ubineer-site

Ubineer's Deep Dive Cites Critical Information 70 Days Ahead of Earnings: Arista Networks

Mon, 18 May 2026 16:14:54 GMT

Ubineer's Deep Dive correctly discovered an anomaly in Arista Networks' public filings some 70 days ahead of earnings, alerting clients to potential pressure on the company's Gross Margin.

The critical insight Deep Dive correctly identified was the considerable delta between the 2024 and 2025 values, combined with the disparity between the Deferred Cost of Goods Sold and Deferred Revenue.

This calculation suggested a potential Gross Margin of approximately ~60.2% waiting to transition into the income statement.

This finding indicated a clear risk that the gross margin could experience further decline on recognition of these deferred line items, especially when the reported GAAP Gross Margin for Q4 and Fiscal Year 2025 was 62.9% and 64.1% , respectively.

May 18, 2026 – Toronto – Earlier this month, Ubineer confirmed that its proprietary Deep Dive product made an impressive discovery well ahead of the earnings, accurately identifying ongoing pressures on Arista Networks ' Gross Margin.

Over two months before Arista Networks' earnings announcement on May 5, 2026, Deep Dive had signalled to clients that there lurked a significant anomaly. The company's artificial intelligence pinpointed a subtle disclosure within Arista's 10-K filing, concentrating its analysis on the Deferred Cost of Goods Sold line item, which made up part of the Arista's Prepaid Expenses and other current assets .

"Discovering anomalies over two months in advance is not just an enormous research advantage for portfolio managers, it also validates Deep Dive 's ability to uncover critical, non-obvious disclosures that human analysts can easily miss," said Andi Kerenxhi , Ubineer's president and founder.

"Our mission is to give our clients an unprecedented edge, and this discovery was utilized by our clients well ahead of the May 5th report. Ubineer's AI-driven technology is redefining the timeline for fundamental analysis." he added.

Deep Dive provided Ubineer’s subscribers with precise and credible market intelligence so they could take appropriate, timely steps to mitigate the gross margin risk.

To learn more, download our Deep Dive slide deck: Arista Networks Q1 2026 .

The Ubineer Advantage

Ubineer's predictive technology offers immediate tangible benefits for portfolio managers, analysts and chief investment officers (CIOs), allowing them to:

Anticipate the Fundamental Future: Obtain an accurate model of a company's performance with a forward view of up to five quarters.
Expect the "Guide": Accurately predict management's future outlook, enabling portfolio managers to set independent targets.
Set a Thesis Margin: Supply an "80% Confidence Interval" for critical metrics like Revenue, Operating Income, and Free Cash Flow, which helps validate or challenge the investment thesis.

Institutional investors are encouraged to contact Ubineer for a personal demonstration of the company's software products and capabilities. We invite you to enquire about companies we currently cover or let us know a publicly traded company you’re interested in or misjudged.

Ubineer's Predictions: Meta Q1 2026 vs Actuals - Another Case Study

Sat, 09 May 2026 19:45:19 GMT

Ubineer's Predictions continues to demonstrate a strong value proposition to CFAs, portfolio managers and Chief Investment Officers. Last time we discussed Celestica . Today, as another case study, we'll be looking at how our AI-powered forecasts performed compared with META's Q1 2026 actual results.

Pinpoint Revenue Accuracy

Ubineer's primary goal is to provide highly accurate models to anticipate how a company will fundamentally perform relative to their guidance. Our AI-driven software demonstrated remarkable precision in forecasting META's total revenue:

Ubineer's Prediction: Q1 2026 Total Revenue was predicted to be $56.058 billion .
Actual Result: META reported actual revenue of $56.311 billion .
Accuracy: The AI’s projection was incredibly close, with a dollar delta of just $252.6 million, representing a marginal 0.44% variance .

Predicting Future Guidance

A unique and powerful feature of Ubineer is its ability to predict management's future outlook. Our models forecast not only the current quarter's results, but also the guidance management issued for the next period.

Ubineer's Prediction: Our software inferred that management would guide META's Q2 2026 Revenue to a midpoint of $59.5 billion (Low: $57.5B, High: $61.5B).
Actual Guidance: Management's actual Q2 Midpoint guidance came in at exactly $59.5 billion .
Management's Quote: " We expect second quarter 2026 total revenue to be in the range of $58-61 billion. "
Accuracy: Ubineer achieved a rather remarkable 0.00% delta on revenue guidance accuracy — essentially a direct hit.

