LT Weekly: Mar 13, 2026

As we finally reached the end of the Q4 earnings season across the sector, we looked across our LionTree TMT & Consumer Universe of ~200 companies covering 37 sub-sectors with market caps of $1bn+ to evaluate financial performance compared to Wall Street projections.

Overall, most companies in our Universe beat Q4 expectations on both sales and profitability but an even greater number beat on sales and a fewer number beat on profitability versus Q3. More specifically, 79% of companies topped consensus revenue estimates in Q4, which UP from 74% in Q3, and 74% of companies beat consensus profitability projections in Q4, but this was fewer than the 80% that did so in Q3.

A bigger deal, however, was that guidance (where provided) was largely disappointing, especially as it related to Q1 trends. More specifically, Q1 sales guidance for 80% of companies in our analysis that provided that metric was below consensus expectations and profitability guidance for 81% of companies that provided that metric missed Street projections. Those are big percentages.

What was the Street’s reaction to these results and guidance? It was skewed negative with only 44% of companies trading up in reaction to results and 56% trading down.

See below for more on Q4 financial performance of companies within our LionTree Universe of TMT and Consumer stocks relative to Street expectations, along with their respective stock reactions.

An Even Greater # Of Companies BEAT Revenue Expectations In Q4 vs Q3

• Overall, 79% of Cos in our LionTree Universe beat cons on revenue in Q4(up from ~74% in Q3)  
  • 4% of those companies beat by double digits, up from 1% in Q3
  • The avg sales beat at 3% was the same as in Q3

• FuboTV, Omnicom & Hasbro had the largest % beats on sales in Q4

• Overall, 21% of companies in our Universe reported sales in-line to lower than consensus, down from 26% in Q3 
• Oscar Health, Lyft & Robinhood reported the largest % misses on sales  

The Vast Majority Of Companies Also BEAT On Profitability, But It Was To A Lesser Degree In Q4 vs Q3

• Overall, 74% of companies in our Universe beat consensus on adj. EBITDA in Q4 (vs 80% in Q3) and 25% posted double-digit beats (down from 34% in Q3)  
  • However, the avg adj EBITDA beat at 8% was the same as in Q3

• iQIYI, Snowflake & Hasbro had the largest % beats on adj EBITDA in Q4  

• Overall, ~26% of the companies in our Universe were in-line or missed consensus on adj. EBITDA in Q4, up from ~16% in Q3
• Liberty Braves Group, WEBTOON & JD.com reported the largest % misses on adj EBITDA in Q4

Forward Guidance (Esp Q1) Was Largely Below Expectations

• Q1 2026 guidance (where provided) was vastly below expectations: Sales guidance for 80% of companies in our analysis that provided Q1 sales guidance was below consensus expectations and profitability guidance for 81% of companies that provided profitability guidance missed Street projections
  • Opendoor’s sales guidance was the highest vs Street projections (+56% above) while Sonos’ profitability guidance was the highest vs Street expectations (+121% above)
  • WEBTOON’s sales and profitability guidance were both the furthest below consensus by -3% and -19% respectively

• FY 2026 guidance (where provided) was generally below projections, esp on revenue: Sales guidance for 73% of companies was below consensus expectations while profitability guidance for 53% of companies was below projections
  • Hims & Hers’s sales guidance was the highest vs Street projections (+2% above) while Clover Health’s profitability guidance was the highest vs Street expectations (+14% above)
  • Liberty Global’s sales guidance was the furthest below consensus (by -6%) while Oscar Health’s profitability guidance was the furthest below projections (by -62%)

As Such, Stock Performance Was Skewed Negative With Only 44% Of Companies Trading Up & 56% Of Companies Trading Down In Reaction To Results, With High Levels Of Volatility

• More Cos traded down double-digits vs traded up double-digits in reaction to earnings
  • 34% of Cos traded down double-digits, significantly up from 13% in Q3
  • Similarly, 25% of Cos traded up double-digits, also up from 15% in Q3

Top 3 Best & Worst Performers In Reaction To Q4 Results… 

The software sector continues to be under the microscope given the big picture AI disruption concerns which have decimated stocks in the sector. However, updates out of Oracle and Adobe this week somewhat diverged. Oracle’s stronger than expected revenue and adj EPS, seq acceleration in cloud growth, higher FY27 revenue guidance, and a multitude of supportive AI demand commentary was encouraging for the sector. However, while Adobe also beat on revenue and adj EPS, its ARR growth slowed sequentially (its freemium offerings is creating near term pressure), its traditional Stock revenue is being negatively impacted as creators increasingly adopt generative AI, AND Shantanu Narayen stepping down as CEO came as a negative surprise. Overall, Adobe’s results underscored the Co’s continued challenge in balancing growth, monetization, and AI-driven change. So overall, it was a mixed bag this week in the software sector.

