Portfolio Dashboard — India & Global

21 Sep 2026 — portfolio change: bought PepsiCo (PEP, Nasdaq) at ~0.6% — entered 🟡 Amber, matching the same-day 4-pillar review (PepsiCo-PEP-Analysis-2026-09-21.md: Business ⚠️ / Financials ⚠️ / Valuation ✅ / Macro ⚠️ → WATCH, medium conviction). This is the fifth consecutive entry at 🟡 or 🔴 (BMW 🔴, ITC 🟡, UMG 🟡, VBL 🟡, PEP 🟡) — acting on Amber is now the settled pattern, not an exception. New US name takes the US sleeve to 15 and Global to 30.
The framing number: PEP has returned −0.36% annualised over five years while Coca-Cola returned +13.46% — a ~14pp-a-year gap between two businesses that are supposed to be twins. ⚠️ Critically, this is NOT a sector de-rating, which was my first hypothesis and the data killed it: KO trades at 26.5× and is +35.7% over twelve months; MDLZ and KDP are up on the year too. The discount is PepsiCo-specific, which is more useful than a sector call because a company-specific problem can have a company-specific fix.
🚨 North America is the cause, and management’s framing does not survive its own segment table. PepsiCo’s release says convenient foods “gained volume market share”. The Q2 PFNA line items: volume flat, effective net pricing −2%, core constant-currency operating profit −8%. They are buying share with price and paying for it out of profit — the same mechanism as COLPAL renting volume with A&P, except through price rather than advertising. Third-party check (Circana, 52wks to 17 May 2026): Frito-Lay $6.4bn, −1.3% in a stagnant category; private label $1.1bn, −0.8%. ⚠️ Note private label is falling too — so this is not mainly share loss to cheaper brands, it is that the pool stopped growing, which is exactly why share gains produce no profit. PBNA: organic +1% but beverage volume −4%. North America is 56% of revenue and 58% of core operating profit.
✅ International is genuinely excellent and carrying the group: Asia Pacific Foods organic +9% / volume +10% / core cc profit +41%; IB Franchise +9% / +5% / +17%; EMEA +6% / +4% / +14%. But 44% growing cannot outrun 56% stalling: total core cc operating profit grew just +1% and core operating margin contracted 40bps.
⚠️ Read the headline carefully: reported revenue +6.4% = 2.4% organic + 2.2% FX + 1.8% acquisitions. Only a third is the business growing; guidance embeds ~1pp FX and ~1pp M&A inside the 4–6% reported range.
Three things a screen will not show you: (1) FY2025 core EPS $8.14 vs $8.16 — dead flat, core cc EPS growth 0%, with reported EPS −14% on $1,993m of intangible impairments. (2) The dividend is barely covered — FY2025 FCF $8,200m against $8,638m returned, i.e. $438m MORE paid out than generated; FCF payout 87%; the raise was +4.0%, the 54th straight, a streak to defend on flat earnings. (3) Tangible book is negative $12.2bn (goodwill+intangibles 1.55× equity) and this is demonstrated, not theoretical — $1.99bn was written off last year. Offsetting: ROIC 19.4% against a 5.2% WACC, net debt/EBITDA 2.25× and falling, interest cover 13.1×.
🚨 The Elliott catalyst is gone — the 8 Dec 2025 settlement gave productivity savings, a core-margin commitment and a ~20% US SKU cut, but no board seats and explicitly no refranchising. Anyone holding for a bottling break-up lost that reason nine months ago. ⚠️ And note the conflict: the settlement promises margin expansion from 2026 while Q2 margin contracted and the NA fix runs through price cuts.
Macro: GLP-1 households went 9% → 21% in sixteen months (PwC/Numerator, May 2026), though Circana finds salty snacks only slightly down and not among the worst-hit categories. ✅ Two risks retired rather than repeated: FDA front-of-pack labelling is still only proposed (final rule slipped to winter 2026), and a federal court VACATED five state SNAP soda waivers in June 2026 — that one is widely quoted as settled and isn’t. ⚠️ Cross-holding clock: India’s FSSAI “energy drink” naming ban hits Rockstar at PEP and Sting at VBL simultaneously.
⚠️ The honest risk — cheapness has been a trap here for four straight years. PEP has been “cheapest in five years” at each of the last four year-ends and fell anyway: market cap $251.8bn (FY22) → $175.6bn, −30%. The multiple re-rates when earnings inflect, not before. Decision point is the Q3 print on 8 Oct — confirmed by company announcement, not a vendor estimate, and only 17 days after entry.

