Offshore
Photo
Fiscal.ai
This is what happens when you're the fastest fundamental data provider in the world. ⚡️⚡️
Soooo many Fiscal AI charts after earnings! https://t.co/TGliCIfcvB
tweet
This is what happens when you're the fastest fundamental data provider in the world. ⚡️⚡️
Soooo many Fiscal AI charts after earnings! https://t.co/TGliCIfcvB
is @fiscal_ai the only platform in the world that already has all of Nvidia's financials? https://t.co/1pRpElwgXn tweet
Offshore
Photo
Fiscal.ai
Nvidia CFO:
"On our balance sheet, inventory grew 32% QoQ while supply commitments increased 63% sequentially. We are preparing for significant growth ahead"
$NVDA https://t.co/oXLBgXPTII
tweet
Nvidia CFO:
"On our balance sheet, inventory grew 32% QoQ while supply commitments increased 63% sequentially. We are preparing for significant growth ahead"
$NVDA https://t.co/oXLBgXPTII
tweet
Offshore
Photo
Dimitry Nakhla | Babylon Capital®
RT @TheShortBear: $GOOGL first to $5T market cap?
Gemini 3 just dropped and I believe it opened the path for Google to become the most valuable company in the world.
The biggest factor of the release was seeing that Gemini 3 success wasn't based on NVDA chips but on their own TPUs.
Google's development of the Tensor Processing Unit (TPU) is the strategic, long-term weapon designed to dismantle the current, high-cost dominance of general-purpose GPUs and position GOOGL as the indispensable, cost-advantaged infrastructure layer for the global AI revolution.
The core of this thesis is a calculated power play:
Shifting the economic center of gravity from $NVDA high-margin, general-purpose silicon to specialized, high-efficiency, proprietary ASICs.
The $GOOGL Full-Stack Advantage: Cost-Control and Commercialization, as per an ex employee:
The Superior Unit Economics of Inference
The world is shifting from costly AI model training to massive, repetitive, and daily inference (running the models across billions of user queries in Search, Ads, and Gemini).
TPUs, which are purpose-built for the tensor operations that define AI, offer significantly better performance per dollar and energy efficiency, a key metric for a hyperscaler.
For internal use (Gemini, Search, etc.), this drives down $GOOGL's largest and fastest-growing CapEx cost, translating directly into higher operating margins for the Google Services and Cloud segments.
A Commercial Challenge to the Ecosystem
The plan is to move beyond the Google Cloud "walled garden" and begin selling TPU capacity externally to other hosters, essentially following the $NVDA playbook. This means the TPU, currently the "closest alternative" to an $NVDA GPU, will be introduced as a viable competitor.
The recent deal with Anthropic to use up to one million TPUs confirms the enterprise-level maturity and ambition of this strategy, driven by the TPU's compelling cost-performance ratio.
While $NVDA's developer ecosystem (CUDA) remains a strong lock-in, the generational improvements in TPUs are reportedly significantly larger than the jumps seen in new GPUs.
Thesis: Consolidating AI Money
Monetization Synergy
$GOOGL's revenue is currently divided into Ad Money, Cloud Money, and AI Money.
By controlling the underlying hardware (TPU) that powers all three segments, $GOOGL effectively consolidates the "AI Money" flowing into its system, turning its own immense AI compute consumption from a cost center into a proprietary competitive moat.
$GOOGL can price its cloud AI services more aggressively because its internal cost of goods sold (COGS) for compute is lower than rivals reliant on third-party silicon.
This competitive wedge accelerates Google Cloud's growth and, critically, expands its operating margin, providing structural long-term leverage.
RESULT
Google ends up taking a large chunk of AWS, Azure and NVDA earnings as an all in one platform to develop, train and host the next generation of AI.
Google cloud will become the place to run it all.
On top, the new release of Gemini makes sure they win the Browser market which will run on the superior models.
All of this will be done at a lower cost then the competition, both through the lower cost AI itself but also through the own manufactured TPU and infrastructure.
