No paywall for this report!

Last week, I came across a viral tweet in which the author claimed that by starting with just $100 in 2012 and then rotating it into the previous year's best-performing S&P 500 stock, your portfolio would now be worth $176K.

For context, the same $100 left in the S&P 500 or Berkshire Hathaway would be worth about $800.

So is that all there is to investing? Can you really compound at 66% a year just by betting on last year’s winner? As the saying goes, there are lies, damned lies, and statistics.

A quick note about AI backtests

Generally, I am skeptical of these AI backtests because they almost always make simple errors. For example, here’s what you will get if you ask Gemini what the best-performing S&P 500 stock in the year 2000 was.

But a simple fact check shows that none of the top 3 stocks mentioned in this list were in the S&P 500 index in 2000. Laboratory Corp got added in Oct’04, DGX in Dec’02, and VRTX in Sep’13.

Here’s Gemini’s response when you call that out:

Pre-AI, you didn’t have to be so skeptical, since any analyst who puts their name on the research will triple-check their work or lose all credibility instantly.

So, is the above backtest correct?

Read more

None of these brands lacks the resources to build its own. Supreme ran a custom-built site for years before moving to Shopify in 2023. And they are far from alone. Millions of merchants in 175+ countries now use Shopify, reaching roughly 900M shoppers.

It started as a side project. In 2004, Tobi Lütke wanted to sell snowboards online under his own brand. He tried Yahoo! Stores, Miva, and osCommerce, but none were good enough. So he built his own storefront. Soon, other merchants wanted one too.

Shopify was built for merchants who don’t want to be just another product among Amazon’s 600M+ SKUs, but want to give consumers a better shopping and brand experience with their own digital storefront. When Shopify started, Amazon was already at a scale of ~$6.9B in sales, selling everything from books to electronics, but that didn’t stop Shopify from becoming one of the fastest-growing e-commerce companies today.

As Shopify scaled, it stuck to the same problem, helping people build their own brands, and expanded from a simple digital storefront to global payments integration, checkout optimization, working-capital credit, tax calculation, inventory management, advertising strategies, and more.

As AI progresses to reshape retail shopping, Shopify is already positioning itself as the leading enabler of independent online retailers. AI-driven traffic to Shopify stores and orders from AI-powered searches grew nearly 3x YoY in 2Q 2026.

The question is whether AI agents make Shopify more essential or easier to replace. Agents don't care about storefront design. They need a catalog they can query, a checkout they can call, and a payment they can settle. Shopify is building all three, and it switched on every eligible US merchant by default 6 months ago.

But at ~110x earnings, the stock already prices in 30%+ GMV growth and margins of 20%+. Below, we break down how Shopify makes money, why agentic commerce could widen its lead, how durable its moats are, and what we think the stock is actually worth.

Subscribe now

Business Model:

Shopify has 2 major revenue streams: subscription solutions and merchant solutions.

Subscription solutions: They first onboard a retail merchant by giving them a basic digital storefront, which includes a website, checkout, and inventory management. They charge a largely fixed subscription fee for managing these services. Their subscription prices have four main tiers, with Basic priced at $29/mo (billed annually), Grow at $79, Advanced at $299, and Plus for large firms from $2,300. This is a high-margin business with ~80% GM.

Merchant solutions: Once a merchant purchases a subscription, Shopify sells additional services like payment processing, shipping, advertising solutions, etc. Shopify makes money in this stream only if merchants make money. They take a simple cut of merchant sales for most of their products here. The gross margin of this business is ~40%.

Payments is their largest contributor to merchant solutions. Shopify processed ~$248B of payments in FY25, or ~66% of total GMV, up from 45% in 2020. This segment has a significant runway, as they launch in new geographies like India and LatAm ex-Mexico, where Shopify Payments isn’t available today. In mature markets like North America, close to 90% of merchants already use Shopify Payments.

Their other services in merchant solutions are still smaller / in early stages:

• Shopify Capital: ~$2.2B loan book, with ~$1.6B deployed in Q2’26 alone.

• Shop Pay Installments: BNPL product, GMV grew by 53% YoY in Q2’26.

• Shopify ads (Shop Campaigns): Merchants running live campaigns tripled YoY in Q1’26, and for some small merchants it drives up to a quarter of GMV.

• POS hardware: GMV grew 32% YoY in Q2’26.

Agentic Commerce & Shopify’s Role:

Consumer AI is graduating from simple chat inference to complex task handling. Meta’s Muse launch has accelerated this process, with Muse being downloaded 5 million times within 22 days vs. the ChatGPT app’s 5 million in 56 days.

Consumers, on average, take ~26 days from their first search on platforms like Amazon to decide on and purchase a product. Consumers have to compare multiple product features, along with prices and payment offers, before deciding on the final purchase.

AI agents excel at exactly these kinds of comparative tasks. But, as with all AI applications, clean data with better context will improve the efficiency and accuracy of these systems.

Traditional digital shopping infrastructure is built around websites optimized for human clicks and eyeballs to improve conversions. This is not the most efficient infrastructure on which AI agentic commerce can run.

Agentic Infrastructure:

An agent can browse a store the way a person does, but it’s slow and brittle. It has to read pages built for humans, and a layout change or bot check can break the purchase. Agents work best when each step of the purchase is a machine call. Search becomes a catalog it can query. Checkout becomes an API, where the merchant still prices tax and shipping. A stored card becomes a payment token capped to one purchase and amount. A saved password becomes account access granted by permission.

For agents to transact with millions of shops, everyone in the stack — the consumer’s agent and the retailer’s systems — needs to speak the same machine language. Two open standards have been built to solve this: OpenAI and Stripe’s Agentic Commerce Protocol (ACP, September 2025) and Google’s Universal Commerce Protocol (UCP, January 2026), co-developed with Shopify.

Both protocols cover the same journey: discover product, build cart, check out, and track order. ChatGPT uses ACP mainly to ingest merchant catalogs, and final payment occurs at the merchant’s checkout. Google uses UCP to close the full loop, as it integrates with agentic payment solutions such as Google Pay and the Agent Payments Protocol (AP2).

Every eligible U.S. Shopify merchant was switched on by default on March 24, 2026, with out-of-the-box access to ChatGPT, Copilot, AI Mode, and Gemini, and zero integration work. A non-Shopify merchant has to set this up through its own platform or build its own AI-readable catalog, or risk losing the demand coming from these agents.

Google, OpenAI, and now Meta are building the surfaces where the asking happens. Shopify is building the catalog, the cart, and the settlement underneath, with Shopify Catalog feeding ChatGPT and Shopify checkout embedded in Copilot.

Agentic Commerce Value Unlock:

Historically, search has favored bestsellers and products with high ad spend. Agents can parse detailed queries against product attributes, so a niche product that better fits the brief can win in the agentic commerce world.

In 2025, 71% of AI-attributed orders on Shopify came from outside its top 100 categories (long-tail orders), compared with about 55% of sales overall. The long-tail share of AI-attributed orders increased to 75% in 2Q26, reflecting a clear benefit for niche products and categories.

Amazon typically charges sellers a 15% referral fee, and many also spend on Amazon ads to stay visible. Niche products will find it difficult to pay such costs at low volumes. Selling direct gives brands more room to offer better value, while agents can find these products more easily than shoppers searching the web.

The AI catalog itself is just another subscription feature for Shopify. The bigger opportunity comes from any additional GMV that agents send to long-tail merchants, from which Shopify makes money through payments, credit, etc.

Business Moats

Shopify’s moats come from its improving product scale, higher switching costs and lower customer acquisition costs:

Scale:

Shopify has been built over 2 decades by refining its product suite. Shopify says its checkout converts up to 36% better than competing platforms, thanks to these optimizations against billions of real transactions and customer behavior data.

A new entrant building a checkout from scratch is starting at zero on this learning curve. The same applies to fraud detection and payment routing — these systems improve with data volume, and Shopify’s data advantage over a new entrant is a significant moat.

As agentic commerce grows, agents can cover a far wider search landscape than humans, but at the end of the day they run on limited compute and cannot look at and compare every product. So companies that help retailers become more discoverable to these agents will gain an advantage and build a moat. Just as earlier data moats were built around human behavior on retail shopping apps, tomorrow’s could be built around agentic behavior with retail SKUs.

As shoppers shift to agentic commerce, some of these traditional scale and data moats can diminish. But Shopify is the leading commerce platform building catalogs for AI agents, and it has a lead in driving better conversion through them: Shopify says AI searches powered by its Catalog convert at 2x the rate of those using scraped data.

Switching costs:

Shopify doesn’t just hold the products and prices customers can buy. A store also carries customer reviews and accounts that a retailer stands to lose in an abrupt switch. Customer passwords can’t be exported at all, so every shopper has to reset their login on the new platform. Hundreds of product listings also need to be migrated and remapped, and the theme and apps built around them rebuilt. This will inevitably lead to some loss of customers and revenue, and the last thing a retailer wants is to lose business.

Add-on services such as inventory financing through Shopify Capital, repaid automatically from a share of the store’s daily sales, make merchants stickier still. When a merchant cancels its Shopify subscription, it has to switch all merchant services: the digital storefront, payments, credit and inventory management tools.

While Shopify’s switching costs are not as high as those of enterprise SaaS, the breadth of its product suite and its integration into nearly every business process of a retailer make it difficult to switch. This is reflected in its retention rate of ~92% among merchants doing over $1 million a year and 97% above $10 million.

Customer acquisition costs:

When a new retailer looks for a digital storefront, Shopify is often the first name they encounter through Search, YouTube tutorials, business media, and word-of-mouth from other merchants.

This is the result of years of brand and community building, combined with the simple fact that millions of successful merchants have publicly credited Shopify with their success. Retailers like Gymshark and Kylie Cosmetics are advertising for Shopify, whether they intend to or not.

Market Opportunity & Competitive Dynamics

The global retail market is estimated at about $31 trillion, and e-commerce adoption is still limited at ~20% of sales, and 12.8% outside China. In the US, e-commerce penetration is 17.1% (Q2’26), up from a ~14.5% average in 2020, and the trend is expected to continue for the foreseeable future.

Shopify has been gaining share within e-commerce as independent and enterprise brands drive more direct volume. Its share of US e-commerce has risen from roughly 8.6% in 2020 to over 14% in 2025.

We think Shopify has a long growth runway, with tailwinds from rising e-commerce penetration, more independent brands and continued share capture. Those tailwinds helped it grow GMV 29.5% in 2025, versus an estimated ~9% for Amazon.

Competitive Dynamics:

Amazon is the only company with the merchandise breadth and merchant relationships to compete with Shopify at scale; together the two account for about half of US e-commerce. But Amazon gives brands on its platform little identity of their own, and that is costing it ground as retailers open their own digital storefronts on Shopify while continuing to sell on Amazon. Shopify’s GMV has grown from 40% of Amazon’s marketplace GMV in 2020 to 66% in 2025.

Wix, Squarespace, WooCommerce and others also help retailers build a digital storefront with some commerce features like basic payments. But none matches the breadth of Shopify’s products. In other words, while competition for storefront design is heavy, few have built the scale to provide everything from payments and ad services to discounted shipping. Wix, for instance, processed $14.3 billion in payments in 2025, versus roughly $248 billion through Shopify Payments.

We believe Shopify can be challenged by these players only at the low end of the long tail and, even if a retailer starts on a low-end rival, it will shift to Shopify as it scales, sells globally and needs credit. Hence we believe Shopify will continue to lead among high-value retail merchants.

At the high end, Shopify also competes with enterprise suites like Salesforce Commerce Cloud and Adobe Commerce. But we see their parents as too diversified to focus on the single problem of the shopping experience, and their products need more plumbing than an end-to-end service like Shopify.

Profitability:

Shopify has a largely predictable Cost of Goods Sold, with ~20% of revenue for its subscription business, as most of the costs here are hosting and support costs. Its merchant solutions COGS is around ~60% of revenue. A large part of this COGS goes toward fees paid to card networks (Visa, Mastercard) and processors like Stripe for supporting payments, etc.

While Gross Margins are largely stable at ~48–50%, they are seeing significant scale benefits in their S&M and R&D expenses. Their S&M expense was $1.66bn in 2025, up 19.4% vs. 2024, while revenue grew by 30.1%. Similarly, their R&D expense was $1.54bn in 2025, up 12.4% vs. 2024.

Over 2021–25, OpEx scale benefits from these two segments enabled operating margins to improve from ~5% in 2021 to 12.7% in 2025. We expect them to continue gaining OpEx scale benefits going forward.

Shopify’s long-term operating margin can be as high as ~25%, compared to mid-teen margins today. This is reflected in its incremental operating margins (profit per $ increase in revenue) of ~30%+, which it achieved in the past.

Shopify also provisions for credit losses and any payment fraud/failures as an operating expense (transaction and loan losses). As its analytical engines improve, its products will better detect losses and fraud, and it will see a lower transaction loss rate. This expense increased from ~1.8% to ~3.5% over the years as payment volumes scaled. We expect it to stabilize and scale down over time as payment adoption reaches ~90% and its analytics products continue to improve.

Valuation:

Shopify valuation depends on 4 major factors:

1. New retailers onboarding, i.e., subscription growth. Subscription growth has historically been ~10–20%, driven by new business openings and additional product purchases/upgrades. We expect this growth to stabilize over time at ~10%, driven by sustained new business growth in the US.

   

2. How much in fees they can collect from customers’ GMV for additional services. Currently, their merchant solutions business collects ~2.40% of GMV, up from 1.86% in 2021. We expect their merchant fee collection to continue to grow as more customers adopt their solutions and payments adoption increases from the current ~68% toward ~90%.

3. GMV growth among retailers using the Shopify platform. Historically, most retailers have been able to grow their GMV by ~10%+. This provides a baseline for GMV growth on the platform. Agentic AI can drive this baseline growth higher and create additional value. In our base case, we assume existing retailers will be able to grow ~20% on aggregate due to better discovery, taking overall GMV growth to ~25%. In the bear case, we maintain ~10% GMV growth for existing retailers, taking total GMV growth to ~15%.

   

4. Operating leverage: Shopify EBIT margin increased from 5.8% in 2021 to 12.7% in 2025. We expect their high growth will enable more operating leverage and expect an EBIT margin of ~20% by 2030 in the base case and ~19% in the bear case.

Based on our valuation model (treating SBC as an expense), we arrive at a base-case valuation of ~$130 using a 30x P/E in 2030 and a bear-case valuation of ~$100 using a 25x P/E.

Today is trading at ~$160 with a ~110x P/E. The market is already pricing in sustained GMV growth of 30%+ and operating margins improving to ~20%+.