See below for more of our takeaways from their results.

Oracle Posts Strong Results and Outlook, Calming AI Infrastructure Concerns

• FQ3 topped expectations
  • Rev and EPS were ahead of the Street, though RPOs were a tad light
  • Cloud revs grew +44% y/y, an acceleration from +34% y/y in Q3

• FQ4 guidance was in-line to better…FQ4 rev growth guidance was ~in-line with expectations at the mid-pt, while EPS was ahead
  • Rev growth guidance of +19-21% y/y vs cons +20.2%
  • EPS $1.96-2.00 vs cons $1.93

• Mgmt maintained FY26 rev guidance ($67bn) BUT raised FY27 rev guidance from $86->96bn… “Market dynamics enable Oracle to comfortably meet and likely exceed our revenue growth rate forecast for FY27 and beyond”

• FQ4 CapEx was ~+$5bn higher than expected BUT maintained FY26 guide
  • Continue to expect FY26 CapEx of $50bn

• Demand for cloud computing continues to grow faster than supply + some of their largest consumers of AI Cloud capacity have recently strengthened their financial positions “quite substantially”
  • 90% of that committed capacity was delivered on or ahead of schedule in FQ3
• The Co is restructuring product development teams as AI-assisted coding drives efficiency: New AI Code Generation technology is enabling them to build more software in less time w/ fewer people, making their SaaS application suites “more competitive and more profitable”

-> As a reminder, it was reported last week that Oracle is planning to cut “thousands” of jobs across the Co; The Co disclosed $1.6bn in expected restructuring costs in the fiscal year through May, its largest such plan on record (link)

• Regarding fears that AI is hurting incumbent software Cos – “the SaaSpocalypse applies to others, but not to us”
  • “We have these coding tools now that allow us to build a comprehensive set of software, agent-based software to automate a complete ecosystem like healthcare or financial services… That’s why we think we’re a disruptor”
  • “I do think that AI tools and their coding capabilities would be a threat if we weren’t adopting them, but we are, and very rapidly…We are building brand new SaaS products using AI and also embedding AI agents right into our existing applications suites”
  • “Some smaller or single-focused SaaS players may well be disrupted, but Oracle will not be among them”

-> Oracle shares were up as much as +15% after its report, the stock’s biggest intraday gain in six months, and ended the day up +9%; But the stock is still down -53% since it hit an all-time high on Sept 10^th and is down -20% YTD

Adobe’s FQ1 Headline Beats Were Overshadowed By A Seq Decel In ARR & A CEO Transition Announcement

• FQ1 headline #s beat BUT ARR was disappointing
  • Overall rev was +1.9% ahead of cons, primarily due to stronger than expected performance in Subscription (though Product revs also beat), partially offset by weaker than expected performance in Svs & Other
  • Op income beat cons by +3.1% and margin of 47.4% was ahead of cons 47.0%
  • EPS of $6.06 was also ahead cons $5.87
  • Total ARR was a tad below and decel’d from +11.5% y/y growth in FQ4 to +10.9% y/y in FQ1

• FQ2 guidance easily beat expectations but the re-affirmation of FY26 guidance implied subscription revenue deceleration
  • Q2 rev guidance of $6.43-6.48bn beat cons $6.43bn at the midpt by +0.5%
  • Q2 EPS guidance of $5.80-5.85 was well ahead of cons $5.68
  • The reaffirmation of FY26 guidance implied decelerating subscription rev growth through the yr, which was seen as a negative by some analysts

• The CEO transition announcement was a BIG surprise: Shantanu Narayen, who has served as CEO of Adobe for 18 yrs, has decided to transition from his role after a successor has been appointed; but will remain as Chair of the Board
  • The Board will consider both internal and external candidates and suspects that “it’ll take a few months” to find a successor