Granular Line-Item Analysis and Operational Analysis

Ubineer is also capable of breaking down META's revenue into specific segments and key operational metrics.

Advertising Revenue:
Prediction: $54.699 billion ($54.7B)
Actual: $55.024 billion
Accuracy: A marginal 0.589% delta

Operating Income [NEW]:
Prediction: $21.911 billion
Actual: $22.872 billion

Ad Impressions and Price Per Ad: Our AI model accurately forecast growth in both key metrics.
Management's Quote: " Ad impressions delivered across our Family of Apps increased by 19% year-over-year ."
Management's Quote: " Average price per ad increased by 12% year-over-year. "

Ubineer Advantage for Investment Professionals

Ubineer offers immediate and tangible benefits to CIOs, portfolio managers and analysts who are willing to explore AI-augmented investment tools.

Anticipate the Fundamental Future: Our predictions provide clients with an accurate model of any company's performance, not just for the current quarter, but with a forward view of up to five quarters.

Expect the "Guide": By accurately predicting management's future outlook, portfolio managers can set their own independent targets and avoid reliance on consensus sell-side estimates or "whisper numbers".

Provides a Thesis Margin: We supply an "80% Confidence Interval" for critical metrics like Revenue, Operating Income, and Free Cash Flow, establishing a clear "floor" and "ceiling" to validate or challenge any investment thesis.

Available for download: META's Q1 2026 Results

If you are an institutional portfolio manager, analyst or CIO, we invite you to contact us for a personal demonstration of our software products and capabilities. Feel free to call us or book an appointment .

Ubineer's Predictions: Celestica Case Study Q1 2026

Wed, 29 Apr 2026 17:11:37 GMT

Ubineer Predictions functions as an AI-driven "foresight engine" designed to identify financial surprises before they happen. Using the Celestica (CLS) Q1 2026 results as a case study, we'd like to present an explanation of what these predictions do and how accurate they were in this specific instance.

Identifying "Sandbagging" (Conservative Guidance)

Ubineer’s primary goal is to determine if a company's management is being overly conservative with their public targets.

What happened: Management guided Q1 2026 revenue at a midpoint of $4.00B .
Ubineer’s View: The AI predicted $4.234B , significantly higher than management's range. It explicitly noted that all of its simulations outperformed management’s full-year guidance.
The Result: Celestica reported $4.05B . While slightly lower than the Ubineer "Best Estimate," the AI correctly identified that management would be forced to raise their full-year guidance (which they did, from $17B to $19B).

Predicting Guidance Raises

A unique feature of Ubineer is that it doesn't just predict earnings; it predicts what management will say about the future.

The Prediction: Ubineer predicted that management would raise FY2026 guidance to the $17.5B – $17.8B range.
The Result: Management actually raised it even further to $19.0B . Ubineer's "Best Model" for the full year was $19.01B —which turned out to be almost a direct hit (99.9% accuracy) on the new annual figure management eventually adopted.

Granular Line-Item Analysis

Ubineer breaks down revenue into specific business segments to see what is driving the growth.

The Prediction: Ubineer highlighted Enterprise revenue as a "key delta" and predicted it would rise. It also predicted CCS (Connectivity & Cloud Solutions) revenue would be $3.19B .
The Result: CCS revenue came in at $3.24B , and Enterprise revenue surged 101% year-over-year. Ubineer correctly identified the specific segment (Enterprise/AI infrastructure) that caused the massive beat.

Confidence Metrics & Risk Assessment

Ubineer provides a Confidence Rating (Medium, in this case) to tell investors how much to trust the AI's "Best Model."

The "Medium" Logic: The AI was confident because its models "consistently beat the midpoint" of guidance, but it flagged a "Medium" rating because the 80% Confidence Interval was wider than previous quarters. This warns the user that while the direction (up) is certain, the exact number has more volatility.

Summary: What Ubineer can do for Professional Investors

Anticipates Beats: It tells you if a company is likely to exceed the numbers they told Wall Street.
Estimates the "Raise": It predicts how much management will increase their future outlook during the earnings call.
Provides a Safety Margin: By giving an "80% Confidence Interval" (e.g., $3.95B to $4.47B), it gives you a "floor" and a "ceiling" for the stock's potential movement.

I n this Celestica example, Ubineer's AI was a leading indicator : it inferred the $17B full-year guidance was too low weeks before management admitted it and raised it to $19B.

If you're an institutional investor, professional analyst or a CIO and would like a personal demonstration of our software products and capabilities, we invite you to reach us with your questions or consider booking an appointment.