Given the market already prices in significant GMV growth partly driven by agentic discovery, we would like to wait for a tactical entry around our base-case valuation of ~$130 if the share price drops due to interest rates or any other short-term factors.

Additionally, we will reassess our entry valuation based on how Shopify’s GMV growth is trending and what portion is attributable to agentic commerce versus macro changes.

Shopify Model

Risks:

Trade barriers: Rising geopolitical tension and higher trade barriers could slow cross-border shopping on Shopify. That would hit both its global volume and its payments network, one of its key moats.

Consumer spending & Interest rates: Shopify is ultimately a discretionary consumption-driven business, and higher interest rates weigh on consumer spending. Its stock also trades at a high growth multiple, so higher rates can hurt the stock through lower growth expectations and the negative impact of higher rates on valuation.

In 2022, as rate hikes took effect and pandemic-era e-commerce unwound, GMV growth slowed to 12% from 47% the year before, and the stock fell ~80% from its November 2021 peak before recovering over time.

Payments disruption: Agentic commerce and payments are evolving fast, and blockchain technology and cryptocurrencies are being floated as the next major shift. Payment processing fees make up the bulk of Shopify’s merchant solutions revenue. A move to public-blockchain and crypto payment rails could affect its payment product pricing.

New-merchant pull: AI is making it easier for retailers to build their own tools. Shopify’s adjacent services (payments, Credit, logistics etc.) keep existing merchants sticky, but its ability to attract new retailers could diminish over time.

Leave a comment

If you enjoyed this piece, PLEASE HIT THE LIKE BUTTON. It’s the only way new readers can discover our reports :)

Disclaimer: Market Sentiment work is provided for informational purposes only, is intended solely for readers in the United States, and should not be construed as legal, business, investment, or tax advice. You should always do your own research.

Despite the headlines, the Nasdaq gained 3%, driven by semis, with the SOXX index up 18% as it recovered from the July drawdown. Meta’s Muse launch helped the semis rally by renewing expectations for AI-driven semiconductor demand.

But beneath all the noise, one thing started to change decisively: interest rates. After a long time, bond yields started to take center stage for all investors, as yields kept rising to new highs, with the current 10-year yield at ~5.25%.

Higher bond yields generally mean one of two things: increasing risk in the overall economy or high demand for the limited capital available. Either way, this will limit companies’ ability to invest in new projects, and the benchmark returns investors demand will keep climbing as yields rise.

Historically, rising rates marked the end of major investment cycles in railways, the internet boom, and housing. With rates now moving higher again, the question is whether today’s AI investments continue to grow despite rising costs of capital.

What’s pushing yields up

Long-term yields are driven by two main factors: inflation expectations and the term premium, i.e., the additional return investors demand for locking in their money over long horizons.

Inflation expectations are rising, with oil hovering around $100, tariffs feeding into prices, and loose fiscal policy, with the deficit at ~7% of GDP. Inflation has been so sticky that the Fed is now expected to raise rates 2 times this year to fight inflation.

The term premium is set by supply and demand for bonds, and both sides are pressuring yields.

On the demand side, some of the largest U.S. bondholders, like Japan and Norway, are looking to offload their Treasuries to fund other agendas. Japan wants to defend the yen against the USD and needs more money to intervene in the market. Norway wants to fund AI capex.

On the supply side, hyperscalers are increasingly competing with U.S. Treasury bonds for the same pool of capital. It can because the S&P rates Microsoft bonds at AAA, which is one notch above the federal govt. itself. AI capex is expected to pull in ~$300B of debt in 2026 and a further ~$400B+ in 2027, with total AI capex reaching ~$1.2T in 2027. This would make hyperscaler debt issuance equivalent to ~40%+ of fresh Treasury bond issuance.

AI debt remains small in aggregate, but on a new-issuance basis, the additional supply is significant. As hyperscalers increasingly rely on debt to fund AI infrastructure, they are competing with Treasuries for the same pool of capital, putting pressure on yields.

How we are positioning:

There is no way out of high yields without less borrowing for AI capex or by the U.S. government, along with lower inflation expectations. But a slowdown in AI capex or government spending would hurt the market.

With AI capex estimates now at $1.2T, our analysis of sustainable capex suggests this should generate ~$2T in revenue for infrastructure companies. As interest rates climb and infrastructure firms slowly ramp revenue, we think we are entering a phase where further AI capex growth beyond 2027 will be limited, if it continues to grow at all.

That said, we are not forecasting an end to the AI capex boom. We see AI as a transformative technology and expect continued investment, but the pace will moderate. The pace of AI capex going forward will be determined by how quickly new applications emerge that use AI. For instance, Muse is one such example.

Therefore, we see this as a good time to slowly shift our portfolio from companies largely driven by demand generated by AI debt (i.e., semis) to other AI enablers that drive the diffusion of the technology into the economy.

To summarize, despite a good September, markets have a tightrope to walk from here, with multiple macro headwinds and little room for further upgrades to the AI capex outlook. Given this evolving situation, we are doubling down on the broadening of AI trade we highlighted earlier and focusing on AI enablers.

Portfolio Update & Position Changes:

As part of our portfolio transition from AI Capex-driven companies to companies enabling AI adoption, we are making the following changes:

Read more

Salesforce and HubSpot are the systems of record for customer and sales data, Workday for employee data, and SAP for finance. Automations built in them work within their own departments, like sales emails in Salesforce or payroll in Workday.

Companies’ workflows, however, span across departments. Take, for instance, a new employee joining a large corporation.

First, HR enters the employee in Workday. Then they raise requests to IT for a laptop, to security for app access, and to facilities for a desk and badge, with her manager chasing all of them.

This is how work gets done in most large companies. Employees bounce across teams, tools, and people, chasing data and permissions to finish a task. Work that moves between these systems has no single owner.

ServiceNow was built for these kinds of workflows spanning across teams. Its founder, Fred Luddy, was inspired by the pink and yellow paper slips that carried approval requests from one department to the next through company mailrooms in the 1990s.

Fred Luddy turned that slip system into software. Once an employee files a request, the platform sends it to the right person to approve or act on it and notifies anyone impacted.

Over time, the platform became an important layer that moves work across internal systems run by a company. Hundreds of ready-made connectors can be plugged into ServiceNow to connect Workday, Salesforce, SAP, etc.

With ServiceNow, the new hire’s Workday entry opens every task at once. IT gets a ticket to ship the laptop, Okta creates her logins, and facilities assigns a desk. Her manager tracks it all in one place. While Workday remains the system of record, ServiceNow runs everything needed for onboarding.

Most enterprise tasks touch the IT department one way or another, and that’s where ServiceNow started. IT provisions logins in Okta, secures laptops in Intune, fixes servers flagged by Datadog, etc. NOW tracks all of those apps, laptops, and servers and how they connect.

From IT, ServiceNow expands department by department. When HR or legal needed software for their own workflows, IT recommended the platform it already had, i.e., NOW. Today, HR cases, customer service, and security incidents run on the same system as IT tickets, and that’s how NOW differentiated itself in the SaaS space.

Business Overview:

ServiceNow’s primary revenue source is its workflow software subscriptions, which account for ~97% of revenue ($12.9B of $13.3B in FY25). The small remainder is professional services, used mainly to implement its workflow products.

The subscription model gives NOW a business with low cyclicality. It starts each year with largely assured revenue unless a customer exits. Its gross revenue retention rate typically holds at ~98%, reflecting how sticky the workflows are once onboarded.

This stable income also gives it enough bandwidth to reinvest in the business and expand its offerings, improving its chances of sustaining long-term growth.

ServiceNow reinvests in two forms:

• S&M: Sales reps continuously engage with customers to understand their workflows and educate them on NOW’s capabilities, acting as a two-way channel between NOW’s product team and its customers. NOW spent 33% of revenue on GAAP S&M in FY25, down from 35% in FY24, and expects it to fall slightly again in FY26 as sales productivity improves.

• R&D: This covers product development and maintenance. The platform needs continuous upgrades to support its growing enterprise scale, and AI is driving further product innovation. NOW spent 22% of revenue on GAAP R&D in FY25.

Business Segments:

ServiceNow has expanded its workflows from core IT services into customer service, HR, legal, finance, etc. It categorizes its workflow business into three broad sections: Technology, Customer & Employee, and Creator workflows.

Technology workflows: This is their core IT service desk workflow segment. They have expanded this to include adjacent workflows such as Security Operations and hardware asset management. This business typically accounts for ~50% of their net-new ACV (Annual Contract Value) and remains a major growth driver.

Customer & Employee workflows: This business consists of workflows for sales teams, employee onboarding, legal, etc. Their customer workflows have crossed $2B in ACV, while their employee workflows have crossed $1B in ACV. This business typically accounts for ~30% of new business and goes head-to-head with other major software firms like Salesforce for customer solutions and Workday for HR solutions.

Creator workflows: This is their low-code and no-code software toolkit, including App Engine and Automation Engine, which allows regular, non-technical employees to build custom business applications and automate tasks without needing to write complex code. This accounts for ~20% of their new business.

Growth Strategy:

ServiceNow’s growth has two parts: landing new customers and expanding by selling more to existing ones.

Landing is close to tapped out. It already serves over 85% of the Fortune 500, and its customer base grew only ~5% in 2025, from 8.4k to 8.8k. There aren’t many large enterprises left to sign.

That makes expansion the primary driver of the ~20% revenue growth it is achieving, and the platform is built for sustained expansion across workflows. This is reflected in its revenue per customer almost doubling from ~$700k in 2020 to ~$1.5M in 2025.

It currently has 658 customers paying more than $5M per year, reflecting the scale to which NOW can grow within an enterprise. ServiceNow continues to grow across all its customer cohorts, including $5M+ customers, as it enables AI capabilities and captures more workflows.

Product & Moat:

ServiceNow’s edge is simple. It can run a single piece of work across many different software tools and track it from start to finish. What makes this possible is its Configuration Management Database, or CMDB.

Think of the CMDB as a live map of a company’s technology. Every server, app, laptop, and service is on it. So is every connection between them and the teams responsible for each one.

Here is how CMDB helps: say a bank’s payments app slows down. ServiceNow logs it as an incident. The CMDB immediately shows which server and services enable the payment app and notifies the owner. It also identifies which customer service depends on the payment app and notifies the customer and others impacted. The tickets go straight to every person responsible for the slowdown and impacted by it.

Without that CMDB map, a problem like this means a dozen people on a call trying to work out what broke. With it, the work finds the right people automatically.

In October 2024, ServiceNow added a second layer called workflow data fabric. While CMDB maps a company’s technology, workflow data fabric reaches into its data. It lets ServiceNow read and use information stored in other systems, like a Snowflake or Databricks data warehouse, without copying it over. This covers neat tables of numbers as well as messy data like emails and documents.

Together, the CMDB and Workflow Data Fabric give ServiceNow’s AI agents what most AI tools lack: context.

Let’s go back to the payments app. The CMDB tells the agent which server failed and who owns it. Workflow Data Fabric lets it pull the failed transactions from the bank’s data warehouse to see how many customers were affected and can notify them as the service comes back online. One tells the agent how things connect. The other gives it the data to act on.

AI agents are only as useful as what they can see, and every new workflow a customer builds on ServiceNow adds to the picture of how enterprise workflows flow. This is a big reason NOW’s AI products crossed $1B in annual contract value (what customers are signed up to pay each year) in under 3 years. Most AI startups would envy that.

None of this works out of the box. The CMDB and workflow data fabric only become valuable once they are wired into the hundreds of apps and data sources a company already uses. That takes time. NOW’s 10-K says each rollout depends on the customer’s integration, data migration, compliance, and security needs. This can take several months to quarters.

That effort is what makes NOW hard to leave. Switching means rebuilding every one of those connections in a new product. And because NOW’s workflows run across departments like IT, HR, and customer service, a company cannot move one at a time to another product without breaking the links between them.

Over time, NOW stops being just another tool and becomes the place where work gets done. This goes well beyond corporations. Nearly all US states use the platform. The City of Raleigh, North Carolina, used NOW’s AI agents to cut its IT help desk costs by 66% (Q2 FY26). If I’m a CIO, I’m not ripping that out to save a bit on licensing.

ServiceNow AI Control Tower:

Earlier, we saw how CMDB and Workflow Data Fabric keep NOW at the center of enterprise AI agent adoption. Its pole position in enterprise workflows makes it a natural place to run those agents, and it has extended this capability into a new product called AI Control Tower (launched in May 2025).

AI Control Tower is a command center for every AI agent and workflow in an enterprise, whether it’s built on ServiceNow or elsewhere. It discovers them and applies standard policies, ensures compliance, and measures what each one delivers.

This is CMDB adjacency, just like a CMDB registers servers/apps and their dependencies. AI Control Tower registers agents, what they can access, and who owns them. Because NOW already holds the map of the workflows these agents need to act on in CMDB, it can identify when agents sidetrack.

With AI Control Tower, NOW positions itself above the AI vendors, entrenching itself as an enterprise AI platform rather than being absorbed by one.

AI impact on NOW’s Business & Pricing:

NOW has traditionally priced its product on a per-seat basis, licensed by fulfiller, i.e., employees who work on the platform auto-remediating tickets and building workflows. Some products, like IT asset management, are priced by devices managed, connectors, data transferred, etc.

AI makes their seats more productive, so fewer of them clear the same work. Street is concerned about the impact of seat compression on NOW’s revenue and their ability to mark up pricing for any loss of seats. The company now trades at $137, ~40% below its last year high.

So it all comes down to whether AI will be a headwind or a tailwind for the company. To find out, let’s dig into:

• How ServiceNow is pricing its AI products

• Which companies can realistically challenge them

• Current profitability and stock-based compensation concerns

• NOW DCF Model and Market Sentiment rating

Read more

The everyday bit is the important piece. Till now, AI usage was limited to replacing a Google search with an AI summary, quick troubleshooting, or something work-specific. With Muse, Meta became the first company to bring agentic AI into the mainstream. Within two weeks of launch, consumers have used Muse to

• Book a haircut with your traditional constraints

• Wait on hold till customer support picks up

• Cancel insurance

• Order groceries

• Schedule an IKEA return

And it pulls this off while looking like a cute little gizmo. This means Muse has now climbed to the No. 1 spot among free apps on Apple’s US App Store, ahead of ChatGPT.

Overnight, Meta went from a company that knew your social media preferences to one with access to your email, calendar, Amazon, and pretty much everything else. Meta is the obvious story here, and we will get to it. The bigger one is what happens to everything else once agents go mainstream.

Tough time to be an intermediary

When you think about it, a lot of multi-billion-dollar companies exist just to reduce friction.

• 90% of business TurboTax does is simple tax filing.

• Booking Holdings pays Google to redirect people to use its travel comparison.