GOOGL becomes the absolute kind, both for retail as a consumer product but also on the backend where everyone will be building. All of this coming at a massive advantage to Search.
Unreal turn of events after 1-1.5years of doubt by the street. GOOGL kept on telling us that their efforts in AI were strong and rooted in investments made a decade ago...
The market doubted them but it will likely make them the first 5T company.
Perhaps that is what Berkshire saw.
tweet
RT @TheShortBear: $GOOGL first to $5T market cap?
Gemini 3 just dropped and I believe it opened the path for Google to become the most valuable company in the world.
The biggest factor of the release was seeing that Gemini 3 success wasn't based on NVDA chips but on their own TPUs.
Google's development of the Tensor Processing Unit (TPU) is the strategic, long-term weapon designed to dismantle the current, high-cost dominance of general-purpose GPUs and position GOOGL as the indispensable, cost-advantaged infrastructure layer for the global AI revolution.
The core of this thesis is a calculated power play:
Shifting the economic center of gravity from $NVDA high-margin, general-purpose silicon to specialized, high-efficiency, proprietary ASICs.
The $GOOGL Full-Stack Advantage: Cost-Control and Commercialization, as per an ex employee:
The Superior Unit Economics of Inference
The world is shifting from costly AI model training to massive, repetitive, and daily inference (running the models across billions of user queries in Search, Ads, and Gemini).
TPUs, which are purpose-built for the tensor operations that define AI, offer significantly better performance per dollar and energy efficiency, a key metric for a hyperscaler.
For internal use (Gemini, Search, etc.), this drives down $GOOGL's largest and fastest-growing CapEx cost, translating directly into higher operating margins for the Google Services and Cloud segments.
A Commercial Challenge to the Ecosystem
The plan is to move beyond the Google Cloud "walled garden" and begin selling TPU capacity externally to other hosters, essentially following the $NVDA playbook. This means the TPU, currently the "closest alternative" to an $NVDA GPU, will be introduced as a viable competitor.
The recent deal with Anthropic to use up to one million TPUs confirms the enterprise-level maturity and ambition of this strategy, driven by the TPU's compelling cost-performance ratio.
While $NVDA's developer ecosystem (CUDA) remains a strong lock-in, the generational improvements in TPUs are reportedly significantly larger than the jumps seen in new GPUs.
Thesis: Consolidating AI Money
Monetization Synergy
$GOOGL's revenue is currently divided into Ad Money, Cloud Money, and AI Money.
By controlling the underlying hardware (TPU) that powers all three segments, $GOOGL effectively consolidates the "AI Money" flowing into its system, turning its own immense AI compute consumption from a cost center into a proprietary competitive moat.
$GOOGL can price its cloud AI services more aggressively because its internal cost of goods sold (COGS) for compute is lower than rivals reliant on third-party silicon.
This competitive wedge accelerates Google Cloud's growth and, critically, expands its operating margin, providing structural long-term leverage.
RESULT
Google ends up taking a large chunk of AWS, Azure and NVDA earnings as an all in one platform to develop, train and host the next generation of AI.
Google cloud will become the place to run it all.
On top, the new release of Gemini makes sure they win the Browser market which will run on the superior models.
All of this will be done at a lower cost then the competition, both through the lower cost AI itself but also through the own manufactured TPU and infrastructure.
GOOGL becomes the absolute kind, both for retail as a consumer product but also on the backend where everyone will be building. All of this coming at a massive advantage to Search.
Unreal turn of events after 1-1.5years of doubt by the street. GOOGL kept on telling us that their efforts in AI were strong and rooted in investments made a decade ago...
The market doubted them but it will likely make them the first 5T company.