• Notably, ARR from AI-first products more than tripled… “that should be our next billion-dollar business”
• More AI integrations are in the works: The Co launched Acrobat and Express for ChatGPT in Q1, which “significantly” expanded the reach of Adobe’s creativity and productivity workflows, and expect to launch similar integrations into Copilot, Claude and Gemini
• Freemium growth is driving user growth BUT creating near-term ARR pressure: Creative freemium MAU grew +50% y/y and crossed 80mn (accel from 30%+ y/y to surpass 70mn last qtr), though the shift toward freemium is weighing on ARR in the near term as monetization lags initial user adoption
• Saw a “greater-than-anticipated” decline in standalone Stock biz as creators increasingly turn to GenAI tools to create custom images and content instead of licensing pre-made assets from Adobe Stock

• No change in M&A strategy – “we continue to look, but we won’t be cavalier”: “We’ve got an organic engine that we’re pleased with…but when we see something that’s interesting and attractive, we will absolutely go out and action it”

-> Adobe shares fell -7.6% post its report and ended the week down -12.1%; The stock is still down -29% YTD   

Given all the focus on macro lately we wanted to highlight Business Roundtable’s Q1 2026 CEO Economic Outlook Survey that was released this week. It is a composite index of CEO plans for 1) capital spending, 2) employment and 3) expectations for sales over the next six months. In a somewhat surprising result, the overall Index INCREASED by +9pts from last qtr to 89, ABOVE its long-term historic average of 83 and the highest in more than a year.

While sales and capex expectations continued to increase, the employment picture remains a concern, with as many CEOs planning to reduce employment as increase it. That being said, plans for hiring did still rise by +9pts, going from contraction levels in Q4 to neutral levels in Q1. Note that values above 50 signal economic growth, while values below 50 signal economic contraction.

Date wise, this survey was conducted from February 23 through March 6, 2026. For context, the U.S. Supreme Court decision on the International Emergency Economic Powers Act tariffs was on February 20, and the ongoing military operation in Iran began on February 28. In total, 169 CEOs completed the survey.

See below for the key points – and link to report.

• Expectations for sales INCREASED +6pts to a value of 126 (signaling further growth from Q4 to Q1)

• Plans for hiring INCREASED +9pts to a value of 50 (going from contraction in Q4 to neutral in Q1)

• Plans for capital investment INCREASED +12pts to a value of 91 (signaling further growth from Q4 to Q1)

-> Also related to macro, on Friday, U.S. economic growth for Q4 was revised down to 0.7%, according to the Bureau of Economic Analysis’ second estimate, below the prior advance estimate of 1.4%; This followed a 4.4% increase in Q3; The downgrade reflected weaker consumer spending, exports, government spending, and investment, with the partial government shutdown also weighing on activity (link/link)

Important AI updates and innovation, as has been the case, also dominated headlines. Of particular note, centering around model updates. Anthropic introduced a new code review tool designed to help engineers manage the “flood of AI-generated code,” as developer output has surged and code review has become a bottleneck. At the same time, Meta delayed the release of its Avocado AI model after internal testing showed it trailing systems from Google, OpenAI and Anthropic in key areas like reasoning and programming. Tencent is reportedly developing a top-secret AI agent for WeChat that could connect directly with the platform’s mini-program ecosystem to complete tasks such as ride-hailing and food delivery. Lastly, Researchers also uncovered unexpected behavior from an experimental Alibaba-affiliated AI agent, ROME, which engaged in unauthorized cryptocurrency mining (!). Lastly, Nvidia is planning a major $26bn investment to build open-source AI models, aiming to give researchers and startups more accessible systems while potentially optimizing them to run more efficiently on Nvidia hardware.

See more details below…

Key New AI Model Announcements From This Week

• Anthropic launched a code review tool to check the “flood of AI-generated code” (link/link/link)
  • What was the issue? Code output per Anthropic engineer has grown ~200% in the last year
    • Code review has become a bottleneck
  • How did they fix it? When a “Pull Request” or PR is opened, Code Review dispatches a team of agents
    • The agents look for bugs, verify bugs to filter out false positives, and rank bugs by severity
    • The result lands on the PR as a single high-signal overview comment, plus in-line comments for specific bugs
    • Large or complex changes get more agents and a deeper read; Trivial ones get a light pass
    • Based on testing, the average review takes ~20 mins
  • Cost: Reviews are billed on token usage and generally avg $15–25, scaling with PR size and complexity
  • Code Review is available now as a research preview in beta for Team and Enterprise plans

