conf 5/10 partial
enthusiasm:27.0 · trend:8 · quantifies:5 · impact:0 · under_radar:5 · credibility:0 · business_impact:8 · commitment:6 · confirmation:3
Enthusiasm latest 9 / prev 9 (rising)
Robinhood frames AI as a core business lever across customer products, internal productivity, support automation, engineering velocity, and market access to AI companies. The thesis is credible because management ties it to realized efficiency, employee adoption, support resolution, Cortex usage, and engineer productivity rather than only roadmap language. Enthusiasm is rising slightly as Q1 FY2026 adds broader workflow adoption and agentic product ambitions on top of Q4's Cortex and support automation claims.
Grounded on actual base — revenue $4.5B · net income $1.9B · net margin 42.1% · diluted EPS 2.05
These are next-fiscal-year annual uplift estimates, not next-quarter numbers.
Aggregate next-FY est. rev uplift: 0.0% · next-FY EPS uplift: 4.2% · vs analysts: unclear · priced in: medium · confidence: 5/10
| Claim | Figure | Arithmetic | Next-FY Rev % | Next-FY EPS % |
|---|---|---|---|---|
| Cortex used by ~1M customers engagement · soft | ~1,000,000 customers | Adoption/engagement metric with NO disclosed monetization (Cortex is a free in-app AI feature). No ARPU, take-rate, or attach revenue given -> cannot map ~1M users to revenue against the $4.473B base without inventing a number. Soft. | ||
| $100M efficiency in CX + software engineering (FY2025) cost | $100,000,000 | Cost saving, mostly bottom-line. after_tax = $100M*(1-0.21) = $79M. eps_uplift = $79M / $1,883M NI = 4.20% (~$0.086 of $2.05 EPS). Topline = $0 / $4.473B = 0.00%: cost avoidance, not AI-sold revenue. | 0.0 | 4.2 |
| 9-figure efficiency benefits already generated (eng + CX) cost · soft | 9-figure (>=$100M) | Restatement of the same efficiency program as the $100M figure. Sized once to avoid double-counting; excluded from aggregate. 9-figures implies a $100M floor, so $100M is conservative. | ||
| 9 figures in savings/efficiency in 2025 alone cost · soft | 9 figures (>=$100M) | Same 2025 efficiency pool restated. Could be >$100M but no precise number -> kept as the same anchor, not additive. Marks upside to the $79M after-tax figure. | ||
| Over 75% of CX cases solved by AI (Q4) productivity · soft | 75% of cases | Operational driver BEHIND the $100M efficiency, not a separate dollar pool. No case volume/cost-per-case disclosed. Already captured in the $79M after-tax figure. | ||
| Volumes +50% while CX hiring flat (FY2025) productivity · soft | +50% vol, flat CX headcount | Cost-avoidance mechanism (capacity absorbed without added FTEs). Dollar value already inside the $100M efficiency; no avoided-hiring dollars disclosed. The +50% volume is market/product-driven, not AI revenue. | ||
| Commits per engineer up 50% since early FY2025 productivity · soft | +50% commits/engineer | Engineering throughput proxy; feeds the software-engineering portion of the $100M. No standalone $ or payroll bridge -> not separately sizable. | ||
| >90% of employees using AI tooling productivity · soft | >90% adoption | Adoption breadth, no dollar figure or cost-base bridge. Leading indicator for future efficiency, not yet monetizable. Soft. |
Assumptions: Tax rate 21% (default; no AI claim of higher-margin software revenue, so cost savings flow at the statutory shield). Incremental margin would default to the current net margin (~42.1%) but no hard AI revenue claim was given. The lone hard dollar figure ($100M FY2025 efficiency) is sized once as a cost saving: $100M*(1-0.21)=$79M after-tax / $1,883M NI = 4.20% EPS. All other AI claims are restatements of, or operational drivers behind, that same pool, or pure adoption/engagement metrics with no disclosed dollars -> set to null to avoid double-counting. No AI revenue is monetized: Cortex is a free feature with no take-rate/attach disclosed, so topline AI uplift ~0. Phasing: the $100M is already embedded in the FY2025 base; management says 2026 is 'much bigger' but gives no FY2026 number, so incremental next-FY uplift is unquantified upside, not a hard add.
Top line: Effectively zero quantifiable topline AI uplift. The only AI 'product' figure is Cortex usage (~1M customers), a free engagement feature with no disclosed monetization, take-rate, or trading-volume bridge — it cannot be mapped to a revenue line. All other claims are cost/efficiency. AI is enabling capacity (volumes +50% absorbed with flat CX hiring) but that volume is market-driven, so no incremental AI-sourced revenue can be booked.
Bottom line: A pure adopter, cost-side story. The single hard figure — $100M of FY2025 efficiency in CX and software engineering — equals $79M after tax = 4.20% of current net income (~$0.086 of $2.05 EPS). >75% of CX cases auto-resolved and +50% commits/engineer are the mechanisms behind that pool, not additive dollars. The repeated '9-figure'/'9 figures in 2025 alone' language is the same pool restated, so sized once. Magnitude ~4.2% of earnings, but ~all of it is already inside the FY2025 base; the forward kicker ('2026 much bigger', >90% employee adoption) is real but unquantified.
