Архив отчётов
Вся аналитика
Страница 51 из 52 · 1238 отчётов
AI Data Security, DSPM, RAG Permission Governance, and Enterprise Knowledge Base Security
Data security is shifting from a compliance afterthought into the primary control plane for putting AI into production—Copilot, Azure AI Search, Unity Catalog, Snowflake Horizon, and Bedrock Guardrails all push ACLs, labels, classification, and retrieval authorization upstream into the call chain. RAG and agents do not rebuild the permission system; they amplify the existing over-sharing baked into SharePoint, email, CRM, and databases, with DSPM, DLP, DDR, access governance, and RAG permissions strung into one chain. Machine identities already outnumber human ones 82:1 (CyberArk), and CrowdStrike's $740 million acquisition of SGNL pulls NHI and AI identity into continuous control. Watch rating: the durable profit pool sits in the control plane (permission graph, label engine, policy engine, retrieval authorization, AI gateway) rather than in scanning, with direct beneficiaries backed by financial evidence including Varonis, CyberArk, Snowflake, MongoDB, Elastic, Trend Micro, Cloudflare, CrowdStrike, Palo Alto, Microsoft, and Google Cloud, while AI-native challengers such as Cyera (valued at $9 billion), BigID, Sentra, Concentric, Securiti, Privacera, Veza, Noma, and Lasso carry strong narratives but thinner revenue proof.
AI Clinical Trials and the Restructuring of the CRO Value Chain
In clinical development, AI is no longer a question of whether it works, but of which layer the value lands on. The first revenue does not flow to fully autonomous clinical AI, but to modules embedded in the workflow that connect to real data and remain auditable: patient screening, trial feasibility, site selection, EDC/eCOA/eTMF, risk-based monitoring, RWD/RWE, pharmacovigilance, and partial document automation. The profit pool sits in workflow control (Veeva subscriptions; Medidata/Oracle/Medable platform fees), data access (Tempus, TriNetX, Truveta, Flatiron, Aetion), and compliance capability. AI clearly compresses operational time (Syneos site startup from months to 24-48h; Flatiron 37 seconds per CRF; Medidata listing generation -90%, review -80%; Medable eCOA build from days to 30 minutes), but the evidence that it shortens biological-uncertainty time or lowers failure rates remains weak. The FDA released its AI decision-support guidance framework in 2025, pushing the sector into a phase of dual regulatory and commercial validation. Rating Overweight: a value chain where data uniqueness, workflow depth, compliance, and clean billing decide who captures the durable profit pool.
AI Content and the Creative Supply Chain
AI content and creativity has moved from "model demos" into "workflow commercialization," and the profit pool flows mainly to platform companies: Adobe posted $23.769 billion in FY25 revenue, $19.2 billion in Digital Media ARR, and over $250 million in Firefly ARR by Q1 FY26; Canva and Figma have embedded AI into the design OS; in a single quarter Meta saw over 1 million advertisers use GenAI tools to produce 15 million+ ads; Kuaishou's Kling booked RMB 340 million in 2025Q4 while AIGC marketing assets drove RMB 4 billion of online marketing spend; Meitu's 2025 imaging/video/design revenue reached RMB 2.954 billion with 16.91 million paying users. Copyright and training-data licensing is rising into a new revenue line: Getty and Shutterstock lead with commercial-use rights plus indemnity plus licensed data, UMG and WMG are shifting from "suing AI" to "conditional licensing," and Disney/Universal/Warner are suing Midjourney while OpenAI has retired its old Sora. Still in pilot: feature-length video generation, open AI-music platforms, standalone consumer AI apps, and AI virtual idols. Top risks: copyright litigation, inference cost and price wars, and platform free-tier substitution. Rating Overweight: the durable winners are integrated platforms that own distribution, workflow, budget access, and copyright governance, not point-solution generators.
AI in Cars and Autonomous Driving: An Investment Study
AI in cars has crossed from demo-grade into revenue-grade, but the revenue structure is sharply tiered. The first to monetize are L2/L2+ ADAS and the "pick-and-shovel" suppliers—GM Super Cruise (FY26 guidance near $400 million, roughly 70% gross margin), Qualcomm automotive (Q2 $1.3 billion, +38%), Mobileye (2025 $1.894 billion, +15%); Robotaxi is genuinely paid but still in the early-scaling phase (Waymo / Apollo Go / Pony.ai / WeRide). Rating Watch: over the short-to-medium term, profit is more likely to flow to ADAS fitment, in-vehicle AI chips/domain controllers, and the few fleet platforms that have already closed the loop.