15 Sep 2026 — COLPAL downgraded 🟢→🟡. Full 4-pillar review (Colgate-COLPAL-Review-2026-09-15.md, the first dedicated file on this name): Business ⚠️ / Financials ⚠️ / Valuation ⚠️ / Macro ❌ → WATCH, hold the 0.5%, don’t add before Q2 (~22–24 Oct). ⚠️ The old Green’s stated reason was factually half wrong, which alone forced the change: Q1 FY27 revenue +12% ✓, but only gross margin expanded — EBITDA margin contracted 146bps to 30.1% on advertising +33.7%. The company is renting volume with A&P (HDFC Sec, 18 Aug: “margins may moderate as higher brand investments drive growth”). FY26 was a down year: revenue −0.3%, PAT −7.8%, 5-yr revenue CAGR just 4.51%. ⭐ The finding worth reusing — a GST rate CUT that is a permanent margin tax. GST 2.0 took toothpaste 18%→5% (eff. 22 Sep 2025), creating an inverted duty structure: under s.54(3) CGST + Rule 89(5), refunds of accumulated ITC are computed on “Net ITC”, which excludes input services. COLPAL’s ad spend alone is 13.7% of revenue, so a slice of credit is permanently unrefundable and expensed — management quantified ~80bps of FY26 EBITDA (≈₹48 cr), ~160bps in Q4, and it recurs every year. ❌ And a hypothesis the evidence killed: I expected the MRP pass-through to have deflated FY26 revenue. It didn’t — Indian net sales are reported net of GST, so full pass-through cuts the consumer MRP ~11% while leaving ex-GST realisation unchanged. Flat FY26 revenue was real demand weakness plus transition destocking. The GST event hit margin, not the top line. (Management also warned Nielsen value growth is distorted by the repricing — track volume.) ⚠️ Dividend cut, and the H2 leg cut twice running: ₹36→₹27→₹24 (FY26 ₹48 vs ₹51). Screener’s “119% payout” is a data error (true 98.5% declared / 104.7% cash). Because cash paid exceeds accounting profit, equity is shrinking — so the 82.7% ROE is partly an artifact of a shrunk base (44% on a ₹3,000 cr base). ROCE 108% is the honest number. ✅ But the balance sheet is the strongest of your five India staples — net cash ₹1,422 cr, zero borrowings, CFO 136% of PAT, clean unmodified audit, and the dividend is cash-covered 1.25× by FCF vs HUL’s 1.01×. 📉 Moat call rests on the competitor’s own audited disclosure, not COLPAL’s: Dabur’s FY26 annual report states it outgrew the category and expanded share, with herbal growing ~2× non-herbal. ⚠️ Do not quote a share LEVEL — the Nov-2025 table everyone cites lists “GSK Consumer 8.8%”, a business that has been Haleon since 2022; the dataset’s company names are 4+ years stale. Direction only: down, from a 43–51% band. Offsetting: still ~3× the #2, and COLPAL’s quick-commerce share exceeds its rest-of-e-commerce share — brand pull is travelling to the channel that was supposed to break incumbent moats. 💰 Valuation is the genuinely interesting part: P/E 36.6× vs a 10-yr median of 44.81×, and within ~3% of its 13-year LOW. For a stock that has never been cheap, that matters — but PEG is 4.1× on consensus and 7.2× on realised 5-yr PAT CAGR, so the de-rating tracks a real halving of growth rather than a mispricing. ✅ FII stopped falling — 24.94%→13.60% over 8 quarters, but Jun-26 was flat, the first non-decline in two years. 🚨 New MD & CEO Manish Anandani (ex-Kenvue) takes over 28 Sep — 13 days out, one month before a print: the classic setup for a guidance rebase. ✅ One risk retired: the 28 Sep Supreme Court FSSAI front-of-pack matter does not touch COLPAL at all — toothpaste is a cosmetic under the Drugs & Cosmetics Act, not a food under the FSS Act. COLPAL is the only one of your five India staples with zero exposure to that hearing. ⚠️ Two items returned unverified and NOT assumed away: promoter pledge % (the only “zero” claim traces to Trendlyne) and the parent royalty rate — both live alongside a ₹267.64 cr transfer-pricing demand. India staples ≈3.7% and, with this, all five are now Amber (HINDUNILVR 1.2 🟡 + DABUR 0.8 🟡 + ITC 0.6 🟡 + VBL 0.6 🟡 + COLPAL 0.5 🟡) — COLPAL was the last Green in the cluster.

11 Sep 2026 — portfolio change: bought Varun Beverages (VBL, NSE) at ~0.6% — entered 🟡 Amber, matching the 7 Sep 4-pillar review (VarunBeverages-VBL-Analysis-2026-09-07.md: Business ⚠️ / Financials ✅ / Valuation ⚠️ / Macro-regulation ⚠️ → WATCH, medium conviction). Note you entered above the review\'s own Buy trigger of EV/EBITDA under ~18x (≈₹300) — bought at ~₹404–411. That is now four of your last seven buys entered at 🟡 or 🔴 (BMW 🔴, ITC 🟡, UMG 🟡, VBL 🟡).
The bull case: PepsiCo\'s bottling appointment extended to 2049 (from 2039) on 21 May with the PepsiCo-only SPV restriction deleted; CY21→CY25 revenue +25.2% CAGR and PAT +42.3%; CRISIL AAA, net debt/EBITDA 0.07x, interest cover 29.5x, promoter pledging 0.0014%; and the stock is −27% from its June-2026 high and −41% from the 2024 peak.
Why Amber, not Green: ⚠️ the cheapness is partly an accounting illusion — P/E 40.4x is 30% below its 5-yr median, but EV/EBITDA 24.3x is exactly at its own 11-yr median of 24.04x; the P/E fell further only because leverage collapsed. ⚠️ Pricing power has gone at the mass end and it is in the P&L: India realisation/case down four straight quarters, the ₹20 pack upsized 250ml→400ml against Campa. ⚠️ Underlying India volume is ~3%, not 14% — Q2\'s +14.4% sits on a base that fell 7.1%. And ROE has halved 34.2%→16.8% post-QIP.
🆕 Two regulatory clocks inside three weeks: the Supreme Court on 10 Sep directed the Centre and FSSAI to produce a firm front-of-pack warning-label timeline within 10 days and listed the matter for 28 September — FSSAI\'s proposed \"HIGHLY SWEETENED BEVERAGE\" hexagon captures CSDs (~74% of VBL volume) in Phase I; and FSSAI\'s ban on the term \"energy drink\" bites ~29 September, hitting Sting. Plus sugar at a 16-year high with a 15-day stock cap on bulk users from 1 Sep.
Cluster: India staples/FMCG goes to ≈3.7% (HUL 1.2 🟡 + Dabur 0.8 🟡 + ITC 0.6 🟡 + Colpal 0.5 + VBL 0.6 🟡) — four of five Amber — though VBL is the genuine diversifier: 18% volume growth at 40x versus ITC shrinking at 17x.

