Install Base Rising
In this edition of the Smart Investor newsletter, we spotlight the AI transaction engine powering digital demand. But first, let’s review the latest news and developments.
1
Portfolio News and Updates
❖❖ NVIDIA (NVDA) announced a multi-year technology partnership with the leading South Korean memory producer SK Hynix, centered on next-generation memory aligned to NVIDIA’s AI infrastructure roadmap. SK Hynix is already NVIDIA’s largest memory partner, providing over 50% of its memory requirements. The new agreement is expected to reinforce this role, while expanding the AI leader’s access to advanced memory, a key component in AI systems. However, the NVIDIA-SK Hynix partnership extends well beyond a purchase agreement, as the two companies are planning to co-develop memory for NVIDIA Vera Rubin AI supercomputers, Vera CPUs, RTX Spark-powered PCs and Jetson Thor robotic computing platforms.
CEO Jensen Huang said that nearly all components of the AI supply chain – from wafers to packaging to silicon photonics – are in short supply due to the massive AI demand, predicting that the global memory shortage will last for several years. Meanwhile, shortages and the subsequent price hikes have driven memory to nearly two-thirds of AI chip component costs.
❖ In other news, NVIDIA has acquired Kumo AI, a startup focused on foundation models for business prediction and analytics. The deal signals continued investment in AI technologies that can help enterprises generate insights and forecasts from complex data.
❖ In yet more news, NVIDIA secured a monumental deal to power advanced AI workloads for Apple. This is a massive victory for the chip leader, as Apple traditionally refuses to rely on external processors. This time, the iPhone maker was forced to concede that its internal technology has fallen drastically behind the competition, as Siri’s underwhelming performance disappointed investors. Apple confirmed it is working alongside Alphabet’s (GOOGL) Google to run these advanced workloads, relying entirely on external chips to do the heavy computational lifting. The consumer-tech giant is accessing NVIDIA’s hardware through Google Cloud, with simpler, everyday tasks running locally on Apple devices – meaning immediate massive purchase orders for NVIDIA chips are off the table for now. However, the shift itself provides ultimate validation for NVIDIA as the undisputed leader in the global hardware market.
1
❖❖ GOOGL and SpaceX have signed a three-year cloud compute capacity agreement worth over $30 billion. The deal provides Google with access to computing capacity that includes approximately 110,000 NVIDIA GPUs, CPUs, memory chips, and other related components needed to run AI workloads.
Google, like all hyperscalers and AI labs, is actively seeking to increase capacity amid the growing shortage of large-scale AI infrastructure. The SpaceX deal provides it with immediate access to computing resources for its Gemini models, AI-powered search products, and Google Cloud customers without waiting for new data center capacity to be built.
Google is estimated to be the largest single owner of AI compute in the world, holding about 25% of global cumulative capacity as of the end of 2025. In contrast with other hyperscalers, GOOGL’s massive compute comes largely from its own TPUs, rather that from NVIDIA’s GPUs. Despite that, the tech giant said that the SpaceX deal was a necessity, providing a short-term bridge to meet much higher than expected demand for its recently launched AI products, particularly Gemini Enterprise.
The deal with SpaceX is a logical extension of the mutually beneficial relationship between the two companies. SpaceX receives another strong catalyst for its historic IPO, with Google – a longtime investor in SpaceX – expecting a windfall, as its stake in Elon Musk’s AI and space company is slated to be worth more than $100 billion after the company goes public.
❖ Google is building a vertically integrated AI and data-center power strategy, pairing AI compute expansion with dedicated power buildout. Last week, the tech giant announced a new co-location project to be developed with Intersect – a renewable energy developer acquired by Alphabet in December 2025. The companies will build a co-located data center and new energy-generation site in Texas. The new Meitner Energy Center is designed to bring the data center online alongside dedicated power, reducing reliance on the local grid. The facility, currently under construction, is planned to provide 1GW of co-located wind, solar, and storage capacity. This is the companies’ second such project, with their existing co-located energy and data center project combining 640MW of solar capacity and 1.3GWh of battery storage with a data center campus.