AI Agent Platforms, Enterprise Automation, Agent Operating Systems, and the Workflow Rebuild
The enterprise Agent moat does not sit in the model; it sits in "permissions-aware data access + business-system entry points + approval/audit/observability." The financial signals already on the ground: Microsoft's AI annualized revenue tops 37 billion, commercial cRPO reaches 627 billion; ServiceNow's 2026 AI contract ACV could break 1.5 billion, with Now Assist large customers up 130% year over year; Palantir Q1 2026 total revenue +85%, U.S. commercial +133%; Salesforce Agentforce has commoditized its pricing across $2 per conversation, Flex Credits, and $5 per seat. Gartner expects 40% of Agent projects to be cancelled over unclear ROI, with "Agent washing" clearly visible. Key names to track: Microsoft / Salesforce / ServiceNow / Palantir / Atlassian / HubSpot / Workday / Google Cloud. Rating Watch: revenue proof is concentrating in a few platform-layer leaders, while adoption pace and ROI clarity—not model capability—remain the swing factors for the rest.
AI Advertising and Marketing Automation: A Value-Chain Study
AI advertising and marketing automation has moved from "showcasing model capabilities" to a "revenue-execution layer." The first profit pool still belongs to traffic-gateway platforms: Google AI Max lifts conversions 14%, Meta's AI ad infrastructure runs at an annualized rate above $60 billion, Amazon Ads totaled roughly $68.6 billion for full-year 2025, plus AppLovin Axon and The Trade Desk Kokai at 5x ROAS. The second profit pool belongs to software platforms that own first-party data, the content supply chain, and workflows: Salesforce Agentforce ARR of $800 million (Agentforce + Data 360 above $2.9 billion), HubSpot Q1 2026 revenue of $881 million, Klaviyo NRR of 110%, Adobe FY26 Q1 AI-first ARR up more than 3x year over year. On the China side, Baidu's AI-native marketing reached ¥2.8 billion in Q3 2025, up 262%, and ¥2.7 billion in Q4, up 110%; Alibaba's "Quanzhantui" penetration is lifting customer-management revenue. On the agency and data side, Publicis acquired Lotame, covering 1.6 billion IDs across 100+ data sources. Standalone, single-point AI creative, AI SDR, and pure generation tools detached from data and channels are the most exposed to being made free by platforms or eroded by price wars. Rating Overweight: the durable profit pools sit with platforms and data-rich software, not with point tools.
Rocket Lab: A Long-Term Owner's Perspective
An evolution from a small-rocket maker into a Launch + Spacecraft + Critical Components + Mission Solutions platform. 2025 revenue was 602 million, with Space Systems at 67%; Q1 2026 revenue reached 200 million at a 38.2% GAAP gross margin, and backlog hit 2.2 billion (+20%). Yet 2025 operating cash flow was -166 million, CapEx 156 million, and free cash flow deeply negative. At $124.77, the diluted market cap of roughly $78 billion sits far above even the optimistic range, leaving no margin of safety. Rating Watch: a genuinely capable space platform priced for a far-off, very large success, with Neutron's mass production and unit economics the decisive variable.
61/100
46Buffett
Palantir глазами долгосрочного владельца
Palantir перешёл от модели "высокий рост при хронических убытках" к модели "высокий рост плюс сильный денежный поток": выручка за 2025 финансовый год составила 4.475 миллиарда (+56%), прибыль по GAAP, приходящаяся на акционеров, — 1.625 миллиарда, свободный денежный поток — 2.101 миллиарда; в Q1 2026 выручка выросла на 85%, а коммерческий сегмент США — на 133%. Тем не менее при текущей цене 133.99 рыночная капитализация в 344.5 миллиарда подразумевает скользящий PE около 150.6x, P/FCF около 128x и скорректированный на SBC P/OE около 176x. Это превосходная компания, оценённая почти под безупречность: ею стоит владеть в долгую, но не стоит приобретать по сегодняшней цене. Три сценария DCF дают 18-30 / 35-60 / 75-110, и текущая цена заметно выше даже оптимистичного диапазона. Рейтинг Наблюдать: качественный компаундер, в цену которого уже заложено десятилетие выдающегося успеха.
43/100
Nebius Group: A Long-Term Owner's Perspective
AI cloud infrastructure growth stock: Q1 2026 revenue of $399 million (AI Cloud is 98%), 2026 guidance of $3.0–3.4 billion, a full-year ARR target of $7–9 billion, contracted power of 3.5GW+ (75%+ owned), a multi-year dedicated deal with Microsoft plus a $12–27 billion long-term agreement with Meta. But Q1 operating cash flow of $2.258 billion was driven mainly by $3.2 billion of deferred-revenue prepayments; CapEx was $2.473 billion, free cash flow is still negative, and conservative owner earnings run about -$1.16 billion per quarter. Currently $219.94, market cap ~$55.8 billion, EV/2026E Sales ~17x. The three DCF cases are 30–70 / 80–150 / 160–240, with the current price sitting in the optimistic range.
