A partial excerpt from an output created by an AI agent using CapRelay
This is a partial excerpt from 160+ rows, generated from a single run by a Codex agent using CapRelay. The Codex agent was provided this prompt:
go through each software company above 2b of market cap and summarize the specific AI bear case for it
| Ticker | Company | Market Cap | AI Bear Case |
|---|---|---|---|
| MSFT | Microsoft Corp | $3.2T | Debate AI Infrastructure Returns: CapEx vs. Revenue Payoff: Amy Hood herself acknowledged that if all the GPUs allocated to first-party apps in FY26Q1 had been directed solely to Azure, Azure constant-currency growth would have exceeded 40% rather than 39%-meaning a meaningful share of capex is not generating near-term Azure revenue. Specific concerns include: Q1 FY26 CapEx was $34.9B in a single quarter. FY25 full-year additions to property and equipment were $64.6B, up from $44.5B in FY24. Amy Hood herself... Bear-case OpenAI Concentration: The Fragility Behind the Backlog: Approximately 45% of the $625B RPO-roughly $280B-is from OpenAI. The non-OpenAI RPO of approximately $345B grew 28% YoY, which is healthy but would be the entire growth story if OpenAI were removed.; OpenAI's contracted Azure spend is approximately 20x its current annual revenue scale, as management acknowledged on the FY26Q1 call. Whether OpenAI can actually consume this capacity depends on its continued... |
| ORCL | Oracle Corp | $498.4B | Debate OCI's Competitive Moat Against Hyperscalers in AI Infrastructure: It is not realistic to assume Oracle can maintain a multi-year networking performance lead against three well-funded companies whose entire business model depends on winning AI workloads. Specific concerns include: AWS, Azure, and Google are each spending well over $50B annually on infrastructure, with enormous engineering teams focused on exactly the bottleneck Oracle is exploiting: networking. It is not realistic to assume Oracle can... Bear-case Thesis Summary: The capital cycle has grown beyond Oracle's traditional financial model. Free cash flow was negative $24.7B for the trailing four quarters ended Q3 FY26. Oracle raised $30B in February 2026 alone and has stated intent to raise up to $50B in calendar year 2026 financing. This is debt-financed growth at a scale Oracle has never attempted, with AI infrastructure gross margins of 30-40% - well below the 80%+ margins of Oracle's legacy license support... |
| SHOP | Shopify Inc. | $164.1B | Debate Agentic Commerce: New Channel or Competitive Threat?: Agentic commerce is a channel where Shopify has no guaranteed privileged position, and significant adversaries could route around it. Specific concerns include: Amazon has its own "Buy with Prime" capability, deep consumer checkout trust, and stronger direct relationships with the leading AI platforms. If Amazon strikes preferential deals with OpenAI or Google for agentic checkout placement, Shopify merchants could be at a disadvantage in the most... Bear-case Agentic Commerce: Real Opportunity, Uncertain Competitive Outcome: Amazon has built Buy with Prime and has deep consumer trust for checkout. If Amazon secures preferential checkout placement in AI platforms - leveraging its existing merchant relationships and consumer data - it could capture disproportionate agentic GMV at Shopify's expense.; Google operates Google Shopping, has its own merchant data infrastructure, and is both a partner and competitor. Google has... |
| INTU | Intuit Inc. | $109.5B | Debate AI Disruption Risk vs. Intuit's Competitive Moat: General-purpose AI is a genuine long-term threat to Intuit's core DIY tax and basic bookkeeping franchises, and the risk is underappreciated. Specific concerns include: The partnerships themselves carry asymmetric risk. Intuit is embedding its workflows and use cases into ChatGPT and Claude - effectively educating two of the most significant technology companies in the world about exactly which tasks Intuit monetizes. If customers start resolving basic tax... Bear-case AI Disruption: A Real Risk Management Has Not Fully Resolved: Intuit's partnerships with OpenAI and Anthropic effectively involve Intuit building its apps into their platforms. Management insists that customer data and domain models stay within Intuit's "four walls," but the partnerships require Intuit to make its workflows visible and accessible to the largest and best-resourced potential disruptors in the world.; The "context versus core" framework Intuit... |
| ADBE | Adobe Inc. | $99.2B | Debate AI Model Competition in Creative Software: Adobe's stock business decline is a preview of a broader AI substitution risk that the bulls are dismissing too quickly. Specific concerns include: Management explicitly acknowledged in Q1 FY26 that the decline in the traditional stock business was "playing out more quickly than we had planned for." This is a concrete, real-time example of AI eroding an Adobe revenue stream - and stock was approximately a $450M book of business.. Bear-case Adobe Stock: AI Substitution Is Already Hitting Revenue: The most concrete evidence of AI substitution in Adobe's business is the stock photography business. |