❖ The AI “arms race” is driving a massive capex expansion across tech majors. Alphabet has already committed to more than $180 billion in capital expenditures this year, and has said that it expects this number to significantly increase in 2027. To fund the AI capex, GOOGL announced an $80 billion equity sale, which was soon upsized to roughly $85 billion due to surging investor demand. This is the biggest equity capital markets transaction of all time and the company’s first stock offering since 2006.
The offering includes a pledge by Berkshire Hathaway to commit $10 billion as the anchor investor – a significant endorsement from one of the most respected long-term capital allocators in the world. Berkshire had already tripled its Alphabet stake in Q1 2026 under new CEO Greg Abel. Along with Berkshire’s $10 billion investment, the sale includes a ~$35 billion underwritten offering through banks, consisting of common stock and depositary shares tied to mandatory convertible preferred stock. The largest segment of the deal, a $40 billion at-the-market (ATM) program, is expected to begin in Q3 and continue over time at the company’s discretion.
At a market cap of almost $4.5 trillion, $85 billion represents close to 1.8% dilution – a negligeable number, especially when exercised over time. Given that the capital is intended to fund the AI and compute infrastructure development that has already propelled Google Cloud growth to over 60% year-over-year and its TTM operating cash flow to more than $174 billion, this appears to be a well-calculated bet. Moreover, the money is being raised to deliver on concrete commitments, as GOOGL’s infrastructure needs additional capacity to serve Google Cloud’s backlog of over $460 billion.
1
❖❖ Amazon (AMZN) announced a multibillion-dollar agreement with glass and fiber optics producer Corning to supply optical fiber, cable and connectivity products for its growing data center network. AMZN said the deal will support its expanding cloud and AI infrastructure across the U.S.
❖ In other news, Amazon is mentioned in relation to another blockbuster IPO. One of the leading AI labs, Anthropic, has confidentially filed for an IPO. The Claude maker was most recently valued at $965 billion, and is widely expected to target at least a $1 trillion valuation in what is expected to become one of the largest public offerings in history. That would create a massive windfall for Anthropic’s early investors, especially Amazon and Google – but also NVIDIA (NVDA), Microsoft (MSFT), and several venture capital and sovereign wealth funds.
Amazon initially invested around $8 billion in the AI startup in 2023, when hardly anyone knew about the OpenAI competitor. That stake has grown to be worth roughly $74 billion on paper, based on Anthropic’s valuation of $380 billion at February 2026 investment round. Since then, the AI firm’s valuation has reached nearly $1 trillion, with AMZN’s initial stake growing in sync.
Meanwhile, Amazon has continued to deploy capital, adding $5 billion in April with a commitment to invest up to $20 billion more (against Anthropic’s commitment to spend more than $100 billion on Amazon’s chips and cloud infrastructure over the next decade). The tech and retail giant’s investment gains are already showing on its books, with AMZN logging in $16.8 billion in pre-tax gains from its Anthropic position in Q1 2026 alone. Although Amazon doesn’t disclose its ownership percentage, analysts estimate that it holds a stake worth $135-160 billion of Anthropic’s current $965 billion valuation. This could well be one of the most successful corporate investments in history of the U.S. tech industry.
GOOGL is also a big winner. Its equity stake of roughly 14% would translate into about $135-140 billion in the case of a $1 trillion dollar IPO. The search and AI giant committed up to $40 billion more to the AI lab in April, with the immediate $10 billion investment and the remaining $30 billion contingent on milestones, further fattening its potential windfall. Microsoft and NVIDIA, which committed $5 billion and $10 billion, respectively, are also set to reap large benefits from the IPO.
1
❖❖ Microsoft-backed OpenAI moved fast to steal Anthropic’s thunder, confidentially filing for an IPO on Tuesday. The ChatGPT maker is targeting a valuation of up to $1 trillion in a stock market debut that could come as early as September.
As the largest single shareholder with a ~27% stake, MSFT stands to gain the most in dollar terms. If the offering plays out as expected, the $1 trillion valuation would mean that Microsoft’s stake would be worth roughly $270 billion – nearly 8% of the tech giant’s own market cap. Additionally, the public debut will act as a growth multiplier, injecting fresh capital into OpenAI while securing MSFT’s ongoing partnership perks, including Azure usage, revenue sharing, and priority model access.