53/100
78Buffett
Kaspi.kz: A Deep-Value Investment Study
A two-sided Payments + Marketplace + Fintech super-app in Kazakhstan, with 2025 net income attributable to shareholders of KZT 1,073.18 billion, a Tier 1 capital ratio of 19.6%, and a Fitch BBB- rating. At the current $87.78 and a market cap of roughly $16.66 billion, the static P/E is about 7.8x and P/Owner Earnings about 10x, a clear discount to Nu/MELI/Block; that discount reflects Kazakh/Turkish sovereign and regulatory exposure, Hepsiburada integration execution, and governance risk. Rating Watch: a good business at a price that is not cheap.
Nasdaq Composite Index: A Deep-Value Investment Analysis
A basket of U.S. innovation leaders (roughly 3,351 constituents; ONEQ samples 1,026 of them); highly concentrated (top ten ~57.81%, IT + Communications + Consumer Discretionary ~77.56%) with a low cash-distribution rate (30-day SEC yield 0.41%). IXIC currently sits at 26,307 / ONEQ at $103.23, TTM P/E 28.93x, with an apparent earnings yield of 3.46% below the 10-year Treasury's 4.59%. A great asset, but not a cheap price, and the margin of safety is faint. Rating Watch: better suited to long-term growth and index investors than to deep-value buyers.
Ethereum: A Long-Term Owner's Perspective
Ethereum carries roughly half of all stablecoins, 44.1 billion in DeFi TVL, and 33.7 billion in rollup-secured assets. Yet supply has still grown a net 1.177 million ETH since the Merge, an annualized +0.266%, and value capture is weak. At a current price of $2,184, ETH already sits inside the optimistic band and needs a strong narrative to deliver. Rating Watch: an important crypto-finance backbone trading at a fair-to-rich price, not a value opportunity with a margin of safety.
51/100
40Buffett
Duolingo: A Long-Term Owner's Perspective
A high-engagement learning platform with an asset-light model, strong cash flow, and a net-cash balance sheet (cash and investments of $1.253 billion in 2026 Q1). But low switching costs, free substitutes, and AI reshaping the boundaries of learning keep the moat only moderate. At $112.06, the headline PE is distorted by a one-time $256.7 million tax benefit booked in 2025. Ideal buy at $85-100, acceptable to hold at $100-125, clearly expensive above $165. Rating Watch: a high-quality business whose moat is not yet wide enough to buy aggressively without a clear margin of safety.
Bitcoin Through a Long-Term Owner's Lens
A powerful scarce asset, but not a strong cash-flow asset: protocol-native owner earnings are roughly zero, and today's $77,993 price needs an ever-rising monetary premium to hold up, leaving no margin of safety. It resembles a high-volatility, no-yield bet on scarcity more than a traditional value investment. Research rating Watch: understandable but not cheap, better observed than bought on classic value logic.
BNB Deep Value Investing Research
BNB is not Binance equity—no dividends, no buybacks, no liquidation rights; value capture works indirectly through "supply-shrinking burns + ecosystem demand." At roughly $645 and a market cap near $87 billion, BSC's P/F of about 778x sits well above Solana (276x) and Tron (91x); the burn-implied annualized "buyback-like yield" of about 4.4% does not clearly beat the 10Y Treasury at 4.63%. Rating Watch: a quality ecosystem asset whose current price offers no meaningful margin of safety—ideal entry $300-450, acceptable $450-650, overvalued above $800.
40/100
AST SpaceMobile: A Value Investing Deep Dive
A leader in the direct-to-standard-phone satellite race, AST SpaceMobile has secured FCC approval for 248 satellites and partnerships with AT&T, Verizon, and FirstNet, guiding to $150-200 million in 2026 revenue with a target of 45 BlueBird satellites in orbit. Yet 2025 operating cash flow was -$71.52 million and investing cash flow -$1.541 billion, still short of the free cash flow stage; at $83.67 and a $24.32 billion market cap, EV/2026E Sales runs about 136-183x, far above Iridium (7x) and Globalstar (39x). Rating Watch: ideal entry $8-14, acceptable hold $15-25, overvalued above $45.
AI Chip Advanced Packaging, HBM, and Test Equipment: A Deep Dive
The real bottleneck in AI chip supply is the compound scarcity of high-end HBM, CoWoS / large-format advanced packaging, N3 / HBM base die, and test/probe capacity. The most certain revenue and margin upside sits with HBM makers, TSMC's 3DFabric, and test and hybrid-bonding equipment leaders. Rating Watch: track TSMC, SK hynix, Micron, ASE, Advantest, BESI, Samsung, Broadcom, FormFactor, and Unimicron, while treating ABF / glass substrate / broad AI PCB names as sympathy plays that still need orders and disclosed breakouts.