Perhaps that is what Berkshire saw.
tweet
Offshore
Photo
Dimitry Nakhla | Babylon Capital®
In less than 5 months, $GOOG has surged from $181 to $300, a +65% gain📈
Shows how powerful sentiment can be
A business with $GOOG moat, growth profile, & AI potential should not have been priced under 20x
Today, $GOOG appears fairly valued https://t.co/WsFxnGCLgv
A quality valuation analysis on $GOOG 🧘🏽♂️
•NTM P/E Ratio: 19.32x
•10-Year Mean: 23.69x
•NTM FCF Yield: 3.79%
•10-Year Mean: 4.18%
As you can see, $GOOG appears to be trading below fair value
Going forward, investors can receive ~18% MORE in earnings per share & ~9% LESS in FCF per share 🧠***
Before we get into valuation, let’s take a look at why $GOOG is a great business
BALANCE SHEET✅
•Cash & Short-Term Inv: $95.33B
•Long-Term Debt: $10.89B
$GOOG has a strong balance sheet, an AA+ S&P Credit Rating & 467x FFO Interest Coverage
RETURN ON CAPITAL✅
•2020: 16.2%
•2021: 27.6%
•2022: 26.1%
•2023: 26.9%
•2024: 32.3%
•LTM: 31.9%
RETURN ON EQUITY✅
•2020: 19.0%
•2021: 32.1%
•2022: 23.6%
•2023: 27.4%
•2024: 32.9%
•LTM: 34.8%
$GOOG has strong return metrics, highlighting the financial efficiency of the business
REVENUES✅
•2014: $66.00B
•2024: $350.02B
•CAGR: 18.16%
FREE CASH FLOW✅
•2014: $12.01B
•2024: $72.76B
•CAGR: 19.73%
NORMALIZED EPS✅
•2014: $1.28
•2024: $8.04
•CAGR: 20.17%
SHARE BUYBACKS✅
•2018 Shares Outstanding: 14.07B
•LTM Shares Outstanding: 12.39B
By reducing its shares outstanding ~12%, $GOOG increased its EPS by ~13.6% (assuming 0 growth)
MARGINS✅
•LTM Gross Margins: 58.2%
•LTM Operating Margins: 33.2%
•LTM Net Income Margins: 30.9%
***NOW TO VALUATION 🧠
As stated above, investors can expect to receive ~18% MORE in EPS & ~9% LESS in FCF per share
Using Benjamin Graham’s 2G rule of thumb, $GOOG has to grow earnings at a 9.66% CAGR over the next several years to justify its valuation
Today, analysts anticipate 2025 - 2028 EPS growth over the next few years to be more than the (9.66%) required growth rate:
2025E: $9.61 (19.5% YoY) *FY Dec
2026E: $10.21 (6.3% YoY)
2027E: $11.55 (13.1% YoY)
2028E: $13.57 (17.4% YoY)
$GOOG has an excellent track record of meeting analyst estimates ~2 years out, but let’s assume $GOOG ends 2028 with $13.57 in EPS & see its CAGR potential assuming different multiples
23x P/E: $312.11💵 … ~17.4% CAGR
22x P/E: $298.54💵 … ~15.9% CAGR
21x P/E: $284.97💵 … ~14.5% CAGR
20x P/E: $271.40💵 … ~12.8% CAGR
19x P/E: $257.83💵 … ~11.2% CAGR
As you can see, $GOOG appears to have attractive return potential IF we assume >20x earnings (a multiple below its 5-year & 10-year mean while assuming a lower 2028E)
At >22x, $GOOG has aggressive CAGR potential & it’s not unreasonable for the business to even trade for ~23x (given its growth rate, moat, balance sheet, & exemplary capital allocation)
Today at $181💵 $GOOG appears to be a strong consideration for investment
Between cloud ☁️ , AI 🤖 , quantum computing ⚛️, $GOOG has a strong growth runway ahead, with the potential for continued margin expansion serving as an additional tailwind
$GOOGL
___
𝐃𝐈𝐒𝐂𝐋𝐎𝐒𝐔𝐑𝐄‼️: 𝐓𝐡𝐢𝐬 𝐢𝐬 𝐍𝐎𝐓 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭 𝐀𝐝𝐯𝐢𝐜𝐞. 𝐁𝐚𝐛𝐲𝐥𝐨𝐧 𝐂𝐚𝐩𝐢𝐭𝐚𝐥® 𝐚𝐧𝐝 𝐢𝐭𝐬 𝐫𝐞𝐩𝐫𝐞𝐬𝐞𝐧𝐭𝐚𝐭𝐢𝐯𝐞𝐬 𝐦𝐚𝐲 𝐡𝐚𝐯𝐞 𝐩𝐨𝐬𝐢𝐭𝐢𝐨𝐧𝐬 𝐢𝐧 𝐭𝐡𝐞 𝐬𝐞𝐜𝐮𝐫𝐢𝐭𝐢𝐞𝐬 𝐝𝐢𝐬𝐜𝐮𝐬𝐬𝐞𝐝 𝐢𝐧 𝐭𝐡𝐢𝐬 𝐭𝐰𝐞𝐞𝐭.