• Fiverr exists so that you can outsource boring, data-intensive tasks to the third world.

Now what happens in an agentic world? What if Muse is capable enough to get all the details to file your taxes? Two weeks after launch, it’s already capable enough to book flight tickets and even get a credit from Delta for a delayed flight. Imagine how much more capable they'll be 2-3 years from now — ChatGPT didn't have live internet browsing three years ago.

And it isn't only thin-wrapper companies at risk. Even Amazon is concerned, and they blocked access to Meta’s Muse last weekend. The logic is simple. No one wants to get commoditized. If I ask my agent to get something via Amazon, Amazon loses the ad revenue it collects while I browse for the product. And it’s no small change: Amazon made ~$70B last year from its ad business, and it’s near-pure profit that hits the bottom line.

Changing Business Models

Beyond the intermediaries, even legacy companies will struggle in an agentic world.

• Moving your bank deposit is annoying, but what if the agent is reliable enough to sweep the idle cash to the highest-yielding account? Also, even a half-decent agent will significantly reduce the $12B a year in overdraft fees paid by inattentive customers.

• An assistant that shops around for the best insurance rate instead of auto-renewing at a slightly higher price will be a massive hit on the insurers. Also, what happens if your agent fights every single denial on your insurance that you were too lazy to fight?

• Even Google search might come under pressure — Google monetizes the results page and gets paid when you click a link. You can argue that Google can create its own agent layer, but monetizing it will be a completely different challenge. Can an agent redirect you to a sponsored listing smoothly without triggering FTC rules?

• It’s estimated that 1/3rd of the loyalty points are never cashed in. Personally, while my AmEx points continue to accrue, I have never bothered to redeem them when booking something because of the hassle. How much more usage would the industry see if agents started optimizing the booking and payment processes?

The trick here is that there are no free lunches. Meta is giving unlimited access to Muse right now because they are in the data collection and customer acquisition phase. Every dollar you save using an agent, or every hassle it removes, is a small lever to keep you on the same platform or turn you into a paying customer.

Which brings us to:

Why is it good to be Meta right now?

As we highlighted last week, platforms with a large consumer base stand to benefit the most as agents go mainstream. And there’s no larger consumer platform than Meta. The stock is up 30% in the last month and has added nearly a quarter of a trillion dollars in market cap after the launch of Muse.

Even after the current run-up, we believe Meta is undervalued based on these 4 factors:

Read more

Grab a cup of coffee for this one. It’s one of our most in-depth reports (and it’s free!).

Successful investing comes down to finding a trend early and knowing when the thesis has become consensus. It’s the second part that trips up so many investors. While riding the trend up, it’s easy to think that it’s going to run forever. All the data and analyst projections point that way. But that’s rarely the case.

Right now, that is the AI infrastructure trade. Capital has crowded into a handful of winners, with valuations increasingly pricing in years of uninterrupted growth. While this has certainly helped our portfolio (+55.3% YTD), many of the obvious AI bottleneck trades have already become consensus.

We are bullish on the ROI AI capex will generate, but skeptical about how much projections will rise next quarter, especially since the ones running the show are clamoring to slow development. Any blip, however minor, will erase years of gains from these stocks.

Plus, the Fed hiking rates will not be great for semiconductor companies — the real worry is not the 25bps hike: it’s a combination of a new hawkish Fed chair, sticky inflation (due to tariffs and the Iran war), and rising national debt.

Who makes the money with transformative technologies?

It’s not the first time investors have bet on a transformative technology to change the world. We have two good historical precedents: railroads and the internet. In both cases, we invested a single-digit % of our GDP into building railroads and telecommunications infrastructure.

While both these technologies ended up changing the world, neither rewarded the long-term investors who funded the buildout. The key being long-term: short-term traders would have made a killing trading the volatility, but the index was more or less in the same place after the hype was over.

The problem with aggressive capital spending is that it’s almost always suboptimal. After the initial euphoria of rising capex driving up stock prices, you will realize you overpaid for everything from the land to the memory to even the fiber-optic cables.

The industry that creates the technology rarely captures the value it creates.

The railroad expansion saw hundreds of railroad companies go bankrupt, and the dot-com bust saw the telecom index collapse by 92%. Yet the infrastructure those failures left behind became the foundation for the winners that came after.

Standard Oil and Sears built their empires on the railroads, and Amazon and Netflix were built on the cheap bandwidth created by the fiber glut. The lasting winners in AI will also be the companies building on top of the infrastructure, not the ones laying it down.

While Cisco's example is beaten to death, it’s still worth comparing it to Amazon. Cisco built the internet's plumbing: the routers and switches every website ran on, and was the world’s most valuable company in 2000. Then the bubble burst, and it took the company ~23 years to regain its ATH.

Yet, even after the recent AI run-up, the company is 1/6th the size of Amazon, which used the same network to become one of the most valuable companies ever!

All this leads to one simple fact:

Here’s something we can guarantee most of you missed during all the AI hype:

In 2026, the Russell 2000 outperformed both the S&P 500 and the Nasdaq 100. The same trend holds over the trailing year.

So why is an index of 2,000 small-cap companies outperforming one built on the very names everyone expects to change the world? We think the answer sets up one of the largest capital rotations in years.

Tracking the Token

Next year, AI Capex is expected to be ~$1.2T+. Based on our capex sustainability model, this should drive ~$2T of incremental revenue. This will be a fraction of the overall incremental revenue generated using AI (as a business, you don’t spend $100 on something unless it makes you back 2-3x that amount).

The question then is: who will be the beneficiaries who capture this value?

The simplest way to find out is to track the value chain. Take the example of Oil. The driller sells it for ~$70, the refiner turns it into jet fuel worth ~$90, the distributor sells it to the airline for $100, and the airline turns that fuel into a $400 seat.

As the commodity moves further downstream, each step adds value.

Commodity conversion doesn’t happen automatically; it requires unique assets. In this case, those assets are refineries, pipelines, and airplanes. These are the companies that benefit from adopting innovations and new tools that enable new commodities.

Apply the same logic to AI tokens, and one industry is massively benefiting: Software.

But not just any software — the ones with moats that AI cannot erode: proprietary data, distribution, workflow lock-in, or system-of-record status. Everything thinner will get squeezed by falling token prices, just like what happened with Fiverr. AI now handles the translation, basic design, and entry-level code that its marketplace was built to broker, and 2026 revenue is set to fall 14 to 17 percent.

So who are the beneficiaries?

1. Cybersecurity

While there is a lot of hype related to “agents escaping sandboxes”, our take is simpler: Frontier models have handed even a mediocre hacker the coding ability of the top 1%. We expect the number of attacks, especially on weak systems, to rise sharply.

The diagram above shows how companies think about this trade-off. Security spending is weighed against the cost of a breach, and total cost bottoms out at an optimal level of defense. More attacks raise the expected cost of failure and push that optimal point toward heavier security.

The problem is that companies cannot respond by banning AI, because the productivity gain is too large to give up. So the balance shifts to a narrower question: give employees access to AI while placing strong guardrails on what data it can and cannot reach.

Which brings us to:

2. Horizontal & Vertical Software Applications

Horizontal platforms such as CRM, IT service management, and ERP run across the enterprise. They own the systems of record, the permission structures, and the workflows that agents must read from and act on. In an agentic environment, that ownership is the advantage. An agent cannot close a deal, resolve a ticket, or post a journal entry without touching these systems, which hands incumbents a durable role.

ServiceNow shows the transition working. Its AI book crossed $1 billion in annual contract value in Q2 2026, with agentic AI in production up 9x in nine months.

Vertical applications, built for a single industry, can hold an even stronger moat. Their workflows are specific, and building useful AI for them requires deep process knowledge that is hard to replicate. Guidewire in insurance, Procore in construction, and Autodesk in design are the type of businesses this favors.

3. Development Infrastructure

As the cost of building software falls to zero, a lot more applications will get built.

As more applications are built and deployed, the infrastructure and tooling needed to develop, connect, secure, and operate them should grow with them.

MongoDB, for example, has expanded beyond its traditional database offering with AI capabilities such as vector search, allowing developers to build AI applications on top of existing data infrastructure. Similarly, companies such as Palantir and Snowflake provide platforms that enable enterprises to develop and deploy AI applications using their proprietary data and workflows.

4. Recommendation Engines

An often overlooked beneficiary is ad targeting and recommendations. Platforms with large consumer bases gain the most, since AI improves with more user data. Reddit and Meta use it to match users with more relevant content and products, boosting both engagement and ad performance.

Unity shows a related version of the same idea: AI lowers the cost of building a game and then improves how those games are discovered and monetized. PocketFM takes it further, using AI-generated content to expand its catalog and scale its audience across global markets at low marginal cost.

5. Low-cost Compute

Not every task needs a frontier model. Writing an email response or updating a deck does not require Claude Fable 5. An open-source model can do the same work at a tenth of the cost or less, and as open models get smarter, the share of workloads running on them will keep expanding.

Usage will keep exploding while the price per token collapses, so the pie grows even as every workload gets cost-optimized. That means serving intelligence becomes a race to the lowest cost per token, the same way Japanese automakers won the US market by delivering most of the performance at a much better price.

If that is the race, hyperscalers have no option but ASICs.

Cheap intelligence into durable margin

Ultimately, it comes down to this — if intelligence becomes abundant and cheap, who stands to benefit? Our answer starts with proprietary data a model cannot reconstruct. But the bar is higher than owning data, and the losers look like winners until they don't. Chegg owned one of the largest databases of solved academic problems in existence, reportedly more than 79 million of them. It did not matter. A free general model answered the same questions well enough; Chegg had no workflow to hold the student.

The result was inevitable:

The second filter is trust but verify, because AI-washing is everywhere. Every management team now claims to be a winner, and the metrics are easy to dress up. Salesforce booked over $1.5 billion in Agentforce ARR this quarter, but only after folding Slackbot and other products into the definition. We will judge adoption by usage, pricing, and margin, not by the earnings call.

This is why we are tilting toward companies that convert cheap intelligence into durable margin, mainly in software and service-heavy sectors, and we will rotate out of some capex-driven semiconductor holdings over time to fund it.

Subscribers will get the full portfolio changes next week (before we execute them). Please consider upgrading your subscription to support our work and to get access to all the reports.

Subscribe now

If you made it till here, I would love to hear what you think:

Leave a comment

Disclaimer: Market Sentiment work is provided for informational purposes only, is intended solely for readers in the United States, and should not be construed as legal, business, investment, or tax advice. You should always do your own research.

No paywall on this one. Share with a friend :)

Markets have sold off twice on the same story. Both times, the trigger was a low-cost open-source model matching frontier models on benchmarks.

The first was DeepSeek R1 in January 2025. NVIDIA fell roughly 17% in a single day. The second was Kimi K3 in July. The SOX index dropped about 2% on the news, then slid 10% over the week as leverage and profit-taking compounded the move.

The reasoning behind both selloffs was the same. Cheaper open-source models mean AI labs can’t earn enough to cover compute costs.

We think that reasoning is too simple. The real impact of open source runs through a set of harder questions.

• Why are open-weight APIs so cheap?

• How are open-model companies staying alive?

• What does this mean for the AI labs?

• And how did open source play out the last time it reshaped an industry?

Open model pricing is a distribution, not a number.

Open model weights are public, and anyone can download them, but running them still means managing large GPU clusters and the software stack. The software stack creates differences between multiple providers serving the same open weights. Inference providers can optimize their software to prioritize price, latency, or accuracy while serving the open weights. Depending on how they optimize, each inference API will be positioned at a particular price-to-spec market.

DeepSeek’s first-party API provides DeepSeek V4 Pro at a blended cost of $0.20 per 1M tokens, while the same model weights are sold at $1.90 by Azure, a ~10x price differential across hosts serving identical model weights. Azure's pricing premium is justified by lower latency and higher accuracy.

On the other hand, Claude Opus-5 costs $3.90 per 1M tokens on every cloud (Azure, AWS, and GCP), even if performance varies slightly, because a single model owner sets the price.

Headline “open is 10x cheaper” comparisons often quote a cheaper version (DeepSeek Cloud) with degraded performance, exaggerating the cost benefits of open weights. For instance, Opus 5 can look ~20x more expensive than DeepSeek, but on the same Azure platform it is only ~2x more expensive.

Deepseek V4 pro max pricing

Claude Opus 5 max pricing

Token price is determined by GPU-hour cost and how many tokens you can squeeze per GPU-hour. GPU costs are similar across firms; they differ in how many tokens they extract from each GPU-hour.

• Precision: Model weights are simply a set of numbers in matrices. Every number is stored using a fixed count of binary digits. Fewer binary digits mean less accuracy when storing a number, which can reduce inference accuracy while saving compute.

• Batching. This packs more concurrent requests onto the same GPU. Each request waits longer as it goes through the batching process, but the GPU does far more total work per hour. This is why low-price hosts are often also slow hosts with higher latency.

  
  The table below shows how latency changes with throughput for serving Deepseek V4 Pro on NVIDIA B300s. Throughput per chip rises ~10x while speed per user falls to ~1/13, as seen in the table below.

• Software optimization. How well the software stack keeps the GPUs busy with custom kernels, speculative decoding, and request scheduling. Unlike the first two levers, a better stack buys speed without giving anything up.

In a closed model, the firm controls the weights and most parts of the serving stack together. Open-sourcing the weights will lead to the same model turning into ten products across a range of price vs performance specs.

How Open Models make money:

Most Open source software is actually free until you are not directly monetizing the software and using it for personal or internal applications. In case of open weight models, weights alone is not a product end user can use. They need AI inference infrastructure management expertise to run the model optimally.

Open labs give the model for free and will let an infrastructure ecosystem be built around the model with multiple companies competing to create a more optimized model serving stack while positioning across price vs performance spectrum.

Once an infrastructure layer is built around a model it becomes expensive to rebuild for a different one (harnesses, caching behavior, conversation-state assumptions, integrations).

As an open model’s volume scales, even a small fee produces meaningful revenue. Built-in switching costs from ecosystem fine-tuning to the model will ensure stickiness once a model scales, limiting market share loss to any new open-weight models.

Successful open-weight models with significant adoption are already playing the second stage, where they charge a fee for anyone selling their model. Moonshot’s Kimi K3 requires any model-as-a-service operator with revenue over $20M to sign a commercial agreement with a fee of up to 30% of revenue. Alibaba shipped its Qwen 3.8 model under a custom license that requires an agreement above $50M in revenue.

Why are Open models priced low:

Open models lag frontier models and have no pricing power. Lower price with a lower cost base is the only way to gain market share. To lower the cost base, Chinese open-source models tend to ride on the leading labs' research and use techniques like distillation, Mixture of Experts (MoE), etc.

