YB new stock pitches (Fri, Jul 17)

Kaspi.kz ($KSPI), a Kazakhstan-based Super App that functions as the digital operating system for the country, trades at just 7.6x PE despite compounding revenue ~30% annually over the past decade and guiding for ~20% YoY growth in 2026, with a reverse DCF implying essentially zero priced-in growth. Founded in 2008 after a Baring Vostok PE investment, Kaspi built a two-sided ecosystem serving ~15.7 million consumers and ~800,000 merchants across three integrated platforms—Payments (QR checkout, B2B, bill pay, biometric Kaspi Alaqan), Marketplace (e-commerce, e-grocery via 90% stake in Magnum, travel, dominant classifieds like Kolesa and Krisha), and Fintech (BNPL, cash/car/merchant loans funded by deposits, approving 99.9% of loans in under six seconds)—plus non-monetized government services that deepen its moat and drive 77 transactions per active consumer per month. TTM segment revenues are ₸2.10T (Marketplace), ₸1.63T (Fintech), and ₸670B (Payments), with 2025 net income of ₸280B, ₸355B, and ₸433B respectively; Q1 2026 saw e-commerce GMV up 41% YoY, marketplace take-rate expanding to 12.1%, a low 0.7% cost of risk, 6.6% NPLs, 92% deposit rollover, and a 64% dividend payout ratio (high-single-digit yield). Margins have compressed due to consolidating recently acquired Turkish e-commerce player Hepsiburada (76% stake, $1.13B, run near breakeven to gain share), higher deposit funding costs from elevated rates, a payments mix shift toward lower-fee QR/B2B (take-rate down to 1.03%), and higher banking taxes (20%→25%) and reserve requirements—though management views these as cyclical/mix-driven and expects recovery as rates fall and value-added services scale. Kaspi dominates competitively with 78-79% payment network share and far higher brand awareness than rivals across all segments (nearest bank comparable is Halyk). Key risks include Kazakhstan macro/political exposure, credit risk on unsecured loans, deposit funding/bank-run risk, Hepsiburada and e-grocery execution risk, and a government-mandated unified QR system by July 2026 threatening its closed-loop payments monopoly (which Kaspi combats via integration with rival banks, AliPay+, Kaspi Alaqan, and government services). Insiders own 46.18% (CEO Mikheil Lomtadze 22.62%, Chairman Vyacheslav Kim 20.78%), Baring holds 23%, and Tencent recently bought a stake alongside the CEO personally adding shares. The author views Kaspi as a one-of-a-kind, fast-growing monopoly priced like a dying business with a dividend floor, attractive for those comfortable with the country risk. NFA.

Read the full article here. Read time: 11 min

Trustpilot Group plc (LSE: TRST), presented as a long at the Sohn Monaco Investment Conference 2026 from a £2.50 entry (~$1.3bn EV, ~22x 2026 EBITDA), is the only horizontal, open review platform at scale, operating a two-sided consumer/business network with a self-reinforcing flywheel that lowers acquisition costs on both sides. The company serves high-consideration purchase categories (money/insurance, travel, tech, health) and has compounded revenue at a 24% CAGR from ~$48m (2017) to ~$260m (2025), with a $50bn TAM and $19bn SAM penetrated below 5%—including $9.7bn North America (only $62m bookings), $1.7bn UK ($116m bookings), and $7.4bn Europe/RoW ($113m bookings)—with the mature UK demonstrating what the underpenetrated, fastest-accelerating US market could become. Competitive advantages include counter-positioning as the sole horizontal+open platform (vs. closed Amazon/Google Maps/Booking/Airbnb and vertical G2/Glassdoor/Capterra), winner-take-most network effects (having overtaken BBB and left Yotpo/BazaarVoice/Birdeye negligible), and scale-based regulatory and anti-fraud moats built over 10+ years using LLM-based NLP, graph neural networks, and sequence models. Contrary to the 'AI is destroying software' bear case, AI makes Trustpilot more valuable via its proprietary data (361m active reviews, 1.3m claimed domains, ~200k new reviews daily) and the SEO-to-AEO shift, becoming the 'truth layer' for answer engines—ranked 5th most-cited domain on ChatGPT in January 2026 (ahead of Reuters, GitHub, Booking.com), heavily referenced in Google AI Overviews, with AI-citation-driven traffic growing from ~75k monthly visits (early 2023) to ~500k+ (early 2026) and search impressions up 45% YoY. Founder Peter Holten Mühlmann retains a 3% stake and board seat, while proven operator CEO Adrian Blair (appointed September 2023) has made substantial insider purchases and scaled buybacks (from zero in 2H FY23 to ~$48m in 2H FY25), retiring ~10% of shares (from ~463m to ~418m) in two years. Underwriting deceleration to a 20% revenue CAGR (~$300m 2026 to ~$650m 2030) and EBITDA scaling from ~$60m (20% margin) to ~$200m (30% margin), the base case (multiple compressing 22x→15x) yields a 2.3x MoM/>30% IRR at a £5.70 price target, and the bull case (25x) yields 3.7x/>50% IRR at £9.20—an attractive asymmetry since growth and margin expansion alone deliver returns even with multiple compression. Recommendation: Long TRST LN.