𝐓𝐡𝐞 𝐢𝐧𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧 𝐜𝐨𝐧𝐭𝐚𝐢𝐧𝐞𝐝 𝐢𝐧 𝐭𝐡𝐢𝐬 𝐭𝐰𝐞𝐞𝐭 𝐢𝐬 𝐢𝐧𝐭𝐞𝐧𝐝𝐞𝐝 𝐟𝐨𝐫 𝐢𝐧𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧𝐚𝐥 𝐩𝐮𝐫𝐩𝐨𝐬𝐞𝐬 𝐨𝐧𝐥𝐲 𝐚𝐧𝐝 𝐬𝐡𝐨𝐮𝐥𝐝 𝐧𝐨𝐭 𝐛𝐞 𝐜𝐨𝐧𝐬𝐭𝐫𝐮𝐞𝐝 𝐚𝐬 𝐢𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭 𝐚𝐝𝐯𝐢𝐜𝐞 𝐭𝐨 𝐦𝐞𝐞𝐭 𝐭𝐡𝐞 𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐧𝐞𝐞𝐝𝐬 𝐨𝐟 𝐚𝐧𝐲 𝐢𝐧𝐝𝐢𝐯𝐢𝐝𝐮𝐚𝐥 𝐨𝐫 𝐬𝐢𝐭𝐮𝐚𝐭𝐢𝐨𝐧. 𝐏𝐚𝐬𝐭 𝐩𝐞𝐫𝐟𝐨𝐫𝐦𝐚𝐧𝐜𝐞 𝐢𝐬 𝐧𝐨 𝐠𝐮𝐚𝐫𝐚𝐧𝐭𝐞𝐞 𝐨𝐟 𝐟𝐮𝐭𝐮𝐫𝐞 𝐫𝐞𝐬𝐮𝐥𝐭𝐬.
𝐈𝐧𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧 𝐜𝐨𝐧𝐭𝐚𝐢𝐧𝐞𝐝 𝐢𝐧 𝐭𝐡𝐢𝐬 𝐭𝐰𝐞𝐞𝐭 𝐡𝐚𝐬 𝐛𝐞𝐞𝐧 𝐨𝐛𝐭𝐚𝐢𝐧𝐞𝐝 𝐟𝐫𝐨𝐦 𝐬𝐨𝐮𝐫𝐜𝐞𝐬 𝐛𝐞𝐥𝐢𝐞𝐯𝐞𝐝 𝐭𝐨 𝐛𝐞[...]
