The Trade Desk
Independent demand-side ad-buying platform across connected TV, retail media, mobile, and the open internet — no owned media, positioned against the walled gardens.
The quarter
Q2 2026 (reported Aug 6): revenue $715.1M (+3% YoY) missed ~$752.4M expected; adjusted EPS $0.34 vs $0.40; adjusted EBITDA $241.3M vs ~$262M. The real damage was the guide: Q3 revenue of at least $650M vs ~$805M consensus, and adjusted EBITDA ~$160M vs ~$340M. The stock fell ~26% on the print. Customer retention remains above 95%.
Recent developments
- Spotify ad exchange: Spotify is piloting its own ad exchange and named TTD its first DSP partner, tying TTD to Spotify video inventory via OpenPath and UID 2.0 (reported Sep 17).
- CEO comp: board approved a 10-year performance option for CEO Jeff Green for up to 7M shares, vesting on sustained price thresholds from $18 to $105 (Sep 14).
- Q2 call: Walmart partnership renewed, Netflix joined Sellers/Publishers 500+, Samsung Ads opened home-screen inventory, Koa AI agents launched with Stagwell.
Bull case
- Independent/open-internet model gains share as CTV and retail media keep growing.
- >95% customer retention with strong margins; AI measurement products are real catalysts.
- Spotify/OpenPath deal shows the open-internet plumbing still matters to big publishers.
Bear case
- Growth has decelerated sharply and the Q3 guide reset expectations hard.
- Amazon, Meta, and Google walled gardens plus rival DSPs keep intensifying.
- Cyclical ad-spend pressure and rising AI capex weigh on the multiple.
Shift4 Payments
Integrated payments and commerce tech for restaurants, hotels, stadiums, entertainment, and travel — acquiring, gateway, POS software, and cross-border commerce in one stack.
The quarter
Q2 2026 (reported Aug 6): gross revenue ~$1.30B (+34% YoY) beat $1.24B expected; gross revenue less network fees (GRLNF) $624M (+50.9%) beat $614.9M; adjusted EPS $1.32 beat consensus; adjusted EBITDA $284M beat $278.2M. The catch: full-year guidance was cut anyway — GRLNF $2.48B–$2.53B, adjusted EPS $5.15–$5.35, adjusted EBITDA ~$1.17B at the midpoint.
Recent developments
- Worldline/Bambora North America acquisition closed March 2026 — 140k merchants and 500+ ISV integrations to cross-sell into.
- Global Blue acquisition announced February 2026, extending the travel/luxury retail footprint.
- CEO transitioned to Taylor Lauber earlier in 2026; integration execution is now the story.
Bull case
- 50%+ GRLNF growth with a vertically integrated stack and real switching costs.
- Large cross-sell runway on acquired merchant bases; international expansion compounding.
- Strong free cash flow generation funds the deal machine.
Bear case
- The guidance cut raises questions about how durable the deal-driven growth is.
- Travel and FX sensitivity — the $25M Middle East hit is a reminder.
- Leverage and acquisition-integration risk; consumer-spending dependence.
Nebius Group
Vertically integrated AI infrastructure spun out of Yandex's international assets — GPU cloud, storage, networking, and in-house data-center design. Nebius AI is ~98% of revenue.
The quarter
Q2 2026: revenue $582.3M (+454% YoY) beat $567.9M expected; adjusted loss of $0.12/share was a big beat vs −$0.67 expected (GAAP loss was −$0.68 — mind the two); adjusted EBITDA $236.2M (~40.5% margin); annualized run-rate revenue hit $3B. Q2 capex was ~$5.7B against $2.3B operating cash flow; nearly 70% of Q2 deals included upfront prepayments, with management expecting >$9B of customer prepayments in 2026.
Recent developments
- Q2 call flagged acquisitions — Eigen AI, Tavily, ClarifAI — extending the platform up the stack.
- August: reported unpermitted gas generators and stop-construction orders at the Vineland, NJ data center — permitting risk to watch, not a final legal finding.
- BofA raised its target post-Q2; Michael Burry disclosed a short position around $212 — the valuation debate is live.
Bull case
- Extraordinary demand with sold-out capacity; long-duration contracts plus prepayments fund the build.
- Vertical integration with ~22-month payback on Q2 deals.
- Microsoft infrastructure agreement (Sep 2025) validates the platform strategically.
Bear case
- Extreme capex ($20–25B/yr) with real dilution risk if prepayments disappoint.
- Dependence on large AI customers and Nvidia hardware.
- Permitting, construction, and power execution risk — Vineland is the exhibit.
Rocket Lab
End-to-end space company: Electron small launch, satellite manufacturing and components, mission systems — with reusable medium-lift Neutron in development and an $8B Iridium acquisition agreed.
The quarter
Q2 2026 (reported Aug 10): revenue $234.1M (+62% YoY) beat ~$232M; backlog $2.36B (+137% YoY); GAAP gross margin 36.1% and non-GAAP 41.5%, both above guide. GAAP loss −$0.08/share; adjusted loss −$0.03, in line with consensus. Adjusted EBITDA loss of $8.8M beat the −$20M–$26M guide. Ended Q2 with ~$2.4B cash.
Recent developments
- Sep 15: completed a $1.94B ATM equity raise, fully funding the Iridium acquisition — dilution risk flagged, funding risk removed.
- Sep 11: 16th Electron mission of 2026 and 95th overall; Neutron targeting Q4 pad delivery.
- New contracts: $266M U.S. Space Force launch deal, $397M SB-AMTI spacecraft contract, $160M+ GEO satellite contracts; >$1B of new contracts already booked in Q3. Raymond James initiated at Outperform/$80; Ark bought ~$45M.
Bull case
- Vertically integrated launch + space-systems platform with record government-backed backlog.
- Neutron and Iridium expand the addressable market and add recurring revenue.
- Management sees standalone adjusted-EBITDA profitability after a successful Neutron test flight.
Bear case
- Continued losses and cash burn; Neutron schedule and technical risk is the swing factor.
- Iridium financing, integration, and dilution overhang.
- High valuation leaves little room for execution error.
Oklo
Advanced-nuclear developer: Aurora fast-fission powerhouses, domestic HALEU fuel recycling, and isotope production. Pre-commercial — the story is cash runway, licensing, and build execution.
The quarter
Q2 2026 (reported Aug 7): revenue $1.21M beat ~$0.09M expected — off a tiny base, mainly acquired engineering revenue — while EPS of −$0.28 missed vs −$0.16/−$0.17 consensus. Ended Q2 with ~$3B cash and marketable securities (including $1.9B raised via 2026 ATM programs); YTD operating cash use $65.5M; capex $126.9M.
Recent developments
- Sep 11: announced a new at-the-market offering for up to $1B — shares fell on dilution risk.
- June: letter of intent with Centrus for domestic HALEU supply for up to five Aurora powerhouses, deliveries from 2029, supporting the planned 1.2GW Ohio campus; MOU with Kiewit for EPC planning.
- Context: January 2026 Meta agreement to support the 1.2GW Pike County, Ohio campus with prepayment/funding for Phase 1. Insider selling notable: 103 insider sales, 0 purchases in the last 6 months.
Bull case
- ~$3B cash runway funds the buildout; AI/data-center power demand is the secular tailwind.
- Meta-scale customer pipeline plus vertically integrated reactor/fuel/recycling approach.
- First criticality at the Groves isotope test reactor shows technical progress.
Bear case
- No meaningful revenue and recurring EPS misses; economics still unproven.
- NRC licensing and construction delays are the defining risk.
- $1B ATM means continued dilution; fuel, supply-chain, and grid-interconnection execution risk.