On (ONON) Signs Mbappé as Nike Leaves Another Door Open

Nike (NKE) left a door open in running, and Swiss brand On (ONON) took its place on store shelves.

Nike left a door open in tennis, and Roger Federer became an On co-owner.

Last week, Nike left a door open in football, and Kylian Mbappé signed with On.

Futbol countries noticed.

Over three days after Mbappé signed, searches for On ran more than three times their usual pace in Italy, France, Spain and Brazil.

Running made On a brand. Football can make it a global phenom…if its products are great (we think they are).

Five billion people follow football, by FIFA's count, more than half of everyone alive. In South America, Africa and Southeast Asia, On's co-chief executive told investors this week, football is "not a category but culture."

Its fans spend $25 billion to $29 billion a year on football shoes, by two research firms' estimates, and Nike and Adidas (ADDYY) collect an estimated 70% of that money.

On is going after that money with Kylian Mbappé, winner of two World Cup Golden Boots, a feat no player managed before him. And it signed him just as it signed Roger Federer: as an owner.

Federer bought into On in 2019 and helped build its tennis shoes. Mbappé now owns a piece of On too, and he will help build its first football boots, due in 2027, with France great Thierry Henry running On's football business.

Most boots On sells will come out of Nike's and Adidas's share.

We’ve seen this play out for ONON before – in fact Nike has a long history of underestimating the underdog…

Read the full article here…

Founders Call: A Massive AI Tipping Point

ICYMI — Check out 3 stocks to watch as META pushes Agentic AI to mainstream consumers…

Infinite Hold Updates

Tesla’s (TSLA) Next Growth Engine Is Hitting the Road 

Tesla just gave us a fresh set of clues about where its next leg of growth could come from. 

Shoppers are coming back. 

And Tesla is building something much bigger on top of that returning demand. Full Self-Driving adoption is climbing, while its young robotaxi business is starting to put real numbers – and real riders – behind a story investors have heard about for years. 

The latest data gives us a much clearer look at where Tesla could go from here. Take a look... 

Tesla Shoppers Are Coming Back 

Let’s start with the business paying the bills today: Tesla’s EVs. 

Over the four weeks through Sept. 17, Tesla consumer demand ran 9% above year-ago levels. Last November, it was 26% below. 

That’s a sharp turnaround. 

And consumers are making it all the way to Tesla’s order page. 

The base Model 3 and Model Y L Launch Series are sold out for 2026 in the U.S. The Model Y RWD and AWD aren’t far behind. 

But car demand is only part of what caught our attention. 

Searches for Full Self-Driving jumped 50% over the 12 months through mid-September – faster than searches for Tesla’s cars. 

Consumers are showing more interest in the technology Tesla wants to build its future around. 

And Tesla already has one enormous advantage there. 

Tesla’s Robotaxi Fleet Is Starting to Scale 

A year ago, Tesla’s robotaxi business was barely a fleet. It had about 10 cars on the road. 

Today, it has 621 – roughly 62 times as many. 

Google-owned Waymo (GOOGL) still operates about seven times as many robotaxis. Tesla has plenty of ground left to cover. 

But Tesla knows how to build cars at scale. 

The latest estimates put the cost to build a Cybercab at roughly $30,000, compared with $150,000 or more for a Waymo robotaxi. 

Tesla is expanding geographically, too. 

It now runs without a safety monitor in Austin, Dallas, and Houston. In July, it reached Miami, Orlando, and Tampa – its first unsupervised service outside Texas. 

More cars on more roads gives us something we’ve been waiting for: Real customers. 

The fleet is still small. The speed of the buildout is what has our attention. 

And as more cars hit the road, we get something even more useful for our research: more real customers telling us what they think. 

Tesla Is Taking on Uber – and Riders Declare a Clear Winner

Tesla isn’t just competing with Waymo for autonomous-driving leadership. 

It’s also going after the millions of people who already use Uber (UBER) to get from place to place. 

That makes Uber an important benchmark. Tesla has to do more than prove a car can drive itself. It has to give riders an experience they’ll choose over the service already sitting on their phones. 

We’re starting to get that comparison. 

Cybercab began carrying paying riders in Austin on Sept. 4. 

