AI & Software Market Review

The S&P 500 and NASDAQ reached all-time highs in Q2, but 2026 has been a difficult year for software relative to the broader market. The broad IGV software index fell 24% in the first quarter, continuing the SaaSpocalypse-driven sell-off of late 2025, before recovering some ground in Q2 on an earnings rebound from select names and the April ceasefire to end the first half down 14%.

Q2 was the NASDAQ's strongest quarter since the 2020 rebound, up 21%, as the market rotated into the beneficiaries of hyperscaler capex: datacenter enablers (+100% YTD) and semiconductors (+60%) led.

The AI trade continues to disperse as investors separate the beneficiaries of AI disruption from the losers: the YTD return spread between datacenter enablers (+100%) and applications (-37%) is nearly 140 points. The mechanics differ by layer. The infrastructure rally is estimate-led, with semiconductor forward earnings growth accelerating to 82%, while the software repricing in both directions has occurred on multiples alone: applications have de-rated to the cheapest layer in the stack and cybersecurity has re-rated to 67x forward earnings, each on essentially unchanged near-term growth expectations.

The defining development of the year so far has been the emergence of the model harness layer, which has made agents broadly available and configurable by nontechnical users. The resulting demand inflection, and a deeper appreciation of AI's potential, are behind the widening divergence in the AI trade. SaaS applications will continue to face disruption, but to be clear they will continue to require a deterministic, transactional core separate from the probabilistic AI layer. The transactional layers are, to widely varying degrees, the moat. This moat can be thin for simple workflows and CRUD applications and low-end of the market use cases where they are more easily replaced vs more complex workflows with high levels of integration and high data gravity. AI needs to be built in starting at the data layer with clean data and governance to feed the probabilistic modeling that adds a layer of value, agentics, and reasoning that makes the new generation apps more robust beyond the precision required of transactional systems. AI is probabilistic and not well suited for recording transactional data, so to assess a SaaS company's moat we need to look at the complexity of transactional workflows and how actively a company is adding value on top with AI, which we explore in a separate document to follow.

The demand inflection for AI has pushed hyperscaler capital expenditures to a projected $739bn in 2026, up 78% year over year and 37% of revenue, and GPU rental pricing up roughly 40% YTD, with on-demand capacity effectively unavailable since February. If we include capital spending on AI more broadly from neocloud vendors and others the figure is expected to be a bit in excess of $1.0T, or about 2% of US GDP. Compute scarcity is driving up inference costs and token economics is becoming a bigger issue. This is demonstrated by Anthropic's push toward metered APIs and usage-based pricing, which has driven its meteoric revenue growth over the past 12 to 18 months. Efforts to drive greater inference efficiency and optimize token economics is a key issue for the industry which we also explored in a follow-on paper.

Private market activity remains robust as AI-native players build to scale at high speed. Capital has concentrated at the top, with late-stage and mega-rounds taking 84% of 2026 dollars, led by OpenAI's $110bn and Anthropic's $65bn raises, while the base remains broad by count. The result is a bifurcated venture market: Series A step-ups sit at record highs while Series C cools and the graduation funnel narrows, raising the question of whether the broad early-stage cohort will secure the later-stage capital it will need.

We are likely entering an IPO supercycle: SpaceX debuted in June as the largest IPO in history, with Anthropic expected to follow within months and OpenAI possibly a bit later. The exit environment is recovering off its trough, with exit volume reaching $213bn in 2025 versus under $60bn a year in 2023 to 2024. Breadth, however, remains narrow. Strategic M&A accounts for roughly three-quarters of exits by count, while public listings (one in eight exits) generally deliver more than half of exit dollars. Should the window reopen in earnest, a deep pipeline of large, high-quality names is waiting.