VYNG

Salesforce reported a record quarter: earnings per share up fifty percent, and a seven-percent raise to full-year guidance. Both numbers are real — and both, underneath, are financed.

In this episode we take apart exactly how a Salesforce business growing about nine percent organically produced a fifty-percent earnings headline — and why the "record" says less than the number suggests.

What's covered:

→ Why two-thirds of the earnings "beat" was a one-time gain on a venture portfolio, not operations

→ How Informatica — an acquisition — supplied roughly a third of the reported revenue growth

→ The buyback math: a $25B debt-funded repurchase that cut the share count 10% (the pizza-and-slices explainer)

→ The tell most recaps missed — an EPS guide raised 7% while the net-income pool behind it fell ~5% 

→ Why Agentforce ARR crossing $1B and 34.8% margins keep this a genuine franchise, not a broken one

→ The VYNG conviction score, and the three things the next quarter has to answer

Every figure is sourced directly from Salesforce's Q1 FY2027 and Q4 FY2026 filings. Full research report with every chart and calculation → vyng.io

VYNG research is editorial and informational — not investment advice. Conviction scores are editorial judgments, not recommendations.

#Salesforceearnings, #CRMstock, #Agentforce, #Informaticaacquisition #WhatChanged #VYNG

NVIDIA just posted a near-perfect quarter — $96.2B in revenue, up 106%, and a Q3 guide of $108B that cleared every bar Wall Street set. So why is that the least interesting thing that happened? Because below the income statement, NVIDIA started guaranteeing its customers' debt — $105 billion of it.

In this episode of What Changed from VYNG, we read past the headline into the fine print of NVIDIA's Q2 FY2027 filing — the numbers that never made a press release but changed the risk behind the whole story.

What's inside: → Why the $108B guide — with zero China assumed — over-delivered on the one metric that was supposed to decide the stock → How purchase commitments exploded from $119B to $279B in a single quarter (and why "primarily memory" matters) → The $105B lease guarantee that turned NVIDIA into the landlord co-signing its customer's lease → Why DSO stretching from 45 to 60 days is the quiet tell on the quality of the beat → The H20 write-down parallel that makes $279B a bet, not a certainty → Our conviction score: 8.0 out of 10, and exactly why it stepped down from 8.4

The full research report — every chart, source, and calculation, sourced directly from NVIDIA's filings — is at vyng.io.

Not investment advice. VYNG's conviction score is an editorial assessment of the evidence, not a recommendation to buy or sell any security.

#NVIDIAearnings, #NVDAQ2FY2027, #Blackwell, #datacenter #revenue, #AIcapex, #OpenAI #HBMmemory #VYNG #WhatChanged.

Amazon reported $62.6 billion of profit last quarter. About $21 billion of it came from selling things to customers. The rest is an appraisal — and only half of it even went through the earnings line.

This episode strips Amazon's Q2 2026 release back to what the business actually earned, and finds a genuinely exceptional operating quarter buried underneath the largest paper gain in the company's history.

What we cover:

→ Why reported EPS of $5.75 becomes roughly $1.91 once you remove $53.4B of non-operating gain

→ The second, larger gain almost nobody will quote — ~$55.7B pre-tax sitting in other comprehensive income

→ $109 billion of pre-tax investment gains in one quarter, against $93.7B of operating income across the entire preceding year

→ The story that got buried: AWS growing 36.8% — fastest in 18 quarters — while expanding operating margin 645 basis points to 39.4%

→ Why free cash flow crossed below zero, and how $67.0B of new long-term debt covered 104% of Amazon's six-month funding gap

→ The circularity worth naming: one counterparty appearing as the asset, the customer, and the investment recipient

→ Why Amazon guided next quarter's operating income below the quarter it just delivered

→ VYNG conviction score: 6.8/10, standalone baseline

Every figure in this episode is sourced directly from Amazon.com, Inc.'s Q2 2026 earnings release dated 30 July 2026. The complete research report — with the full delta ledger, segment split, bias flags, and sourcing method — is at vyng.io

full report also here: https://www.vyng.io/reports/what-changed-amazon-q2-earnings

#Amazonearnings, #AMZNQ22026, #AWSgrowth, #Anthropic #investment #equityresearch, #earningsanalysis #whatchanged #vyng

Not investment advice. VYNG conviction scores are editorial assessments of how much confidence the available evidence deserves — not buy, sell, or hold recommendations.

