Investment Decision Journal

FnGuide activates Korea coverage, but A+ status is withheld

Decision: Accumulate selectively at ₩18,810, using the 14 August 2026 close, with a modeled 17%–22% base-case three-to-five-year KRW IRR. Do not call FnGuide A+ yet.

The change is research-, fundamentals- and valuation-driven. The bottom-up model separates recurring financial-information revenue from FnIndex royalties and deliberately normalizes linked ETF AUM below the roughly ₩78tn Q2 peak. FY2026 is modeled at about ₩55.2bn revenue, ₩26.7bn operating profit and ₩1,940 EPS, implying roughly 9.7x earnings. Corporate Value-up is treated as a market-structure tailwind through Korean equity values, ETF participation and customized shareholder-return indices rather than an automatic rerating. The main offset is governance: the Hwacheon/Kwon block controls roughly 48% and has substantial skin in the game, but the broader group's historical record shows conservative excess-cash retention and only middling minority-shareholder capital returns. Improve conviction if linked AUM stabilizes above ₩65–70tn, implied index monetization stays near the modeled 5bp level and the 2026–2027 payout/cancellation actions establish a repeatable shareholder-return framework. Reverse the decision if linked AUM remains below ₩50tn, monetization compresses materially or related-party/cash-allocation behavior disadvantages minorities.

Korea company coverage starts without a one-name ranking

Decision: Publish FnGuide as the first current KRX company, add a Korea market section and keep the Korea ranking explicitly pending.

A one-company universe cannot answer a capital-allocation ranking question. Future Korea candidates must be refreshed on comparable price dates and scored with an explicit governance and Corporate Value-up framework. A FnGuide-versus-S&P Global comparison has been added only as Needs review because the existing SPGI price date is 28 July 2026 and a stale cross-market relative conclusion would violate the repository's publication rules.

Food Empire keeps SGX rank one, but moves to selective accumulation

Decision: Retain Food Empire as the leading published SGX idea at S$2.42, but reduce the base-case IRR to 11%–14% and change the action from Accumulate to Accumulate selectively.

The change is fundamentals- and valuation-driven. 1H2026 revenue rose 15.0% and normalized NPAT increased 12.2%; Central Asia grew 33.6% and the Kazakhstan plant creates a visible 2H growth leg. Against that, operating profit grew only 10.2%, operating margin slipped to about 15.0% from 15.6%, Southeast Asia slowed sharply and US$26.1m of operating cash flow did not cover US$30.1m of capex. The thesis remains intact, but the next return leg must come from earnings and cash conversion rather than multiple expansion. Improve conviction if Kazakhstan lifts segment profit, operating margin recovers above 15.5% and Vietnam reaccelerates. Reverse the decision if margins keep lagging despite commodity relief or working capital and capex continue to absorb most operating cash flow into FY2027.

Food Empire comparisons withheld after both review triggers fired

Decision: Move Food Empire vs Sembcorp and Food Empire vs UltraGreen.ai from Current to Needs review rather than carry their old relative preferences forward.

Food Empire has been refreshed after 1H2026, while UltraGreen reported on 12 August and Sembcorp on 13 August. Both comparisons explicitly depended on those new results and current relative prices. The prior Food Empire preference remains in each file's revision history but is not presented as current until the alternative company is rebuilt on the same price and forecast date.

UltraGreen.ai enters current SGX coverage at rank three

Decision: Accumulate selectively at S$1.53, with a 10%–14% base-case three-to-five-year IRR. Prefer Food Empire for new SGX growth capital at the dated prices.

The change is research- and valuation-driven. FY2025 continuing revenue grew 28%, gross margin held near 85%, normalized NPAT was US$63.8 million and year-end net cash was US$176.1 million. The thesis is the regulated ICG consumables franchise, not an assumed software rerating: UGDS produced about US$0.4 million of FY2025 external revenue and a segment loss. Reverse the decision if underlying vial growth stays below 5%, gross margin falls below 80%, cash conversion remains weak or capital is deployed into adjacent projects without commercial proof.

Deep research moves to a linked evidence and FAQ layer

Decision: Keep the company page as the concise investment decision and place durable product, history, organisation and moat questions in a separately versioned note linked from the thesis.

The UltraGreen note records verified answers on the UltraLinQ disposal, the legal software-development footprint, what the AI claim does and does not establish, and remaining disclosure gaps. This preserves useful discussion without duplicating valuation, scenarios or monitoring.

S&P Global enters current US coverage; broad AI-casualty thesis rejected

Decision: Watch / accumulate selectively at $439.83, with an approximately 11% base-case five-year IRR and a preferred 15% hurdle entry around $370–$380.