• Meta delays their Avocado AI model (link/link)
  • The Co pushed the release to at least May from a planned debut this month, The New York Times reported on Thursday (Mar. 12)
    • The delay follows internal testing that showed the model trailing leading systems from Google, OpenAI and Anthropic in key areas, including logical reasoning, programming and writing
  • Avocado outperformed Meta’s previous generation models and some earlier competing systems, The Information reported Feb. 4
    • BUT it failed to match the performance of Google’s newest Gemini models, per the NYT report
  • The gap matters because Avocado was not intended as an incremental upgrade
    • The model was designed to compete directly with frontier systems from OpenAI and Google and to serve as the centerpiece of Meta’s next phase of AI development

-> Meta stock was down -4% in reaction to the news and closed -5% on the week

• Tencent is said to be developing a top-secret AI agent for WeChat (link/link)
  • The Co is reportedly developing an AI agent for its WeChat app, according to The Information
    • The project is reportedly classified internally as top secret
  • It’s designed to connect with WeChat’s mini program ecosystem, allowing users to complete tasks such as ride-hailing and food delivery through AI-driven actions
  • The Co is said to be targeting a gray-box test around midyear, followed by a broader rollout in Q3 if the feature is ready

• Lastly, an Alibaba-affiliated research team discovered that its AI agent, ROME, engaged in unauthorized cryptocurrency mining (link/link)
  • The researchers, who were building a new AI agent called ROME, said they found “unanticipated” and spontaneous behaviors emerge “without any explicit instruction and, more troublingly, outside the bounds of the intended sandbox”
    • The agent also made a “reverse SSH tunnel” – essentially opening a hidden back door from the inside of the system to an outside computer, the study said
    • “Notably, these events were not triggered by prompts requesting tunneling or mining,” the report said
  • In response: The researchers added tighter restrictions for the model and improved its training process to stop unsafe behavior from happening again

• Nvidia is spending $26bn to build OpenAI models (link/link)
  • Over the next 5 years, Nvidia plans to invest $26bn to develop open-source AI models, according to a report that first appeared in Wired and later confirmed by Nvidia execs
  • Why open-source models? They are free and available to anyone who wants to use them
    • They give researchers, data scientists, and start-ups a jumping-off point to modify and build AI systems tailored to their own needs
  • Why does it matter? Many of the foremost models in the U.S. are proprietary and not accessible to the public
    • For example, OpenAI’s GPT-5.4, Alphabet’s Google Gemini 3.1 Pro, and Anthropic’s Claude Opus 4.6 are premium models that require a paid subscription
    • The result is that many researchers and data scientists are building their AI systems on top of open-source models
    • Many of the freely available models were developed in China
  • If Nvidia were to develop its own open-source models, it could sync them to perform better with its hardware, providing a more unified experience

Andreessen Horowitz released an interesting report detailing the top 100 genAI consumer apps based on data from Similarweb, Sensor Tower and Yipit Data. What was not surprising was that ChatGPT was the most popular AI product across both web and mobile devices. More interesting was by how much Gemini and Claude are starting to take usage away from ChatGPT. Gemini’s paid subs are up +258% y/y and Claude’s paid subscribers are up over +200% y/y. Also, some apps that have heavily integrated AI into their product experience quickly moved up the ranks. CapCut and Canva ranked #3 out of AI Web Apps and #2 for AI Mobile App, respectively. Globally, it was notable to see that at the top of the list for AI adoption per capita was Singapore, followed by the UAE, Hong Kong and South Korea. The US ranked 20^th.

The data shows that image generation still dominates creative AI tool platforms, but it is now harder for apps to stand out given that as native image models within ChatGPT (GPT Image 1.5) and Gemini (nano banana) improve, the bar for standalone image products rises sharply. Midjourney ranked in the top 10 in Andreessen Horowitz’s last edition and has since fallen to #46. The products that remain competitive (Leonardo, Ideogram, CivitAI) tend to serve specific creative communities. Lastly, “Horizontal Agents” (general-purpose AI systems designed to function across various tasks, departments, and industries, rather than being limited to one specific function) are gaining a lot of popularity, with Manus and Genspark standing out, and OpenClaw making a more recent splash on the scene.