11 Sep 2026 — 5paisa Capital (5PAISA, 1.5%) downgraded 🟡→🔴 after a full 4-pillar review (5paisa-5PAISA-Review-2026-09-11.md). Business ❌ / Financials ⚠️ / Valuation ⚠️ skewed red / Regulation ❌. The Amber was set on "steady core, flat profit" and missed four things. (1) NSE active clients 3.31 lakh, ~0.75% share, rank ~16, −14% YoY while the industry bottomed and recovered — it is now smaller than IIFL Capital, its own promoter group's other broking arm, and only 6.3% of its 52.6 lakh registered clients are active. (2) The April rights issue took net worth ₹650 cr → ₹1,131 cr (+74%) while quarterly PAT went ₹11.0 → ₹11.6 cr, so ROE is now ~3.9% — below the Indian risk-free rate. (3) ⚠️ The screen P/E is wrong by ~35%: vendors divide TTM profit by the pre-rights weighted-average share count, so they print 25.5x where the true figure on shares outstanding is 34.8x — the dearest of its six peers on the lowest ROE of the six. P/B 1.36x is genuinely near its 13-yr low, but only because the book earns 3.9%. (4) SEBI's Closing Auction Session (3 Aug) took 22% out of F&O turnover in one month, on top of index option contracts −52% in FY26, with another consultation pending.
✅ The genuine counterweight, and it is not small: the promoter subscribed the 1:2 @ ₹300 rights ~17.55 lakh shares above entitlement (32.76% → 36.50%, ~₹206 cr), FIIs also went above pro-rata, and retail under-subscribed. At ₹328 the stock is only 9.4% above what insiders paid five months ago. Clean FY26 audit, D/E ~0.34x. This is a shrinking-relevance and returns-on-capital call, not a solvency or fraud one.
Giskard/Trendlyne now resolved: 5paisa is buying 100% of Trendlyne's parent for ~₹207 cr = ~13× FY25 sales and 13.5% of its own market cap; ballot passed 99.99% for; allotment still pending SEBI approval. Three things to keep on file: a paragraph saying promoters would buy Giskard shares was withdrawn as "inadvertent" on 29 Jul; the sellers of the 58.68% cash leg are undisclosed while IIFL is reportedly an early Trendlyne backer; and the 1 Sep AGM carried 12 separate IIFL related-party approvals.
⚠️ Could not verify and am not assuming: promoter pledge %. Also open — the ₹626 cr contingent-liability composition, and a ₹33.03 cr tax demand raised 4 May 2026 under s.158BC, the block-assessment section (implies a search action). Reverts to 🟡 on Q2 FY27 PAT above ~₹18–20 cr or two months of stabilising client counts. Q2 date not yet announced (LODR deadline 14 Nov).

9 Sep 2026 — portfolio change: bought back the Infosys NYSE ADR (INFY) at a provisional 0.5%, having sold it on 29 Jul — a ~6-week round trip out and back in. Added as 🟡, mirroring the NSE line: Q1 FY27 revenue +14% but only +2.4% constant-currency, and FY27 guidance trimmed to 1.5–3.0% cc (top end cut from 3.5%). Nothing material since that print — the ₹18,000 cr buyback was FY26, completed Dec 2025. Book now 76 holdings; Global 29 / 48.8%, cash ~12.7%.
⚠️ This restores a duplication: Infosys is again held twice (NSE 0.7% + ADR 0.5% ≈ 1.2%), alongside HDFC Bank's NSE+HDB pairing (~1.9%). With FLXI (1.1%), roughly 4.2% of the portfolio is duplicated or overlapping India exposure, with a USD/INR layer on the ADR portions.
7 Sep 2026 — portfolio change: booked profit in GMM Pfaudler (GMMPFAUDLR), sold at 0.6% (was 🟡). Its last print was the tell: revenue +16% and order backlog ₹2,289 cr (+20%) looked strong, but EBITDA actually fell 7% with margin down to 10.1% — the +118% PAT was finance-cost and FX driven, not operational. Selling into that is consistent with the earnings-quality read. Book now 75 holdings — India 47 / 38.5%, Global 28 / 48.3%, cash ~13.2%. (Also reconciled a stale figure: the stats panel had been showing cash at 15.6% while the header said 12.6% — both now read 13.2%.)
4 Sep 2026 — LULU Q2 FY26: the 3 Sep binary resolved, and it resolved badly. Signal stays 🔴 (no change — it was already Red).
Headline looked like a beat; it was not. Q2 (ended 2 Aug): revenue $2.416bn, −4% (−5% cc), comps −9% (−10% cc), EPS $2.92 vs $3.10 LY. But $0.86 of that EPS was a one-off IEEPA tariff refund ($134.5m + $4.1m interest). Strip it out and the quarter is: underlying EPS $2.06, −34%; gross margin 54.9% not 60.5% (the refund was worth 560bp, so underlying margin fell ~360bp); operating margin ~13.2% vs 20.7% LY, not the 18.8% printed. SG&A deleveraged 400bp (41.7% of sales vs 37.7%).
🚨 The Q3 guide is what took the stock down ~20%. Revenue $2.29–2.32bn = −10 to −11%, and EPS $0.93–0.98 against $2.59 last year — a 62% collapse. Q2's −4% was survivable; Q3 says the decline is accelerating, not troughing. FY26 cut for the second time this year: revenue $10.35–10.50bn (−5 to −7%) from $11.00–11.15bn; EPS $9.48–9.73 — but that includes the $0.86 refund, so underlying $8.62–8.87 vs a prior guide of ~$11.05: a 21% cut to real earnings in one quarter, and −34% vs FY25's $13.26.
What broke: leggings −20% — the core franchise; accessories −13%; Americas comps −12% (US rev −8%, Canada −11%). China stopped being the growth engine: revenue +4% reported but −2% cc, comps −8% cc, driven by brand sentiment — sustained negative social/media commentary compounded by backlash over their Great Wall event, plus Tmall not repeating 618 and LULU refusing to discount into it. Australia turning promotional while LULU holds full price. Costs rising into falling sales (H2 marketing being increased deliberately).