Read the full article here. Read time: 5 min

PlasCred Circular Innovations ($PLAS.CN), a C$25M fully diluted Canadian microcap trading at $0.24, is a pre-revenue mixed-plastic pyrolysis company claiming to have solved the 'Holy Grail' of converting unsorted Types 1-7 plastics into condensate/naphtha feedstock via a proprietary process operating at just 350°C and 10 psi (vs. peers' 500°C+ and, for Aduro, ~2,500 psi), using catalysts and a two-phase (liquid then gaseous) reactor design that minimizes coking, toxic byproducts, and waxes while producing a clean, sulfur-free, olefin-containing oil at $34/bbl processing opex for 72% gross margins against a 5-year binding Trafigura offtake for 2,000 bbl/d at $120/bbl (Trafi, Shell, Exxon, and Dow all conducted DD and offered offtakes); the company has a Circular Materials feedstock deal (with pre-processing and potential tipping fees that management says could cut opex near zero) and a valuable logistics partnership with CN (Canadian National Railway), leasing land at Scotsford, AB for its first commercial plant. The tech is validated by a 400 kg/day, 2 bbl/day Primus pilot running in batch since May 2023; next is Neos (100 tonnes/day feedstock, 500 bbl/d, $25M capex, $11M+ EBITDA, 77% pre-tax IRR, 2.3-year payback, targeted mid-2027 commissioning, FID expected Q4 2026), followed by Maximus scaling in three phases to 2,000 tonnes/day and 10,000 bbl/d. PlasCred needs ~$6M more to bring Neos online (funding sought through 1Q28, possibly non-dilutive; ~$9M warrants callable above $0.40 could remove the need), and CEO Troy Lupul (two prior successful water/wastewater exits) owns ~22% (~14% FD) while running a tight ~$50k/month burn. A conservative bull model (Neos ramping to 85% utilization, Maximus phases funded half-debt/half-equity at 15x EV/EBITDA) yields a path to $200M+ EBITDA by 2033 and a 100x+ current fully diluted market cap, with upside optionality from plastic credits, tipping fees, faster growth (multiple US states courting PlasCred), and a favorable comparison to Aduro (C$676M, ~25x PlasCred, lacking offtake, feedstock deal, pilot, and unit economics). Key risks, earning a 4.5/5 'Spicy Rating,' include the massive 250x scale-up from Primus (2 bbl/d) to Neos (500 bbl/d)—where prior pyrolysis efforts have failed—the shift from batch to continuous processing, problematic real-world contaminated feedstocks, unsecured Neos/Maximus financing and future dilution, the June 2026 departure of CTO Wayne Monnery (attributed to no longer needing a chief chemist in the build phase), no patent protection yet (broad application filed, trade secrets withheld), industrial accident risk, and the critical fact that Neos is the company's only shot—if it fails, the stock is worth zero, so position sizing must be sensible.

Read the full article here. Read time: 15 min