• Distillation: Frontier labs spend billions discovering what works through thousands of failed runs, human-annotated data, etc. Open labs compress that discovery process into training data by distilling frontier-model outputs. Anthropic has recently raised concerns about mass bot-driven distillation from a Chinese lab.

• Architecture: Sparse MoE designs activate only a fraction of the total model parameters for each token. DeepSeek pioneered this approach, but labs now use it widely. Labs also use techniques such as FP8 training to reduce training costs with only marginal efficiency compromises.

Cheap training for a model means less needs to be recouped through inference to stay afloat. This makes aggressive pricing more sustainable. Further, Revenue = Price × Volume; a marginally lower price does not shift volume from frontier models to open-source models. To scale revenue, the only way is to price significantly lower to attract higher usage volume and drive higher overall revenue.

Open Source adoption and impact on AI ecosystem:

Open weights have overtaken proprietary models on the leading API platform OpenRouter. They now account for over 50% of all token volume on the platform, rising from ~33% in late 2025 and under 2% for Chinese open models in late 2024.

As per GS, open model API revenue accounts for <20% of total API revenue generated via OpenRouter. This large gap indicates API calls for open weights are largely driven by the lower end of the price-performance band, while proprietary models continue to lead in the higher price-spec category.

Latest US AI lab models GPT-6 Astra and Claude Fable 5.1 are priced at $10/$50 and $10/$50 per million tokens, in a similar band as GPT-4 ($30/$60) and Opus 3 ($15/$75) at the time of their release. For us, the key takeaway is that there is a separate market for leading specs, and AI labs can sustain their pricing in the market.

How AI labs and Infrastructure firms can respond:

What matters for AI labs is retaining and leading in $ market share. For now, their superior models are doing the job for them. However, we don’t think they will hesitate to change their business models if they see a threat to their $ market share.

If AI labs lose their model lead and begin to lose the premium market, we expect them to increasingly match open-source pricing and potentially move toward open-weight models themselves. OpenAI, for instance, has already open-sourced some of its models in the past; it released GPT-OSS in August 2025, its first open-weight model since GPT-2.

Lower AI token costs don’t necessarily mean AI labs will go bankrupt. Their underlying infrastructure and GPU costs are broadly similar to those faced by open-weight models. The bigger impact would be on margins and consequently, equity valuations. In other words, falling token costs are more likely to erode AI labs' valuations than to erode the value of the infrastructure investments that power AI.

Hyperscalers are model-agnostic infrastructure distributors. They already host multiple open-source models, including DeepSeek, Qwen, GLM, MiniMax, etc. Furthermore, AWS Bedrock and Google’s Model Garden allow enterprises to build their own models beyond simply using open-source models.

Irrespective of the model, hyperscaler platforms will continue to capture inference-servicing margins, with their edge being their ability to host thousands of GPUs efficiently by dynamically matching demand and supply. This frees enterprises from large upfront investments and continuous infrastructure maintenance requirements - essentially the same value hyperscalers have historically provided through the cloud.

Hyperscalers also benefit from export controls and regulations that require customers to run models within the US, limiting competition from China. That is why we continue to hold infrastructure firms and believe open source will not negatively impact them.

What we’re watching

History has run the open-source experiment before with two major products: Linux and Android.

In the early days of computing, multiple operating systems competed, including Unix, Microsoft Windows, and IBM’s OS/2. Linux differentiated itself not by being a simple, feature-rich operating system, but by being more customizable and adaptable to different applications. Over time, it built a large community and became the default open-source operating system. Microsoft retained its more lucrative consumer market through ease of use and a broader ecosystem, while sophisticated customers with specific needs moved to Linux. Most importantly, Linux succeeded while remaining free to use because the cost of developing and maintaining it was distributed across a large community.

Android took a different path. Google built an open-source operating system while retaining control over much of the broader ecosystem and monetizing services around it. This is broadly similar to the model companies such as Moonshot (Kimi) and Alibaba are pursuing with open-weight models today.

Open weights are currently a fast-moving market, with multiple Android-like systems emerging, but there is no equivalent Linux-like, community-driven system at the frontier. Truly community-driven models, such as AI2’s OLMo and EleutherAI’s Pythia, remain well behind the most capable models. The key barrier is training cost: Linux required distributed engineering effort, while training a frontier model requires GPUs worth millions of dollars. Community effort alone cannot easily fund these training runs, so most competitive open-weight models have a company behind them. Those companies ultimately have incentives to monetize the models or the ecosystems around them.

Either way, open weights limit AI labs’ pricing power by creating a wide range of price-performance options and making infrastructure increasingly important. Similar to operating systems, we expect the open-weight ecosystem to converge toward a few winners over time.

However, unlike operating systems, models are relatively cheap to switch between, so durable monopolies are harder to sustain, and leadership could continue shifting among a small number of winners. We will continue tracking token consumption and the number of infrastructure providers to understand where open weights converge and identify potential opportunities created by that convergence.

Leave a comment

If you enjoyed this piece, PLEASE HIT THE LIKE BUTTON. It’s the only way new readers can discover our reports :)

Disclaimer: Market Sentiment work is provided for informational purposes only, is intended solely for readers in the United States, and should not be construed as legal, business, investment, or tax advice. You should always do your own research.

The semis leverage unwind ended with Korean regulators tightening leveraged-ETF rules and forced sellers like Situational Awareness (4x leverage) getting cleared out. Korean Finance Minister Koo Yun-cheol apologized in parliament for launching the products without thinking through how they would trade, as the regulatory body plans to increase restrictions on leveraged ETFs and stocks.

August also saw a shift in Fed rate expectations. With core PCE stuck at 3.3% and increasingly hawkish comments from Fed officials and Chair Warsh, we see higher rates as the likely next move. Fed rates disproportionately affect the valuation of growth companies, i.e., most AI names.

As markets move from pricing rate cuts to rate hikes, the growth rally becomes more selective, and the bar for stock momentum rises from reasonable beats to blowout earnings. We believe this is the time to stick to high-conviction names where we expect sustained earnings beats and have valuation comfort.

Subscribe now

Improved AI Capex Outlook & Broadening of AI trade:

The semis index rallied ~10% in early August and gave it all back as markets gauged the longevity of the capex cycle. Nvidia’s earnings offered a partial answer in the final week.

Nvidia delivered another beat-and-raise quarter; more importantly, it guided to ~70% revenue growth in FY28, higher than what analysts had modeled. Nvidia’’s outlook increases our confidence in AI capex-driven names into 2026–27.

The AI trade is expanding beyond semis to AI beneficiaries. Salesforce’s Agentforce ARR crossed $1.5B, up 240% YoY, and it deepened its commercial partnership with Anthropic. The stock rallied 22%, one of its best days ever, on strong AI revenue.

In August, IGV rallied ~16% (vs SOXX up ~0.7%) as markets started to re-rate SaaS companies (Adobe, NOW, WDAY, etc.) as AI beneficiaries rather than businesses being disrupted by AI. While we don’t want to call CRM a clear AI winner yet, Salesforce’s earnings gave a clear view of how a well-executed AI transformation can turn a SaaS company into an AI firm.

For us, software companies benefiting from AI and the corresponding SaaS rally are a good AI-ecosystem indicator. They address the AI revenue question and justify the AI capex buildout.

However, in the near term, finding new AI beneficiaries can drive capital rotation out of semis and into companies (SaaS and others) benefiting from AI adoption.

We see this as an opportune moment to add companies that benefit from AI, extending beyond the enablers providing hardware infrastructure. As part of this, we have added Uber, and we will cover more direct AI beneficiaries going forward.

Exiting August, we are more comfortable backing AI infra builders, with better visibility on AI capex through 2027. We are also more cognizant that the market is starting to reward AI beneficiaries beyond the infra builders. We will shift the portfolio accordingly over the next few months.

Now moving to MS portfolio performance

Portfolio Update:

Our portfolio returned ~6% in the month, in line with NASDAQ and ahead of SOXX. Performance was helped by a stronger recovery in memory names, our largest allocation, along with some other high-performance-compute bottlenecks.

Our best performers in August were

Read more

They spent more than $30B over 10 years before turning an operating profit in 2023.

But once it did, the company became a free cash flow machine:

Just as the company was finally hitting its stride, it now faces an existential threat with Autonomous Vehicles. The assumed risk is simple — a company like Waymo can create its own app, sidestepping Uber, and provide the service at a lower cost than Uber can with its human drivers.

But in our view, this is simplistic thinking that doesn't account for why Uber lost all that money during the building phase — ride-hailing is not a global network effect. If you are a marketplace like Amazon, you can reliably service a few states or even the entire country with a massive warehouse once you have built it up. But for Uber, the network you have built in New York isn't useful to someone trying to get a cab in Boston. In the U.S. alone, Uber has spent nearly a decade building 8,000+ micro-markets to match supply with demand.

From an Autonomous Vehicle (AV) developer's perspective, it’s much better (operationally and financially) to plug into the existing Uber ecosystem than to build a system independently (more on this later).

How Uber makes money

To get a sense of how big Uber has become, they have 200mm+ active users and enabled ~13B trips (rides + delivery) in 2025 alone. It also runs a B2B delivery arm, but it accounts for a small fraction of the overall business and adds minimal value.

Let’s go from the least to most important segment.

Freight:

Uber Freight is a digital freight brokerage that matches shippers with truck carriers. Rides and freight book differently. In rides, one clearing price is visible to both sides, and Uber takes a disclosed cut. In freight, shippers and carriers negotiate prices separately. Uber contracts as principal, committing a price to each side, so it books the entire gross booking as revenue.

Delivery:

Launched in 2014, it quickly scaled to become the second-largest local delivery business globally ex-China, with $85B in total GBV in 2025, closely trailing the largest player, DoorDash, at $100B. Last month, Uber announced a cash offer for Delivery Hero at an equity value of $14.8 billion. Once the acquisition closes with the necessary approvals, Uber will be the largest delivery business, with ~$135 billion in GBV.

Rides :

This is Uber’s largest and most profitable business, with $100B in gross bookings value (GBV) and $7.5B in EBITDA in 2025. That profitability shows up in its high take rate of ~30% (revenue retained after driver payouts) and EBITDA of 7.6% of GBV.

Rides GBV has sustained 20% growth for 15 years because Uber keeps innovating on its product; it now operates in 10+ categories, from motos to luxury cars, tailored to local demand. Product innovation, rising demand density, and falling cost per mile for EVs (and eventually AVs) are enabling Uber to expand into suburban markets and increase overall market penetration.

Every Uber ride breaks down into three major costs.

1. Driver income. The median Uber driver grosses $21.18 an hour, or about $12.50 per trip. That is roughly 70% of the rider fare. After fuel and maintenance, the driver keeps just over 50%.

2. Insurance. This is the second highest cost. Most US states mandate high-value commercial coverage for rideshare trips, which can run up to 30 times a normal policy. Depending on the state, that is 10% to 30% of the fare. We assume about $2 per trip.

3. Vehicle operating cost. AAA puts fuel and maintenance at about 25 cents per mile. That works out to roughly 15% of the fare.

Autonomous Vehicles | Landscape, Unit Economics & Adoption

AV Landscape:

AVs face a high bar to launch: a robotaxi must be meaningfully safer than a human driver. But the bar is fixed, unlike GPUs, where performance keeps doubling every few years, helping the likes of NVIDIA stay ahead of the competition. AVs have a finish line on safety and reliability.

More than a dozen companies globally are racing to reach this finish line. Waymo got there first in the US, but several well-funded companies (Zoox, Tesla, etc.) are in advanced trials with launch timelines in the next few years.

Additionally, Mobileye and NVIDIA are building the tech stack to make AVs plug-and-play technology that any automobile company can license or buy. Stellantis and Volkswagen are already taking this route to build AVs.

With so many players racing to build a technology that has a finish line, we believe the AV tech stack will be commoditized.

As the technology gets commoditized, AV hardware is also getting cheaper. Waymo’s Jaguar I-PACE fleet cost $150,000–200,000 per unit, ~$75,000 for the vehicle and ~$125,000 for autonomous hardware. Its replacement, the Zeekr-built Ojai, costs ~$60k, with a ~$38k base vehicle and AV hardware under $20k, using 42% fewer sensors.

AV scaling is supported by declining cost per mile as hardware costs fall and AV companies compete. AV companies’ success now depends on their ability to scale and capture existing demand — and Uber is in the best position to provide that demand.

AV unit economics & utilization:

AVs carry a large share of operating costs as fixed overhead — fleet management, maintenance, depreciation, etc. To cover these costs, AVs need high utilization rates. As trips scale from 10 per day to 30 per day, fixed costs can drop from ~30% of revenue to ~10% of revenue, improving AV profit margins.

Currently, AVs break even at ~20+ trips a day. AVs launched on Uber’s network are already operating in this zone of mid-to-high 20s to low 30s trips per vehicle per day.

Uber has reported that AVs on its platform generate 30% more trips than on proprietary apps. Consumer ETAs are also 25% lower on Uber, incentivizing riders to keep coming back (the classic platform effect). The only way to break Uber’s grip on demand is to repeat what Uber itself did: heavy consumer discounts and enough supply to hold ETAs down.

This utilization gap on Uber versus proprietary apps drives Uber’s take rate. 30% higher utilization means 30% more revenue on the same AV fleet. Even after Uber takes a 15% cut, AVs still come out ahead.

Uber's take rate is bounded by the incremental utilization it generates for the AV. When an AV player is fully dependent on Uber, Uber can likely charge a flat ~20–30% take rate.

China offers a preview of the central role aggregators play in driving demand and utilization for AVs. T3 Chuxing, China’s #3 ride-hailing operator, saw its orders from aggregator platforms climb from 61.5% to 77.5% to 85.9% over three years. T3 owns and operates its cars, but demand increasingly flows through aggregators.

AV adoption rate & Market size:

Waymo’s cost per mile in the US is estimated at ~$2 in 2025, comparable to that of human drivers on Uber. By 2030, AV cost per mile is expected to fall below $1 as hardware costs decline and fleet management gains economies of scale.

AVs are under 1% of the total cab market today, but within their operating markets they have captured up to a 20% share. Uber has commented on increasing market share in these markets, showing early signs of how AV platforms drive adoption and benefit from it.

As cost per mile declines and AV fleets expand with multiple competing players, AVs are expected to reach ~20% of the US cab market by 2030. With cabs at ~2% market share of personal vehicle miles, lower-cost AVs have room to grow the cab market itself, not just take share within it.

How AV’s will affect Uber’s Effective take rate

Uber’s rides business today has a headline take rate of ~30%, but its incremental contribution margin is only ~10%. If we look deeper at where the other 20% of take rate goes, most of it falls into insurance and driver incentive programs during peak demand. These two components consume 15% of the headline take rate, leaving Uber’s effective take rate at ~15%.

