Rohit Mittal
CEO of Helium Ventures, VC-Backed Company Acquisitions
Cloud SaaS repricing is not stopping anytime soon because AI is moving so fast that no industry is safe. Security stocks such as Cloudflare, Palo Alto Networks, CrowdStrike and Zscaler fell 25% in a few days after Mythos was released, and better models from others are coming soon. Cloud software stocks are down 30% this year and 50% over five years. x.comÂ
Buying software stocks is the way to position against AI hype, since the AI narrative has devalued them. AI looks massively overhyped and its valuations cannot prove out, but shorting is impractical with volatility premiums this high. Agreeing with @RealZoltanT that SaaS is the best long if you want to be short OpenAI and Anthropic, SaaS is set to rip. x.com x.com x.comÂ
Software companies face a multi-year repricing because AI has removed the scarcity that made software valuable. For 20 years building software required engineers, letting SaaS firms charge monthly fees forever; when anyone can build a working product in a weekend with Claude, software stops being the moat. Atlassian is down 75%, HubSpot 69% and Figma 86%, and layoffs show headcount no longer scales with output. Survivors will be those with distribution, data and trust underneath. x.comÂ
Software stocks look set to benefit as the market signals that open-weight models will help software and that AI extinction risk was overblown, with AI instead a net positive for growth and margins in coming years. The cloud software ETF is becoming one of the best-looking charts in the market and is close to clearing $37 resistance and a volume shelf. x.comÂ
Cybersecurity stocks' large run-up is at risk from a cheaper, open Chinese AI model. ValsAI reports that DeepSeek V4.1 Flash 1017 beat GPT 6 Sol on its offensive cybersecurity task; DeepSeek is far cheaper to run than a closed US model. x.comÂ
Usage-driven AI stocks are less exposed to frontier-model companies as the AI trade separates winners from losers. All rose, all will fall, and only some will recover; favor usage-driven over frontier-biased stocks. DigitalOcean is GAAP profitable; its token usage is a nonlinear revenue proxy. SiTime is GAAP profitable; robotics, defense and self-driving are each growing 50% year over year. CrowdStrike and Rubrik also do not rely on frontier models, though valuation is a concern. x.com x.comÂ
Cybersecurity stocks such as CrowdStrike and Palo Alto Networks are set to de-rate in 2027 as Anthropic and OpenAI aggressively productize frontier models into cybersecurity. The whole cyber industry would face competitive risk and lose budget share. Frontier models have a tangible edge over apps in cyber, and the labs need such revenue to justify their valuations. Palo Alto's Nikesh Arora and CrowdStrike's George Kurtz may talk big now; their demeanor 12-18 months from now is the test. x.com … Show more
Cybersecurity stocks such as Palo Alto Networks, CrowdStrike and Zscaler face a sea of red once pod shops absorb the AI-disruption risk, given the sector's valuations. Investors at Slow Ventures say they are increasingly convinced AI's first major victims will be these cybersecurity incumbents, whose security stack may not survive an AI-native threat world. x.comÂ
Zeta Global, ServiceNow, Datadog and CrowdStrike look increasingly attractive among software names as AI disrupts applications: businesses with proprietary data that build on AI and deliver intelligence and measurable outcomes could win. Cybersecurity valuations remain a concern. x.comÂ
SaaS will have its time in the sun again, and now, while it is still hated, is the time to pick favorites and build positions. Bearish sector narratives don't last forever, and the view that all SaaS is being disrupted should ease somewhat this summer, though multiples may not return to prior levels. Some SaaS names have little moat against Anthropic and peers, but many, especially cybersecurity and Palantir, are strengthened by the agentic push. x.com x.com x.comÂ
Security software is arguably more of an AI bubble than semiconductors. Firms are bid up to nosebleed multiples on moderate AI-related revenue acceleration while facing medium- to long-term risks from Microsoft, Google and the AI labs and a 10-year Treasury yield above 5%. Their forward P/Es rest on non-GAAP earnings that exclude large stock compensation. A reversal could come from results missing high buy-side bars, security share gains by Microsoft or Google, or AI lab security launches. x.com … Show more