| CDNS | Cadence Design Systems Inc | $91.9B | Debate AI-Driven EDA Demand vs. AI Cannibalization Risk: AI-driven productivity gains could eventually compress the number of engineers and tool licenses customers need, and monetization is still unproven at scale. Specific concerns include: Management itself acknowledges it takes two contract cycles to convert AI adoption into higher ACVs. This is a multi-year lag, and if customers use AI to do the same design work with fewer licenses, the revenue impact would be embedded in backlog before showing up in... Bear-case The Agentic AI Question: Amplifier or Substitute?: Management's argument is compelling: agentic AI runs more experiments in parallel, calling underlying tools at higher frequency, and chip complexity is growing exponentially. These factors should increase, not decrease, EDA consumption.; But Cadence itself acknowledges that monetization of AI tools takes two contract cycles to flow through to higher ACVs. That lag means any softening in AI tool adoption - or any... |
| DDOG | Datadog, Inc. | $45.8B | CapRelay current debate section AI-Native Customer Concentration and Revenue Volatility gives a direct AI bear case: The AI-native cohort is Datadog's single biggest risk, and management has already telegraphed it. The cited downside is Datadog's FY26 revenue guidance explicitly models its largest customer growing slower than the rest of the business, implying a meaningful deceleration from this single customer that could drag on reported growth all year. Management modeled this conservatively by design.; The AI-native cohort is inherently lumpy. These customers grow explosively, then optimize usage and... CapRelay current debate section Competitive Moat vs. AI-Driven Disruption gives a direct AI bear case: AI is also creating new attack surfaces on Datadog's competitive position that shouldn't be dismissed. The cited downside is General-purpose LLMs are already capable of anomaly detection, log analysis, and RCA on unstructured data. |
| MDB | MongoDB, Inc. | $20.4B | CapRelay current debate section Competition from PostgreSQL and Emerging Rivals gives a direct AI bear case: Postgres is capturing the market that MongoDB needs, and well-funded new entrants are raising the stakes. The cited downside is Postgres is winning the default database decision for the majority of new application builds, particularly in the AI-native startup community in Silicon Valley - precisely where the next generation of high-growth workloads is being built. MongoDB's own management concedes it "neglected" the Bay Area and had to launch a "reclaim the Bay" campaign to rebuild mindshare it should... CapRelay current debate section Enterprise AI Application Workload Ramp gives a direct AI bear case: Enterprise AI remains stuck in experimentation, and the AI revenue contribution keeps getting pushed out. The cited downside is MongoDB's CEO stated clearly, as recently as the FY26Q4 earnings call, that... |
| FROG | JFrog Ltd | $5.5B | Debate Security Suite Adoption and Competitive Displacement: OpenAI recently launched a security tool that replaces SAST capabilities - one of JFrog's own features - which signals that AI-native security tools could undercut JFrog's pitch. Specific concerns include: JFrog is competing against deeply entrenched vendors with dedicated security sales forces, large R&D budgets, and established CISO relationships. Palo Alto, CrowdStrike, and others are converging on software supply chain security with significant... Bear-case The Core DevOps Franchise Is Maturing: Enterprise customers are explicitly citing AI infrastructure cost unpredictability as a reason to delay cloud commitment decisions. Rather than accelerating cloud migration, AI is extending the decision cycle. CIOs told JFrog in FY25 they are shifting from "mega cloud migration" initiatives to "fit-for-purpose hybrid" deployments - which is structurally better for JFrog's hybrid positioning, but worse for cloud consumption... |
| GTLB | Gitlab Inc. | $3.7B | Debate Competition with GitHub and AI-Native Coding Tools: GitHub's Microsoft ecosystem advantage is structurally hard to overcome, and AI is creating new attack surfaces on GitLab's turf. Specific concerns include: OpenAI models power GitHub Copilot, and Microsoft's close partnership gives GitHub preferential access to frontier AI. GitLab, by contrast, sources its models from third parties and holds no exclusive AI partnerships.; New entrants like Cursor and Claude Code are capturing developer mindshare at the... Bear-case GitHub's Structural Advantages Are Widening in the AI Era: GitHub Copilot launched before GitLab Duo and has a much larger installed base of AI users. Microsoft's Azure and Office 365 ecosystem gives GitHub a natural enterprise entry point that GitLab cannot replicate.; GitHub benefits from close ties to OpenAI, giving it preferential access to frontier model capabilities. GitLab relies on third-party model vendors with no exclusive partnerships. |
| BRZE | Braze, Inc. | $2.9B | Debate AI Competitive Differentiation vs. New Entrants: Braze's AI lead is recent, can erode quickly, and faces better-resourced competitors. Specific concerns include: Braze's AI features - Operator, Agent Console, Decisioning Studio - are largely new. Operator and Agent Console just reached general availability. First-mover advantages in software AI features tend to be short-lived when better-resourced competitors, like Salesforce with Agentforce or Klaviyo with its Shopify-integrated AI stack, accelerate... |