✅ Third-party share data — checked, and it does NOT exonerate management. Their story is "silhouette shift, we're still market leader." Independent data says both things are true, and the share loss is the bigger one. PassBy foot-traffic (pub. 2 Apr 2026): LULU category traffic share 86.5% → 77.5%, −9.0 pts; Alo Yoga 1.4% → 8.6% (+7.2); Vuori 9.2% (−0.8). Same-store visits YoY: Alo +7.5% (88 stores), Vuori +1.9%, LULU −3.1% (395 stores). LULU declined in all three formats — enclosed mall −3.7%, open-air −4.75%, standalone/street −7.9%. That standalone number is the most diagnostic figure in the whole file: standalone stores live on destination intent, so −7.9% means people are choosing not to make the trip. An assortment can be fixed in two seasons; eroded destination intent cannot. Alo's trade areas also index 204 vs LULU's 158 — it is winning in richer catchments.
But the silhouette story is real too, and worth crediting: Particl (pub. 23 May 2026, data through Apr 2026) shows wideleg silhouettes leading women's revenue at Vuori as well ("Halo Essential Wideleg" its top style) — so the move away from tight bottoms is a genuine category shift, not a management excuse. Same report is a counter-signal on Vuori though: its Q4 2025 volume ran below the Q4 2024 peak and early-2026 volume is soft. So the share is going to Alo specifically, not to "the challengers" generally.
⚠️ Two honest caveats on that data. (1) PassBy's comparison window is roughly 2025 vs 2024 — it predates this Q2 collapse, so if anything it understates where things are now. (2) Alo's percentages flatter it off a tiny base (~100 stores vs LULU's 395 same-stores), and foot-traffic share is not revenue share. ⚠️ Also logged: the widely-recirculated Earnest Analytics athleisure share numbers are April 2024 vintage — 2.5 years stale — and search engines still surface them as current. Do not use them.
Sell-side after the print: Morgan Stanley $83 (from $93, Underweight) explicitly expecting new management to go "shrink to grow" and drive another round of EPS cuts; UBS $106 (from $120). Both PTs are below the pre-print $121.77 close.
What is still intact — this is not a solvency story. $1.39bn cash, no funded debt, inventory −1% in dollars and −7% in units (no markdown bomb building), still ~13% underlying operating margin, 825 stores, operating cash flow $589m in H1 vs $210m LY. Note they repurchased 2.7m shares for $330m in Q2 at ~$122 average — above today's price.
Read: the turnaround thesis is not dead, it is pushed out at least two more quarters, and the next print is a brand-new CEO's first — Heidi O'Neill starts the week of 7 Sep and management has already said she will "take a deep dive into the strategy and action plan." That is the standard setup for a Q3 rebase. Management called Q3 the trough; they said much the same in Q1 and were wrong. Do not average down before her first print. Position is only ~0.5%, so the cost of this is roughly 0.1% of the book — keep it sized like the option it is.
⚠️ Cluster read-across: LULU's problems are largely company-specific (own product miss, own China brand incident), so this is not automatic confirmation of a sector thesis. But the US-consumer and tariff legs are shared. NKE fiscal Q1 (~29 Sep) is the next test, then ONON Q3 (~11 Nov).

28 Aug 2026 — portfolio change: bought ONON + NKE; UMG folded in; 3 signal changes; 1 rendering bug fixed.
Bought On Holding (ONON) and Nike (NKE) at ~0.6% each, one tranche each with intent to add, and Universal Music (UMG, 0.7%, bought 27 Aug) is now on the board. Global goes to 28 names, the book to 76 holdings, and cash falls to ~12.6%. Both new names had been analysed before the buy, so the analysis→entry loop held — but ⚠️ you bought the name the 22 Aug three-way scored AVOID and skipped the one it scored BUY. That comparison put adidas at BUY (🟢🟡🟢🟢), On at WATCH / buy on weakness (🟢🟢🟡🟡) and Nike at AVOID (🟡🔴🔴🔴). Nike enters as the lowest-scoring position in the book, below BMW. Not a reason to reverse it — but the four reds are now the things to track, and 29 Sep is the first test.
⚠️ New correlated theme — sportswear is now ~1.7% (ONON 0.6 + NKE 0.6 + LULU 0.5), built entirely inside one month. Same US consumer, same Vietnam/China supply chain, same tariff regime, and On competes with Nike head-on in running. Treat the three as ONE position for risk, not three — and note that Dick’s Sporting Goods’ 25 Aug guidance cut hit all three on the same day, which is exactly what correlated exposure looks like in practice.
🔻 Downgrade — Pernod Ricard 🟡→🔴. FY26 (27 Aug) came in below its own guide: net sales €9,404M, organic −3.9%; profit from recurring operations €2,423M, organic −5.2%; EPS €5.85, −19%. USA −14%, China −19%. Net debt €10,662M = 3.7× EBITDA. The dividend was held at €4.70 with a scrip option. The real damage is to the algorithm: FY27 organic only ‘broadly stable’, the FY27–29 framework cut to ‘close to the lower end’ of +3–6%, and the €1bn efficiency plan pushed to FY28. A 7.4% yield on 3.7× leverage is the market pricing dividend risk.
🔻 Downgrade — HDFC Bank ADR 🟢→🟡. Governance, not earnings. A US securities fraud class action was filed covering 17 Jul 2023 – 26 May 2026, alleging internal-control failures that let deposit inducements be routed through the marketing budget; lead-plaintiff deadline 12/13 Oct. The bank’s numbers are unchanged and the India line stays 🟢, but the claim is about the bank’s own disclosure, not an ADR artefact — so it should not sit unflagged on a core holding.