Ubineer Joins CAPIS for March 2026 Research Briefing

Wed, 01 Apr 2026 18:15:38 GMT

President Andi Kerenxhi was pleased to join Ed O'Dowd of Capital Institution Services (CAPIS) in a podcast to outline our perspective, platform and new product offerings.

Ubineer invites you to listen in full or check the transcript summary below for details.

00:00 — Introduction: AI-Driven Research and Focus on Mega-Cap Tech

Ed O’Dowd opens the March CAPIS Monthly Research Briefing, introducing Andi Kerenxhi of Ubineer. He frames today’s session as a continuation of CAPIS’ ongoing series featuring differentiated research perspectives, highlighting Ubineer’s use of AI to enhance fundamental analysis.

01:24 — AI CapEx: A “New Industrial Revolution”

Andi describes the current AI investment cycle as “insane—in a good way,” likening it to a new industrial revolution. He outlines Ubineer’s approach of converting financial data into predictions, identifying anomalies to reduce manual research and testing whether investment theses are already priced into markets. He emphasizes that AI is transforming how institutional investors validate conviction and assess forward-looking expectations.

Andi explains that today’s presentation will continue with his previous analyses of NFLX, META, GOOG, MSFT and AMZN, analyzing Ubineer’s estimates versus actual performance and then taking a look at Nvidia and its role in market structure. .

05:30 — Netflix: Content Strategy, Pricing Power, and Growth Outlook

Andi revisits his prior Netflix thesis: increased amount of content drives value, and monetization comes through an influx of users and higher pricing.

Key takeaways:

Revenue growth broadly tracked expectations (17.4%), though not perfectly.
Strategic change hinges on content acquisition (e.g., attempted Warner Bros. Discovery deal). The deal failure did not alter the core thesis.
User growth appears to be plateauing.

Forward-looking insight:

The next major lever is pricing power, with potential price hikes expected within the next 2–3 quarters.
Rising streaming obligations signal continued investment in platform value.

9:00 — Meta: Ad Pricing, AI Spend, and Escalating CapEx Commitments

For Meta, Andi shares his thesis that revenue growth would be driven by the number of ads shown and price per ad.

Both prior predictions held: Advertising growth accelerated over the last two quarters (~25%+) and CapEx continued rising alongside revenue.

Total obligations surged from ~$80B to ~$240B, reflecting aggressive long-term AI investment. Meta is intentionally reinvesting cash flow into AI infrastructure as long as it continues driving revenue growth. Andi also reiterates the longer-term “call option” on Meta through AI-enabled hardware (glasses).

13:10 — Alphabet (Google): Cloud Growth and CapEx “Catch-up”

Andi highlights Google Cloud as the previously overlooked “sleeping giant.” Cloud growth accelerated from ~30% to ~47%. Google issued CapEx guidance for the first time ever (~$175–185B). Growth is being driven by signed contracts (remaining performance obligations). RPO surged from sub-$100B to ~$250B, indicating strong forward demand.

16:00— Microsoft & Amazon: Cloud Demand, Upgrades, and Cash Flow Deployment

For Microsoft, revenue growth was driven by deat count, cloud + AI spend and customer upgrades (365 E5 → E7 licenses). Seat count concerns exist due to AI, but no clear deterioration yet. ~$625B+ in unmet demand highlights capacity constraints.

For Amazon, Andi’s predictions for Q4 and Q1 were that world e-commerce would be a key driver for revenue growth, as well as cloud and AI spend. Today’s numbers show e-commerce stabilizing and AWS growth accelerating. Free cash flow expected to turn negative due to heavy CapEx. RPO jumped in one quarter from 20% to ~ 40%, supporting continued investment.

21:20 — NVIDIA and the Scale of AI Infrastructure Spending

Andi shifts to NVIDIA as the central beneficiary of AI investment. Total AI CapEx projected to rise from ~$443B → ~$755B YoY. ~40–50% of that spend flows to NVIDIA.

NVIDIA’s data center revenue expanded from < $20B → ~$200B post-ChatGPT. CEO commentary suggests ~$1T in future demand visibility. Its greatest risk is supply chain constraints. However, inventory (+100% YoY) and purchase obligations (~$95B) suggest NVIDIA has secured supply capacity.

25:50 — Why Mega-Cap Tech is Lagging: Capital Deployment vs. Cash Accumulation

Andi addresses a central market question: Why is large tech (the Mag 8) underperforming relative to global markets? His view is that these companies are now spenders, not accumulators. Capital is flowing outward to suppliers, infrastructure providers and global markets. An additional factor is uncertainty around future free cash flow durability.