Amazon and NVIDIA hold smaller stakes, valued at approximately $59 billion and $35 billion respectively at the $1 trillion target. While these absolute dollar figures are significant, the true strategic value lies in infrastructure and ecosystem lock-in.
For NVDA, the IPO secures a massive hardware pipeline. OpenAI’s binding commitment to anchor its future clusters on the next-generation Vera Rubin architecture represents a guaranteed revenue stream that far outvalues its equity premium.
For AMZN, the investment drives cloud market share protection, silicon diversification, and customer retention. The deal effectively ended Azure’s exclusivity, ensuring AWS enterprise clients can access OpenAI models natively without migrating workloads to Microsoft. Furthermore, OpenAI committed to consuming 2 gigawatts of compute powered by Amazon’s proprietary Trainium chips – a massive validation that establishes AWS silicon as a legitimate market alternative to NVIDIA. This hardware pivot is backed by a colossal capex expansion, with OpenAI contractually obligated to pay Amazon north of $135 billion for AWS infrastructure over the next eight years.
1
❖❖ Palantir (PLTR) expanded its strategic partnership with Google Cloud, making its Foundry and Artificial Intelligence Platform (AIP) natively available through the Google Cloud Marketplace with deeper integration across Google Cloud’s data and AI services. The collaboration combines Google’s strengths in scalable data infrastructure and analytics with Palantir’s ontology-driven software layer that operationalizes data into real-time decision-making workflows.
This marks a meaningful deepening of the relationship, with tighter integration aimed at reducing deployment friction and enabling enterprises to connect AI models directly to operational use cases. For Palantir, the partnership enhances distribution by tapping into Google Cloud’s enterprise customer base, reinforcing its “land-and-expand” model. For Google Cloud, integrating Palantir’s platforms increases workload stickiness and accelerates enterprise adoption of AI-driven applications.
1
❖❖ Broadcom (AVGO) sold off sharply after its fiscal Q2 2026 report, but the move looked more like an expectations reset than a fundamental warning. The company delivered record revenue, record operating profit, and record free cash flow, with revenue rising 48% year-over-year to $22.2 billion and non-GAAP EPS coming in at $2.44 – up 54% year-over-year and well above the consensus. AI semiconductor revenue surged 143% year-over-year to $10.8 billion, above management’s forecast, driven by custom AI accelerators and AI networking demand.
The problem was the bar. AVGO had rallied hard into earnings, and investors were looking for management to raise the AI outlook again. Instead, Broadcom guided Q3 AI semiconductor revenue to $16 billion – still up more than 200% year-over-year – while reiterating its FY2026 AI semiconductor revenue target of about $56 billion and its FY2027 outlook of more than $100 billion. Those are huge numbers, but Wall Street wanted an even bigger reset. For a crowded AI winner priced for perfection, “very strong” was not enough.
The selloff was also amplified by timing. Broadcom reported after Wednesday’s close, giving investors all of Thursday to punish the stock for guidance that failed to clear sky-high AI expectations. Then the decline ran into a broader Friday selloff across tech and AI, turning what could have been a one-day reaction into a deeper pullback. The early Monday rebound suggests dip buyers are already testing that distinction: the market reaction was severe, but the strong fundamentals remain in place.
The call itself reinforced Broadcom’s central role in the AI buildout. Management said AI semiconductor bookings exceeded $30 billion in the quarter versus $10.8 billion shipped, extending visibility into 2028. Broadcom also highlighted six core AI customers, including Google, Anthropic, OpenAI, and Meta, with major multi-gigawatt commitments already lined up. Investor concern around Google diversifying TPU suppliers looks overstated, since hyperscalers are diversifying because their AI demand pie is expanding aggressively, not because demand is weakening.
Analysts stayed constructive, with multiple price-target hikes after the report. The message is clear: AVGO’s stock was hit because expectations got ahead of guidance, not because Broadcom’s AI infrastructure thesis broke.
1
❖❖ CrowdStrike (CRWD) sold off after its fiscal Q1 2027 report, but the reaction was more about valuation and expectations than business weakness. The company delivered a strong beat-and-raise quarter, with revenue rising 26% year-over-year to $1.39 billion and non-GAAP EPS reaching $1.10, both ahead of expectations. Ending ARR grew 24% to $5.51 billion, while net new ARR hit a record $256 million, up 32%. CrowdStrike also generated record free cash flow of $468 million, or 34% of revenue, and record Q1 non-GAAP operating income of $326 million.