🔻 Downgrade — Deepak Nitrite 🟢→🟡, and this one is a judgement call. ICRA (27 Aug) reaffirmed AA(Stable)/A1+ and raised lines to ₹990cr, but disclosed that Phase-1 capex has gone to ₹3,016cr from ₹2,400cr — a 26% overrun, with the ₹10,500cr polycarbonate programme still to come and an explicit downgrade trigger at DSCR <2.5×. Q1 itself was 🟢-quality (revenue +36%, EBITDA margin doubled to 21%). But the whole bull case is the polycarbonate build, and this is the first hard evidence it is running over budget before the larger leg starts. Contained at the Q2 print and it goes straight back to 🟢.
🚨 Real Brokerage — the deal is DONE, and the risk has changed shape. The combination closed 24 Aug, the 10-for-1 consolidation went through, and the stock resumed as Real REMAX Group Inc. on 25 Aug. Cash elections hit the $80M cap, so cash electors were prorated to ~$4.33 + 0.3535 shares; RE/MAX holders took 39.6%, Real holders 60.4%. It was funded with a new $550M secured term loan, and a $450M buyback was authorised at close — so a debt-free balance sheet became a levered one in a single day. The market is not applauding: $21.10, −10.7% on 28 Aug, a third straight down session. ⚠️ Aggregators still show the pre-merger share count, so any market cap you read there is understated by ~40%.
🆕 India — two genuinely large items. TCS won a five-year ~€1.25bn AI engagement with Porsche and is buying MHP, Porsche’s IT arm (€320m cash, €742m CY25 revenue, ~4,500 staff) — the largest Indian IT deal in years and its first carve-out since 2008. Reliance received SEBI’s final observations for the Jio Platforms IPO on 28 Aug (fresh issue only, no OFS, sized in the press at ~$3.8–4bn) — the last regulatory gate on the value-unlock. Also: ACC’s merger vote is set for 29 Sep (328 Ambuja per 100 ACC), so there is now a hard date on ACC ceasing to be listed.
⚠️ Tariff correction worth having on the record: the current US reciprocal rate on Indian goods is 18%, in force since the Feb 2026 interim agreement — not the 50% you will see all over the wires this week. That figure is 27 Aug 2025 content being republished on its anniversary. Pharma and electronics stay exempt; textiles, gems and engineering goods carry the full 18%. The live catalyst is the unsigned phase-one trade agreement, not the rate.
🔧 Fixed: the summary and valuation cells were writing HTML as plain text, so every <b> and <br> added since late July showed up as literal tags. They now render properly — older refreshes will suddenly look correctly formatted.

22 Aug 2026 — portfolio change: sold Uber, +10% in 44 days.
Booked profit in Uber (UBER), previously 0.5% and scored 🟢 Green — +10% over 44 days, ≈106% annualised. Your best annualised exit of the month by a wide margin (Gartner was +10% over ~6.5 months, ≈19% annualised). ⚠️ Worth being clear about what this was: Uber was 🟢 on fundamentals — Q2 gross bookings +24%, first-ever $10bn TTM FCF, adj. EBITDA +33% — so this is a valuation/trade exit, not a thesis break, and nothing in the last print argues against owning it again lower. The US book drops to 11 names and cash rises to ~14.5%. That makes eleven exits against three buys since the start of August.

9 Sep 2026 — DABUR downgraded 🟢→🟡. Full 4-pillar review (Dabur-DABUR-Review-2026-09-09.md): Business 🟢, Financials 🟢, Valuation 🟢, Macro/regulation 🔴 → WATCH, hold 0.8%, don't add before Q2 (~28 Oct). Nothing broke at the company — Q1 FY27 was its best quarter in three years and the dividend is covered 1.5× by FCF. The downgrade is macro: ~48% of domestic sales are rural, the highest rural skew in large-cap FMCG, into a monsoon that failed (−13.8%), while inputs inflate (Brent ~$97.7, Hormuz shut) and the FY27 plan is explicitly price-led. Non-obvious finds: the quoted "EBITDA margin 24.3%" is not the consolidated figure (19.7% is); FII has fallen 20.5%→9.7% over 3.5 years, the same pattern as HUL; and the FSSAI stay (now 16 Dec) was won on procedure, not merits, on a claim ("100% Natural/Pure") that is the honey/ghee proposition itself. India staples ≈3.1% now runs three Ambers of four — HINDUNILVR 🟡 + DABUR 🟡 + ITC 🟡 + COLPAL — all on the same three drivers; treat as one position.

22 Aug 2026 — earnings refresh: 1 new print, 1 signal change, and one action item that needs attention before Monday.
🔻 Downgrade — Flowers Foods 🟡→🔴. Q2 (20 Aug) missed on both lines and the FY26 guide was cut for the first time this year. Net sales -4.0% on volume -5.8%; adj. EBITDA -19.2%, margin 9.3% from 11.1%; adj. EPS $0.21 vs $0.30. ⚠️ The telling detail: Branded Retail took price/mix +3.8% and lost 7.6% of volume — that is elasticity, not a soft comparison. FY26 adj. EPS cut to $0.75–0.85, EBITDA to $453–481M. With $399.9M of debt current against $52.8M cash, ~3.5x leverage and goodwill exceeding equity, the June dividend halving now reads as necessary rather than precautionary. Shares $6.95, sitting on the 52-week low, market cap -55.6% y/y. Offsets: gross margin held, the reset dividend is covered, 6.6x EV/EBITDA, and Tastykake is up for sale (Reuters, 10 Aug). Hold, don't average down.