While AVs might carry a slightly lower headline take rate, Uber doesn’t have to pay peak-demand incentives, and insurance costs will be minimal because AVs prove safer & AV operators are likely to absorb this cost themselves as part of fleet negotiations. Thus, even if Uber accepts a lower take rate of 15% on AV rides, it remains competitive with the effective take rate of human-driven trips.

Uber Model and Market Sentiment Rating

To summarize our findings:

• Robotaxis won’t be a winner-take-all market and will most probably end up a split race with multiple players clearing the bar.

• From the supplier side, they will look to maximize utilization, and from the consumer side, users will look for the fastest taxis — both of which Uber is perfectly positioned to service.

• Directionally, AVs should improve Uber's net take rate. And the deflationary pricing cuts both ways: lower prices expand the market by pulling more users into AV rides.

Read more

No paywall on this one. Share it with a friend :)

The elephant in the room in every AI discussion is whether hyperscaler capex is sustainable. The bulls are happy as the numbers keep getting revised higher every quarter.

On the other end, bears look at the $1T number and call it unsustainable.

What surprised us is that we couldn’t find anyone who actually worked through the numbers to figure out who’s right. While certainty is impossible in investing, just like building a DCF model, working through the numbers gives us a plausible estimate of what conditions are needed to sustain the current spending.

This is crucial as nearly 80% of the Market Sentiment portfolio and ~50% of the S&P 500 is directly exposed to AI or AI-adjacent spending. Any pullback here will bring a brutal drawdown.

So get a cup of coffee and let’s dig into what is probably the most important investing question of the decade! To answer this, we need to model

1. What goes into a datacenter, and how long it lasts.

2. What the new capacity earns, and for how long.

3. What’s the running cost, and how would the margins evolve over time?

Join 46,000 other investors to make sure you don’t miss our next briefing.

Subscribe now

What goes into an AI datacenter?

Today, a 1GW datacenter is expected to cost ~$40B. This capex splits into two categories defined by asset life.

1. Chips — compute and memory

Traditionally, chips last 5 to 6 years, after which a more efficient generation makes them uneconomic to run, and reliability begins to drop. While cloud companies are trying to extend this life by moving older chips to lower-intensity inference workloads, for our model we are assuming a 5-year life (more on this later).

CoreWeave’s A100s (a 2020 release) are still fully booked, and a batch of Nvidia H100s (2022) that came off contract was immediately re-let at ~95% of the original price. CoreWeave CEO Mike further noted (Q1 2026) sustained pricing across the older fleet as inference pushes demand up for the older-generation devices.

2. Datacenter shell — building, power equipment, networking

The useful life of buildings runs up to 30 years. Google commented on its recent earnings call that ~60% of capex went to servers and ~40% to datacenters and networking equipment. The shell’s share of the mix has come down from ~50%+ a few years ago as semis have become more expensive.

Given rising prices with each new GPU generation, we assume a semis-to-shell capex ratio of 2:1. So for every $100B spent on a new datacenter, $67B goes to chips and $33B goes to the shell.

We take the top-of-the-range shell life because the shell is chip-agnostic. The building, power, and cooling serve successive generations of chips without major changes. Also, the permitted sites are one of the scarcest assets in the buildout, and the owners would want to push as much mileage out of them as possible.

This capex drives the entire AI semis complex. Microsoft, Amazon, and Google plan to spend ~$600B on capex in 2026, versus ~$300B in 2025. This spending gets redirected one way or the other into the chip side (Nvidia, AMD, TSMC, SK Hynix, etc.) and the shell side (Vertiv, GE, Schneider Electric, Caterpillar, etc.).

What the capacity earns?

A datacenter earns by renting compute. Two variables set that revenue: how much of the capacity gets rented, and the rate per GPU-hour.

The first bit is easy for now: all data points to the fact that capacity is sold before it is even built. AWS backlog reached $496B in Q2 2026, and Amazon has said the lion’s share of its 2027 capacity is already reserved. Every other hyperscaler shows the same trend. Microsoft closed June 2026 with $678B of commercial RPO, up 84% year over year; Google Cloud’s backlog hit $514B in Q2 2026, up more than $50B in a single quarter.

The rate per GPU-hour is the tricky bit, as it keeps falling as the GPU ages. For example, the one-year committed pricing for H100s has fallen from ~$3.00 to ~$1.80-2.40 per GPU-hour over three years.

The average H100 rental rate over its first three years was ~$2.25/hr. At 90% utilization, that is ~$18k of revenue per GPU per year. A $100B datacenter puts ~$67B into chips at the 2:1 ratio, which at ~$32k per H100 buys ~2.1 million GPUs.

So roughly, a $100B datacenter will generate ~$37B of revenue a year over its first three years.

What it costs to run?

A datacenter has two major operating costs: energy and maintenance.

Electricity: Datacenters are measured in power; as the name suggests, a 1 GW datacenter consumes 1 GW of power. But there are efficiency losses between power drawn from the grid and power actually used by the compute, measured as PUE (Power Usage Effectiveness). Datacenters have an estimated PUE of 1.14, so a facility with 1 GW of IT load draws 1.14 GW from the grid. Assuming a compute utilization rate of ~90%, total annual consumption comes to 1 GW × 90% × 1.14 × 8,760 hours = 8,987 GWh (~9.0 TWh).

US average industrial power cost is 8.66 cents per kWh as of April 2026, or $86,600 per GWh. Datacenters are built where power is cheapest, with low-cost states running below that; for our model, we assume the US industrial average. This gives a total power cost of 8,987 GWh × $86,600 per GWh = $778mm, or approximately $800mm per year for a 1 GW datacenter at a 90% utilization rate.

Maintenance: Maintenance costs are primarily spread across property taxes, labor, and repair work. As per Epoch AI, average maintenance for a 1 GW datacenter stands at ~$300mm across these three categories.

So, all in, the annual cost of running a 1 GW datacenter stands at ~$1.1B with $800mm in energy at today’s prices and $300mm in maintenance. As per a Morgan Stanley report, this can go as high as $2B depending on power costs and property taxes at the location. For our model, we will be conservative and take $2B as the annual cost of operating a 1 GW datacenter.

A 1 GW datacenter is estimated to cost ~$40B to build. With $100B in capex, we can build 2.5 GW, leading to total operating costs of ~$5B per year.

The Model

Bringing it all together, we can now model the cash generation of a $100B datacenter.

• At a 2:1 semis-to-shell ratio, $67B goes to chips and $33B to the shell.

• At 90% rental occupancy, the datacenter will generate $37B in annual revenue for the first 3 years and then $24B for years four and five (assuming a 1/3rd decrease in GPU rental pricing).

• Operating costs hold at ~5B per year across 5 years.

Now assume $100B is invested in datacenters every year, and let’s see what steady-state revenue it will generate.

Every $100B of capex buys chips with a five-year life. For the first three years, that cohort generates $37B of revenue annually. In years four and five, revenue drops to $25B as pricing falls with chip age — the decay we saw earlier.

Now spend $100B every year and run it forward. By year five you’re operating five vintages at once: three young cohorts earning $37B each and two aging cohorts earning $25B each.

(3 × $37B) + (2 × $25B) = $161B

That’s steady state. From year five onward, one cohort retires as a new one comes online, so revenue holds at ~$160B while capex stays at $100B.

So a $100B annual capex will generate ~$160B of annual revenue over the long run — a revenue-to-capex multiple of ~1.6x.

Based on our model at this 1.6x multiple and $5B annual costs, a cloud company generates an IRR of ~21%.

It is worth remembering our conservatism here. We modeled a 5-year semiconductor life, while CoreWeave currently runs up to 6+ years; hyperscalers also have better negotiating power and run a mix of GPUs and more efficient ASICs, both of which support lower operating costs. Adjusting for these, the IRR should land even higher.

Share

Benchmarking against real data

The model implies chip spend is recovered in roughly three years. Now let’s check that against what an H100 purchased in 2023 would actually have earned, based on its purchase price and realized rental rates. An H100 cost an estimated $25–40k in 2023.

Based on SemiAnalysis’s rental price tracker, the average H100 rental rate over its first three years (2023-26) was ~$2.25/hr, peaking at ~$3.00/hr in the early period. Assuming 80% utilization and a 75% cash margin, that generates ~$11.8k of annual cash flow, implying a payback of ~2.7 years. With cloud companies already locking in capacity for 2027-28, realized utilization may run above our 80% base case.

_{H100 ROI Model | Source: Market Sentiment Research}

Morgan Stanley’s unit economics model for a 1GW datacenter built on Nvidia’s latest GB300 chips assumes a materially shorter payback in the current pricing environment. The model implies $23B of semis capex, generating annual cash profit of ~$15B at the low end, leading to a payback of ~1.5 years on the chip spend.

For reference, if we take Morgan Stanley’s pricing (payback period of 1.5 years) for the first 2 years and a 1/3rd price decline every year after that and plug it into our model, we will have an IRR of ~30%. This is how attractive current compute pricing is.

Can revenue get there?

The top three cloud companies are expected to spend ~$600B on capex in 2026, against ~$300B in 2025. Applying the 1.6x revenue-to-capex multiple to that base, they need to generate ~$960B in cloud revenue to sustain the current level of spend.

Today’s combined run rate is ~$380B, growing at 40%+ and still accelerating.

So rather than debating whether the current semiconductor drawdown is a correction or the start of a bubble bursting, the question can be reduced to something simpler: can cloud companies grow their revenue to $960B?

We think they have a reasonable chance. Every major cloud is accelerating on an already large base:

• AWS is at a ~$169B run rate, with growth accelerating to 37% from 17.5% a year ago.

• Google Cloud has crossed a ~$99B run rate, with growth jumping to 82% from ~32%.

• Azure is past $100B and still accelerating, to 43% from ~39%.

At the current ~40% growth rate, cloud revenue reaches a ~$1T run rate in three years and ~$1.5T in four. Amazon CEO Andrew Jassy has said he expects AWS alone will likely scale to a $1T revenue business over time.

Further, when thinking about their capex, it helps to remember internal capex needs for R&D and first-party products like Search, YouTube, Office and Prime, which would require spending regardless of AI for cloud services. Cloud companies spent ~$100B of capex in 2022, before AI began scaling for consumers.

All in, we think the 2026 capex of ~$600B is sustainable as revenue scales. We will revisit this next year to test whether the Street’s ~$1T estimate for 2027 holds up against actual revenue and bookings growth.

Where it can break

While current growth rates and bookings are great, we see three major risks to cloud demand.

1. The compute demand does not show up

Everything above assumes enterprises and consumers keep using more AI. Usage is growing today, and enterprise spend is rising with it.

Now that the token-maxing era has passed, companies like Coinbase and Uber are optimizing AI budgets by placing monthly spend limits per employee.

But while some firms trim employee AI credits, aggregate spend keeps climbing as AI penetrates internal workflows. In BCG’s survey of 2,600 C-suite executives, only 6% plan to cut AI spend, while average budgets are expected to double in 2026 versus 2025

2. One big player blows up, and there goes the backlog

Over 40% of cloud backlog traces to two model labs: an estimated ~$700B to OpenAI and ~$330B to Anthropic.

Anthropic last reported a ~$47B run rate as of May 2026, up from ~$9B at the end of 2025, with roughly 80% of revenue from enterprises. Enterprise revenue is stickier and more sustainably monetized, which makes the $330B commitment look better covered.

OpenAI’s run rate is at $25B since February 2026 compared to 2025 revenue of $13.1B and a $20.9B operating loss. It filed a confidential S-1 in June 2026 but has since leaned toward delaying the listing to 2027. We view the OpenAI backlog as carrying more risk.

The right question, though, is what actually happens if OpenAI disappears tomorrow. As a ChatGPT user, I won’t stop using AI. I will switch products, and that platform’s compute demand will absorb OpenAI’s.

3. Decline in demand from VC-funded AI companies

The third risk is a pullback in demand from VC-backed AI startups. Venture funding rose to ~$650B in 1H26 from ~$350B in 2023. Look deeper, though, and a large share concentrates in the top five companies, essentially the AI labs. Exclude those and 1H26 funding is ~$220B, against ~$300B for all of 2023, a difficult year for venture capital.

We therefore don’t see venture investment as a primary driver of AI compute demand. A decline in venture funding would matter at the margin, but ex-labs venture investment has not scaled meaningfully with AI in the first place.

None of this makes it a sure thing. Every bank in the street underwrote the fiber buildout in the dot-com bubble. It worked till it didn’t.

But with this, we have a sneak peek into the math that the hyperscalers and the banks that are underwriting the multi-billion-dollar bonds are looking at. While we are working with public data, they would have access to at least 100x more data points flowing directly from the datacenters.

This is also exactly why we are comfortable positioning MS portfolio across these three categories and have conviction to hold through the market drawdowns.

ASICs (~20% of portfolio)

ASICs enable hyperscalers to improve profitability on the current revenue base and unlock new use cases by pushing down token cost.

Memory (~33% of portfolio)

Memory has seen a sharp correction because of unwinding leveraged trades. With sustained capex of $600B and a large part of it going to inference rather than training, we see memory as a capacity-constrained market for the next one to two years and remain bullish in the near term.

Advanced Packaging (~25% of portfolio)

This category contains manufacturing lines (TSMC, Intel, Amkor) whose revenue is directly linked to chip demand; hence, we remain bullish on them. Capital equipment firms (LRCX, AMAT, KLAC) depend on growth in capex, not sustainability of capex. Sustained capex means no need for new manufacturing lines, which will directly hit their revenue growth. We are least bullish on these firms and will trim our positions as capex guidance peaks at $1T in 2027.

If you made it till here, I would love to hear what you think:

Leave a comment

That’s it for now. As you can see, hundreds of hours of research go into each of these reports. Please consider upgrading your subscription to support our work and to get access to all the reports.

Subscribe now

Judging by DMs we received last week, a lot of you are running portfolios wildly outside your risk tolerance. While we cannot recommend custom portfolio strategies, a simple trick I have learned over more than half a decade is to build a sleep-well portfolio.

The setup is simple: your portfolio should be structured in a way that you can sleep through the market open without missing a beat. Most of us are not fund managers, and there is no good reason to check your portfolio every hour (and it's detrimental to your happiness irrespective of how the portfolio performs). If you are all-in, or worse, 4x levered like Leopold on names that have gone up 2-5x in the last year, you are asking for trouble. Having a reasonable exposure will also help you think rationally, which is exactly what we were doing a few days back (with the thesis playing out literally the next day).

With that, let’s get into the big developments over the past month, how our portfolio performed, and how we are positioning over the next few months.