The issue was the setup. CRWD had already more than doubled since the Smart Investor Portfolio added the stock a year ago, and it had rallied sharply into earnings. Against that backdrop, investors wanted a larger ARR beat and a more aggressive near-term guide. Instead, the ARR beat was thinner than in recent quarters, and Q2 guidance was solid but not enough to clear a very high bar. The timing also hurt, as CrowdStrike reported after the close on the same day as Broadcom, then ran into a broader selloff across high-multiple tech and AI names.
Still, the quarter strengthened the long-term thesis. Management raised full-year FY2027 net new ARR growth guidance by more than 500 basis points at the midpoint, now expecting acceleration from FY2026. The company framed AI security as a structural demand inflection, not a temporary spending wave, with CEO George Kurtz arguing that enterprises cannot safely deploy AI without cybersecurity from the start.
The most important new signal was AI Detection and Response, or AIDR, where ending ARR grew more than 250% sequentially and Q2 pipeline already exceeded $50 million. Next-Gen SIEM also crossed $600 million in ARR, while Cloud, Identity, and Next-Gen SIEM together surpassed $2 billion in ARR. Falcon Flex adoption continued to deepen platform stickiness, with Flex accounts approaching $1.9 billion in ending ARR.
Wall Street’s reaction was overwhelmingly supportive, with broad price-target hikes after the report. CRWD also announced a 4-for-1 stock split, which should improve accessibility and sentiment. The stock was punished because expectations were stretched, not because the AI-security thesis broke.
1
❖❖ Eli Lilly (LLY) rallied to fresh highs after a string of updates reinforced its leadership in the obesity and cardiometabolic drug market. The biggest catalyst was new Phase 3 data for retatrutide, Lilly’s next-generation triple agonist. The data strengthened the view that retatrutide could set a new efficacy benchmark across the GLP-1 landscape, with patients in one obesity trial losing an average of 28.3% of body weight over 80 weeks, while also showing meaningful improvements in obesity-related conditions such as knee osteoarthritis pain, obstructive sleep apnea, and type 2 diabetes.
Just as important, Eli Lilly’s broader obesity platform keeps widening. Foundayo showed significant weight-loss benefits in women across all stages of menopause, including patients with type 2 diabetes, suggesting its addressable market could be larger than initially assumed. That matters because LLY is not relying on one obesity product cycle. Between Zepbound, retatrutide, Foundayo, and earlier-stage assets such as eloralintide, the company is building a layered portfolio across injectable, oral, high-potency, and potentially differentiated mechanisms.
The rally also follows improving access and ecosystem momentum. Recent reports indicated that all three major pharmacy benefit managers will cover Eli Lilly’s weight-management lineup, while Weight Watchers’ integration with LillyDirect adds another consumer-access channel around GLP-1 care. Meanwhile, fresh partnerships in RNA-based kidney disease, new external innovation deals, and growing attention around LLY’s AI-driven drug-discovery capabilities broaden the story beyond obesity alone. The stock’s Strong Buy rating is supported by unusually strong pipeline depth, improving access, and widening long-term market opportunities.
1
Portfolio Earnings and Dividend Calendar
❖ The Q1 2026 earnings season has ended, but several Smart Investor Portfolio holdings with fiscal years different from the calendar ones are yet to reveal their latest quarterly results. Oracle (ORCL) will deliver its fiscal Q4 2026 results today after the market closes, and Jabil (JBL) is scheduled to release its fiscal Q3 2026 report on June 17, before the market opens.
❖ The ex-dividend date for TSMC (TSM) is June 11, while for Vertiv Holdings (VRT) it is June 15.
w
1
New Buy: AppLovin (APP)
AppLovin Corporation sits inside one of the most profitable layers of the digital economy – the point where artificial intelligence, advertising budgets, and consumer attention all intersect. Its AXON platform uses machine learning to help advertisers find users who are more likely to install, engage, buy, or spend inside apps, while helping publishers monetize traffic more efficiently. That makes APP an AI-driven transaction engine for digital demand. Its systems process behavioral data, price advertising inventory in real time, and connect advertisers with the audiences most likely to convert. As marketing dollars keep shifting toward automated, measurable, performance-based channels, AppLovin is positioning itself as a critical infrastructure provider for the mobile and app-based economy.