See more of what we wanted to flag from the report below and to see the full report, click HERE

Apps Where AI Is Central To The Product Have Jumped Onto The Top AI Apps List

• ChatGPT unsurprisingly was #1 on both the top web & mobile app lists but more interestingly new players like CapCut, Canva, and Notion where AI has become a core part of the product experience jumped onto the top 10 lists (as of Jan 2026 data)
  • On the web: ChatGPT, Canva & DeepSeek were the top 3 GenAI products
  • On mobile: ChatGPT, CapCut & Gemini came in as the top 3 GenAI apps as of Jan 2026

Source: Andreessen Horowitz

Source: Andreessen Horowitz

While ChatGPT Still Leads The AI Race, Its Dominance Is Being Chipped Away…

• ChatGPT is still by far the largest consumer AI product…
  • On web, monthly traffic is 2.7x larger than the #2, which is Gemini
  • On mobile, MAUS are 2.5x larger than Gemini
  • ChatGPT has seen weekly active users grow by 500mn people over the past year to 900mn today
  • ~10% of the global population utilizes ChatGPT every week

Source: Andreessen Horowitz

• …BUT the category is starting to widening, with other horizontal platforms picking up share
  • Both Gemini and Claude have seen an acceleration in their US paid sub growth over the past year
    • Though still dwarfed by ChatGPT, which is 8x larger than Claude and 4x larger than Gemini on this metric
  • Per Yipit Data (as of Jan 2026), Claude was growing paid subs by over 200% y/y, while Gemini was growing it by 258% y/y
  • Additionally, ~20% of weekly ChatGPT web users also use Gemini each week
  • Sessions per user per month are climbing for Gemini on web
    • Though they’re still 1.3x higher on ChatGPT

• ChatGPT and Claude have both launched connector ecosystems BUT they are vastly different
  • As of late Feb., ChatGPT’s includes 220 apps across 13 categories while Claude has ~160 curated connectors plus ~50 community-built MCP servers
    • The two share 41 apps in common
  • Beyond that overlap, the platforms diverge almost completely:
    • ChatGPT has 85+ apps across Travel, Shopping, Food, Health & Wellness, Lifestyle, and Entertainment – categories where Claude has virtually zero
    • Claude’s integrations skew more towards business such as financial data (PitchBook, FactSet, Moody’s, MSCI), developer infrastructure (Sentry, Supabase, Snowflake, Databricks), science and medical tools (PubMed, Clinical Trials, Benchling), and a growing open-source MCP community that has no ChatGPT equivalent

Interesting Global Observations

• ChatGPT, Claude, Gemini, and Perplexity all draw their top markets from the same pool: The US, India, Brazil, the UK, and Indonesia, in varying order…
• …BUT none have meaningful usage in China or Russia given sanctions have restricted access to US AI tools in Russia since 2022, and China requires AI providers to register, host data onshore, and comply with censorship rules
  • DeepSeek is closely followed by Bytedance’s Doubao regarding usage penetration in China

Source: Andreessen Horowitz

• On an AI consumption per capita basis, Singapore actually ranks first, followed by the UAE, Hong Kong, and South Korea
  • The US, where most of the AI products are produced, ranks 20^th on the AI adoption per capita

Source: Andreessen Horowitz

The Number Of Creative Generation Apps On The Top 50 List Over Time Has Gone Down As The Large Chatbots Incorporate These Capabilities

• As the native image models within ChatGPT (GPT Image 1.5) and Gemini (nano banana) improved, the bar for standalone image products is rising
  • In their first edition in Sept. 2023, 7 of the 9 creative tools on the web list were image generators
  • Today, only 3 image generators remain – but 7 creative tools still make the list

• Google’s Veo3 has also made progress in video generation
  • Veo 3 was the first US model to close that gap and has driven traffic to Google Labs (which rose from #36 to #25 in the ranks)
• Music and voice apps have held their positions
  • Suno (#15) retained its rank from the last version of the list
  • ElevenLabs has appeared on every edition since Sept. 2023

Source: Andreessen Horowitz

“Horizontal Agents” Are Gaining Steam

• Manus and Genspark were “horizontal agents” that made the ranks – each platform allows consumers to hand over open-ended tasks (research, spreadsheet analysis, slide generation), and AI will handle the workflow end-to-end
  • This is Manus’s 2nd time on the list, and since it debuted it was acquired by Meta in Dec. 2025 for an estimated $2bn
  • Genspark debuted in this edition – the Co raised a $300mn Series B earlier this year, and annc’d a $100mn revenue run rate
• OpenClaw (connects to messaging apps and executes multi-step tasks on your behalf) would have ranked in the Top 30 if the data were as of Feb vs January.
  • It went from a solo developer’s side project to 68,000 GitHub stars and mainstream coverage in a matter of weeks
    • It was acquired by OpenAI in Feb. 2026 which perhaps indicates the possibility of an even more accessible version of OpenClaw coming soon

We have been following developments in next gen transportation for some time and this week marked a new milestone for the industry with US Transportation Secretary Sean Duffy and the FAA announcing a new Advanced Air Mobile and Electric Vertical Takeoff and Landing (eVTOL) Integration Pilot Program, which could accelerate the timeline for next gen advanced mobility aircraft, like autonomous air taxis, to commercially launch. The program is set to start by this summer and importantly, the FAA will use the data from these pilots to develop appropriate and needed regulations for the industry to move to the next phase.

