YB new stock pitches (Thu, Aug 20)

Dong A Eltek (088130.KQ), a Fairlight Alpha Fund portfolio holding, is a Korean manufacturer of deposition equipment that layers organic molecules onto OLED and OLEDoS screens used in TVs, smartphones, laptops, tablets, and automotive displays, competing mainly with Canon Tokki, while holding a monopoly (in commercial scale and quality) in the OLEDoS market for VR/AR headset screens, which have pixels 10x smaller than OLED. It operates through its 47%-owned subsidiary Sunic System (market cap KRW 714 billion) and also holds large cash balances, a VC arm with significant related-technology businesses, and its original OLED/OLEDoS testing business. A sum-of-the-parts valuation yields KRW 420 billion versus DAE's market cap of just KRW 100 billion, an extreme holdco discount even by Korean standards. In 2025, Sunic System had revenues of KRW 516 billion (DAE's share KRW 243 billion), giving DAE total revenues of KRW 312 billion and earnings of KRW 96 billion; 2026 looks similar, with Q1 2026 showing contract liabilities of KRW 134 billion and total contract value of KRW 612 billion, comparable to early 2025. Additionally, DAE is building a new facility to seamlessly switch between OLED and OLEDoS machine manufacture and add optionality in Perovskite solar panel production, which uses similar deposition technology to absorb the bluer spectrum and could raise efficiency from 21% to above 30%.

Read the full article here. Read time: 2 min

On Holding AG (ONON), a premium Swiss sportswear compounder trading at $31, is a long-term buy (the analyst owns shares from ~$38, bought in early August) sized as a volatile mid-cap consumer cyclical, with a fair value corridor of $42 (Bear, 17x exit P/E), $62 (Base, 25x), and $74 (Bull, 30x), placing the current price 26% below even the Bear Case and inside the $29–42 accumulation zone. Despite reporting Q2 2026 results with 21.6% constant-currency sales growth (+13.5% reported to CHF 850.3M), its highest-ever gross margin of 65.4% (best in sportswear), net income of CHF 105M (vs. a CHF 40.9M loss YoY), and raised full-year gross margin guidance to at least 65%, the stock fell 14% on the report and is down 33% YTD—driven by ~8 points of FX translation loss from a strong Swiss franc, a modest sales miss, weak reported Americas growth (+4.5%, or +13.0% cc), deliberate wholesale discipline (+4.8%), and low-20% full-year cc guidance implying H2 deceleration. The bull case rests on strong DTC growth (+34.3% cc, now 45.7% of sales), Asia-Pacific +54.7% cc, apparel +47.7% and accessories +88.3% growth, an early-stage brand (90+ countries, only 30% global awareness, 50M+ pairs sold ever), a narrow but real moat from full-price discipline and innovation (CloudTec, LightSpray robotic manufacturing, SURREAL foam) plus elite athletes, ~$0.8B net cash, 13.6% FCF margin, 12.3% net margin, rising 17.1% ROIC (up from ~11% in 3 years) and 24% ROE, and cheap valuation at 17.3x forward P/E (vs. 53x IPO-era mean, near the -1 SD band), 2.2x P/S, and a PEG of 0.70—the cheapest in a competitor group (Nike 23.7x/PEG 1.80, Adidas 15.3x, Deckers/HOKA 12.1x/PEG 1.69, Lululemon 11.1x) despite ~25% consensus 5Y EPS growth (3x faster than peers); consensus revenue is projected to reach $6.18B by FY2028 with EPS of $2.57, and analysts rate it Strong Buy (avg target $46.15, +48%). Key risks and bear points include a discretionary US consumer/recession exposure, CHF and tariff headwinds, competition from a recovering Nike, a bigger HOKA, and Adidas (with fashion-cycle risk on the Cloudtilt), 6.4% short interest betting the slowdown is real, slight ongoing dilution, and no dividend or buyback to support the stock in drawdowns—with the realistic bear case being dead money (mid-teens cc growth and sideways movement with 20% swings) rather than a crash; the analyst advises judging the thesis on constant-currency growth and gross margin rather than share price.

Read the full article here. Read time: 15 min

CuriosityStream ($CURI, short) is a declining streaming service masquerading as an AI content-licensing play, whose recent Q2 'profitability' (50% QoQ revenue growth, gross margins up 56%→72%, operating profit inflecting from a $1.5m loss to $9.2m profit) is a mirage built on accounting gimmicks. In Q2, CURI sold three brands (Curiosity University, Curiosity Audio, Catholic Stream) to newly-formed Curiosity Brands—a possible shell with 'limited operating history and independent financial resources'—for just $300k, while simultaneously recognizing $8.5m upfront (the NPV of a 30-year, $10m licensing deal paid in four annual installments starting June 2027, discounted at 7%) as licensing revenue despite receiving no cash today, effectively dressing up an asset sale as high-margin licensing revenue. Additionally, CURI inflates revenue via 'barter' transactions, swapping content rights for other companies' content (recording revenue at zero cost while capitalizing acquired content), with barter making up $5.3m of just $5.5m in Q2 content additions—meaning it isn't organically replenishing its library (three-quarters of content assets are now licensed from others), undermining the core thesis that its library is uniquely valuable for AI licensing. Stripping out the $5.3m barter revenue and $8.5m one-time deal, real quarterly licensing revenue collapses from $14m to just $200k (down from $2.2m in Q1); Q2 operating cash flow was -$4m despite zero organic content spend and marketing cut from $3.2m to $1.9m, and subscription revenue fell 5% YoY. Guidance for 2H revenue of $38-41m and full-year adjusted EBITDA of $18-22m is misleading: 2H adjusted EBITDA guidance is only $6-10m (down sharply from $11.4m in Q2) and ignores ~$15m of trailing stock-based comp (a real cost on a $210m market cap), while the $17-22m year-end 'cash and investments' target (vs. only $10.8m cash/debt securities) is padded by equity investments including Nebula ($3.7m) and German subsidiary Spiegel Venture (adding ~$2m paid and ~$3m revaluation of the previously $0-held 32% stake in Q3). Meanwhile, the company pays a 10% dividend funded from balance-sheet cash and SBC, and insiders have sold over $2m of shares in the past year, continuing recently. Following the 30%+ (author saw 40%) pop on 'nonsense' earnings that actually reinforced the failed licensing narrative, the author views CURI as a very good short at the $210m market cap; it is currently one of his largest single-stock short positions at ~0.7%.

Read the full article here. Read time: 5 min