In less than 5 months, $GOOG has surged from $181 to $300, a +65% gain📈
Shows how powerful sentiment can be
A business with $GOOG moat, growth profile, & AI potential should not have been priced under 20x
Today, $GOOG appears fairly valued https://t.co/WsFxnGCLgv
A quality valuation analysis on $GOOG 🧘🏽♂️
•NTM P/E Ratio: 19.32x
•10-Year Mean: 23.69x
•NTM FCF Yield: 3.79%
•10-Year Mean: 4.18%
As you can see, $GOOG appears to be trading below fair value
Going forward, investors can receive ~18% MORE in earnings per share & ~9% LESS in FCF per share 🧠***
Before we get into valuation, let’s take a look at why $GOOG is a great business
BALANCE SHEET✅
•Cash & Short-Term Inv: $95.33B
•Long-Term Debt: $10.89B
$GOOG has a strong balance sheet, an AA+ S&P Credit Rating & 467x FFO Interest Coverage
RETURN ON CAPITAL✅
•2020: 16.2%
•2021: 27.6%
•2022: 26.1%
•2023: 26.9%
•2024: 32.3%
•LTM: 31.9%
RETURN ON EQUITY✅
•2020: 19.0%
•2021: 32.1%
•2022: 23.6%
•2023: 27.4%
•2024: 32.9%
•LTM: 34.8%
$GOOG has strong return metrics, highlighting the financial efficiency of the business
REVENUES✅
•2014: $66.00B
•2024: $350.02B
•CAGR: 18.16%
FREE CASH FLOW✅
•2014: $12.01B
•2024: $72.76B
•CAGR: 19.73%
NORMALIZED EPS✅
•2014: $1.28
•2024: $8.04
•CAGR: 20.17%
SHARE BUYBACKS✅
•2018 Shares Outstanding: 14.07B
•LTM Shares Outstanding: 12.39B
By reducing its shares outstanding ~12%, $GOOG increased its EPS by ~13.6% (assuming 0 growth)
MARGINS✅
•LTM Gross Margins: 58.2%
•LTM Operating Margins: 33.2%
•LTM Net Income Margins: 30.9%
***NOW TO VALUATION 🧠
As stated above, investors can expect to receive ~18% MORE in EPS & ~9% LESS in FCF per share
Using Benjamin Graham’s 2G rule of thumb, $GOOG has to grow earnings at a 9.66% CAGR over the next several years to justify its valuation
Today, analysts anticipate 2025 - 2028 EPS growth over the next few years to be more than the (9.66%) required growth rate:
2025E: $9.61 (19.5% YoY) *FY Dec
2026E: $10.21 (6.3% YoY)
2027E: $11.55 (13.1% YoY)
2028E: $13.57 (17.4% YoY)
$GOOG has an excellent track record of meeting analyst estimates ~2 years out, but let’s assume $GOOG ends 2028 with $13.57 in EPS & see its CAGR potential assuming different multiples
23x P/E: $312.11💵 … ~17.4% CAGR
22x P/E: $298.54💵 … ~15.9% CAGR
21x P/E: $284.97💵 … ~14.5% CAGR
20x P/E: $271.40💵 … ~12.8% CAGR
19x P/E: $257.83💵 … ~11.2% CAGR
As you can see, $GOOG appears to have attractive return potential IF we assume >20x earnings (a multiple below its 5-year & 10-year mean while assuming a lower 2028E)
At >22x, $GOOG has aggressive CAGR potential & it’s not unreasonable for the business to even trade for ~23x (given its growth rate, moat, balance sheet, & exemplary capital allocation)
Today at $181💵 $GOOG appears to be a strong consideration for investment
Between cloud ☁️ , AI 🤖 , quantum computing ⚛️, $GOOG has a strong growth runway ahead, with the potential for continued margin expansion serving as an additional tailwind
$GOOGL
___
𝐃𝐈𝐒𝐂𝐋𝐎𝐒𝐔𝐑𝐄‼️: 𝐓𝐡𝐢𝐬 𝐢𝐬 𝐍𝐎𝐓 𝐈𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭 𝐀𝐝𝐯𝐢𝐜𝐞. 𝐁𝐚𝐛𝐲𝐥𝐨𝐧 𝐂𝐚𝐩𝐢𝐭𝐚𝐥® 𝐚𝐧𝐝 𝐢𝐭𝐬 𝐫𝐞𝐩𝐫𝐞𝐬𝐞𝐧𝐭𝐚𝐭𝐢𝐯𝐞𝐬 𝐦𝐚𝐲 𝐡𝐚𝐯𝐞 𝐩𝐨𝐬𝐢𝐭𝐢𝐨𝐧𝐬 𝐢𝐧 𝐭𝐡𝐞 𝐬𝐞𝐜𝐮𝐫𝐢𝐭𝐢𝐞𝐬 𝐝𝐢𝐬𝐜𝐮𝐬𝐬𝐞𝐝 𝐢𝐧 𝐭𝐡𝐢𝐬 𝐭𝐰𝐞𝐞𝐭.
