company.

We focus on events where:

• investor attention is high;

• uncertainty is meaningful;

• the stock has an active options market;

• competitors have provided useful evidence;

• the result could materially change the investment narrative.

Earnings Intel is not a service that claims to know exactly where a stock will trade after earnings.

No serious person can promise that.

It is not a list of AI-generated stock predictions.

It is not a transcript summary.

It is not another earnings calendar.

It is not a recommendation to blindly buy calls, buy puts, or trade every event.

And it is not a substitute for understanding your position size and risk tolerance.

Earnings Intel provides a structured view of the event:

• what the market expects;

• what could invalidate those expectations;

• what related companies have revealed;

• which outcomes appear most important;

• whether the current options pricing looks consistent with the evidence.

We do the research, but you make the decision.

Retail investors have access to more information than ever, but most of it arrives without context.

One platform shows analyst estimates.

Another shows the options chain.

Another provides news.

Another lists historical earnings.

Social media supplies hundreds of contradictory opinions.

The investor is then expected to assemble all of this into one coherent conclusion.

Earnings Intel is designed to perform that synthesis.

Not by pretending uncertainty can be eliminated.

But by making the uncertainty understandable.

Before every major earnings report, there are several possible futures.

Our job is to show you:

Which futures the market is pricing—and which ones it may be overlooking.

That is Earnings Intel.

Subscribe now