🚨 ACTION — Real Brokerage: 10-for-1 reverse split lands 4:01pm ET Monday 24 Aug. The BC Supreme Court granted the final order on 21 Aug; the RE/MAX combination closes 24 Aug and Real REMAX Group Inc. begins trading 25 Aug under a new CUSIP. Every 10 REAX shares become 1 at ~10x the price — position value unchanged, but fractional entitlements below ½ a share are cancelled with no payment. Worth checking that your 2.6% holding is a clean multiple of 10 before Monday. RE/MAX cash elections were oversubscribed against the $80M cap, so electors are prorated to ~$4.33 cash + ~0.3535 shares — marginally more dilution, marginally less cash out.
⚠️ Correction — Rajesh Exports Q1 FY27 was filed on 14 Aug and the 16 Aug refresh missed it. Now logged, and it stays 🔴. Consolidated revenue +82.7% with PAT swinging to ₹46.92 cr from a loss looks like a turnaround — but that consolidated line is the exact figure SEBI's 3 Jun interim order alleges is fabricated (~₹15.15 lakh cr across FY21–FY25, 97–99% booked through unverifiable overseas subsidiaries), and the pre-tax margin on it is 0.02%. The verifiable standalone book shrank: revenue -7.6%, PAT -29.7%. New in the filing: an auditor Emphasis of Matter now covering the SEBI order, ED searches and a newly initiated SFIO investigation.
Checked and clear: a full sweep of all 74 holdings found no other earnings, trading update or profit warning between 15 and 22 Aug — the window fell in a genuine gap between the 9 Jul–14 Aug reporting run and Pernod Ricard on 27 Aug. Pernod's FY26 date is now confirmed (27 Aug, 07:30 CEST). Nomad's entry gains the actual reiterated ranges (organic -2% to -5%, cc adj. EBITDA -5% to -10% — declines, not flat).

19 Aug 2026 — portfolio change: two new positions, the first buys since 11 Aug.
Bought Tencent (700, HKEX) at ~1.0% — added 🟢 Green. The 25 Jul analysis scored it WATCH-leaning-Buy and was upgraded to BUY after the 12 Aug interim: revenue +11%, operating profit +12% with margin stable at 33%, and ex-New-AI-Products operating profit +19% with margin up 3pp to 42%. ⚠️ The widely-reported "profit miss" (IFRS net +0.7%) is a ~RMB14.5bn swing in associates, not Tencent's own business. Ads +22% and accelerating. Entered ~8.7% below the July price. Honest negatives: net cash −22% YoY as capex ran +176%, and the investment portfolio shrank RMB60bn in a quarter. Forward P/E ~14.3x — use forward, not the 17.5x trailing.
Bought ITC (NSE) at ~0.6% — added 🟡 Yellow, matching its 16 Aug review (three greens, one red). A buy-the-transition entry: the 1 Feb 2026 tax rewrite pushed cigarette tax incidence from ~50–60% to ~65–80% of MRP and cigarettes PBIT fell 31.5%. ⚠️ Do not read the "+27.6% revenue" headline as growth — net revenue ex-excise was −11.1%; the topline moved on an Ind AS reclassification. Stock is 1.2% off its 52-week low, 35% below the high, at 17.6x vs a 23.9x 10-yr median — but forward P/E is above trailing, so Q2/Q3 FY27 get worse before better.
Overlap to watch: Tencent is the top holding of FLXC (0.9%) and Prosus/PRX — analysed 18 Aug — is essentially a Tencent proxy, so effective exposure exceeds the 1.0% headline; China/HK is now ~2.7% across 700 + 9618 + FLXC. ITC lifts India FMCG to ~3.1% (HINDUNILVR 1.2 + DABUR 0.8 + ITC 0.6 + COLPAL 0.5). Book now 74 holdings; cash ~14.0% — first redeployment after ten exits in nine sessions.

16 Aug 2026 — 12 results folded in, 7 signal changes. Q1 FY27 season now essentially complete.
Upgrades: Indigo Paints 🟡→🟢 — revenue +19.7%, PAT +61%, volumes up double-digit across all four categories; best quarter in two years. · Titagarh 🟡→🟢 — PAT +70% on a real mix shift into passenger rail (+197% segment), OPM +300bps, order book ₹26,635 cr. · Valiant Organics 🟡→🟢 — PAT +254%, OPM 18.0% vs 12.2%, pharma segment turned profitable. · Bata 🟡→🟢 — PAT +23% on margin execution plus a ₹25/share interim dividend; stock +9%. · JD.com 🟡→🟢 — the profit inflection: food-delivery losses halved, group operating income swung positive, retail margins at records. · Sammaan Capital 🔴→🟡 — earnings still falling (revenue -31%, PAT -27%), but NNPA 0.15%, ratings upgraded to AA+ by all three agencies under IHC, borrowing cost -85bps. No longer deteriorating.
Downgrade: Repco 🟢→🟡 — PAT +5.5% and asset quality better YoY, but the credit-cost tailwind reversed (+0.2% after two years of write-backs), coverage fell to 54.5%, and GNPA ticked up sequentially.
Held: Laxmi Dental 🟢 (record quarter, profit growing 1.7x revenue) · Relaxo 🟡 (margins flat, not expanding) · NMDC Steel 🟡 (the +97.6% PAT is optical — EBITDA actually fell 2%; the govt stake sale is the real story) · Asian Granito 🟡 (growth is low-margin subsidiary volume; standalone PAT -57%) · Nomad 🟡 (volumes -6%, EPS guide cut on financing not operations).
⚠️ Data flags: Asian Granito has two conflicting year-ago bases — exchange filing implies PAT +8.8%, the company's own release implies -28.4%; unresolved. Repco's widely-quoted "NII +4.3%" uses Q4 as the base (correct: +10.2%). Titagarh's "+197%" is the passenger segment only, not total revenue. Flowers Foods is 20 Aug, not 14 Aug.
Only Rajesh Exports (🔴, SEBI matter) is still on Q4 FY26 — expected Aug/Sep.