The change is research- and fundamentals-driven. Mobility was separated on 1 July, so historical consolidated earnings were replaced with the post-spin recast. Q2 2026 Market Intelligence revenue grew 6%, adjusted operating profit rose 10% and margin expanded to 36%, while Ratings and Indices remained the dominant high-margin profit engines. AI risk is concentrated in desktop interfaces and distribution economics rather than the entire enterprise. Reverse the decision if Market Intelligence growth remains below 4%, seat and pricing pressure overwhelms API revenue, standalone free cash flow persistently lags adjusted EPS or the benchmark franchises suffer structural rather than cyclical deterioration.

US company coverage activated; ranking and peer comparison withheld

Decision: Publish the first fully refreshed US company page without labelling S&P Global rank 1 and without creating an artificial peer comparison.

A US ranking requires several companies with aligned prices, normalized earnings, free-cash-flow treatment and scoring. A current SPGI-versus-Moody’s, MSCI, LSEG or FactSet comparison also requires those alternatives to be refreshed first. The absence of a comparison is deliberate publication discipline, not an omission of stale work.

Alpha recovery validated; move from waiting to selective accumulation

Decision: Accumulate selectively at S$0.520, with a preference for an ex-distribution price below S$0.50 and a 10%–12% base-case IRR.

The change is fundamentals-driven: 1H2026 occupancy reached 95.0%, NPI rose 10.8%, distributable income increased 10.4%, financing cost fell to 3.76% and the full 2.03-cent DPU was declared. The reported 19.4% DPU growth is normalised to about 10% because the prior period included retained cash. Near-NAV pricing means future returns should come mainly from the 7%–8% cash yield and modest DPU growth. Reverse the decision if occupancy falls below 92%, FY2027 renewals weaken materially or development capex fails to create per-unit accretion.

Japan Elevator Service Holdings enters current JPX coverage

Decision: Accumulate selectively at ¥1,677.5, with a 9%–12% base-case IRR over three to five years.

The conclusion is supported by recurring maintenance, about 11% share of a still OEM-dominated market, route-density leverage and modernization growth. The change is research- and valuation-driven: a 34–36x base exit range is more consistent with the company’s history than the earlier 30x assumption. Reverse the decision if organic additions remain below 7%–8%, operating margin falls below 18%, a systemic safety issue emerges or adjacent expansion dilutes returns.

JPX company coverage activated; ranking withheld

Decision: Publish the first fully refreshed Japan company page without labelling it rank 1.

A ranking requires a comparable multi-company universe. Additional Japan companies must use aligned price dates, normalized forecasts, scoring and portfolio roles before a capital-allocation order is published.

Multi-market publication structure established

Decision: Keep SGX research active while preparing separate JPX and US market sections.

Existing research from prior chats will be refreshed before publication. The homepage will show migration status rather than naming stale or artificial placeholder picks.

Rankings separated by market

Decision: Do not publish a global ranking until SGX, JPX and US research has comparable freshness and price assumptions.

Quality, valuation and expected return remain separate. Cross-market capital allocation can be added later as a deliberate portfolio decision.

Initial SGX ranking established

Ranking: Food Empire → Sembcorp → Singtel → Stoneweg Europe → Alpha Integrated REIT → AIMS APAC REIT.

Food Empire ranked first for balance-sheet-adjusted organic compounding. Sembcorp ranked second for valuation-driven upside. Singtel was classified as the lower-risk core option.

Next research work

• FnGuide: recheck the 14 August price and primary disclosures during preview review, then monitor linked ETF AUM, implied index monetization, FY2026 operating profit and the ≥25% payout framework.
• Refresh additional Korea companies before publishing a KRX ranking; include controlling-shareholder governance and Value-up implementation in the score.
• Refresh S&P Global to a common date before converting FnGuide vs S&P Global from Needs review to Current.
• Food Empire: verify 2H Kazakhstan revenue and segment-profit contribution, operating-margin recovery above 15%, Vietnam reacceleration and operating cash flow versus capex.
• Refresh UltraGreen.ai's 12 August 1H2026 results before restoring the Food Empire comparison.
• Refresh Sembcorp's 13 August results and current price before restoring the Food Empire comparison.
• S&P Global: monitor Market Intelligence organic growth, API production conversion, seat and pricing commentary, standalone free-cash-flow conversion and repurchase prices.
• Alpha Integrated REIT: monitor FY2026 DPU, FY2027 renewal economics and New Tech Park Phase 3.
• Japan Elevator Service: review FY2027 first-quarter results on 6 August 2026.
• Refresh additional Japan companies before publishing a JPX ranking.