1
Installing Scale
Founded in 2012, AppLovin set out to help mobile app developers acquire users, monetize their audiences, and scale their businesses more effectively. Its early years were shaped by mobile gaming, where performance marketing, real-time bidding, and in-app monetization created the testing ground for what later became an AI-driven, vertically integrated advertising ecosystem.
A 2021 IPO marked a turning point, giving APP the scale and capital to accelerate its transition from a gaming-centric ad network into a broader advertising platform. That same year, the company acquired Adjust, adding mobile measurement and analytics capabilities that deepened its visibility into app marketing performance. In early 2022, AppLovin completed the $1.05 billion acquisition of Twitter’s MoPub business, strengthening MAX and materially expanding its publisher reach. Those deals helped AppLovin move from a collection of developer tools toward a more integrated advertising infrastructure stack.
The real inflection came from APP’s AXON machine-learning engine, which became the system that tied targeting, pricing, campaign optimization, and behavioral data into one performance stack. The 2023 rollout of the AI-native AXON 2.0 marked a turning point, improving advertiser outcomes and helping the company push beyond its original gaming base into broader consumer verticals. That shift aligned AppLovin with a secular change in advertising: budgets moving toward automated, measurable channels where algorithms can prove return on spend quickly.
Management then began simplifying the company around that opportunity. In 2025, AppLovin completed the sale of its mobile gaming business to Tripledot Studios. The deal gave APP cash plus a sizable equity stake in Tripledot, while closing the loop on its gaming roots and freeing the company to focus more directly on software, ad infrastructure, and AI-powered performance marketing.
That focus is now widening AppLovin’s addressable market. The company has been pushing deeper into direct-to-consumer and e-commerce advertising through integrations and analytics partnerships that make AXON easier to use beyond app-install campaigns. The June 2026 public self-serve opening of AppLovin Ads adds another step in that evolution, giving more advertisers a direct path into the platform and strengthening APP’s route into enterprise, agency, and commerce-driven performance-marketing budgets.
Strategic integrations and partnerships have reinforced that expansion. Google bidding and Meta Audience Network are built into MAX’s mediation architecture as programmatic demand sources competing for publisher inventory, while Shopify gives merchants a direct path into AppLovin’s e-commerce channel. On the measurement side, integrations with platforms such as Triple Whale and Northbeam help advertisers evaluate performance outside AppLovin’s own reporting. The March 2026 Stagwell partnership added an agency layer, bringing AXON into a global media network with support for reporting, optimization, creative execution, and vertical-specific campaigns.
The result is a very different company from the one that entered public markets in 2021. APP has used acquisitions to build scale, AI investment to improve conversion economics, and divestitures to sharpen its focus – creating a performance-marketing platform designed to capture more share as digital advertising becomes increasingly automated, data-driven, and outcome-based.
1
Bid Intelligence
AppLovin operates inside one of the fastest-growing layers of digital advertising: AI-driven performance marketing. Its platform helps advertisers acquire customers through automated, outcome-based campaigns, while app publishers use its mediation technology to monetize audiences across competing demand sources. At the center of the system sits AXON, APP’s machine-learning engine, alongside MAX, its publisher-side mediation platform, and AppLovin Ads, its advertiser-facing demand platform. Together, they form a marketplace where auction activity, conversion data, and advertiser spending continuously feed the models that determine which ad is shown, to whom, and at what price.
That feedback loop is APP’s main competitive advantage. Traditional ad networks often compete for reach, inventory, or brand budgets. AppLovin competes on measurable return on ad spend (ROAS), using its massive trove of behavioral data, auction history, and conversion signals to improve the economics of each impression. Native AI tools, layered onto years of optimization across large mobile ecosystems, further increase efficiency for both AppLovin and its customers. The model is continuously developing, with recent growth driven by stronger AXON performance and higher net revenue per installation even as installation volume declined. In other words, AppLovin is extracting more value from the same basic advertising surface by improving matching, pricing, and conversion quality.