11 Aug 2026: Amara Raja 🟡→🟢 — released a day early (10 Aug, not 11). Consolidated revenue +23.9% to ₹4,215 cr, the strongest top line in years; PAT ₹191 cr +15.9%; new-energy revenue +73% with segment loss halving to ₹22 cr; AP pollution-board closure orders revoked. Stock +7.9%. ⚠️ EBITDA margin fell to 9.6%, below 10% for the first time — growth is volume/price-led, not margin-led. (Headlines quoting ₹4,041 cr / PBT ₹272 cr are STANDALONE.)
Portfolio changes: sold Boston Scientific (BSX) on 10 Aug (was 0.6%, 🟡) — held ~11 days, bought 30 Jul. Bought BMW (11 Aug) — added as 🔴 Red on its latest print: Q2 automotive EBIT margin collapsed to 2.3% vs its own 8–10% target, group pre-tax profit -35%, and FY26 guidance is just a 1–3% auto margin. A deep-cyclical value entry, not a healthy compounder. Booked profit in Meta (0.5%, 🟡, held ~12 days), PayPal (1.6%, 🟢) and Accenture (1.0%, 🟡) — all sold 11 Aug. Book now 72 holdings; Global 25 / 45.9%, cash ~15.6%. Ten exits in nine sessions against one buy.
⚠️ Auto exposure now 5.1% across three names — Stellantis 3.4% 🟢, VW 1.1% 🟡, BMW 0.6% 🔴 — all three fighting the same US-tariff and China-demand pressures. Correlated bets on one macro thesis.
Still pending as of 13:30 IST — board meetings held today, filings expected this evening: Bata (interim dividend to be declared, record date 19 Aug), Repco, Asian Granito, Laxmi Dental (call 12 Aug). Then 13 Aug: Sammaan, Indigo Paints, Relaxo, NMDC Steel, Nomad, JD.com; 14 Aug Flowers Foods; Titagarh 11–14 Aug.
10 Aug 2026 — earnings catch-up: 16 results folded in, 6 signal changes.
Upgrades: ESAF SFB 🟡→🟢 — PAT ₹80 cr vs a ₹81 cr loss; NNPA collapsed to 0.8% from 3.8%, slippages -84% (flattered by a loss-making base, GNPA still 5.4%). · Aarti Drugs 🟡→🟢 — EBITDA +30%, margin +120bps; the "-7% PAT" is a tax-refund base effect, ex-that +29%. · HT Media 🔴→🟡 — back to profit (₹43.5 cr vs -₹11.4 cr) on a genuine print recovery, though other income doubled and supplies ~half the swing.
Downgrades: Chambal 🟢→🟡 — revenue -11.8% on plant shutdowns; most outlets ran standalone PAT +10.3%, but consolidated PAT fell 4.6% on an IMACID JV loss. · Real Brokerage 🟢→🟡 — revenue +30% and adj. EBITDA +38%, but swung to a GAAP loss on RE/MAX deal costs with gross margin compressing.
Held: Occidental 🟢 — big beat (adj. EPS \$2.40 vs ~\$1.88), debt down to \$11.8bn, dividend +8%, best FCF since 2022. · Uber 🟢 — bookings +24%, first-ever \$10bn TTM FCF (GAAP EPS flattered by a \$1.6bn revaluation gain). · LIC 🟢 — VNB +61%, margin +750bps, and the government OFS clears the sell-down overhang (~5 margin points came from rate assumptions, though). · GMM Pfaudler 🟡 — order book +20%, but PAT +118% is finance-cost/FX driven while EBITDA fell 7%.
Second sweep (prompted by a check on Inox Wind — my "mid-Aug" dates were stale for six names): Deepak Nitrite 🟡→🟢 — revenue +36%, EBITDA margin doubled to 21%, plus a ₹2,500 cr Bisphenol-A project. · Sheela Foam 🟢 — record quarter, first ever above ₹1,000 cr revenue (the "+839% PAT" is a low-base artefact; +45% EBITDA is the real read). · Inox Wind 🟡 — order book jumped to 4.4 GW (from 3.1), but Q1 revenue was FLAT against a +75% FY27 guide — execution credibility is now the thesis. · Electrosteel 🟡 — second straight sharp fall (PAT -46%, volumes -26%); on watch. · Ola Electric 🔴 — revenue -45% YoY, still ₹336 cr of loss. · 🚩 Coffee Day 🔴 — the material item isn't the P&L: auditors issued a DISCLAIMER OF CONCLUSION with explicit going-concern doubt, ₹3,357 cr MACEL receivable of uncertain recoverability, covenant breaches and lenders recalling loans.
Date corrections: Relaxo 13 Aug (not 5 Aug); Nomad full Q2 13 Aug; Bata, Asian Granito, Amara Raja, Repco all 11 Aug; Laxmi Dental 12 Aug; Sammaan, Indigo Paints, NMDC Steel ~13 Aug; Titagarh 11–14 Aug; Valiant ~mid-Aug.
7 Aug 2026 — portfolio change: sold Dassault Systèmes (DSY) (was 1.0%, 🟢 — Q2 had shown 3DEXPERIENCE/cloud +14% and margins up). Book now 75 holdings — India 47 / 38.5%, Global 28 / 49.0%, cash ~12.5%. That's six exits in five sessions (Atul Auto, Gartner, SAP, Sundram Fasteners, Dassault; SMTI earlier) with only LULU bought — cash is now larger than any position by a wide margin, and the EU book is down to 10 names.