Apple posted the best June quarter in its history and crossed 50% gross margin for the first time. Two of those records came from a tariff refund — and Apple said so itself, in the second sentence of its own release.

In this episode of What Changed, we strip the refund out and find a company that hit its own April guidance to within six basis points. Then we go to the balance sheet, where $9.2 billion of intangible assets appeared with no matching cash outflow.

Covered:

→ Why a 7% EPS beat is really a 1% beat once you remove $0.11 of tariff refunds

→ The six basis points between Apple's underlying margin and its own guidance midpoint

→ Services missed the Street and its own guidance — and why that matters more than a 22% iPhone quarter

→ Why the September guide is a supply story, not a demand story

→ Inventory nearly doubling to 18.5 days — and why that's Apple agreeing with its own guidance

→ The $9.2B of intangibles with no cash trail, and what it might say about how Apple is really paying for AI

→ VYNG conviction score: 6.6 / 10 — a standalone baseline, our first published score on AAPL

Every figure is sourced directly from Apple's Q3 FY2026 press release and condensed consolidated statements. Where we've inferred something the filings don't state, we say so out loud. The complete research report — with every chart, source and calculation — is at vyng.io

full report here: https://www.vyng.io/reports/what-changed-apple-earnings

Not investment advice. Editorial analysis only. Conviction scores are editorial judgments about evidence quality, not buy/sell recommendations or price targets.

#Appleearnings, #AAPLQ32026, #TimCook, #JohnTernus, #equityresearch, #newiPhone, #financialanalysis #AIcapex #whatchanged #vyng

Meta generated $31.86 billion of operating cash flow in Q2 2026 — up 25% year on year — and converted just $784 million of it into free cash flow. That's a 91% collapse, on $60.8 billion of revenue.

The headlines are arguing about a 14.5% earnings miss. That argument is a decoy: strip out the $2.40B legal charge and $1.18B of severance Meta disclosed itself, and adjusted EPS lands near $7.35 — above consensus. Meta didn't miss. It beat.

The real change is on the cash flow statement, and it's structural.

In this episode:

→ Why the EPS "miss" dissolves under a fair comparison — and what Meta's GAAP-only reporting has to do with it

→ How capital expenditure jumped 57% sequentially to $31.08B, the biggest single step of the AI buildout

→ The $24.9B debt raise, three straight quarters of zero buybacks, and a share count that has started rising again

→ Why R&D absorbed 88% of every incremental revenue dollar this quarter

→ The counterintuitive finding: Reality Labs honoured its guidance while Family of Apps margin fell from 53.0% to 38.8%

→ Ad impressions decelerating from +19% to +14% while price per ad holds at +12% — and why a price-led growth story is the harder one to compound

→ The VYNG conviction score, and the three questions Q3 will answer

Every figure in this episode traces to Meta Platforms' 8-K Exhibit 99.1 earnings releases. Adjusted EPS and segment margins are VYNG calculations from disclosed line items — Meta publishes no adjusted earnings figure, and we say so. The complete research report, with every chart, source, and calculation, is at vyng.io

Full report also here: https://www.vyng.io/reports/what-changed-meta-q2-earnings

Clarity, not clutter.

Not investment advice. The VYNG conviction score is an editorial assessment of how much confidence today's evidence deserves — never a buy or sell recommendation.

#Metaearnings, #METAQ22026, #MetaPlatformsresults, #Metacapex, #AI #RealityLabs #Metabuyback #earningsanalysis #equityresearch #podcast, #VYNG #whatchanged

Merck and Moderna announced the first successful Phase 3 trial of an mRNA cancer therapy in history. The release quoted hazard ratios precisely — for a completely different trial.

For the Phase 3 itself, the confirmatory study in 1,137 melanoma patients, there was no hazard ratio, no confidence interval, no p-value and no survival curve. The result was described in words. For the older, smaller Phase 2b, the same release printed HR 0.51 and HR 0.411 to three decimal places with 95% confidence intervals.

In this episode of What Changed, Vikas unpacks what that gap actually costs an investor.