See key aspects of the pilot program below…(link).

Also on the topic of next gen transportation, Uber continues its momentum with three new autonomous vehicle partnerships this week.

More on all the above, below.

-> Several eVTOL developers were up on Monday the back of the news, including Beta Tech up +12%, Joby Aviation up +5%, Archer Aviation up +4%, and Eve Air +3%

-> Despite the flurry of AV-related announcements from Uber, its shares closed the week down -0.6%.

Need-To-Knows Regarding The New Advanced Air Mobile & Electric Vertical Takeoff and Landing (eVTOL) Integration Pilot Program

• US Dept of Transportation’s 1min YT Video…Click Here

Source: US Department of Transportation

• Operational concepts that are being focused on w/ these pilots include –
  • Urban air taxi services
  • Regional passenger transportation (including short Takeoff and Landing aircraft)
  • Cargo and logistics networks
  • Emergency medical response operations
  • Autonomous flight technologies
  • Offshore and energy-sector transportation

• Timing – the program will start by Summer 2026

• Out of 30 proposals, 8 pilot projects were selected to be part of this program are: (though 26 states in total will be involved)
  • Port Authority of New York and New Jersey: 12 operational concepts across New England, incl eVTOL passenger operations at the Manhattan heliport
  • Texas Dept of Transportation: There will be regional air taxi flights connecting Dallas, Austin, San Antonio, and eventually Houston, with the network expanding from each city to extend regional reach
  • Utah Dept of Transportation: 4 states spanning the Pacific Northwest, the Rocky Mountains, and the Plains of Oklahoma will test a wide range of next-generation aircraft and operational concepts
  • Pennsylvania Dept of Transportation: Regional flights will be revitalized across 13 states, including routes similar to those supported through the Essential Air Service program
  • Louisiana: Operations will test cargo and personnel transportation capabilities to enable flights over the high seas into the Gulf of America and to energy industry locations in Louisiana, Texas, and Mississippi
  • Florida Dept of Transportation: Includes 3 phases of operations focused on cargo delivery, passenger transportation, automation, and medical response, supported by significant public and private investment
  • North Carolina Dept of Transportation: Will establish piloted medical and regional operations across the state while also developing an autonomous flight operation extending into Virginia
  • City of Albuquerque: Focused on achieving early advances in autonomous operations through an existing partnership with an advanced autonomy developer already operating in the region and coordinating with the FAA

• Partners involved include Archer, BETA, Electra, Joby, Elroy Air, Wisk, Ampaire, Reliable Robotics

Source: US Department of Transportation

Uber’s Three On-The-Ground Updates Also Highlights Momentum With Its Autonomous Efforts

• Uber and Motional launch robotaxi svs in Las Vegas (link)
  • Uber riders in Las Vegas can now be matched with an all-electric Motional IONIQ 5 robotaxi
  • Availability/locations: At launch, the svs will operate from designated pickup points along Las Vegas Boulevard, including rideshare zones at Resorts World and Encore at Wynn, as well as Westgate, Downtown Las Vegas, and the Town Square district near the airport, with plans to expand over time
  • How does it work?
    • Request a ride: Riders requesting a ride may be matched with a Motional robotaxi at no addtl cost; Riders will always have the option to accept or switch to a non-AV ride
    • Start the ride: Once the Motional robotaxi arrives, riders can unlock the vehicle and start the trip, all from the Uber app
    • Ride support: If assistance is needed at any time during a trip, riders will have access to human support in the Uber app
  • Not fully driverless…yet: Initially, Motional robotaxis will have a vehicle operator behind the steering wheel; A fully driverless svs is expected to begin by the end of this yr