6 Aug 2026 — payments day, and both prints were bad. Shift4 🟢→🔴 — the one signal change. Q2 beat on every adjusted line (GRLNF $624M +51%, adj. EPS $1.32 vs $1.24) and then cut FY26 adj. EPS to $5.15–5.35 from $5.50–5.70. The damage isn't the cut, it's the cash: adjusted FCF fell 82% to $21M, true FCF was −$9M, and FY26 adj. FCF is now guided flat at $465–475M vs FY25's $464M while EBITDA grows ~20%. Organic GRLNF growth has halved to 11%. Net debt $4.16B sits at ~3.6x against their own 3.75x ceiling, and the buyback was throttled from $295M to $25M. 🚩 Separately: the SEC-filed shareholder letter still contains unfinished draft placeholders ("TBU - TAYLOR LETTER", "TBD logos") and the leverage chart is blank — no net leverage figure disclosed for a second straight quarter, following the CFO change (Aug'25) and CAO resignation (Mar'26). Valuation stays the green pillar (8.8x fwd, ~11% FCF yield, cheapest ever) — this is a hold, not an add. Stock −13%. · Fiserv stays 🔴 and gets worse: adjusted revenue $4.96B −4%, a miss (ignore the "$5.29B beat" — that's GAAP revenue against an adjusted consensus), Financial Solutions organic deteriorated −6%→−8%, and FY26 adj. EPS was cut to $7.20–7.40 from $8.00–8.30 — the third cut in 18 months. Stock −11.4%. Post-Q2 note: avoid for new money, reduce the 3.0%. · Net effect: three of your four payments names are now 🔴/🟡 on a ~13.8% theme weight.
5 Aug 2026: Vaibhav Global — Q1 out (4 Aug): revenue +13%, PAT +50%, EBITDA +37% (margin 11%), digital +21%, interim dividend ₹1.50: stays 🟢, margin expansion continuing. · Occidental — Q2 lands tonight (after US close); preliminary 8-K shows realized oil $96.78/bbl but negative US gas realizations (-$1.48/Mcf) and a $156M hedging cost. · Uber Q2 also due today (not out at refresh time). · Portfolio change: sold SAP (was 1.2%, 🟢) and Sundram Fasteners (was 0.4%, 🟢 — record FY26, non-auto diversification working). Book now 76 holdings — India 47 / 38.5%, Global 29 / 50.0%, cash ~11.5%. That's five exits in three days with no redeployment; cash is now comfortably your largest single line.
3 Aug 2026: Escorts Kubota — Q1 out: revenue +28%, PAT ₹386 cr. The "-72% YoY" doing the rounds is a base effect (last year's Q1 held the ₹1,601.7 cr railway-division sale); like-for-like on continuing operations it's +4.5%. Strong topline, near-flat profit → margin compression: stays 🟡. No other portfolio results since 31 Jul. · Portfolio changes: sold Atul Auto (was 0.4%, 🟢) and Gartner / IT (was 1.1%, 🟡 — exited at +10% over ~6.5 months, ≈19% annualised, and ahead of its 4 Aug Q2 print). Book now 78 holdings — India 48 / 38.9%, Global 30 / 51.2%, cash ~9.9%.
31 Jul 2026 — actual weights loaded. All holdings re-ranked on your 31 Jul statement; weights are now % of total portfolio (India 39.3% · Global 52.3% · cash ~8.7%), replacing the earlier per-book estimates. Sanara MedTech (SMTI) removed — absent from the new list, treated as sold. LVMH (0.5%), Meta (0.5%) and BSX (0.6%) now carry confirmed weights.
Portfolio change: bought Lululemon (LULU) — re-entry; added to Global as 🔴 Red on its latest print (Americas comps -5% for a 5th straight quarter, FY26 EPS guidance cut to $10.95–11.15, tariff drag) — a deep-value turnaround bet rather than a healthy compounder. Weight provisional 0.5%.
Results: AWL Agri — Q1: revenue +17.5%, PAT +47.7%, margin +130bps: upgraded 🟡→🟢. · Worldline — H1: EBITDA beat, net debt nearly halved, but FY revenue guidance CUT on Financial Services weakness: stays 🟡. · Prada — H1: +5% organic but Versace dilution cut EBIT margin ~520bps, net income -15%: stays 🟡. · Titagarh date correction: Q1 was never due 30 Jul (stale 2024 source) — now expected ~11–14 Aug.
30 Jul 2026: Microsoft — FQ4 beat, Azure +43%: was 🟢 (position since sold). · Dabur — Q1 PAT +15%: stays 🟢. · Bajaj Housing — PAT +23%, NIM squeeze: stays 🟢. · Stellantis — H1 swung to €670M profit, guidance only reaffirmed: stays 🟢. · Fannie Mae — Q2 net income +20%: stays 🟡. · Portfolio change: sold Microsoft, bought Meta (META) & Boston Scientific (BSX) — Global back to 31, both 🟡.
29 Jul 2026: PayPal — Q2 beat & raised FY EPS guide: 🟡→🟢. · Colgate India — Q1 rev +12%, PAT +6.9%: 🟡→🟢. · Suzlon — Q1 rev +23%, book 6.1 GW; PAT -6%: stays 🟢. · Kering — H1 Gucci slowing (Q2 -2%): stays 🟡. · Sold Infosys NYSE ADR — Global now 30 holdings (Infosys still on NSE).
28 Jul 2026: HUL — Q1 rev +10.1%, USG best in 13 quarters; PAT -3% on prior-year tax base: stays 🟢. · Sold Capgemini, bought LVMH (MC) — added to Global as 🟢.
25 Jul 2026: ACC — Q1 PAT -61%: stays 🟡. · Volkswagen — H1 PAT -31%, revenue outlook cut: stays 🟡.
24 Jul 2026: Infosys — FY27 guidance top-end trimmed: 🟢→🟡 (India + ADR). · Suryoday SFB — PAT +113%: 🟡→🟢. · SAP — cloud +24%: stays 🟢. · Solara — PAT +55%: stays 🟡.
23 Jul 2026: Tanla 🟢 (calls paused); Dassault 🟡→🟢; added valuation/next-results/catalysts/insights/notes.
22 Jul 2026: Dashboard created — baseline; Bandhan Q1 (→🟡).