𝐓𝐡𝐞 𝐢𝐧𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧 𝐜𝐨𝐧𝐭𝐚𝐢𝐧𝐞𝐝 𝐢𝐧 𝐭𝐡𝐢𝐬 𝐭𝐰𝐞𝐞𝐭 𝐢𝐬 𝐢𝐧𝐭𝐞𝐧𝐝𝐞𝐝 𝐟𝐨𝐫 𝐢𝐧𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧𝐚𝐥 𝐩𝐮𝐫𝐩𝐨𝐬𝐞𝐬 𝐨𝐧𝐥𝐲 𝐚𝐧𝐝 𝐬𝐡𝐨𝐮𝐥𝐝 𝐧𝐨𝐭 𝐛𝐞 𝐜𝐨𝐧𝐬𝐭𝐫𝐮𝐞𝐝 𝐚𝐬 𝐢𝐧𝐯𝐞𝐬𝐭𝐦𝐞𝐧𝐭 𝐚𝐝𝐯𝐢𝐜𝐞 𝐭𝐨 𝐦𝐞𝐞𝐭 𝐭𝐡𝐞 𝐬𝐩𝐞𝐜𝐢𝐟𝐢𝐜 𝐧𝐞𝐞𝐝𝐬 𝐨𝐟 𝐚𝐧𝐲 𝐢𝐧𝐝𝐢𝐯𝐢𝐝𝐮𝐚𝐥 𝐨𝐫 𝐬𝐢𝐭𝐮𝐚𝐭𝐢𝐨𝐧. 𝐏𝐚𝐬𝐭 𝐩𝐞𝐫𝐟𝐨𝐫𝐦𝐚𝐧𝐜𝐞 𝐢𝐬 𝐧𝐨 𝐠𝐮𝐚𝐫𝐚𝐧𝐭𝐞𝐞 𝐨𝐟 𝐟𝐮𝐭𝐮𝐫𝐞 𝐫𝐞𝐬𝐮𝐥𝐭𝐬.
𝐈𝐧𝐟𝐨𝐫𝐦𝐚𝐭𝐢𝐨𝐧 𝐜𝐨𝐧𝐭𝐚𝐢𝐧𝐞𝐝 𝐢𝐧 𝐭𝐡𝐢𝐬 𝐭𝐰𝐞𝐞𝐭 𝐡𝐚𝐬 𝐛𝐞𝐞𝐧 𝐨𝐛𝐭𝐚𝐢𝐧𝐞𝐝 𝐟𝐫𝐨𝐦 𝐬𝐨𝐮𝐫𝐜𝐞𝐬 𝐛𝐞𝐥𝐢𝐞𝐯𝐞𝐝 𝐭𝐨 𝐛𝐞[...]
Offshore
Dimitry Nakhla | Babylon Capital® In less than 5 months, $GOOG has surged from $181 to $300, a +65% gain📈 Shows how powerful sentiment can be A business with $GOOG moat, growth profile, & AI potential should not have been priced under 20x Today, $GOOG appears…
𝐫𝐞𝐥𝐢𝐚𝐛𝐥𝐞, 𝐛𝐮𝐭 𝐢𝐬 𝐧𝐨𝐭 𝐠𝐮𝐚𝐫𝐚𝐧𝐭𝐞𝐞𝐝 𝐚𝐬 𝐭𝐨 𝐜𝐨𝐦𝐩𝐥𝐞𝐭𝐞𝐧𝐞𝐬𝐬 𝐨𝐫 𝐚𝐜𝐜𝐮𝐫𝐚𝐜𝐲. - Dimitry Nakhla | Babylon Capital® tweet
Offshore
Photo
EndGame Macro
Nvidia’s Blowout Isn’t the Whole Story And Here’s What’s Hiding Underneath.