→ Why "the trial met its endpoint" tells you the sign of an effect, not its size

→ The sealed exam result — an analogy for why "passed" is compatible with wildly different outcomes

→ How the most bullish analyst on the street ended up estimating the hazard ratio himself, at 0.6–0.7 — weaker than the number the release did print

→ Why recurrence-free survival is not overall survival, and why that distinction decides reimbursement

→ The 2024 regulatory refusal almost nobody is mentioning this week

→ The manufacturing problem behind a therapy built individually for every single patient

→ VYNG's conviction score on $MRNA: 5.8/10, and exactly why it isn't higher

This is delta-first equity research: individualised neoantigen therapy, adjuvant melanoma, intismeran autogene, KEYTRUDA, pembrolizumab, INTerpath-001, mRNA oncology, biotech earnings analysis, hazard ratio interpretation, surrogate endpoints, and how to read a topline press release for what it leaves out.

Every chart, source and calculation behind this episode is in the full research report at vyng.io, sourced from Merck and Moderna's joint release, Moderna's own filings and the trial registry.

Clarity, not clutter.

Editorial commentary for information and education only. Not investment advice, not a recommendation to buy or sell any security, and no account taken of any listener's objectives, financial situation or needs. The VYNG conviction score is an editorial assessment of disclosed evidence, not a rating or price target.

 #Moderna #MRNAstock #Merck #intismeranautogene #mRNA-4157 #V940 #KEYTRUDA #pembrolizumab #INTerpath-001 #melanoma #adjuvanttherapy #individualised #neoantigen #therapy #cancervaccine #hazardratio #recurrence-freesurvival #biotechinvesting #equityresearch #earningsanalysis #WhatChanged #VYNG

Microsoft just closed the most profitable year in its history, and the cash the business actually generated went down. Both are true, and the reason why is the most useful thing in this quarter's filing.

Microsoft reported Q4 FY2026 on 29 July: revenue of $90.0 billion, ahead of both consensus and the company's own guidance, with Azure accelerating to 43% growth as it crossed $100 billion in annual revenue. Headline earnings per share grew 32%. Operating income grew 18%. This episode is about that gap — and about two accounting changes announced on the call that almost nobody covered.

In this episode:

→ Why 45% of the increase in pre-tax profit came from below the operating line, including a $3.2 billion gain on Microsoft's Anthropic stake

→ The useful-life extension on data centres, from 15 years to 25, and what it does to future depreciation

→ How a finance-to-operating lease reclassification cut the calendar-2026 capital expenditure guide by roughly $15 billion while management described the underlying spending as unchanged

→ The $678 billion backlog that grew 84% — or 25% excluding a single customer — and why the same correction makes bookings look better, not worse

→ Microsoft Cloud gross margin's fourth consecutive quarterly decline, from 68% to 65%

→ Free cash flow down 23% in the quarter and 6.5% for the full year, Microsoft's first annual decline of this cycle

→ What's genuinely, unambiguously good: Azure accelerating at $100B scale, 18% revenue growth on a 2% smaller headcount, and Copilot past 30 million paid seats

→ VYNG's conviction score, with the reasoning behind it

#Microsoftearnings #MSFTQ4FY2026 #Azuregrowth #Microsoftcapex #AIinfrastructurespending #MicrosoftCopilot #hyperscalerearnings #cloudgrossmargin #equityresearch #earningsanalysis #whatchanged #vyng

The complete research report — every chart, source and calculation, sourced directly from Microsoft's filings — is also at vyng.io

here is full report link : https://www.vyng.io/reports/what-changed-microsoft-earnings

Clarity, not clutter.

Coca-Cola reported +16% EPS growth and raised full-year guidance — a clean beat on the surface. Underneath, the growth engine quietly switched pedals: for the first time in years, Coke grew on volume, not price.

In this episode of What Changed, we take the same $1.03 of earnings and tell it four ways — from the +16% headline down to the +9% that actually describes selling beverages — and show why the gap is currency and paper securities gains, not the business.

Covered in this episode: → Why +16% EPS decomposes to +9% real operating growth (currency + one-off securities marks)

→ The volume-for-price switch: unit case volume accelerated to +5%, every segment positive, while price/mix stalled at +2%

→ Why the guidance raise is higher-quality than it looks — the cleanest operating line went up as the currency assumption was cut

→ Asia Pacific's paradox: +8% volume, −9% price/mix, and profit barely back to flat

→ The quiet turn in EMEA operating profit, and the World Cup marketing question for the second half

→ VYNG conviction score: 6.8/10 — and what would move it

This is our first VYNG conviction score on Coca-Cola. Keywords: Coca-Cola earnings, KO stock, Q2 2026 earnings, organic revenue, currency-neutral EPS, unit case volume, price/mix, affordability, beverage stocks, guidance raise, consumer staples, dividend stocks.

