Published: July 22, 2026 | Reading Time: ~11 minutes | Channel: business
Tonight, two companies worth a combined $4.7 trillion report earnings after the market closes. By tomorrow morning, one of them could be worth 10% less — and we won't know which one until the numbers hit. Alphabet and Tesla kick off the Magnificent Seven earnings parade, and the market is already wobbling: Nasdaq futures are down 1.05%, oil just surged past $88 on an 11th straight night of US airstrikes on Iran, and the VIX is creeping back above 17.
But none of that is the real story. The real story is what IBM already told us, and what the market still refuses to hear.
Two weeks ago, IBM missed revenue expectations by 3.7%. Three point seven percent. The punishment? A 25% single-day crash — the worst in the company's 115-year history. Forty billion dollars of market cap vaporized. Not because IBM is collapsing. Not because of a fraud. Because the market has quietly, violently repriced the penalty for earnings disappointment in the AI era.
That's the setup. Alphabet and Tesla don't just need to beat estimates tonight. They need to demolish them — or the IBM treatment is coming for someone else.¹
Let's get the figures on the table, because you can't bet on what you can't measure.
LSEG consensus has Alphabet posting Q2 revenue of $116.8 billion, up 21.1% year-over-year.² That's the third straight quarter above $100 billion. The engine? Google Cloud, projected at $22.2 billion — up 63.3% — the fastest growth among major cloud providers.³
But here's what the headline misses: Alphabet's Q1 net income of $62.6 billion was a mirage. $37.7 billion of it came from unrealized mark-to-market gains on minority stakes in Anthropic and SpaceX. Strip that out, and Alphabet missed the $2.63 consensus EPS by a cent.⁴
The capex story is even more jarring. Alphabet guided FY2026 spending at $180–190 billion — up 107% year-over-year in Q1 alone. They raised $84.75 billion in equity in June specifically for AI infrastructure. Free cash flow fell 47% in Q1.⁵
So the question tonight isn't "Is Google growing?" It's "Is Google growing fast enough to justify burning $190 billion a year on AI infrastructure that hasn't yet proven it generates a dollar of profit?"
Wall Street has 28 buy ratings, 5 holds, and zero sells. The average price target is $428.43 — 15.5% above the July 15 close.⁶ Options markets are pricing in a 5.3% move. That sounds manageable. It isn't. When IBM was pricing in a "modest" move before its crash, nobody priced in a 25% wipeout either.
| Metric | Q2 2025 Actual | Q2 2026 Estimate | YoY Change |
|---|---|---|---|
| Total Revenue | $96.4B | $116.8B | +21.1% |
| Google Cloud | $13.6B | $22.2B | +63.3% |
| Google Services | $82.5B | $94.3B | +14.3% |
| Operating Income | $31.3B | $40.4B | +29.1% |
| Operating Margin | 32.4% | 35.3% | +2.9pp |
Source: LSEG Data via IG⁷
Tesla delivered 480,126 vehicles in Q2 — up 25% year-over-year, its best Q2 ever, and roughly 74,000 above the analyst consensus.⁸ It was the company's first year-over-year delivery growth in two years. On the surface, that's a clean win.
But peel back the hood and it gets complicated fast.
Wall Street expects revenue around $26.4 billion, with EPS of approximately $0.53. The range on earnings is absurd: estimates run from $0.27 to $0.74 — a spread so wide it tells you analysts are basically guessing.⁹ Deutsche Bank is at $0.36, barely above last year's $0.40.
Here's the tension: Tesla produced only 451,758 vehicles but delivered 480,126. That means it drew down 28,000+ vehicles from inventory. Good for the balance sheet — but was it profitable? If automotive gross margin ex-credits held at Q1's ~12.5%, the quarter is genuinely strong. If margins slipped, Tesla just proved it can move metal only by slashing prices and offering cheap financing. Volume without profit.
And the regulatory credit cushion? It's vanishing. Tesla booked just $439 million in credit sales in Q2 2025, already down more than 50% year-over-year. The $7,500 federal EV tax credit expired September 30, 2025. That faucet is off.¹⁰
Options traders are pricing in a 7.6% swing in either direction. For a $900 billion company, that's roughly $68 billion of market cap hanging in the balance tonight.
Energy is the bright spot: 13.5 GWh of storage deployed, up 40% from Q2 2025.¹¹ Quietly, Tesla's energy business is becoming the most consistent growth engine it has — even if the market still values the stock on robotaxi fantasies.
Here's where we get contrarian.