Deep Research on the AI Optical Communications Supply Chain
Within AI optical communications, the themes that have already converted into orders and revenue rank, in order, as 800G pluggables, early-stage 1.6T ramp, AI switches and switch chips, EML/CW lasers, and key optical components and high-end optical manufacturing; CPO/NPO/Optical I/O are mid-term architectural directions whose contribution to financials over the next twelve months remains weaker than 800G/1.6T pluggables. Priority names to track are Arista, Cisco, Broadcom, Marvell, Coherent, Lumentum, Fabrinet, Innolight, Eoptolink, T&S Communications, and Accelink. Rating Watch: the demand is real, but the order in which profit is distributed across the chain will keep shifting, so the highest-conviction value sits with switch chips, key lasers, and the system software layer rather than with commoditized module assembly.
Deep Research on AI Server ODMs, Rack-Scale Systems, Liquid Cooling, and Data Center Power Infrastructure
The second landing point of AI CapEx is shifting from a chip-centric model toward a rack-and-facility-centric one: GPUs and HBM carry the largest value but their profit leverage is already priced in, while order leverage and the profit pool are moving toward rack-scale integration, liquid-cooling core components (CDU, cold plate, QD), power distribution (transformers, PDU, UPS), high-power connectors, and short-reach copper cables. Priority names to track are Vertiv, Eaton, Schneider, Amphenol, HPE, Supermicro, Wiwynn, Quanta, Hon Hai, and Delta. Rating Watch: durable order and profit leverage is migrating from the chip layer to the boundary between the system layer and the facility layer.
AI Storage, Data Platforms, RAG and the Agent Data Layer: A Deep Dive
The AI bottleneck is shifting from "compute" to "data supply, data movement, and data governance." The segments with the greatest direct revenue sensitivity: enterprise SSD / high-capacity HDD, AI storage systems, consumption-based lakehouse revenue, and search / retrieval / data streaming / governance subscriptions. Priority names to track: Microsoft / Oracle / Micron / Dell / MongoDB / Elastic / WD / Seagate / Palantir / Montage Technology; among private companies, watch Databricks / VAST Data / WEKA / MinIO / Qdrant. Rating Watch: a real, monetizable demand shift whose winners are concentrated in storage hardware under tight supply and platform software with sticky governance moats.
AI Application Layer, Copilot, Agents, and Vertical SaaS: A Deep Dive
The AI application layer is moving from "feature enhancement" toward "standalone revenue." The tracks that have already proven a revenue model: AI coding (GitHub Copilot/Cursor), AI customer service (NICE/Five9/Intercom Fin), legal and tax verticals (Thomson Reuters/RELX/Wolters Kluwer), and vertical healthcare plus developer tools. Business models are bifurcating across three billing layers: seat + usage + outcome. Rating Watch: prioritize researching Microsoft / ServiceNow / Salesforce / Intuit / Oracle / Thomson Reuters / Wolters Kluwer / NICE / GitLab / Tempus.
AI Agents: A Deep Dive into the Enterprise Execution Layer, Governance, and Monetization
Agents are moving from "answering" to "acting"—becoming the enterprise system-of-action execution layer. The clearest revenue evidence on the ground is concentrated in Salesforce Agentforce (ARR 800 million, +169%), Microsoft 365 Copilot (20 million paid seats), NICE AI ARR +66%, HubSpot outcome-based pricing, Palantir Q1 +85%, and UiPath ARR 1.853 billion. The hardest layer to displace is not the model itself but permissions, the workflow entry point, connectors, governance, and distribution. Key names to track: Microsoft / Salesforce / ServiceNow / NICE / HubSpot / Palantir / UiPath / Pegasystems / Appian / SAP. Rating Watch: revenue proof is concentrating in a handful of platform and application-layer leaders, while adoption pace—not model capability—remains the swing factor for the rest.
44/100
Zhipu: A Long-Term Owner's Perspective
China's largest independent large-model company, with 2025 revenue up ~132% to RMB 724 million; yet R&D of RMB 4.1 billion, a net loss of RMB 4.7 billion, and a Reuters price-to-sales ratio of ~554x mean HK$1,040 is paying for an extremely distant, extremely optimistic future with no evidence of free cash flow. Rating: Avoid.
48/100
56Buffett
Xiaomi Group Long-Term Owner-Oriented Research
Xiaomi is now a platform hardware company built around smartphones, AIoT, internet services, and a rapidly scaling EV business. 2025 revenue reached RMB 457.29 billion, profit attributable to owners reached RMB 41.64 billion, EV revenue rose 223.8% and moved close to breakeven, while internet services carried a 76.5% gross margin; however, the HK$30.7 share price already pays upfront for stable smartphones, EV scaling, and AIoT expansion. Research rating Watch: a stronger business, but Apple-like moat depth and a clear margin of safety are still absent.