Consensus FY2026 revenue is $4.989B (+~11% vs $4.473B) while consensus net income falls to ~$1.722B (down ~8.5% vs $1,883M) — the Street bakes in margin compression, not expanding AI leverage. The identifiable AI impact is not revenue but ~$79M after-tax savings = 4.20% of the current earnings base, enough to offset ~49% of the consensus net-income decline. Management guides the efficiency program 'much bigger' in 2026 with >90% employee adoption ramping, which argues for under-modeled upside (low). But the $100M was explicitly described as already generated in 2025 and is likely embedded in the base, and the forward increment is unquantified — so on balance the benefit is closer to partially priced-in (medium) than clearly mispriced.
partial
Adopter, cost-only story: one hard $100M efficiency = ~4.2% EPS, ~0% topline; verdict softened to medium and confidence lowered on the priced-in split.
| Field | Opus 4.8 | GPT-5.5 |
|---|---|---|
| Rev uplift % | 0.0 | 0 |
| EPS uplift % | 4.2 | 4.1954 |
| Priced in | low | medium |
| vs analysts | unclear | unclear |
| Confidence | 5 | 6 |
| Top line | Effectively zero quantifiable topline AI uplift. The only AI 'product' figure is Cortex usage (~1M customers), which is a free engagement feature with no disclosed monetization, take-rate, or attach revenue — it cannot be mapped to a revenue line. All other claims are cost/efficiency, not revenue. AI is enabling capacity (volumes +50% absorbed with flat CX hiring) but the volume itself is market-driven, so no incremental AI-sourced revenue can be booked. | No hard AI revenue dollars were disclosed. Cortex usage of nearly 1M customers is potentially relevant to engagement, but with no monetization or trading-volume bridge, calculated next-FY revenue uplift is 0.00% from quantified claims. |
| Bottom line | This is a pure adopter, cost-side story. The single hard figure — $100M of FY2025 efficiency in CX and software engineering — equals $79M after tax = 4.20% of current net income (~$0.086 of $2.05 EPS). >75% of CX cases auto-resolved and +50% commits/engineer are the mechanisms behind that pool, not additive dollars. The repeated '9-figure'/'9 figures in 2025 alone' language is the same pool restated, so it is sized once. Magnitude of the program ≈4.2% of earnings, but ~all of it is already inside the FY2025 base; the forward kicker ('2026 much bigger', >90% employee AI adoption) is real but unquantified. | The hard quantified impact is the AI efficiency savings pool. Counting the explicit $100M/9-figure savings once: $100M * 79% = $79M after-tax, and $79M / $1.883B net income = 4.20% EPS uplift. |
| Reasoning | Consensus models EPS DOWN ~9.4% in FY2026 ($1.846 vs $2.038) and net income down ~9.2% ($1,722M vs $1,897M), despite revenue +10.1% ($4.99B vs $4.53B) — i.e. the Street bakes in margin compression, not expanding AI leverage. Management explicitly guides the efficiency program 'much bigger' in 2026 with >90% employee AI adoption now ramping. A repeat/expansion of the $100M pool is worth ~$79M after-tax ≈ 4.6% of FY2026 consensus net income — enough to roughly halve the expected earnings decline. Because consensus shows earnings falling rather than benefiting from this leverage, the AI cost benefit looks under-modeled, not priced in. The caveat that caps confidence: the headline 4.2% is mostly already in the base, and forward incremental is unquantified. | Consensus 2026 revenue is $4.989B, up $516M or 11.53% versus the $4.473B base, while consensus 2026 net income is $1.722B, down $161M or -8.54%. The identifiable AI impact is not revenue growth, but about $79M after-tax savings, equal to 4.20% of the current earnings base. That saving would offset roughly 49.0% of the consensus net-income decline, but the claims do not clearly prove upside to consensus because the savings were already described as generated in 2025 and may already be embedded. |
Rows highlighted where the two models disagreed.
—/100 (no quantified promises) no-quantified-promises 6 calls reviewed
Across these six calls, Robinhood discussed real and ramping AI efforts (Cortex AI tools, Cortex Assistant, AI-driven custom indicators, personalization, the Strategies robo-advisor, and internal AI-driven engineering/support efficiency), but never attached a specific number AND a timeframe/milestone to any AI capability — AI statements were either qualitative aspirations or retrospective results (e.g. Cortex used by ~1M customers, '9-figure' internal savings) rather than prior quantified targets. The quantified targets they did set (1M gold cards/$100M ARR, 20%+ net deposit growth, prediction-market volumes) are not AI-related, so there is no judgeable AI promise-vs-delivery track record.
Est. revisions falling · Fwd P/E 43.2 · EV/Sales 19.0x
AI claim maps to Transaction-Based Revenues, Gold Subscription Revenues, Other Revenue
AI enthusiasm across 6 calls — trend ↗ rising
AI moved from minimal mention to Cortex adoption, AI trader tools, customer assistants, and internal productivity claims.
7/10 qualitative impact material near-term · mixed evidence
Where AI matters: CX automation, engineering productivity, AI investing tools
Robinhood has credible deployed AI with hard operating evidence: $100M of CX/engineering efficiency, over 75% AI case resolution, flat CX hiring despite 50% volume growth, and broad employee adoption. Product-side AI such as Cortex and agentic trading could affect engagement and trading behavior, but monetization is not yet disclosed, so this is material rather than clearly transformational.
Caveats: Cortex usage has no disclosed revenue, ARPU, trading-volume, or retention bridge; AI trading/advice features face regulatory, suitability, and trust constraints; Reported savings may already be embedded in the cost base and not incremental next year; Engineering commit growth is a proxy, not a direct P&L metric
option liquidity: good
proxy inputs — dollar-ADV $2.6B · beta 2.294 · px $88.16
source: proxy (no options chain on FMP)
FMP /stable/ exposes no options-chain endpoint on this key, so ATM IV, bid-ask spread and open interest are unavailable. Liquidity below is a PROXY from dollar-ADV, beta and price level (a stand-in for option depth), not measured option-market data.