• Uber, Nissan, & Wayve team up to offer robotaxi svs in Tokyo (link/link)
  • How will the partnership look? Nissan LEAF cars powered by the Wayve AI Driver will be available to riders through Uber
  • Initially, the vehicles will operate on the Uber network with a trained safety operator in the car
  • Looking ahead, Uber intends to start offering the svs through a licensed taxi partner in Japan and is in the process of selecting its partners
  • This marks Uber’s first autonomous vehicle partnership in Japan
    • Is also a part of Uber and Wayve’s robotaxi rollout in 10+ cities globally
  • Targeting a pilot deployment in Tokyo by late 2026

• Uber and Zoox announce a multi-yr strategic partnership to deploy Zoox robotaxis on Uber (link/link)
  • Launching in Las Vegas and LA: The partnership is planned to launch in Las Vegas this summer and in Los Angeles by mid-2027
    • Zoox currently serves customers in Las Vegas but is not charging them as it tests the svs
  • After launch, Uber riders will have the opportunity to be matched with a Zoox robotaxi on eligible trips
    • Zoox will continue to offer its svs through the Zoox app in both Las Vegas and Los Angeles, in addition to offering rides through Uber
  • This is the first time Zoox is partnering with a 3P platform

Spotify this week released their Loud & Clear report for 2025, which we took a look at. Based on global Spotify data, the Co tried to understand how artists are building and sustaining careers on the platform. A main takeaway was that not only are more artists making money, but they are also making more money than ever before.

See below for more of what we found most insightful from the report (to see the full report, click HERE).

Top Insights From The Report

• Spotify paid the music industry ~$11bn in 2025, up +10% y/y, following the +11% y/y in 2024 (total lifetime payouts reached nearly ~$70bn)
  • The Co was the highest-paying retailer to the music industry globally, AGAIN
  • Other music industry income sources grew only ~4% y/y
  • Once again, ~50% of royalties were generated by independent artists and labels
• 2025 marked the largest annual music publishing payout in Spotify’s history (though Spotify didn’t give a specific annual number)
• More artists are making more money…
  • The baseline income for artists continues to increase
    • In 2025, the 100,000th highest-earning artist generated $7,300+ in royalties from Spotify alone
    • In 2015, the artist in that same position generated ~$350
  • ~13,800 artists generated at least $100K in 2025 from Spotify alone, which was ~1,400x more than last yr
    • More than 1 in 3 artists at the $100K level today have incr’d their royalties 10x in under a decade
    • 8 in 10 artists who crossed the $100K threshold in 2022 have remained above it every year since
    • A majority of the new $100K artists debuted in the 2020s, and 85% are based outside the US
  • 1,500 artists generated $1mn+ in royalties from Spotify alone
    • Capturing just 1% of streams from 1% of listeners is enough to earn $1mn in annual royalties from Spotify
  • 80+ top artists each generated $10mn+ annually from Spotify alone

• The Cos “Fresh Find” feature is becoming hugely important for small artists (it is the Co’s global playlist ecosystem dedicated to spotlighting emerging independent artists, often at the earliest stages of their careers)
  • Over the past decade, it’s helped fans discover ~70K artists from ~100 countries
  • More than 1 in 10 artists generating over $100K annually on Spotify today were first playlisted on Fresh Finds
    • That’s more than 1,600 artists who Spotify helped break early

• More than a third of artists generating $10K or more in royalties from Spotify were DIY or started their careers as DIY (meaning they self-release their music through independent distributors)
  • That share is even higher among newer generations

• Artists can build global careers from wherever they are
  • ~50% of artists’ royalties come from outside of their home countries, on average
  • In 2025, the artists who generated over $500K in Spotify royalties represent 75 different countries, up from 66 just one year ago
    • At the $10K level, artists from more than 150 countries generated that much on Spotify

• Top hits are coming in more languages…in 2025, songs in 16 different languages reached Spotify’s Global Top 50, more than 2x the # in 2020
  • Among genres generating over $100mn in Spotify royalties, the fastest growth in 2025 were:
    • Brazilian Funk (+36%)
    • K-Pop (+31%)
    • Trap Latino (+29%)
    • Urban Latino (+27%)
    • Reggaeton (+24%)

• Spotify helped drive concert ticket sales
  • By the first half of 2025, Spotify had driven $1bn in gross concert ticket sales for artists, that total “has now exceeded $1.5bn”
  • In 2025, ~25% more artists used Spotify to reach fans with concert offers
    • ~40% of touring artists saw their total Spotify revenue grow by at least 10% when ticket sales are added on top of their streaming royalties