Overall, it was quite a bad month for AI bulls with the SOXX semiconductor ETF down 21% and memory names taking a beating (DRAM was down ~30% last month).

We see this correction as unwinding of leverage and portfolio concentration that’s been building in semis. Korea’s leveraged ETFs had ~$10B AuM at the beginning of 2026. By June it ballooned to ~$50B with SK Hynix-linked leveraged products alone at ~$23 billion.

On July 16, South Korea’s Financial Services Commission announced stricter regulatory rules for single-stock leveraged ETFs, raising the minimum margin requirement from 10 million Korean Won to 30 million Korean Won and limiting purchases to a maximum of 20 shares per person per transaction.

During the stock price decline phase, the mandatory end-of-day rebalancing mechanism of leveraged ETFs triggered programmatic automatic selling, instantly creating a sharp selloff. That day, SK Hynix fell another 11%, Samsung slumped over 8%, and the wave of panic quickly swept through Europe and the US. Leopold’s $45 billion AI fund blowing up last week added fuel to the fire.

While the correction can be largely attributed to normal market volatility amplified by leverage, there were quite a few material events:

The launch of Kimi K3

Moonshot released Kimi K3 on July 16, the largest open-source model with 2.8 trillion parameters. AI capability benchmarks have placed it on par with OpenAI 5.5 and Opus 4.8.

Kimi follows a dozen open-source, low-cost models China has released since DeepSeek R1. In our view, AI adoption gains far outweigh open-source AI efficiency gains, as reflected in Nvidia’s sharp rebound post the first DeepSeek event.

But companies are currently in the token-maxxing era. Every employee, whether they need it or not, has access to frontier models with a nearly unlimited AI budget. There will come a time when companies see the AI bill and ask if it’s really providing the ROI. This does not mean they will stop using AI. They will just optimize it better.

There’s no need to use Claude Fable 5 to rewrite your email or update your deck for the 20th time. Open source models can do the same at 1/100th of the cost. Serving intelligence becomes a race to the lowest cost per token, the same way Japanese carmakers won the US market by delivering most of the performance at a much lower price.

CXMT had a monster IPO

CXMT is a Chinese memory company holding 8% of global DRAM capacity. CXMT listed on Shanghai on July 27, raising $8.6B in Asia’s biggest IPO this year. It popped 466% on debut, valuing it at ~$500B+ as one of the most valuable China-listed companies.

China’s semiconductor capability buildout is well known; however, it lags the leading edge by a few nodes, and CXMT itself doesn’t have HBM capability.

New memory capacity by Korean players

Samsung and SK announced investment of ₩800T (~$516B) in four new fabs to double memory capacity within five years. This is the most significant news for memory names with questions on disciplined supply. However, new capacity comes online only in 2H27–28, keeping memory in tight supply for the next 4Qs.

These plants are backed by major contracts for Korean memory names. SK has long-term contracts of ~$700B ($500B with Nvidia, Microsoft, Anthropic) while Samsung has ~$200B (Broadcom MOU).

While the Capex addition is real, the quantum of Capex should be taken with a pinch of salt, as these announcements were made as part of a South Korean government super-industrial plan and not a corporate strategy event. Historical precedent shows that memory companies will delay Capex if they don’t see supportive pricing (2022-23 delayed wafer starts). Thus, we see this capacity addition plan having no impact for the next 1-2 years, and the supply-demand dynamic at that point will be tested based on AI adoption.

MS Portfolio Update

Read more

FYI — Live tracker is only available to paid subscribers and can be accessed at the end of this report.

First things first.

Judging by DMs we received last week, a lot of you are running portfolios wildly outside your risk tolerance. While we cannot recommend custom portfolio strategies, a simple trick I have learned over more than half a decade is to build a sleep-well portfolio.

The setup is simple: your portfolio should be structured in a way that you can sleep through the market open without missing a beat. Most of us are not fund managers, and there is no good reason to check your portfolio every hour (and it’s detrimental to your happiness irrespective of how the portfolio performs). If you are all-in, or worse, 4x levered like Leopold on names that have gone up 2-5x in the last year, you are asking for trouble. Having a reasonable exposure will also help you think rationally, which is exactly what we were doing a few days back (with the thesis playing out literally the next day).

With that, let’s get into the big developments over the past month, how our portfolio performed, and how we are positioning over the next few months.

Overall, it was quite a bad month for AI bulls with the SOXX semiconductor ETF down 21% and memory names taking a beating (DRAM was down ~30% last month).

We see this correction as unwinding of leverage and portfolio concentration that’s been building in semis. Korea’s leveraged ETFs had ~$10B AuM at the beginning of 2026. By June it ballooned to ~$50B with SK Hynix-linked leveraged products alone at ~$23 billion.

On July 16, South Korea’s Financial Services Commission announced stricter regulatory rules for single-stock leveraged ETFs, raising the minimum margin requirement from 10 million Korean Won to 30 million Korean Won and limiting purchases to a maximum of 20 shares per person per transaction.

During the stock price decline phase, the mandatory end-of-day rebalancing mechanism of leveraged ETFs triggered programmatic automatic selling, instantly creating a sharp selloff. That day, SK Hynix fell another 11%, Samsung slumped over 8%, and the wave of panic quickly swept through Europe and the US. Leopold’s $45 billion AI fund blowing up last week added fuel to the fire.

While the correction can be largely attributed to normal market volatility amplified by leverage, there were quite a few material events:

The launch of Kimi K3

Moonshot released Kimi K3 on July 16, the largest open-source model with 2.8 trillion parameters. AI capability benchmarks have placed it on par with OpenAI 5.5 and Opus 4.8.

Kimi follows a dozen open-source, low-cost models China has released since DeepSeek R1. In our view, AI adoption gains far outweigh open-source AI efficiency gains, as reflected in Nvidia’s sharp rebound post the first DeepSeek event.

But companies are currently in the token-maxxing era. Every employee, whether they need it or not, has access to frontier models with a nearly unlimited AI budget. There will come a time when companies see the AI bill and ask if it’s really providing the ROI. This does not mean they will stop using AI. They will just optimize it better.

There’s no need to use Claude Fable 5 to rewrite your email or update your deck for the 20th time. Open source models can do the same at 1/100th of the cost. Serving intelligence becomes a race to the lowest cost per token, the same way Japanese carmakers won the US market by delivering most of the performance at a much lower price.

CXMT had a monster IPO

CXMT is a Chinese memory company holding 8% of global DRAM capacity. CXMT listed on Shanghai on July 27, raising $8.6B in Asia’s biggest IPO this year. It popped 466% on debut, valuing it at ~$500B+ as one of the most valuable China-listed companies.

China’s semiconductor capability buildout is well known; however, it lags the leading edge by a few nodes, and CXMT itself doesn’t have HBM capability.

New memory capacity by Korean players

Samsung and SK announced investment of ₩800T (~$516B) in four new fabs to double memory capacity within five years. This is the most significant news for memory names with questions on disciplined supply. However, new capacity comes online only in 2H27–28, keeping memory in tight supply for the next 4Qs.

These plants are backed by major contracts for Korean memory names. SK has long-term contracts of ~$700B ($500B with Nvidia, Microsoft, Anthropic) while Samsung has ~$200B (Broadcom MOU).

While the Capex addition is real, the quantum of Capex should be taken with a pinch of salt, as these announcements were made as part of a South Korean government super-industrial plan and not a corporate strategy event. Historical precedent shows that memory companies will delay Capex if they don’t see supportive pricing (2022-23 delayed wafer starts). Thus, we see this capacity addition plan having no impact for the next 1-2 years, and the supply-demand dynamic at that point will be tested based on AI adoption.

MS Portfolio Update

Read more

The jump in performance per watt was incredible. M chips did the same tasks Intel chips did at roughly a quarter of the power draw. The result: the M1 Air managed 15 to 18 hours of real use versus barely 8 hours on the Intel Air with the same battery capacity.

Apple achieved this with custom silicon. Intel could never build this chip for Apple because Intel had to serve everyone. Its processors carried decades of baggage and had to work in every Windows machine from a $300 Dell to a gaming rig, so they could not be optimized for one laptop or one operating system.

Apple’s decade long investment into custom silicon paid off — Mac revenue which was stagnated at ~$25B from 2017 to 2019 jumped 60% and touched ~$40B by 2022 before normalizing to $33B for the latest year.

This is not about laptops; It is that owning your silicon converts directly into performance per watt, and performance per watt converts directly into a product nobody else can build.

Nowhere is this now more important than AI. Our estimates show that for every $100 of inference cost, hyperscalers currently hand roughly half to Nvidia. This eye-watering 85% gross margin Nvidia charges for its AI chips is currently the single largest tax on the AI buildout, and it is a tax you can escape by designing your own silicon.

Market Sentiment dives deep into the bottlenecks of the AI world. Join 46,000 other investors to make sure you don’t miss our next briefing.

Subscribe now

What is an ASIC?

ASIC stands for Application Specific Integrated Circuit. It is a chip designed by one company for its own specific use case, rather than sold on the open market. Examples include:

• The Apple M series chips we just discussed

• Google’s TPU AI processor

• Amazon’s Trainium 3 AI processor

• Nokia’s ReefShark base station processor

If you are wondering how these differ from Nvidia or AMD GPUs: a GPU is a Swiss Army knife whereas an ASIC is a scalpel.

Nvidia builds its GPUs to run anything. Training, inference, any model architecture for text, vision, protein or physics, in any precision format. That flexibility comes at a price. A meaningful share of every GPU die is spent on circuitry your specific workload will never touch.

An ASIC flips the trade. You strip out everything except what your workload needs and pour the entire silicon budget into exactly that. Bitcoin mining chips are the purest example. They compute one hashing function billions of times a second and can do literally nothing else. Not even a calculator app would run on one.

The AI versions apply the same logic to neural networks. The result is 2 to 3x better performance per watt and real savings on silicon cost: Google's TPU7x matches NVIDIA's B200 on raw compute at roughly a third of the price: ~$13,000 per TPU vs $35,000.

Artificial Analysis benchmarked open source Llama models across hardware. The results below are clear: Google's TPUs squeeze roughly 2x more tokens out of each TFLOP than any Nvidia chip.

Your Margin Is My Opportunity

With AI going mainstream, the nature of AI compute is changing. Inference is starting to take a larger component of overall compute. Inference workloads are predictable and repetitive: exactly what custom chips are built for. Projections show that just the inference demand is the fastest growing segment at a 35% CAGR and will need 90GW of data center by 2030.

The catch with ASICs is that they are expensive to build. A leading-edge custom chip costs $500M+ in design, IP licensing, and tape-outs before a single unit ships. The math only works at scale. You need enough deployment volume that the per-chip savings outrun the upfront design cost.

Every hyperscaler has now crossed that threshold and is building their own chips.

Google has been at it the longest, and its TPUs are the most mature ASICs in production. Amazon is not far behind with Trainium and Inferentia, which we believe will boost AWS returns in the coming years. Meta is the newest entrant: its MTIA chips already run recommendation inference internally, and in February it signed a multibillion-dollar deal to lease Google’s TPUs, with talks underway to install them in Meta’s own data centers from 2027.

The race to serve models at the lowest cost

In our view, companies are currently in the token-maxxing era. Every employee, whether they need it or not, has access to frontier models with a nearly unlimited AI budget.

This will not last, for two reasons.

First, the C-suite is starting to demand ROI for the spend. Uber burned through its entire 2026 AI budget in four months on coding tools, and its COO now admits it is very hard to draw a line between token consumption and useful product output. Microsoft reportedly canceled most of its direct Claude Code licenses this month and moved engineers to GitHub Copilot.

Second, not every task needs a frontier model. Writing an email response or updating a deck does not require Claude Fable 5. An open source model can do the same work at a tenth of the cost or less, and as open models get smarter, the share of workloads running on them will keep expanding.

Note that this does not mean AI spend shrinks. Usage will keep exploding while the price per token collapses, so the pie grows even as every workload gets cost-optimized. What it means is that serving intelligence becomes a race to the lowest cost per token, the same way Japanese carmakers won the US market by delivering most of the performance at a much better price.

If that is the race, hyperscalers have no option but ASICs.

Investing in Accelerating ASIC Adoption

The custom ASIC market is projected to grow from 8% of AI accelerators today to 19% by 2033, a $118 billion segment compounding at 27%. Given how fast the open source models are improving, we believe this projection might be understated.

We have identified 5 companies that sit directly in this flow, and we are backing them with ~20% of the MS portfolio.

Read more

Quite a few of you have joined us over the past week (we were #5 in the Rising Finance Bestsellers) and here’s a guide on what we do at Market Sentiment and how to use the subscription effectively.

Investing in asymmetric opportunities

There are only two ways to make consistent alpha in the market:

1. Be early

2. Be contrarian

It’s futile to look at a company that’s covered by 20 analysts and expect to find something that all of them have missed. Our focus at Market Sentiment is to be early to trends that are going to define the next few years.

Whether you like it or not, that trend right now is AI. We are pumping a mind-bending amount of capital into compute. To put things in perspective (all figures adjusted for inflation):

• Every month, we are spending more on AI infrastructure than the total cost of the Manhattan Project.

• Every quarter, we now allocate more to AI than the entire 14-year Apollo program.

• Just last year, we spent more than what it took to build our entire Interstate highway system over three decades.

AI is the consensus trade now. But you can still be early to what it depends on. Capital at this scale never flows evenly. The railroad mania of the 1800s bankrupted most railroads but made fortunes for the companies selling steel and rights of way. Every infrastructure boom in history worked this way.

The AI buildout will be no different. The hyperscalers spending hundreds of billions will not be where the returns concentrate — that trade is long over. The returns will concentrate at the bottlenecks that these companies can’t sidestep. Our focus is to find them before the market prices them in.

Thematic Coverage

We have built five thematic baskets to date. Three of the five have already re-rated. Energy and Agentic AI have not, which is why they are worth a closer look right now.

1. Energy — AI needs energy. A lot of it. And more importantly, it needs it now. The power demand for data centers is expected to grow 10x the historical generation capacity growth. This is the one basket that’s underperformed (4% vs 15% for the S&P 500), and the reason was a simple mistake: we underestimated how fast the price of power would turn political, and over-allocated to companies exposed to electricity prices. The fix is just as simple: strip those out, and the rest of the basket handily beats the market (more on this below).

2. Optics — We were one of the first to cover the Optics trade last year. While the tech behind is complex, our thesis was simple: As models become more powerful and AI pushes data transfer speeds higher, the industry has no choice but to adopt optical interconnects. The basket has now rerated and is up 75% vs. 16% for QQQ.