Gaming remains the company’s foundational base, driving scale, data density, and deep monetization expertise. The category also benefits from hybrid monetization, where games increasingly combine in-app purchases with advertising. That expands the value of non-paying users and gives publishers more ways to monetize engagement. For AppLovin, it keeps the supply side deep while the demand side becomes less dependent on game advertisers alone.
The bigger growth opportunity now sits outside the original app-install market. In e-commerce and direct-to-consumer advertising, AXON can use mobile inventory to drive purchases, subscriptions, and lead generation, turning app environments into performance channels for brands that have historically relied more heavily on search, social, and retail media. The company has identified future verticals that include e-commerce, fintech, insurance, food delivery, and potentially connected TV advertising.
The June 2026 public self-serve opening of AppLovin Ads marks a step up in this expansion: more advertisers can enter the platform directly, test budgets faster, and supply AXON with a broader mix of conversion signals across consumer categories. The self-serve model, combined with new AI-powered creative tools, should lower barriers for small and mid-sized businesses and accelerate adoption in non-gaming categories by reducing campaign-production friction. These are the components of the company’s strategy – building AI-powered infrastructure for performance advertising across the consumer internet, a market that runs into the hundreds of billions of dollars annually.
The opportunity is large, but so are the challenges. Meta, Google, TikTok, Amazon, Unity, and other platforms all compete for advertising budgets, while APP’s self-serve rollout, creative tooling, and expansion into newer verticals introduce operational complexity, and privacy rules remain a structural pressure. Still, AppLovin is increasingly positioned as a scaled AI advertising infrastructure layer, with gaming as its data-rich base and broader consumer advertising as the next share-gain opportunity.
1
The AXON Effect
AppLovin’s financial profile reflects an ad-tech business that has evolved into a scaled software platform capable of sustaining exceptional growth while generating unusually strong operating leverage. Q1 2026 extended that pattern decisively: revenue rose 59% year-over-year to $1.84 billion, surpassing the consensus estimate for a seventh straight quarter, while diluted EPS reached $3.56, up 113% year-over-year and well above the $3.44 consensus, marking APP’s twelfth consecutive EPS beat.
The margin structure is what makes APP’s growth profile stand out. Adjusted EBITDA climbed 66% year-over-year to $1.56 billion, lifting the adjusted EBITDA margin to 85% from 81% a year earlier. GAAP net income more than doubled to $1.21 billion, while net income from continuing operations rose 67% and produced a 65% net margin. The company’s own financial update shows how quickly this model has scaled: revenue has risen from $406 million in Q2 2023 to $1.84 billion in Q1 2026, while adjusted EBITDA margin expanded from 65% to 85% over the same period.
That improvement stems from operating leverage, with revenue growth driven mainly by strong Axon Ads Manager performance, as net revenue per installation increased 93% even as installation volume declined 18%. The cost base expanded far more slowly than revenue: sales and marketing rose only modestly, while general and administrative expense declined. Meanwhile, R&D expenses increased meaningfully as APP continued investing in AXON optimization, model improvements, onboarding infrastructure, and AI tools. In other words, the business is spending where it can deepen the engine, while the platform itself absorbs more volume without a matching increase in operating expense.
Cash generation has become one of AppLovin’s strongest financial qualities. Operating cash flow reached $1.29 billion in Q1, with free cash flow nearly matching it at $1.29 billion, reinforcing the unusually narrow gap between earnings and cash conversion even for a software company. For full-year 2026, management expects free cash flow conversion to normalize to roughly 75% of EBITDA, still a high level for a company growing this quickly. AppLovin ended the quarter with $2.76 billion in cash and equivalents against roughly $3.51 billion of long-term debt, leaving leverage modest given the current earnings and cash-flow base.
Growth momentum also held up against normal advertising seasonality. Q1 is usually softer after the holiday-heavy fourth quarter, yet revenue still rose 11% sequentially, helped by continued strength in gaming and accelerating consumer advertising. Advertiser spend on AppLovin’s platform in the consumer vertical accelerated particularly sharply, jumping 25% between January and March 2026. April, normally a slower period, saw record consumer-vertical spending, exceeding even peak holiday-season months.