Using AI agents simulating investor behavior, Andi demonstrates that no solid consensus exists on future cash flow expectations for FY 2026 nor 2027. This lack of clarity is contributing to market hesitation.

29:23 — Q&A: Strategic Risks Across Big Tech

Entering the Q&A portion of the MRB, Ed asks which companies face the most strategic challenges moving forward. Andi highlights Microsoft (seat count uncertainty) and Meta (must justify massive AI spend if revenue growth slows).

31:05 — Oracle and AI Infrastructure Monetization

Andi pushes back against the “AI bubble” narrative, arguing that the market is still in its early stages.

On Oracle:

Strong monetization of data center capacity
Growth tied to remaining performance obligations vs. purchase obligations
NVIDIA may be enabling new competitors by allocating supply strategically

32:47 — Regulatory Risks & Impact Relative to AI Tailwinds

Addressing recent court cases against Meta and Google, Andi states that financial penalties are immaterial relative to the revenue scale, and the core driver remains user engagement (time spent on platforms).

34:14 — What the Market Needs to See Next: Proof of Monetization

Andi outlines the key condition for market confidence: Continued growth in cash flow from operations. There’s evidence that AI spend is generating real revenue, not subsidized usage. He emphasizes that this is the core debate in markets today, and if proven, current investment levels will be validated.

35:50 — Closing Remarks and April MRB Preview

Ed closes the session, thanking Andi and previewing the next CAPIS briefing with Markets Policy Partners, who will return on Thursday, April 16 to discuss macroeconomic and geopolitical themes.

Source: https://capis.com

Ubineer Product Road Map: Pillars of Alpha Architecture

Mon, 23 Mar 2026 19:49:08 GMT

The world of investment is constantly accelerating, demanding not just speed, but a new class of precision and Artificial Intelligence autonomy to stay ahead. At Ubineer, we've engineered a product stack designed to move beyond traditional analysis, enabling the investor to truly become an Alpha Architect .

The Investor as an Alpha Architect

Our last post outlined the four core pillars defining the new era of professional investing: Hyper-Speed Data Synthesis, Predictive Modeling Precision, Insights at Scale and the redefined role of The Investor as the Alpha Architect.

This shift is not merely about adopting new technology. It's about fundamentally redesigning the investment process to capture asymmetric alpha opportunities that are invisible (or extremely difficult to detect) using traditional approaches.

At Ubineer our entire product roadmap is engineered to operationalize these four elements. Our solutions serve as the necessary infrastructure, empowering your team to transition seamlessly into the role of an advanced Alpha Architect. With that said let's jump into the 4 Pillars.

Pillar 1: The Foundation of Actionable Data

The most brilliant models are useless without clean, quality data. As the saying goes Garbage in, Garbage out. Our proprietary Data Engine delivers a Hyper-Speed Data Ingestion and Synthesis , which converts fragmented, unstructured data from public SEC sources into a unified, immediately actionable data fabric.

Information Automation is the key output of this engine. By transforming raw, messy text into precise line items, custom formulas our engine is able to discover anomalies that even the best analysts would miss. This converts traditional, slow investment research into an autonomous engine for insight and anomaly detection discovery, giving you and your investment team back the most valuable resource: time.

We accomplish this with a product called Layers .

Pillar 2: The Engine of Predictive Advantage

Our Predictions product is our advanced machine learning environment focused on Predicting Company Key Performance Indicators (or KPIs) . Think of it as a meeting of quants and portfolio managers creating models that:

Forecasting KPIs: Our models are built to out predict Wall Street on the key metrics that move a public company. Our models forecast KPIs at least five quarters in the future from revenue line items all the way to free cash flow margin. On average we outperform Wall Street between 70-90% per company per quarter. We also identify inflections before they are reported, delivering the predictive precision required to stay ahead of the curve.
Autonomous Anomaly Detection: The market often misses hidden risks and structural issues buried deep in public filings. We use AI to uncover these hidden signals , identifying reporting risks and structural anomalies in company data long before human analysts or the broader market can react.

Pillar 3: The Crucible of Conviction

One of the key questions we get from customers is: "Has the market priced in my Investment thesis?"

So we built a product that uses Generative Artificial Agents to stress test any KPI specific Investment Thesis. Our OmniOra AI—a specialized, multi-agent system—attempts to simulate the probability of your investment thesis is priced in while identifying your investment thesis blind spots and rigorously critiques your thesis logic.