3. Advanced Packaging — Once transistor shrinks stopped being cheap, the industry stopped chasing smaller nodes and started gluing chips together. Advanced packaging became the way to keep Moore's law alive, and the bottleneck moved from the wafer to the package. The basket is now up 102% vs 15% for QQQ.

4. Memory — Everyone knows memory stocks are cyclical and have burned investors before. But that’s the wrong question. The right question is whether we’re in a structural shift — and the data says yes. We made memory our highest allocation (34%) and were handily rewarded with a 134% weighted return (we bought the dip during the Iran war).

5. Agentic AI — We are at the inflection point for Agentic AI. Most AI conversations right now are about coding assistants, but that’s really only useful for the ~25 million developers out there. With Claude now generally available in Microsoft Office, AI can meaningfully support the 1.4 billion knowledge workers who spend their days buried in email, spreadsheets, and jumping between a dozen tools.

MS Portfolio

Our portfolio is where we put capital behind these emerging trends. Below you'll find every position, its current buy/sell/hold rating, and all the changes we've made along the way. If you're just starting out, begin by gradually allocating to our buy-rated names — these are the companies we still consider undervalued.

Market Sentiment Portfolio

Two important notes:

1. Many of you have DM’d us asking how you should allocate. As much as we’d like to help, we’re not financial advisors and can’t give customized allocation advice. What we can share is exactly what’s in the sheet: our positions, allocations, and buy/sell/hold ratings, plus how we think about each of them in our deep dives.

2. Starting Monday, this sheet moves behind the paywall. Our full portfolio, including positions, allocations, and ratings, will be available exclusively to paid subscribers. If you’ve been on the fence, this week is a good time to upgrade.

Subscribe now

What’s coming?

Which brings us to the good stuff. There are two big themes that we have been tracking for a while and will be sharing with you over the next few weeks.

1. ASICs

ASICs (Application Specific Integrated Circuits) are projected to overtake GPU unit shipments by next year.

ASICs are chips custom designed for one specific workload. Because these chips are tailored at the transistor level, they are highly efficient at the task its designed to do.

Google’s TPU7x matches NVIDIA’s B200 on raw compute at roughly a third of the price: ~$13,000 per TPU vs $35,000.

The catch here is that ASIC development is a multi-year, multi-billion-dollar project that only hyperscalers like Google and Amazon can afford. But that’s exactly who is building them. As model winners emerge and applications standardize, ASICs lead the next wave of compute growth.

Here’s how we look at the ASIC supply chain: design tools (Cadence, Synopsys), design services (AVGO, Marvell), and licensed IP blocks (Arm, Aspeed).

The real prize sits with the designers. Complex AI ASIC design is effectively an AVGO–Marvell duopoly. This lets them charge fixed development fees plus a variable price per chip.

Given ~95% unit growth and per-chip pricing power, AVGO and Marvell earnings should see superior earnings growth over the next 2–3 years. AVGO expects its AI semiconductor revenue to grow from approximately $20 billion in FY25 to over $100 billion in FY27 primarily driven by these ASICs.

When your data-center investments run into hundreds of billions, spending a few billion on a custom ASIC is a no-brainer.

2. Energy

One mistake we made with the energy basket: we expected power prices to keep rising as demand scaled. Demand did scale. But power prices became a political problem rather than a supply-demand equation. In January, the governors of all 13 states in the PJM grid region, along with the Trump administration, jointly declared that power prices needed to be reined in and that data centers should pay for the grid costs they create. The price ceiling on capacity auctions was then extended to 2030, taking the best-case scenario we had modeled for these companies off the table.

The fix is to separate the demand thesis from the price thesis. The demand thesis was right and remains right. The price thesis is now capped by regulators until 2030.

So we are restructuring the basket around companies that get paid on volume, not price. That means the companies selling the physical building blocks of the grid: gas turbines, transformers, switchgear, high-voltage cables, backup power, and cooling systems, along with the contractors who install all of it. A transformer sold to an Amazon data center faces no price ceiling, because voters don’t pay for it.

That’s it for now. Leave a comment if you have an idea/theme that you think we all should be looking at.

Leave a comment

Disclaimer: Market Sentiment work is provided for informational purposes only, is intended solely for readers in the United States, and should not be construed as legal, business, investment, or tax advice. You should always do your own research.

What’s more interesting is the shift in mainstream sentiment. It has moved from “AI investment is setting cash on fire” to “maybe these trillion-dollar companies aren’t spending for nothing.”

Take the chart below from the FT:

While estimates should always be taken with a bucket load of salt, it’s worth digging into what they are projecting:

1. The CapEx is expected to plateau (not drop) — Once you get past 2027, the CapEx is expected to stop growing at 70 to 80% a year that we are seeing now and settle into a maintenance plus refresh level.

2. Revenue will catch up quickly — If inference, cloud, and ad monetization compound while CapEx flattens, operating cash flow keeps climbing.

3. Finally, CapEx is paid upfront but counts against profits gradually in the form of depreciation over the life cycle of the asset. As the accumulated asset base grows, the depreciation add-back in the FCF bridge gets very large.

If this plays out and free cash flow explodes, the multiples will re-rate immediately. These names now carry a discount due to CapEx uncertainty. When quarterly prints start confirming that capex is flattening and cash is converting, the uncertainty discount will lift and the multiple will expand.

While all 3 are projections, we have some signs of revenue catching up. Bloomberg reported last week that AI revenue has now caught up with the quarterly CapEx depreciation. If the revenue trend holds while capex plateaus, it confirms that the trillions hyperscalers poured into AI capex were worth it.

Global AI sales, excluding China, reached $25 billion in the first quarter of 2026, exceeding the industry’s estimated $21 billion in depreciation costs tied to investments in data centers and chips for the second consecutive quarter. — Bloomberg

Market Sentiment dives deep into the bottlenecks of the AI world. Join 46,000 other investors to make sure you don’t miss our next briefing.

Subscribe now

That brings us to the MS portfolio. We are more or less in line with where we were at the beginning of last month, with the MS portfolio up 75% YTD. Memory continues to be our strongest performer, with SK Hynix, Samsung, and Micron all performing exceptionally well.

MU -6.25%↓ trades around $1,050, roughly 8–9x forward earnings — versus Nvidia’s ~30x trailing multiple, for a company that’s just as levered to AI capex. The gap exists because investors still expect memory prices to crash as they did before.

The Strategic Capacity Agreements (SCAs) are Micron’s direct answer to that skepticism: floor pricing through 2030 for a significant share of its supply, signed by the same hyper-scalers and automakers. We believe the market is still not pricing this long-term structural shift in memory demand and profitability for the next 3-5 years.

As we highlighted last week, we will continue to hold these names and expect them to rise further as the market digests the impact of long-term contracts and becomes less concerned about the cyclicality of memory stocks.

Update on Optics

Our Optics basket was a bet on AI data centers hitting the “Copper Wall”, a physical limit beyond which copper cables cannot transmit data fast enough over a given distance.

Existing high-speed copper cables can transmit data at 112 Gbps over a distance of 2 meters. The next generation of servers is projected to run at twice that speed, and at that point traditional passive copper will no longer reliably carry a signal across a server rack.

The basket was built intentionally to get exposure to both approaches being developed to overcome this problem:

1. Add signal boosters to existing copper data cables (Active Electrical Cables)

2. Convert the electrical signal into an optical signal, achieving near-perfect transition (Co-Packaged Optics – CPO)

Since its launch (Dec’25), our Optics basket is now up 72% vs 14% of QQQ. This outperformance was driven by AEC volumes scaling faster than planned: in the most recent quarter, Credo Technology Group reported ~150% YoY revenue growth and Astera Labs ~90%.

While AEC volume scaled, our optical bets also got several validating events:

• Nvidia announced a multibillion‑dollar purchase commitment for Co-Packaged Optics along with a $2B investment in Coherent.

• Arista Networks also launched XPO-MSA, a switch that uses removable optical connectors. It’s a stepping stone toward denser optical connections while the industry works toward fully integrated optics (CPO).

AEC companies benefit from mature products and are expected to grow by ~60%-80% through 2027, with a strong outlook through 2030. CPO companies are expected to grow ~30%-40% in 2027 as their initial product scales, with potential for better growth over the long run.

Post this strong rally, most portfolio companies in our basket are trading above 100 TTM P/E. Given this, we are updating the optics basket from Buy to Hold.

MS Portfolio Holdings Update

Read more

No paywall on this one. Share it with a friend :)

This one’s a bit different from our usual reports. I just saw an incredible stat from the Nomura research team, and it felt like a good time to revisit our research on Leveraged ETFs from two years ago.

Leveraged ETFs have grown nearly tenfold in the past decade. AUM now sits at an all-time high — this surge is a key driver of the rising volatility we are seeing in markets. The more concerning part is what sits underneath it. Most of this money is retail chasing performance in semiconductors. Over $50B in leveraged ETF AUM is now concentrated in semiconductor funds.

Investors buy 2X and 3X leveraged ETFs expecting 2X and 3X returns. It rarely plays out that way. The risk of capital loss runs far higher than intuition suggests. Here is what happened to the 2X-leveraged MicroStrategy ETF, a fund with over $1B in AUM.

Market Sentiment dives deep into the bottlenecks of the AI world. Join 46,000 other investors to make sure you don’t miss our next briefing.

Subscribe now

To understand what’s going on, we have to first understand how leverage works in investing:

Let’s say that you had $10K in your brokerage account. You have a high-conviction stock you think is going up and want to invest more than $10K in it. You can take another $10K as a loan from your brokerage, staking your entire position as collateral.

Assume the stock went up 10% over the next month, and you exit the position.

This is the magic of leverage. You used a 2:1 leverage on your investment and doubled your return1

What’s important to realize is that this works the other way, too — If the stock went down 10% and you had to exit the position, the 2:1 leverage would literally double your losses.

Having a large amount of leverage is like driving a car with a dagger on the steering wheel pointed at your heart. If you do that, you will be a better driver. There will be fewer accidents but when they happen, they will be fatal - Warren Buffett

Understanding Leveraged ETFs

It’s vital to understand how leveraged ETFs work — class action lawsuits were filed (and later thrown out) because investors did not understand how they are structured.

For example, let’s take a quick look at the TQQQ fact sheet:

The fund is only trying to 3x the daily return of QQQ and not the long-term return.

To understand how the index gains and index volatility affect the total return, consider this — Let’s say you invested $100 into QQQ and TQQQ. On day 1, the index went up 10%. On day 2, the index went down 10%.

Here’s how both funds would have performed2

• Day 1:

  • Index (QQQ): 100 x (1 + 10%) = $110

  • 3x Leveraged Index (TQQQ): 100 x (1 + 3 x 10%) = $130

• Day 2:

  • Index (QQQ): 110 x (1 - 10%) = $99

  • 3x Leveraged Index (TQQQ): 130 x (1 - 3 x 10%) = $91

• Overall Return (after 2 days):

  • Index (QQQ): (99 - 100)/100 = -1%

  • 3x Leveraged Index (TQQQ): (91 - 100)/100 = -9%

Share

Leverage vs. Concentration

Investors prefer concentration over leverage when maximizing their expected return.

This means most of us will allocate larger and larger portions of our portfolios to stocks, expecting higher returns. In contrast, you would have gotten a better risk-adjusted return if you had a well-diversified portfolio and then leveraged that.

One key reason for increasing concentration rather than leverage was aversion to leverage. Investors believed that increasing concentration was the “natural or conventional” path to taking leverage, which felt speculative.

What’s the optimal leverage?

While there is no one right answer, leveraging 10x on the market rarely works out. That’s because if you are levered 10x and the market drops 10% (relatively common), you will be wiped out.

In this thought experiment by Elm Wealth, the impact of using leverage in the S&P 500 from 1927 to 2020 was evaluated.

Notice that 1.5x leverage leaves you with the most money at the end, but at some point, you’d have experienced a drawdown of 95%.

Above 1.5x leverage, return goes down and risk goes up, driven by the volatility drag we’ve been discussing.

At 4x leverage, after suffering more than a 99.99% drawdown, you’d eventually only be left with the $1 you started with (just $0.09 inflation-adjusted), and at 5x leverage you’d have been fully wiped out on October 19th, 1987.

Most studies recommend up to 2x leverage, and there are few good reasons to go above 3x.

TQQQ during the Dot-com bubble and the Global Financial Crisis

The biggest concern regarding leveraged ETFs is that most of them were not present during the dot-com bubble and the Global Financial Crisis.

For example, TQQQ was launched in 2010.

So what happens if we create a simulation to evaluate how a hypothetical TQQQ portfolio would have performed?

The 3x-leveraged QQQ would have been wiped out twice — once during the dot-com bubble crash and again during the global financial crisis.

Leave a comment

$100 invested in TQQQ at the end of 1999 would have grown ~10x by the top of the dot-com bubble in 2001. But in the crash that followed, the portfolio would have dropped to <$1. While in our simulation, the portfolio does make a comeback, we would argue that, given the transaction and borrowing costs (which we ignored here), a real fund would have been shut down.

This also corroborates our earlier research, which recommends that optimal leverage be <2. Let’s take another example of a real leveraged ETF — SSO (2x leveraged S&P 500), which was launched in 2006 (before the Global Financial Crisis).

$100 invested in SSO at its launch would drop to ~$20 during the Global Financial Crisis. While this drawdown was brutal, it was not a scenario so extreme that the fund had to shut down. During the bull market that followed, the SSO portfolio overtook the S&P 500 by 2014, and by 2025, your investment in SSO would be 2x that of the S&P 500.

Researchers at Vanderbilt University had similar findings.

When they ran 10,000 Monte Carlo Simulations on leveraged ETFs based on the S&P 500, they found that over a 20-year period, a 2x-levered ETF had only a 0.5% probability of shutting down, and a 3x-levered ETF had a 7.6% probability of shutting down.

It’s also interesting to note that if you were 2x short the S&P 500, you had a 94% chance of the fund shutting down over a 20-year investment horizon.

Should you be investing in leveraged ETFs now?

Not if you are looking at individual stocks and especially semiconductors. Not to be “that guy”, but our MS portfolio has nearly doubled YTD without touching leverage.

There is no reason to add additional risk by using 2x or 3x leverage on an already volatile name. That’s how you end up in situations like this:

Finally, you have to be aware that for leveraged ETFs to make money, it’s not enough for the market to go up; it should do so with relatively low volatility.

Volatility Drag

While most investors are worried about the fund shutting down, the real risks with leveraged ETFs lie in the volatility drag. Take our earlier example of the fund's +10% and -10% moves over 2 days. The 3x levered ETF declined 9% compared to the underlying index's 1% drop.