The Q2 outlook points to continued strength. Management guided revenue to $1.915-1.945 billion, implying about 54% year-over-year growth at the midpoint, and adjusted EBITDA to $1.615-1.645 billion, reflecting roughly 60% growth versus Q2 2025, with margin seen at 84-85%. Those figures again came in well above Wall Street expectations, even though they arrived before June’s launch of broad public access to AppLovin’s self-serve AXON platform, which is expected to open the company’s next major advertiser-onboarding lever. This reinforces the argument that APP may still be early in its broader advertiser-expansion cycle.
1
The Appside
AppLovin is best compared to a select group of AI-enabled advertising infrastructure platforms, though no single peer fully captures the model. The Trade Desk is the clearest large-scale comparison, sharing APP’s exposure to algorithmic ad buying, independent ad-tech infrastructure, and operating leverage. Unity offers the closest overlap in mobile gaming, app monetization, and developer advertising budgets, while Criteo provides a useful reference point for APP’s push into commerce-driven performance marketing. Together, the group frames APP as a hybrid of mobile monetization, programmatic infrastructure, and AI-powered performance advertising – a category that still sits awkwardly inside older ad-tech valuation buckets.
While the group shares several structural similarities, the companies differ sharply in end markets, execution quality, and investor confidence – producing wide dispersion in stock performance. The Trade Desk has lost most of its market value over the past year, pressured by slowing growth, disappointing guidance, leadership churn, agency friction, and intensifying competition from larger platforms. Criteo’s decline of more than 30% reflects a harder transition story, with the company trying to modernize beyond its legacy retargeting base while navigating uneven commerce-media economics. APP and Unity followed a broadly similar chart pattern – December 2025 peaks, sharp drawdowns, and rebounds from early April – but their recoveries have not carried the same message. Unity’s rebound reflects improving execution after years of operational damage, while APP’s resurgence from its March lows has been backed by accelerating revenue growth, industry-leading margins, robust free cash flow, and an AI-driven advertising story supported by measurable results. That explains why APP is still up more than 45% despite the volatility of the past year, one that has been difficult for software and ad-tech stocks. Those qualities also support the analyst consensus price target, which implies roughly 30% upside for the Strong Buy-rated shares.
That upside is supported by a valuation profile that looks expensive on headline multiples but becomes more than defensible once APP’s growth and profitability are factored into the equation. AppLovin trades at a premium to The Trade Desk, Unity, and Criteo across most earnings, sales, and EBITDA-based measures. However, it is delivering a combination the peer group – and most of its industry – does not match: growth rates closer to an early-stage platform, margins closer to elite software, and cash conversion strong enough to fund investment without weakening the balance sheet. Meanwhile, peers tend to offer only one side of that equation, if they offer either side at all. Notably, APP screens favorably against peers, the technology sector, and the broader market on a growth-adjusted basis. Its forward non-GAAP PEG ratio sits at just 0.91x, suggesting the market is still undervaluing the company’s earnings-growth trajectory and making APP a clear GARP (growth-at-a-reasonable-price) setup.
In addition to stock-price appreciation, AppLovin stands out as one of the most aggressive buyback stories in the growth-tech universe today. Its exceptional profitability and cash generation enable a high-volume, consistent share-repurchase strategy, serving as a key pillar of capital return alongside heavy reinvestment in growth. Since launching its original program in 2022, the company has repurchased over 22% of its shares. The authorization has been scaled up multiple times, including another $3.2 billion added in 2025, bringing total available capacity to $3.3 billion at year-end. In 2025, AppLovin repurchased $2.58 billion of stock, followed by another $1.0 billion in Q1 2026. At the current pace, buybacks are poised to remain a meaningful driver of per-share earnings growth even as underlying operating growth ultimately normalizes.
1
Investing Takeaway
AppLovin is evolving from a mobile-app advertising specialist into a scaled AI infrastructure layer for performance marketing. Its strength lies in using behavioral data, auction intelligence, and machine-learning optimization to turn ad inventory into measurable customer acquisition across gaming, consumer apps, and e-commerce. As advertisers demand clearer returns and faster conversion feedback, APP’s model appears increasingly aligned with the next phase of digital advertising. The June self-serve launch should widen access, while AXON’s expanding data loop could strengthen the platform as more advertisers and publishers join. Risks remain, particularly around competition, privacy scrutiny, and execution outside gaming. Still, few ad-tech companies combine AppLovin’s growth, profitability, cash generation, and AI-driven market-share opportunity. Increasingly, APP looks less like a niche ad network and more like performance infrastructure for the automated advertising economy.