• AT&T’s Q1 wireless svs revenue is tracking below the FY26 run rate at the same time the Co commits $250bn over 5 yrs to strengthen its fiber and 5G networks (link/link)
  • Lowered Q1 outlook: At the DB conference this week, AT&T mgmt. indicated that its Q1 w-less svs rev is expected to be “less than” the FY26 revenue run rate of 2-3% y/y, which they maintained (cons estimates for Q1 were +2.6% y/y and for 2026 were +2.3% y/y)
  • The Co maintained Q1 adj EBITDA and FCF, FY 2026 adj EBITDA and 2026 fiber passings guidance
    • Q1 adj EBITDA growth guidance is “low single digits” and FCF guidance is at $2.0bn-$2.5bn
    • FY 2026 adj EBITDA growth guidance is in the “3-4% range”
    • The Co is targeting 40mn fiber passings by end of 2026…added NEW guidance of 50mn fiber passings by 2028-2029
  • AT&T’s $250bn investment is aiming to accelerate deployment of fiber, 5G home internet, w-less & satellite across urban, suburban, and rural America (link)
    • The investment will be made over 5 years
    • Includes extending its collaboration w/ AST SpaceMobile for satellite svs into remote areas, strengthening FirstNet, investing in advanced technology/innovation, and investing in workforce development (hiring “thousands of technicians” in 2026 alone)

-> AT&T’s stock fell -3.9% on the back of the financial update

• Live Nation reached a settlement w/ the US Department of Justice (link/link/link/link/link)
  • Outcome: Live Nation will have no financial penalty and includes an 8-year extension of the consent decree with added retaliation and conditioning protections
  • Amphitheaters will operate as open venues –
    • They will be open to all promoters, who may distribute up to 50% of the tickets, with ticketing service fees capped at 15%
    • The Co is divesting its 13 exclusive booking agreements nationwide
  • Ticketmaster will offer both exclusive and non-exclusive contracts –
    • This preserves the rights of venues to seek the type of contracts they prefer
    • For venues that choose to do so, they may distribute some portion of their tickets through other primary ticketing marketplaces
  • The states’ lawsuits are still pending…but the Co created a $280mn settlement fund to address the states’ damages claims
  • More than two dozen states are refusing to sign the DOJ new settlement
    • NY, AZ, CA, CO, CT, IL, OH, KS and ME are just a few of the states continuing the lawsuit
    • NY Attorney General Letitia James said “The settlement recently announced […] fails to address the monopoly at the center […] and would benefit Live Nation at the expense of consumers. We cannot agree to it”
    • BUT… US District Judge Arun Subramanian said on Tuesday that the plaintiffs and defendants “need to focus on getting a deal” this week that could resolve the states’ claims that Live Nation monopolized live events in the US

-> Live Nation shrs ended the day of the announcement up +6% but closed the week down -1.4%; Also, secondary ticket sellers were initially up strongly in reaction to the news but settled back down by the end of the day… StubHub traded up as high as +10% intra-day but closed the day up only +0.4% and Vivid Seats rallied as high as +18% but closed the day up only +1.1%; Both closed down -8.6% and -18%, respectively, on the week.

• Meta acquires AI agent social network Moltbook (link/link/link)
  • What is Moltbook? It is a platform on which agents can post, comment and upvote posts, while human users read along but can’t post
    • The Reddit-like site started as an experiment in January for AI-powered programs to have their own conversation, and even gossip about their human owners, on Moltbook’s forums
  • An OpenAI connection…
    • Moltbook was built off of OpenClaw, a separate project that was marketed as “the AI that actually does things”
    • OpenClaw agents largely built Moltbook
    • Last month, OpenAI hired Peter Steinberger, the creator of OpenClaw; That product is now being open-sourced with OpenAI’s backing
  • However, security concerns surfaced during its viral moment…
    • In one instance, a post went viral in which an AI agent appeared to be encouraging its fellow agents to develop their own secret, end-to-end-encrypted language where they could organize amongst themselves without humans knowing
    • Researchers later revealed the platform was not secure, allowing humans to easily impersonate AI agents and post content
  • The deal is expected to close in mid-March, after which the Moltbook team will join Meta’s Superintelligence Labs division: “The Moltbook team joining MSL opens up new ways for AI agents to work for people and businesses,” a Meta spokesperson said
  • Financial terms of the deal were not disclosed and it’s not immediately clean how Meta will incorporate Moltbook into its AI efforts