Every figure is sourced directly from Coca-Cola's Q2 2026 filing. Full research report with all charts and calculations at vyng.io

also available here:https://vyng.io/reports/what-changed-coca-cola-q2-earnings

VYNG is editorial and informational — not investment advice, not a recommendation. Conviction scores are editorial judgments. Do your own research.

#CocaColaearnings #KOstock #Q22026earnings #organicrevenue #currencyneutralEPS #unitcasevolume #price/mix #affordability #initiatives ·#beveragestocks #guidanceraise #consumerstaples #dividendstocks #PepsiCovsCocaCola #fearandgreed #WhatChanged #VYNG

Intel reported the largest quarterly loss in its history: $11 billion. The single biggest reason for it is that Intel's own share price went up.

That isn't a joke or a rounding error — it's how the accounting genuinely works, and once you see the mechanism the entire quarter reads differently.

In this episode of What Changed, we take Intel's Q2 FY2026 earnings apart layer by layer:

→ Why shares held in escrow for the U.S. Commerce Department under the CHIPS Act create a liability that rises when Intel's stock rises — and how that produced a $12.53B non-cash charge, up 11.5x from $1.09B a quarter earlier

→ The cleanest test in the filing: add back that one item and nothing else, and a $(2.16) loss per share becomes +$0.29

→ Why GAAP operating income was positive $1.80B — the first positive operating quarter in eight — while revenue grew 25%, the fastest in more than 15 years

→ The Data Center & AI inflection: revenue +59%, but operating income +291%, with segment margin moving from 16.1% to 39.5%

→ The counterintuitive finding: Intel's largest segment grew revenue 15% sequentially while operating income fell 7%, compressing margin more than six points into a guided-flat PC quarter

→ Why $7.0B of operating cash flow and $(8.4)B of adjusted free cash flow appear in the same table — and which one describes what the factories actually cost

→ VYNG's conviction score, and the three questions the next quarter has to answer

Every figure is sourced directly from Intel's Q2 FY2026 and Q1 FY2026 earnings releases. Complete research report, with all charts, sources and calculations, at vyng.io

also available here: https://vyng.io/reports/what-changed-intel-q2-earnings

#Intelearnings, #INTC #stockanalysis, #CHIPS #Actaccounting, #escrowedshares #marktomarket, #IntelFoundry, #DataCenter #AI, #semiconductor #earningsanalysis, #GAAP #non-GAAP, #Intel18A, #earnings #breakdown #podcast.

Alphabet reported EPS up 294% to $9.11 — and it's one of the most misleading numbers of earnings season. More than two-thirds of it is a single non-cash SpaceX mark-up. Strip it, and the operating quarter was a whisker light.

In this episode of What Changed, we decode the gap between the headline and the truth — and find the change that actually matters, three statements deep.

Covered:

→ Why the $9.11 headline is a "house-next-door" paper gain, not a cash one

→ The real inflection: Google Cloud revenue +82%, margin ~21%

→36%, backlog past $460B

→ The identity shift — Alphabet stops returning cash and starts raising it (~$85B in equity, preferred & debt; buybacks to zero)

→ The first negative free-cash-flow quarter in memory, and why capex doubling to $44.9B is a choice, not a constraint

→ The behavioral trap of "commitment escalation" on a $180–190B build

→ VYNG conviction: 6.5 → 7.2/10, and the three questions next quarter must answer

Every figure is sourced directly from Alphabet's Q2 FY2026 filing. Full research report, charts and calculations at vyng.io

Full report here: https://www.vyng.io/reports/what-changed-google-q2-earnings

Editorial content — not investment advice. VYNG's conviction score is an assessment of how much confidence today's evidence deserves, not a buy or sell recommendation.

#Alphabetearnings, #Googlestock, #GOOGL, #GoogleCloud, #AIcapex, #SpaceX, #Gemini, #earningsanalysis #vyng #whatchanged