One feature keeps grabbing their attention: There’s no steering wheel. 

Nearly half of Cybercab posts on X mention it. Those posts run positive over negative by about 7 to 1. 

And Cybercab is starting to compete for attention beyond the people who have ridden one. 

During one week in early September, U.S. search interest in Cybercab surpassed search interest in Uber’s own brand. 

Search interest gets people in the door. The experience determines whether they come back. 

Tesla Robotaxi and Waymo both average about 4.4 stars in App Store reviews. Uber and Lyft sit near 3.7. 

The one-star reviews make the difference even clearer. 

About one in four Uber reviewers gives the app one star. For Tesla Robotaxi, it’s one in eight. 

We see another big difference when riders talk about safety. About 39% of Uber posts we track describe feeling unsafe, versus 10% of Robotaxi posts. 

These are still early days for Tesla’s robotaxi business. But this gives us something far more useful than promises about what Cybercab might become. 

We can watch consumers make the comparison themselves. 

What It Means for Our TSLA Position 

Step back and look at how much has changed. 

Last November, Tesla consumer demand was running 26% below the prior year. Now it’s 9% above.  

Tesla had about 10 robotaxis a year ago. Now it has 621. 

And the Full Self-Driving fleet has logged 14.6 billion miles of real-world driving. 

 Those numbers show us a company with renewed consumer demand today and a much larger opportunity taking shape around autonomous driving. 

The business keeps expanding beyond the car – and the consumer signals we can measure remain strong. 

TSLA is an Infinite Hold. We have no plans to let go. 

Amazon (AMZN) and Shopify (SHOP) Pick Different Paths on AI Shopping 

Meta Platforms’ (META) new Muse AI agent can do more than recommend what to buy. 

It can navigate online stores, pick out products, and complete a purchase for you. And it quickly climbed to No. 1 among free apps after launching Sept. 8. 

Now Amazon and Shopify are taking very different approaches to it. 

Amazon began blocking Muse from its retail site this week. It already blocks other shopping agents, including ChatGPT, Gemini, and Perplexity’s Comet.

The stakes are especially high for Amazon’s ad business.

Amazon brought in $76 billion in advertising revenue over the 12 months through June. Sponsored listings work because shoppers visit Amazon and scroll through products.

 An AI agent can skip much of that. Tell Muse what you want, and it can go hunting for the product itself.

Shopify is going the other direction. 

It opened its stores to Muse and will let shoppers pay through Shop Pay. That checkout system passed $400 billion in lifetime accelerated gross merchandise volume in June. 

Shopify doesn’t have to convince people to start their shopping on Shopify.com. Its merchants can sell wherever customers show up, while Shopify powers the transaction behind the scenes. 

SHOP jumped 7% Monday after the partnership was announced, then gained nearly 8% more Tuesday. 

For us, the consumer behavior from here is more interesting than the two-day stock move. 

If people are willing to hand more of their shopping over to AI agents, Amazon could face a new challenge to the way it attracts shoppers and sells ads. Shopify could gain another way to funnel transactions through its platform. 

We own both stocks – AMZN as an Infinite Hold, and SHOP in the Core Conviction portfolio. As AI agents start doing more of the shopping themselves, we’ll be watching where consumers actually spend their money. 

Google (GOOGL) Is Helping Squeeze More Power From Two Nuclear Plants

Google needs more electricity for its growing AI infrastructure. Now it’s helping Georgia Power get more out of two nuclear plants already on the grid.

The companies signed an agreement to support upgrades at Georgia’s Vogtle and Hatch nuclear plants that could add about 96 megawatts of generating capacity.

Instead of building new reactors, Georgia Power plans to upgrade equipment including turbines, pumps, motors, and cooling systems. The goal is to produce more electricity from the nuclear plants it already operates.

Google will subscribe to that added capacity. The deal still needs approval from the Georgia Public Service Commission.

AI companies have poured billions into chips and data centers. But none of those investments work without enough electricity to keep the machines running.

Google isn’t waiting around for that power to show up. It’s helping fund upgrades that can squeeze another 96 megawatts out of existing nuclear plants.

That tells us where the AI buildout is headed next: Google has the chips. Now it’s going straight to the power plants to make sure it can run them.