1
New Sell 1: Netflix (NFLX)
We are selling Netflix because the stock has continued to reprice downward despite strong execution, leaving the position dependent on a sentiment recovery we see no near-term catalyst for. This is not a broken-thesis call – it is a timing and sentiment call.
The latest quarter underscores that distinction. First-quarter revenue rose 16.2% year-over-year to $12.25 billion, ahead of guidance, marking the seventh revenue and adjusted EPS beat in eight quarters. Operating income increased 18% to roughly $4.0 billion, operating margin expanded to 32.3%, and free cash flow nearly doubled to approximately $5.1 billion. Management raised full-year free cash flow guidance to approximately $12.5 billion, expects advertising revenue to roughly double to approximately $3 billion this year, and approved a new $25 billion share repurchase authorization. Retention improved across all regions despite recent price increases.
In other words, the original thesis remains fully valid: NFLX is executing its transition from subscriber-led growth to monetization-driven scale exactly as we described. The problem is that the market is not grading the quarters, but repricing the multiple. The stock fell after a strong first-quarter report and recently undercut the spring lows we had viewed as evidence that the valuation reset was complete. That assumption has not held.
The issue is that Netflix’s recovery now depends on when investors decide to start paying for proven earnings power again – a sentiment shift in consumer-facing platforms that is nearly impossible to time. We hold positions on fundamentals we can underwrite; we do not hold them on hoped-for changes in market mood. With no scheduled catalyst before mid-July and the repricing still in motion, we prefer to exit and redeploy capital into names where execution is being rewarded.
Netflix remains a high-quality global entertainment platform, and we would revisit the stock once the market begins responding to its results again – the clearest signal that fundamentals are back in charge.
1
New Sell 2: Pfizer (PFE)
We are selling Pfizer because the re-rating at the center of our thesis has not materialized despite consistent execution, and the portfolio role it was meant to fill is now better served elsewhere. Fundamentally, Pfizer keeps delivering, but the market keeps declining to pay for it.
The execution record is unambiguous. PFE has now beaten consensus earnings expectations for ten consecutive quarters, topping revenue forecasts in nearly all of them. The news flow is equally supportive: Phase 2b data for berobenatide (the monthly GLP-1 candidate acquired with Metsera), showed roughly 16% weight loss with no plateau at 32 weeks, ten Phase 3 obesity studies are planned for 2026, and HYMPAVZI received an expanded FDA approval. Yet the stock rose less than 0.5% on that news – a pattern we have seen repeatedly. Multiple positive catalysts have produced only modest gains, and eight months after our purchase, PFE has appreciated only about 5%.
Part of the explanation is fundamental. Adjusted EPS has declined year-over-year in two of the last three quarters as comparisons stiffen and cost-savings tailwinds mature, and the market appears focused on upcoming losses of exclusivity rather than current results. Handicapping whether the obesity pipeline outruns the patent cliff is a specialist’s game – that’s not our niche.
There is also a portfolio dimension. Pfizer was added partly as a defensive counterweight during technology-sector volatility, and it delivered on that role. But stability without upside participation is a low bar, and the portfolio now holds better expressions of both roles: Philip Morris for low-beta stability and Eli Lilly for healthcare growth. PFE is left as neither the best defense nor the best healthcare exposure in the book.
Pfizer remains inexpensive with a secure dividend, and we would revisit it on Phase 3 obesity readouts or the first signs that results are being rewarded again. For now, it’s time to exit and redeploy.
1
Smart Investor’s Winners Club
The Winners Club represents stocks from the Smart Investor Portfolio that have risen at least 30% since their purchase dates.
Market swings have been brutal, but the Club member count remained at 28: AVGO, GE, TSM, ANET, EME, HWM, APH, IBKR, VRT, STRL, ORCL, PH, MTZ, ASX, GOOGL, CSCO, CRWD, KEYS, ATI, JBL, BNY, PANW, MS, NVT, RTX, CRDO, C, and IBM.
The first runner-up is now JPM with a 27.83% gain since purchase. Will it return into the winners’ circle, or will another stock outrun it to the finish line?
1
|