In extreme cases, even if the index is moving up, you will lose money due to volatility. From February 2020 to January 2023, QQQ was up 30.76%. But, TQQQ declined 11.37%. This was due to

1. ~60% drop due to the COVID-19 crash

2. +200% run-up during the everything bubble in 2021

3. Another 50% drawdown during the 2022 rate hikes.

This incredible amount of volatility caused the 3x levered index to move very differently compared to the underlying index.

Leverage — The secret behind Buffett’s returns

Once again, it all comes down to your skill as an investor. While leveraging and investing in speculative stocks rarely works out, you can make leverage work for you if you know what you are doing.

While Buffett eschews debt publicly, a large chunk of his returns can be attributed to the smart use of leverage.

Berkshire Hathaway has realized a Sharpe ratio of 0.76, higher than any other stock or mutual fund with a history of more than 30 years.

In the famous paper Buffett’s Alpha, researchers analyzed Buffett’s portfolio to understand how he was able to generate such a high risk-adjusted return. They found that Buffett used the cheap leverage available to him from his insurance float and invested it in high-quality value stocks to generate market-beating returns.

We estimate that Buffett’s leverage is about 1.6-to-1 on average.

Buffett’s returns appear to be neither luck nor magic, but, rather, reward for the use of leverage combined with a focus on cheap, safe, quality stocks.

The real alpha might be in realizing that very few of us have the patience and the skill of Buffett.

Share

Market Sentiment is now fully reader-supported. A lot of work goes into these articles, and if you enjoyed this piece, PLEASE HIT THE LIKE BUTTON. It’s the only way new readers can discover our reports :)

Footnotes

2

Ignoring borrowing costs and transaction fees. Here is the exact mechanics of return compounding: (source)

• Quarterly revenue: $41.5B (up 340% yoy)

• Guidance for next Quarter: $50B (Estimate was $43B)

• 16 long-term contracts worth $100B+

Since the beginning of the year, we have consistently highlighted that memory has shifted from a cyclical product to a structural need for hyper-scalers, and Micron’s latest earnings prove it.

Why AI needs memory?

Reiterating our demand drivers from our memory bottleneck post,

• Agentic reasoning generates a long sequence of steps before giving the final answer. This means that the AI model should now retain all prior steps in memory so it has context for what it’s expected to do.

• Images, audio, and video generation consume far more memory than mere text generation. For example, to ensure characters are consistent throughout an AI-generated clip, the model must retain the entire frame in memory rather than just a few words.

• Finally, since model providers are pushing universal assistants that know everything about you, they now simultaneously have to keep the current conversation, general information, and old conversations to make sure the experience is seamless, which pushes the memory usage even further.

Nvidia’s own bill of materials tells the same story. The 2025 GB300 NVL72 rack carries about $374K of memory — roughly 9% of its ~$4M total cost. Just a year later, next-gen VR200 packs roughly 3x the memory capacity at roughly 2x the price per GB, pushing memory cost to ~$2M — a 435% jump that brings memory to ~26% of AI server cost.

New demand vectors

In the earnings call, Micron CEO Sanjay indicated memory demand for a robot is 10x that of an autonomous car. And these robots are expected to grow exponentially as the price per unit falls and they can tackle general-purpose tasks.

Powered by these tailwinds, we believe Memory will stay in a supercycle for years. That conviction is shared. Sixteen companies, including two hyperscalers, signed long-term agreements with Micron to lock in supply for three to five years. Traditional memory deals run two to three quarters at most — the shift reflects a long-term supply crunch.

These contracts cover advanced memory products and account for roughly 25% of Micron’s supply volume and 40% of its revenue at current prices. This is bullish for Micron as it secures prices and profit margins for the next three to five years.

The biggest issue with memory stocks was their cyclicality. Just 4 years ago, Micron’s gross margin went from the high 40s to near -31%, with Micron selling memory below cost due to oversupply. These long-term contracts prevent a similar downturn — The largest agreements carry a ceiling at the current CQ2 market price and a floor that holds through the term.

Even at the floor, Micron expects a gross margin well above its prior peak quarterly margin of roughly 61%, which puts the floor north of 65% on our estimate.

How are we positioning?

Read more

or in the “there is no rational reason for you to pay ~100x revenue multiple for a loss-making company.”

But what if there was a middle ground? What if you could trade this IPO without ever buying the trillion-dollar story? The setup is in the mechanics, not the fundamentals, and that's what we are going to find out today. But before we get there, let’s take a closer look at the company and how they make money.

Business Overview

While you can read the full 400-page prospectus here, SpaceX has 3 main business segments:

• Space — sells launch services. Revenue comes from two streams: (i) Launch Services, deploying payloads to orbit for commercial and government customers on the reusable Falcon 9 and Falcon Heavy, and (ii) Launch and Development, building spacecraft and providing launch/mission services for government space programs (Falcon 9, Falcon Heavy, Starship, and Dragon).

• Connectivity — is the Starlink business; revenue comes from broadband and Starlink Mobile service across 156 markets, split into Consumer Broadband, Enterprise Solutions, Government Solutions, and Mobile.

• AI — is the newly acquired xAI/X business. Revenue comes from advertising on the X platform plus AI solutions and infrastructure: subscriptions, data licensing arrangements (access to X's real-time and historical data), and API access to the Grok models.

Looking at the numbers,

• Starlink is the strongest business where revenue grew ~50% ($11B now), operating income more than doubled, and adjusted EBITDA rose 86%.

• Even though revenue improved for launch services, operating profit flipped from $21M profit to $657M loss — the driver here is Starship development, which increased R&D costs from $1.7B to $3B.

• AI revenue grew 22% (arguably small base of $2.6B), but now the operating loss is $6.3B. Once again, R&D is driving losses, as the company spent nearly $5B on training the Grok model and building out AI compute capacity.

This also allows us to build a rough valuation model to assign a value to each division.

SpaceX Model

Here’s how the numbers look:

• Starlink — We modeled subscribers to grow at a 25% CAGR from 9 million in 2025 to 84 million by 2035, a number that only holds if direct-to-cell scales alongside fixed broadband. Blended ARPU falls from $81 to $48 a month, because cheap direct-to-cell plans dilute the premium home base. At a blended 15x multiple, it carries $167 billion in intrinsic value.

• Launch Services — External Falcon 9 is anchored to third-party demand, near the 58 launches seen in 2025, rising into the late 2020s then declining as Starship absorbs the heavier external missions. Assuming a successful transition to Starship, modeled revenue reaches $25B by 2035, bringing the valuation for launch services close to $60B.

• xAI / AI — is the big gamble. As of now, the numbers are based on real contracts, but we are assuming a growth rate of nearly 50% for 10 years. 80% of the 2035 revenue comes from contracts that do not exist yet. Even with this optimistic projection, AI will add only $343B in value.

What’s interesting is that even with the most optimistic projections for each business segment, the company's total value comes to only $568B ($43 per share), a far cry from the asking price of $1.75 trillion ($135/share).

The kicker? If you exclude the AI division, the company's total value is around $220 billion — which is what it nearly traded at in private markets before the IPO hype and the xAI merger.

How to trade the world’s largest IPO?

While by valuation, SpaceX will be listed among the top-10 companies, it will trade like a mid-cap due to its small float. Broadcom, which is similar in size, has $1.7T in tradable float compared to only $75B for SpaceX.

Now that we know that the stock is going to be volatile and priced way above what it should be priced at by any reasonable estimates, here’s how you can trade it (based on whether you got the IPO allocation or not)

Read more

This is a guest post from . We launched Rebound Capital last June as our equity research division after more than 4 years of research at Market Sentiment. Here’s the one year update (and a special offer for Market Sentiment readers who make it till the end)

On June 5, 2025, we launched Rebound Capital. It was the culmination of more than 4 years of research at Market Sentiment, where we evaluated everything from investing in companies more than 100 years old, to economic moats, to tracking insider trades by members of Congress.

There’s this one idea we always came back to: investing in quality companies in a drawdown.

Here’s what we wrote about Meta in 2022:

No matter how you look at the company, the financial metrics are incredible. 80% gross margin and 30% ROE with growing revenue. But the stock is down 55% in the last 1 year.

I am not pitching it as a great business. They certainly have some incredible challenges ahead of them in terms of privacy and attracting a young user base. No one knows directly the solutions for their hardest problems—that’s why they’re the hardest ones.

But writing off a company based on two bad quarterly reports is just too short-sighted.

The same theme kept recurring:

• Netflix lost 3/4th of its value because it did not grow subscribers one quarter.

• ASML lost ~1/2 its market cap on tariffs and export restrictions.

• Amazon was down 56% after its first quarterly loss since 2015.

So we finally pulled the trigger and launched Rebound Capital, putting our own capital on the line. We followed Charlie Munger’s advice of “take a simple idea and take it seriously” and spent the past year going through hundreds of stocks to find the optimal rebound setups.

Fast forward a year, and we are outperforming the S&P 500 while holding top 3 positions in relative value names such as Amazon, Constellation Software, and MercadoLibre. To put this in context, 79% of all active large-cap U.S. equity funds underperformed the S&P 500 last year.

While it’s easy to talk a good game when your portfolio is outperforming, sticking to the strategy when it’s underperforming is the hardest thing to do as a public facing equity research platform. Q1 was hard on the Rebound Portfolio, where we underperformed the market by 9%. We continued to share our portfolio performance publicly throughout the period of underperformance (here are our Feb, Mar, & April updates). But sticking to our strategy helped us double down on quality names that rebounded strongly after the March dip.

Not to name names, but there are a lot of “gurus” who have Excel spreadsheets where the portfolio never goes down, and they magically exit at the right time. For us, we have maintained the policy of informing you of all our trades 24 hours in advance, and honestly, our biggest success till now is that nobody has ever said anything bad about us.

Overall, we are now in a good position with our portfolio compounding faster than the index while trading at essentially the same valuation. The headline numbers (weighted average of our portfolio holdings):

• Revenue growth, next 3 years: ~18%

• EBIT growth, next 3 years: ~19%

• EV/EBIT: ~27x, in line with the S&P 500

Here is the unusual part. We own a basket whose EBIT is set to grow ~19% per year over the next three years, compared with <13% for the S&P 500, and we pay no premium on the multiple to do so.

Divide valuation by growth, and the portfolio sits at ~27/19 = ~1.4, versus 27/11 = ~2.5 for the index (lower is better). In short, our portfolio compounds faster than the market while costing about the same, which on a growth-adjusted basis makes it materially cheaper than the index it is built to beat.

Your support has been incredible, and it’s what makes this work feel meaningful. While most creators are hush-hush about how much they make, we are going to show you our real numbers. After all, this is the easy bit after sharing our portfolio performance :)

As of today, we have more than 21,000 investors who subscribe to us, of which 969 investors pay to read our research, which translates to revenue of $358K (ARR). The spike you see in December is when Michael Burry became a paid subscriber to Rebound Capital and endorsed our note on Birkenstock.

Consider this your nudge to put your own work out in public. You never know who’s reading.

Finally, of all the metrics at Rebound Capital, our retention rate is the one that we are most proud of. ~90% retention at 6 months is unheard of in the newsletter space and is the best testament to the quality of the research we produce and the value you derive from it.

Right now, the market is in love with AI. Money is pouring into a narrow group of names at prices that already reflect the next 5 years of growth. To be clear, we are bullish on what AI can do over the long run. But when an equipment vendor that has always been cyclical gets priced as if the cycle has ended, the reward for chasing it is small, and the cost of being wrong is high. With many indexes changing their decade-old rules to accommodate hyped-up IPOs like SpaceX, indexing is no longer the safe, neutral choice. The case for doing your own research only gets stronger from here.

And this is where being independent matters. A fund manager has to answer to a committee, hug the benchmark, and own the hot name so they don’t look wrong at the next quarterly review. We answer to no one but you. We never have to chase the story of the moment and can sit in an out-of-favor name and wait for the market to catch up, because no one is going to pull our capital for being early.

While investors flock to one part of the market, high-quality companies are being ignored. Some of the best compounders we know are at their lowest multiples in years. We will continue to focus here. The crowd overpays for what it loves and ignores the good businesses on sale. But the market eventually catches on.

So we will keep researching the names nobody wants to talk about. The rotation back to high-quality names will come. Our job is to be patient enough to hold the right names when it arrives.

Anniversary Offer from Rebound Capital

To celebrate one year of Rebound Capital, we are offering 33% off to all current subscribers of Market Sentiment. This offer is valid only for the next 24 hours and if you are an existing reader of Market Sentiment.

Subscribe to Rebound Capital

With how the markets are right now, we are right in the middle of the good old days. In case you missed the everything bubble of 2021, this is right about how it felt back then.

Every thesis was proven in days, and random companies were doubling up every week. Hell, we were even trading pictures of rocks at the same value as that of an actual waterfront property in Florida.

While we are not that crazy yet (as the price rise is somewhat supported by earnings improvement), when we did our last portfolio update a month back, the MS portfolio was up 33% YTD.

Today, we are at ~78%!

The problem is that even though demand is real, there's a lot of crowding into semiconductor stocks. "Long Semis" is now the most crowded trade, and investors are already discounting multiple years of strong growth.

As Rebound Capital highlighted, there is literally panic buying in semiconductor stocks.

AMD is up roughly 330% over the trailing 12 months and now trades at 50x forward earnings, compared with Nvidia at 24x. The semicap names have rerated even harder: ASML at roughly 52x trailing earnings versus 38x at the end of 2024, Lam Research at 60x against a ten-year average of 20x, and Applied Materials at 42x against a ten-year average of 18x. These are cyclical equipment vendors being valued as if the cycle no longer exists.

The memory and optics complex is where the melt-up is cleanest. SK Hynix is up about 248% year to date and Micron 226%, with both joining Samsung in the trillion-dollar club. In optics, Applied Optoelectronics is up roughly 335% YTD, Lumentum 120%, and Coherent close to 80%. Some of the other Tier 2/3 suppliers are up similarly.

With the bad news out of the way, here’s what’s happening with the fundamentals:

Demand for semiconductors is nothing like what we’ve seen before. Until now, investors expected that we were at peak demand, with hyperscalers reinvesting almost all of their cash flow into AI data centers. With Google raising $80B from stock sales to fund its AI buildout, there is now theoretically no upper limit on the money that can be pumped into the ecosystem.

Still, even with the parabolic increase in revenue and profits, we're slightly jittery as the stock price tends to drop well before the slowdown data comes through in earnings reports. Based on the previous cycles, by the time the estimate cuts arrived, IT stocks had already fallen an average of nearly 30% — and individual names like Intel, Nvidia, and Micron fell considerably more.

With this level of crowding and hype, any hiccup, however temporary, will lead to a sharp sell-off. The optimal strategy here would be to take a step back, dig into names and sectors outside the current hype cycle, and invest at more palatable valuations.

Read more