For two years, the bull case on AI stocks has rested on a simple observation: unlike the dot-com era, today's leaders generate real cash flows. NVIDIA posts actual profits. Alphabet prints money from Search. S&P 500 valuations sit at ~22x forward earnings — below the 25x-plus threshold usually associated with true bubbles.
This argument addresses valuation. It says nothing about whether the earnings themselves are sustainable.¹²
Johns Hopkins economist Steve Hanke calls it a "dual bubble." The first bubble is classic: prices outrunning earnings. The second — the dangerous one — is an earnings bubble, where the profits themselves are inflated by circular AI investment cycles, easy credit creation by private banks, and capex spending that companies book as each other's revenue.¹³
BCA Research's Peter Berezin has been warning for months that today's AI trade is "primarily an earnings bubble rather than a valuation bubble." His historical analogs are unsettling: pre-2008 banks, pandemic-era work-from-home stocks, cyclicals like semiconductors. When earnings bubbles burst, they leave behind real excess capacity — data centers, chip fabs, server farms — not just paper losses.¹⁴
And here's the detection problem: analysts only cut profit estimates after stocks have already fallen. There's no early warning system. Berezin's research shows Wall Street analysts are "not particularly good at predicting when earnings bubbles will burst." Stocks fall first. Estimates follow.¹⁵
IBM just proved this in real time. The company missed revenue by 3.7%, and BofA cut its price target to $280 from $330 — after the stock had already cratered 25%. Not predictive. Reactive.¹⁶
If the earnings bubble thesis is right, IBM wasn't an isolated incident. It was a preview. And tonight's reports from Alphabet and Tesla are the first real test of whether the market still believes the AI profit story — or whether it's quietly stopped.
Alphabet's $180-190 billion annual capex is breathtaking. For context, that's roughly the GDP of Hungary. Free cash flow fell 47% in Q1. If tonight's earnings call doesn't demonstrate that AI monetization is actually scaling — not just "positioning" or "investing for the future" but generating real incremental revenue — the capex narrative flips from "ambitious" to "reckless" overnight.¹⁷
Google Cloud's 63% growth is impressive, but it comes with a question: how much of that is customers renting compute to build AI products, versus companies using AI products that generate sustainable business value? If it's mostly the former, the cloud revenue is circular — AI companies paying cloud providers to run AI training, which generates revenue that justifies more AI spending. That's not a business model. That's a daisy chain.
Tesla delivered 480,000 cars. Great. But if gross margins ex-credits fell below 12%, Tesla essentially bought those deliveries with discounts — right as regulatory credits fade and the federal tax credit disappears. The market has been valuing Tesla on autonomy and robots for years, but tonight's numbers are about cars. If the car business can't hold margin at record volume, the narrative that "Tesla isn't an automaker anymore" gets tested in the most uncomfortable way possible.¹⁸
IBM is also reporting after the close. The company already pre-announced the miss, so the numbers themselves shouldn't surprise anyone. But management's commentary on the earnings call — particularly around whether clients are still "pausing on new deals" because of AI disruption — will either confirm or defuse the contagion thesis. If Krishna sounds worried, the entire enterprise software sector should brace.¹⁹
I'm not going to tell you to "stay calm" or "think long-term." That's what every wirehouse advisor says while your portfolio gets shredded. Here's what actually matters tonight:
1. If you own Alphabet or Tesla directly, know your exit. The options market is pricing a 5.3% Alphabet move and a 7.6% Tesla move. That's not abstract. On a $100,000 position in Tesla, that's $7,600 of potential overnight loss. Have a stop-loss or a hedge. Don't wake up tomorrow morning wondering what happened.
2. The contagion risk is real — and it's broader than you think. IBM fell 25% on a 3.7% revenue miss. If Alphabet's cloud growth decelerates even modestly, or Tesla's margins disappoint, the selling won't be contained to those two names. Microsoft and Meta report next week (July 29), and Amazon and Apple the day after (July 30). A bad read-through tonight could pre-price disappointment across the entire Magnificent Seven.²⁰
3. Cash is a position. With the 10-year Treasury at 4.63% and the VIX at 17.63, you're being paid nearly 5% to wait. If you're uncomfortable with the binary risk of earnings season, moving 15-20% of your tech exposure to cash or short-duration Treasuries between now and July 30 costs you almost nothing in opportunity and buys you enormous flexibility.