The quarter is objectively massive. No one sees $57B in revenue and a $200B after hours surge and thinks that’s normal. But once you look past the headline and actually sit with the numbers, the picture gets more complicated, not in a conspiratorial way, just in the way late cycle booms usually get complicated when you read the details closely.
The Print Was Enormous But Uneven Underneath
Nvidia delivered everything the market wasn’t positioned for…Revenue up 62% YoY, margins holding in the mid!70s, data center revenue up 66%, and guidance raised again. All confirmed across pages 1–3 of the earnings report.
That’s why you get a candle straight up. The market went in expecting “great but slowing,” and instead it got “still accelerating.”
But parts of the financials don’t perfectly line up with the clean narrative.
Where the Numbers Don’t Feel As Clean
1. Inventories nearly doubled while management says everything is “sold out.”
Page 6 shows inventories rising from $10.1B in January to $19.8B now. Page 8 shows $4.8B of that increase this quarter.
If every GPU is truly spoken for, tight inventory would make sense, not a warehouse filling up. The helpful explanation is timing: systems built faster than hyperscalers can rack them. That happens in big ramps.
But it also hints at something else: some of this surge is tied to hardware already produced but not yet deployed, not pure real time consumption.
2. Accounts receivable jumped by roughly $10B while the story is “unlimited demand.”
Page 6 shows AR rising from $23.1B to $33.4B year to date.
If Nvidia has all the leverage, you’d expect faster cash conversion, not a larger pile of unpaid invoices. Hyperscalers always pay, that’s not the issue. The point is that this boom is partly running on extended terms and concentrated customer relationships. It’s the kind of detail that doesn’t matter until something in the cycle shifts.
3. Margins that look almost unreal for a hardware heavy business
Gross margin at 73.4% and guided to 74.8% (pages 1–2). Operating margin around 63%.
These margins reflect a temporary monopoly window. Nvidia is capturing value at a level almost no hardware company can sustain over time. Nothing about it is shady, but it’s fragile. When margins sit this high, the distance between perfect and normal becomes a long fall. The risk isn’t today; it’s how the market reacts when margins eventually normalize.
4. The buyback engine is doing real work
Page 8 shows $12.5B in buybacks this quarter, and page 2 shows $37B returned over nine months.
That’s meaningful. Buybacks magnify EPS, reduce float, and amplify upside reactions. Combined with massive passive inflows from Vanguard, BlackRock, State Street, Fidelity, and JPM, it explains why the stock can move like a much smaller company despite its size.
How This Happens Despite a Softening Global Economy
This is the strange part: the world is slowing, yet Nvidia looks untouched by it.
And honestly, it is.
•Hyperscalers are in a capex arms race, not a normal business cycle.
•Governments are pushing money into AI, defense computing, and infrastructure.
•Companies are cutting elsewhere so they can keep spending on compute.
This is capital concentration. Most businesses are tightening. A handful are building the future as fast as possible. Nvidia happens to sit in front of that firehose.
The Real Takeaway
The quarter is extraordinary. Nvidia delivered something few modeled, and the market reacted exactly how it reacts when something rare happens.
But the deeper details like inventories rising, receivables swelling, margins at peak levels, buybacks boosting optics are reminders this boom isn’t spotless.
It doesn’t diminish what they accomplished. It just highlights the gap between Nvidia’s world and the broader economy and historically, that gap is where future volatility tends to form.
BR[...]
Nvidia’s Blowout Isn’t the Whole Story And Here’s What’s Hiding Underneath.
