Mid-Week News (January 8th, 2025)
Photo by Markus Winkler / Unsplash

Mid-Week News (January 8th, 2025)

15 min read

Table of Contents

1. Uber (UBER) – Confidence, Nvidia & More

a. Nvidia

I will cover Nvidia CEO Jensen Huang’s Consumer Electronics Show (CES) keynote on Saturday. It was not all that eventful, but there were some things to mention.

Uber and Nvidia announced a new “collaboration” this week. The two will work together to expedite development of autonomous vehicle (AV) technology for the industry. That doesn’t really sound like a company in fear of this inevitable transition. Uber will provide its world-class mobility data to bolster the training depth and timelines for Nvidia’s new Cosmos platform. Cosmos offers a managed environment for developers to access and customize models and tokens. It’s purpose-built for “physical AI” use cases such as AVs or Tesla’s Optimus bot; it helps companies access world-class assets with greater efficiency and lower cost than doing so alone. Cosmos is a somewhat similar idea to hyperscalers renting out managed computing capacity to customers… but this time for physical AI. The partnership also involves NVDA’s DGX Cloud platform, which includes the company's latest and greatest chips, networking equipment and software. Per the release, there are “more details coming later in the year.”

With the Waymo partnership already secured for Uber, this is the most encouraging new relationship Uber could announce, aside from Tesla or maybe Amazon’s Zoox. Nvidia is providing the foundation of essentially every major AV program. So? If this highly capable company sees value in partnering with the current mobility leader to expedite development, that bodes well for Uber’s perceived value proposition.

This is not an “all clear” for me to resume adding shares like a formal Tesla partnership would be. Still, it is a positive step in rebuilding my stronger conviction in this name to eventually justify making it a larger piece of my portfolio.

b. Confidence

Uber signed an accelerated share repurchase agreement with Bank of America to expedite $1.5 billion in buybacks. This is part of its current $7 billion buyback plan, rather than in addition to. The move is a byproduct of confidence in the underlying business. Here’s what CFO Prashanth Mahendra-Rajah had to say about it:

“We are entering 2025 with considerable momentum and expect to continue scaling our free cash flows significantly, enabling us to return meaningful capital to shareholders while still investing in growth. Our stock is undervalued relative to the strength of our business.”

CFO Prashanth Mahendra-Rajah

My increased level of caution had (and has) nothing to do with 2024 or 2025 results and had nothing to do with Uber’s current financial execution at a highly compelling price. What was it based on? Uber is moving from a guaranteed market share king of ride-hailing and food delivery, to a future where that dominance is less certain. I’ll keep saying it.

I am adamant that Uber will be a large piece of the AV transition, as its drivers can supplement fleets so providers like Tesla and Waymo don’t need to overbuild from the start. Overbuilding is quite expensive for hardware like this and Uber can perfectly complement these fleets as they grow and mature. Uber CEO Dara Khosrowshahi sees it taking 10 years before 50% of Uber’s rides are autonomous; regulatory obstacles and technological hurdles such as making these cars work in extreme weather will take more time.

But what happens once that process is complete? What happens when the majority of rides in the USA and across the globe are unmanned? That’s where the uncertainty and added risk comes from here. And that’s why sentiment has turned sour for a fundamental darling. Uber is the most powerful demand aggregator on the planet in this business by two country miles. It is perfectly positioned to help professional fleets and individual cars plug into massive, stable, visible demand to raise occupancy rates and operational profitability. It’s also extremely well-equipped to take the headache out of tedious fleet management. The firm boasts routing algorithms trained on more data and time than anyone else has, and professional support (internally and through partners) for other tasks like fleet cleaning and maintenance. That’s why Waymo is exclusively partnering with Uber in Austin and Atlanta to only offer rides through that app. And while Waymo is experimenting with other business models in Miami, they’re doing so with Moove… a company that Uber owns a sizable stake in with multiple board seats.

And while all of this is encouraging, there is one risk to Uber claiming a large piece of this pie that has forced me to lighten up. That risk is one vendor, such as Waymo, Zoox or Tesla, owning the entire market. Controlling most of a ride-sharing duopoly today and the coinciding network effect is somewhat meaningless when you’re partnering with other monopolies. They can easily build their own network, based on having a product offering nobody else can match. Uber’s long term niche and value is somewhat reliant on more perfect competition unfolding in the AV race. That’s how it can become the Expedia-like demand aggregator for all AV fleets. If multiple vendors succeed here, then those vendors choosing to operate on their own will be at an inherent disadvantage vs. the others that use this reliable traffic. 

I find the monopoly or duopoly outcome to be unlikely, as many others like May Mobility and Avride are also making rapid progress in this field. They will probably also debut in U.S. cities this year. And there are more. Still, I’m not certain, as the world has never seen a transition like this one. Anyone telling you that they are 100% confident is being overzealous. I will continue to hold what I own, respect this risk, and use hints like partnerships and technological advancements (outside of Google and Tesla) as reasons to grow more confident on this name once more.

Today, 3% of holdings makes sense to me.

c. New AV-Based App Features

The Wall Street Journal published an article this week on the work Uber and Lyft (LYFT) are doing to cater to AV operators.

Per the article, Uber is developing a few new capabilities. All tools, some infrastructure-based and some feature-based, are meant to make Uber a better partner for future AV fleets. On the infrastructure side, the firm is expanding its car storage footprint for large fleets. It’s also equipping these facilities with needed data processing capacity (10x typical scale) and chargers. Furthermore, Uber is also now teaching employees how to maintain the stored hardware. Per Uber SVP of mobility Andrew Macdonald, “this level of nitty-gritty takes years to build.” For app features, it will debut tools to make autonomous cars honk on command, as well as open trunks on command. 

While I think maximizing utilization rates will always be the largest value driver for Uber with AV fleets, that doesn’t mean it cannot complement that utility with smaller add-on offerings to entice vendors. That’s what I see this news as.