4. Watch free cash flow, not EPS. Alphabet's Q1 EPS was inflated by $37.7 billion of mark-to-market equity gains. Tesla's earnings have been propped up by regulatory credits for years. Free cash flow — actual money generated by operations minus capital expenditures — is the metric that can't be gamed. If Alphabet's FCF continues declining, and Tesla's automotive margin ex-credits is below 12%, the earnings-beat headlines are meaningless.
5. The oil wildcard. WTI crude at $88.64 (+4.21% today) on the 11th straight night of US airstrikes against Iran is not priced into tech earnings. If energy costs stay elevated or spike further, it hits everything: shipping, component costs, consumer spending. The $37.5 billion the US has spent on the Iran conflict is real money competing with the AI capex boom for capital and attention.²¹
1. The "buy the rumor, sell the news" trap. Alphabet and Tesla are both up meaningfully from recent lows. If their numbers are merely "good" rather than "spectacular," the selloff could be brutal — not because the companies are failing, but because expectations have run ahead of reality. We've seen this movie before: NVIDIA beat earnings by 11% in August 2024 and the stock dropped because the whisper number was even higher.
2. The circular AI revenue problem. If Google Cloud's growth is driven primarily by AI startups raising venture capital to buy cloud compute, the revenue is only as durable as the VC funding environment. When venture funding cycles (and they always do), that cloud revenue contracts. Alphabet won't break out this breakdown. You'll have to read between the lines on the earnings call.²²
3. The currency and tariff overhang. Trump's new 25% tariff on Brazil took effect today. A potential 100% tariff on generic drugs is being floated. The dollar's strength against a basket of currencies means multinational earnings take a translation hit. Alphabet generates roughly 53% of revenue outside the US. A strong dollar alone could shave 2-3% off reported growth.
4. The JPMorgan paradox. While IBM cratered 25%, JPMorgan just posted $21.2 billion in quarterly profit — the highest for any bank in US history.²³ Record bank profits alongside a historic tech crash is not a normal market signal. It suggests credit is still flowing freely through the system, simultaneously inflating both asset prices and the reported earnings that justify those prices — right up until it doesn't.
The market is walking into tonight's earnings with the same confidence it had before IBM's 25% crash — and that's the problem. Alphabet growing at 21% and Tesla delivering a record quarter should be enough. But in an environment where a 3.7% revenue miss costs a company $40 billion of market cap, "good" isn't good enough anymore. The earnings bubble thesis says the profits themselves may be less durable than they look, and tonight is the first major test of whether the market is ready to believe that — or punish the doubters. Either way, someone's getting hurt.
The only question is who.
Yahoo Finance — Stock market today: Dow, S&P 500, Nasdaq futures slide with Alphabet, Tesla earnings on deck. https://finance.yahoo.com/markets/live/stock-market-today-wednesday-july-22-dow-sp-500-nasdaq-alphabet-tesla-083644887.html
IG International (LSEG Data) — GOOGL Q2 2026 earnings preview: 63% cloud growth eyed. https://www.ig.com/en/news-and-trade-ideas/alphabet-q2-2026-earnings-preview-260716
TipRanks — Stock Market News Today, 7/22/26: Futures Slip as Oil Jumps; Alphabet, Tesla Earnings Ahead. https://www.tipranks.com/news/stock-market-news-today-7-22-26-futures-slip-as-oil-jumps-alphabet-tesla-earnings-ahead
Electrek — Tesla (TSLA) Q2 2026 earnings preview: strong deliveries, murky profits. https://electrek.co/2026/07/21/tesla-tsla-q2-2026-earnings-preview/
CNBC — IBM stock craters 25%, the worst day on record, after company issues second-quarter earnings warning. https://www.cnbc.com/2026/07/14/ibm-warns-second-quarter-earnings-fell-short-of-expectations.html
Fortune — Why IBM just suffered its worst stock crash of all time — and what it says about the market's 'dual bubble'. https://fortune.com/2026/07/15/why-did-ibm-stock-crash-earnings-bubble-ai-spending/
interactive investor (FactSet Data) — US earnings season Q2 2026. https://www.ii.co.uk/investing-with-ii/international-investing/us-earnings-season
All claims verified against Gold-tier (LSEG/IG, FactSet, Reuters-sourced data) and Silver-tier (CNBC, Yahoo Finance, TipRanks, Electrek, Fortune) sources. Each source URL was scraped and confirmed accessible. Last verified: July 22, 2026.
The market has priced in perfection. When perfection arrives, it's called meeting expectations — and in this market, meeting expectations might be the most dangerous outcome of all. 🎯