The quarter is objectively massive. No one sees $57B in revenue and a $200B after hours surge and thinks that’s normal. But once you look past the headline and actually sit with the numbers, the picture gets more complicated, not in a conspiratorial way, just in the way late cycle booms usually get complicated when you read the details closely.
The Print Was Enormous But Uneven Underneath
Nvidia delivered everything the market wasn’t positioned for…Revenue up 62% YoY, margins holding in the mid!70s, data center revenue up 66%, and guidance raised again. All confirmed across pages 1–3 of the earnings report.
That’s why you get a candle straight up. The market went in expecting “great but slowing,” and instead it got “still accelerating.”
But parts of the financials don’t perfectly line up with the clean narrative.
Where the Numbers Don’t Feel As Clean
1. Inventories nearly doubled while management says everything is “sold out.”
Page 6 shows inventories rising from $10.1B in January to $19.8B now. Page 8 shows $4.8B of that increase this quarter.
If every GPU is truly spoken for, tight inventory would make sense, not a warehouse filling up. The helpful explanation is timing: systems built faster than hyperscalers can rack them. That happens in big ramps.
But it also hints at something else: some of this surge is tied to hardware already produced but not yet deployed, not pure real time consumption.
2. Accounts receivable jumped by roughly $10B while the story is “unlimited demand.”
Page 6 shows AR rising from $23.1B to $33.4B year to date.
If Nvidia has all the leverage, you’d expect faster cash conversion, not a larger pile of unpaid invoices. Hyperscalers always pay, that’s not the issue. The point is that this boom is partly running on extended terms and concentrated customer relationships. It’s the kind of detail that doesn’t matter until something in the cycle shifts.
3. Margins that look almost unreal for a hardware heavy business
Gross margin at 73.4% and guided to 74.8% (pages 1–2). Operating margin around 63%.
These margins reflect a temporary monopoly window. Nvidia is capturing value at a level almost no hardware company can sustain over time. Nothing about it is shady, but it’s fragile. When margins sit this high, the distance between perfect and normal becomes a long fall. The risk isn’t today; it’s how the market reacts when margins eventually normalize.
4. The buyback engine is doing real work
Page 8 shows $12.5B in buybacks this quarter, and page 2 shows $37B returned over nine months.
That’s meaningful. Buybacks magnify EPS, reduce float, and amplify upside reactions. Combined with massive passive inflows from Vanguard, BlackRock, State Street, Fidelity, and JPM, it explains why the stock can move like a much smaller company despite its size.
How This Happens Despite a Softening Global Economy
This is the strange part: the world is slowing, yet Nvidia looks untouched by it.
And honestly, it is.
•Hyperscalers are in a capex arms race, not a normal business cycle.
•Governments are pushing money into AI, defense computing, and infrastructure.
•Companies are cutting elsewhere so they can keep spending on compute.
This is capital concentration. Most businesses are tightening. A handful are building the future as fast as possible. Nvidia happens to sit in front of that firehose.
The Real Takeaway
The quarter is extraordinary. Nvidia delivered something few modeled, and the market reacted exactly how it reacts when something rare happens.
But the deeper details like inventories rising, receivables swelling, margins at peak levels, buybacks boosting optics are reminders this boom isn’t spotless.
It doesn’t diminish what they accomplished. It just highlights the gap between Nvidia’s world and the broader economy and historically, that gap is where future volatility tends to form.
BR[...]
Offshore
EndGame Macro Nvidia’s Blowout Isn’t the Whole Story And Here’s What’s Hiding Underneath. The quarter is objectively massive. No one sees $57B in revenue and a $200B after hours surge and thinks that’s normal. But once you look past the headline and actually…
EAKING: Nvidia, $NVDA, stock rises +5% after announcing record quarterly revenue of $57 billion.
That's +$205 BILLION of market cap on this move. https://t.co/m0t1J3cKCT - The Kobeissi Letter tweet
That's +$205 BILLION of market cap on this move. https://t.co/m0t1J3cKCT - The Kobeissi Letter tweet