Pillar 4: The Control Center for Execution

Finally, all of the data, modeling, insight discovery and investment thesis testing converge in a unified workflow interface. It's the place where you seamlessly integrate the synthesized data, the predictive models, and the stress-tested insights into final, strategic action. More on this pillar soon.

Ubineer doesn't just offer tools, we offer a transformation. We provide the integrated intelligence required to eliminate information lag, predict market movements with greater accuracy, and architect portfolios designed for sustained alpha.

The AI Era of Professional Investing

Sun, 22 Mar 2026 19:19:20 GMT

AI has ushered in a seismic shift in predictive software and market intelligence, and Ubineer has been on top of it right from the start.

A New Era of Professional Investing

The investing world is undergoing a seismic shift, one that promises to redefine the very essence of professional investing.

For decades, the image of a successful investor was defined by the quick-witted analyst, or the seasoned portfolio manager pouring over dense annual reports and other sources.

While experience, intuition and fundamental analysis will always play a valuable role in allocating capital, AI has become a powerful and disruptive partner entrant, and it's maturing in at an alarming rate.

The AI Era of Professional Investing is not undergoing a gradual evolution. It's a fundamental switch in how capital is managed, risk is assessed, and opportunities are identified.

We are in the early stages of an era where the amount of data an analyst will need to sift through to convert to information will increase by 10-15 times. None will matter more because it's all about how many valuable decisions can be made with respect to unpriced insights.

This will be the Alpha driver of any fund in the future. This is what Ubineer aspires to solve.

From Human-Centric Investment Intel to Human-Augmented

The transition is marked by a profound change in the primary driver of investment decisions. The world is switching its operational paradigm from one heavily reliant on human cognitive limits to one empowered by artificial intelligence scale and precision.

In this new era, the professional investor doesn't merely compete with other humans, they compete by wielding more AI algorithms that help identify Insights. Their value will be driven by their decision-making process, and that process is singularly focused to capture Alpha.

What Defines This New Era?

Hyper-Speed Data Synthesis: AI can ingest, process, and correlate petabytes of structured and unstructured data. This capability far surpasses any human capacity.
Predictive Modeling Precision: Beyond simple forecasting, machine learning models can identify non-linear relationships and subtle market anomalies that are invisible to any humans. This leads to alpha generation driven by granular, data-validated insights, rather than broad macroeconomic assumptions alone.
Insights at Scale: AI systems can dynamically discover insights that a human has no ability to do because they lack the right context, or are not able to spot the anomaly in the data.
The Investor as Architect: The role of the professional investor is transforming. They are moving away from being hands-on stock pickers and towards becoming sophisticated decision architects. Their value lies in overseeing the insights these intelligent machines discovers and more imprortantly making the asymmetric decisions to capture alpha that is yet to properly be priced in.

The AI future of professional investing is not about replacing human insight and intervention, but rather it's about augmenting it with exponential power.

It is an exciting, challenging, and inevitable transition that promises a more efficient, sophisticated, and potentially much more rewarding financial landscape for those ready to examine its benefits and features.

We hope you and your investment team join us in this ubiquitous and pioneering movement.

Ubineer's Predictions KPI software has consistently outperformed Wall Street consensus estimate with more than 80% accuracy.

Q4 2025 Mega-Cap Trends: AI Driven Insights from Ubineer

Mon, 06 Oct 2025 19:48:03 GMT

In the September 2025 edition of CAPIS’s Monthly Research Briefing, Ubineer founder Andi Kerenxhi joined Ed O'Dowd to examine how AI is shaping the investment landscape through the lens of fundamentals.

Andi shared data-driven insights into mega-cap tech names like Netflix, Meta, Google, Microsoft, Amazon and NVIDIA , highlighting CapEx signals, demand drivers and why he believes we’re still in the earliest innings of a potential AI investment cycle.

00:00 – Introduction

Ed welcomes Andi Kerenxhi, a Toronto-based investor focused on helping PMs validate investment hypotheses using AI tools trained on company fundamentals. Ubineer uses proprietary data ingestion and conformal prediction models to assess the likelihood of earnings outcomes, CapEx trends, and KPI shifts.

03:02 – AI’s Role in Investment Decision-Making

Andi explains that Ubineer doesn’t provide investment recommendations, but rather surfaces probabilities and insights from public filings, KPI projections and spending patterns. Their tools are designed to supplement traditional fundamental research, not replace it.

