YB new stock pitches (Mon, Jul 6)

IRIDEX Corporation (IRIX), a medtech company trading at less than 0.5x EV/Revenue, presents a straightforward thesis: a genuine chance to emerge this year as a profitable company, with underlying Q4 profitability likely significantly higher than the full-year run-rate suggests. Q1 results beat all assumptions with higher revenue, a higher gross profit margin, and a higher adjusted EBITDA than expected. The investor's calculation uses management's full-year revenue guidance of $51-53 million, an assumed gross profit margin of 37-43% (noting the 37% reported two quarters ago was hurt by one-time inventory write-downs), and full-year operating expense guidance of $19-19.5 million (excluding D&A and stock-based compensation), viewing this as a conservative baseline given multiple margin-expansion initiatives underway, such as contract manufacturing. The path to profitability, after decades of it seeming impossible, is driven by a newly favorable reimbursement environment, strong clinical data, and a management team moving past legacy constraints, with a strong recent earnings call reinforcing the story. After opportunistically selling half his position on an after-hours algo-driven spike and partially buying it back the next day, the investor increased his position; portfolio is +31.2% YTD.

Read the full article here. Read time: 2 min

Kawasaki Heavy Industries (KWHIY) fell more than 7% (ADR opening near $6.64, Tokyo shares at a five-month low) after reports it was finalizing a ~¥200 billion (~$1.23 billion) capital raise via new shares and convertible bonds aimed at overseas institutions, triggering dilution fears; however, the raise—executed at record earnings (sales ¥2.31 trillion, net income ¥108 billion, EPS jumping from ¥105 to ¥129, with guidance for another record year)—is offensive growth capex into four inflecting themes the market pays up for elsewhere. First, aircraft engines and aerospace, leveraging Japan's largest postwar defense expansion (spending doubling toward 2% of GDP, FY2026 budget over ¥9 trillion, ¥43 trillion/$275 billion committed through 2027) and April's scrapping of the lethal-weapons export ban, making its C-2, P-1, and licensed engines exportable, plus participation in the UK/Italy Global Combat Air Program (GCAP) fighter; second, semiconductor manufacturing robots as picks-and-shovels exposure to the global fab buildout (TSMC Arizona, Samsung Texas, Rapidus, China), where it holds a near-monopoly niche and is expanding capacity as demand outruns supply; third, AI/humanoid robotics via an Nvidia partnership and planned Silicon Valley center, offering free optionality given peers like Figure AI ($40B) and Chinese names (Unitree, UBTech) trade at revenue multiples; and fourth, hydrogen infrastructure (world's first hydrogen-fueled large commercial vessel engine, Canada-Japan and Daimler Truck partnerships), a Japanese national policy priority though economically the weakest bucket. Kawasaki trades at ~20x earnings versus pure-play defense peers Rheinmetall (~40x) and BAE (doubled), while also generating shipbuilding, rail, and motorcycle cash flows; the thesis is that summing the four themes as pure-plays, less a conglomerate discount, suggests the dilution is buying accelerated growth cheaply, though Japanese industrial structural challenges and unpredictable re-rating timing make it 'not obviously buy' but 'definitely not sell the dilution.'

Read the full article here. Read time: 5 min

Kioxia ($KXIAY), the 'SanDisk of Japan,' is a pure-play NAND flash memory and SSD manufacturer and top-three global NAND producer, spun off from Toshiba (which invented NAND in 1987), acquired by a Bain Capital-led consortium in 2018 for ~$18B, and IPO'd on the Tokyo Stock Exchange in December 2024 (ticker 285A); it trades at ¥76,260 with 546.3M shares outstanding for a ~¥41.66 trillion ($256B) market cap. The thesis centers on the May 15, 2026 inflection when Kioxia blew past guidance, delivering FY2025 revenue growth of 37% (vs. 27.7%–33.0% guided) and operating profit growth of 92.7% to ¥870.4B (vs. 57.1%–77.0% guided), with operating margin expanding from 13.1% in Q1 to 59.5% in Q4 (full-year 37.2%, up from 26.5% in FY2024); forward Q1 FY2026 guidance calls for revenue of ¥1,750.0B (+74.5% QoQ) and non-GAAP operating profit of ¥1,300.0B (+117% QoQ, ~74% margin), more than the entire prior fiscal year's record profit, with results due July 31, 2026. The growth is driven by a structural shift from cyclical consumer electronics (flat/declining smartphone and PC demand) toward AI data centers and enterprise (targeting 60%+ of revenue medium-to-long term), as SSDs increasingly serve as extended GPU memory layers to solve KV cache overflow (via CM Series, NVIDIA CMX/Storage-Next) and RAG bottlenecks (GP Series) plus massive AI output storage (LC Series); management forecasts total flash memory demand growing from 997 exabytes (2025) to 1,807 EB (2028) at 22% CAGR, data center demand growing at 46% CAGR (295 EB to 909 EB, from 30% to 50% of total), and inference workloads growing at 86% CAGR vs. 16% for training, with tight supply-demand persisting through FY27 and NAND market revenue estimated to be ~4x larger in 2026 than 2025. Kioxia's proprietary BiCS FLASH 3D NAND technology (8th-gen 218 layers, 10th-gen 332 layers, using CBA architecture and a dual-axis strategy of vertical stacking plus 2D shrink) targets ~10% annual front-end cost-per-gigabyte reductions and 1.5x–1.6x density jumps per generation, driving margin expansion regardless of NAND pricing. It shares R&D and capex 50/50 through a 25+ year manufacturing JV with SanDisk at its Yokkaichi and Kitakami plants (extended January 29, 2026 through December 31, 2034, with SanDisk paying $1.165B cash), together accounting for 30% of global NAND production. Management plans ~¥470B annual capex (60% jump, capped at ~20% of revenue) and ~¥230B R&D (8–9% of revenue) over three years, uses long-term customer agreements (LTAs) targeting mid-20% operating margins, boasts trailing-12-month ROIC exceeding 60% (targeting highest among manufacturers), expects to reach net cash positive by end of Q1 FY2026, and plans progressive dividends and potential buybacks starting FY2027 (up to 50% of excess cumulative free cash flow). The author invested in Kioxia in May 2026 after previously holding SK Hynix, seeing continued upside.

Read the full article here. Read time: 12 min