Netflix

07:00 – Netflix’s Capital Allocation

Andi starts with Netflix, noting the company’s recent spike in unrecorded streaming obligations—now the fifth highest on record—indicating aggressive content investment. About 35–40% of spend goes to in-house production versus licensing, and any change here would be the earliest sign of a strategic shift.

9:45 – Revenue Growth Drivers

Netflix’s Q3 guidance calls for 17.3% revenue growth, but Andi expects at least 17.4%, which could trigger a market reaction. He also highlights the growing role of U.S. users in driving incremental revenue, especially since Netflix’s price hikes earlier this year. Andi points to emerging Q4 seasonality tied to live events, such as the Paul vs. Tyson fight and NFL games, which could support a stronger-than-expected Q4 performance. He encourages investors to track Netflix’s content mix and capital distribution closely.

Meta

12:43 – Meta’s Historic Ad Revenue Spike in Q2

Meta added $8.2B in Q2 ad revenue, outpacing even its typical Q4 holiday surge. Andi suggests this could be a function of increased ad load on user feeds and broader monetization strategies.

14:13 – CapEx Spend

CapEx is rising rapidly, with growth visible across Meta’s buildings, data centers and server purchases. Construction-in-progress and server deployments are accelerating. “All other obligations” now top $80B—up from $20B in Q1 2024—suggesting multi-year infrastructure buildout.

16:50 – Meta Hardware

Andi discusses Meta’s new wearable devices as a potential iPhone moment. Priced at $799, these glasses could create an entirely new product category. Q1–Q2 2026 will be crucial in validating whether this is a transformational shift or just another moonshot.

Google

18:42 – Search and YouTube Ad Trends

Despite fears that AI might erode Google’s core business, ad revenue from search and YouTube is still climbing. Google Network revenue is shrinking, but the ad engine shows resilience—especially as YouTube expands its GenAI roadmap.

21:37 – Cloud Business Acceleration

Google Cloud (GCP + Workspace) is experiencing a breakout quarter, with RPO (remaining performance obligations) growing ~40% YoY. Price hikes to Workspace users helped, but growth is largely organic—suggesting strong enterprise demand for GCP.

23:00 – Data Center Leasing

“Leases not yet commenced” spiked from <$1B to $24B in just two quarters. This level of forward-committed infrastructure spend reinforces Andi’s view that the AI cycle is still in its early phase.

Microsoft

24:55 – Cloud and Productivity Suite Momentum

Microsoft’s Azure and Microsoft 365 Commercial businesses are both contributing to a $4.25B revenue spike, with 365 upsells driven by GenAI integration. The current growth rate has jumped to ~28% YoY.

37:50 – RPO Acceleration

RPO growth in 2025 (fiscal year-end June 30) hit new highs, with clients signing large, multi-year contracts at a faster pace than ever. Andi sees this as further evidence of durable enterprise demand—not just hype.

Amazon

28:56 – E-Commerce Still the Core Driver

Amazon’s world e-commerce business (excluding Whole Foods and physical stores) hit nearly $140B last quarter, with U.S. and international segments both accelerating. Andi references Fed data on retail e-commerce trends as confirmation.

31:20 – AWS

While AWS is larger than Azure or GCP, its RPO growth is slower due to scale. Andi expects a near-term revenue spike to close the gap between reported sales and future commitments.

32:50 – The AI “Bubble”

Andi argues that true bubbles are fueled by debt, and today’s AI CapEx is largely funded by free cash flow. Most spending is still “first breath in the balloon,” not late-cycle mania.

35:07 – Audience Q&A

Andi takes questions from the audience regarding AI trends:

He highlights emerging players like CoreWeave to show that hyperscalers may not have lasting moats in GenAI training workloads. OpenAI’s willingness to shift from Azure to Oracle is proof that compute can be commoditized if price is right.
He adds that NVIDIA doesn’t manufacture its own chips (TSMC does), but it excels at integrating chips, servers and racks into ready-to-deploy data center stacks. This capability is smoothing out the CapEx curve for cloud buyers, extending the demand cycle.
He reiterates that until debt becomes widespread and unsustainable, we’re not in a bubble. Valuations may run hot, but the underlying spending is healthy and rational—for now.

44:49 – Wrapping Up

Ed thanks Andi for the unique insights and reminds attendees to join next month’s Market Research Briefing on Thursday, October 23, featuring returning guest Markets Policy Partners for a deep dive into policy, the Fed and macro forces.

Source: https://capis.com