What we do not own, and why.

Kotak Mahindra Bank is the obvious omission. Our model rates it REDUCE: a 12% ROE bank at 2.0x core book, priced for a 17% ROE it does not earn, with CASA sliding toward 40% and a CEO succession still ahead. The argument the house has used for years — a 16–17% ROE bank cannot compound above that without dilution, so 4–5x book is unsustainable — applies with more force to a 12% one. IndusInd is a governance rebuild we do not need to underwrite; Federal Bank and AU Small Finance are good franchises but neither has IDFC First's deposit engine or its operating-leverage runway at a comparable price. Among public-sector banks, Bank of Baroda and Canara are cheaper on paper but have weaker CASA, thinner capital and none of SBI's subsidiary optionality; SBI's premium (1.25x vs 0.9–1.0x) is the price of a franchise that survived the AQR and emerged with the best retail book in the system.

Nestlé India is a superb business at 60x+ with the least GST benefit and the slowest volume growth of the group; ITC is cigarettes-led (we specifically exclude cigarettes from the volume thesis) and now a hotels demerger story; Dabur and Godrej Consumer have weaker execution and portfolio issues (honey/chyawanprash seasonality; GCPL's African drag). Marico is a copra-cost story, not a volume story. Tata Consumer pays 60x for tea. In discretionary, V-Mart and ABFRL lack Trent's unit economics and balance sheet, and Shoppers Stop is the wrong price point for a transfer-led recovery.

HDFC Life is the quality benchmark but grew VNB only 2% in FY26 and trades at a premium for it; ICICI Prudential Life has a weaker bank channel. LIC is cheap for structural reasons — product mix, agency cost and a 57% share that is only going one way. In general insurance, Go Digit is a 4–5x-book growth story without ICICI Lombard's underwriting record, Star Health has the worst-in-class loss ratio in retail health and New India Assurance has a combined ratio above 110%.

Cholamandalam is an excellent vehicle financier but at 4x+ book already prices its growth, and its ROE (18–20%) is on a more bank-like, lower-yield book; M&M Financial has a decade of volatile credit costs; Sundaram Finance is conservative to the point of not growing. HDB Financial lists at a premium to its own economics on the HDFC halo. We hold no gold-loan or microfinance NBFC in the PMS (Manappuram appears only in the AIFs): the gold-loan regulatory reset and the FY25–26 microfinance stress cycle are exposures we would rather take through IDFC First and Shriram's diversified book.

Sun Pharma is the quality leader but at 35x+ already prices its specialty franchise and is the most exposed to the lenalidomide cliff; Cipla faces a Goa-plant remediation and a respiratory pipeline dependency; Lupin has re-rated to Sun-like multiples on a handful of US launches; Zydus is the closest competitor to our GLP-1 thesis but priced for it. We own no hospital (Apollo, Max, Fortis at 60–80x EBITDA-normalised P/E) and no CDMO at 70x+ (Divi's, Syngene) — the sector can be attractive while individual valuations still matter. Apollo Pharmacy is buried inside a hospital valuation; MedPlus is the pure play.

We do not own TCS, HCL Tech or Wipro because if we are going to hold a small tactical position in a sector with a capped upside we want the cheapest large franchise with the cleanest AI narrative — Infosys. Dixon (2.6–3% margins, PLI 1.0 expired, ROCE falling) is volume without value; Syrma and Amber are priced higher than Kaynes on trailing earnings with lower margins. PB Fintech is priced at 90x+ for an insurance-distribution model; Eternal is a quick-commerce cash-burn story we hold only in the AIFs. Persistent and Sagility lack Indegene's domain moat.

Ambuja/ACC is the other consolidator but the Adani group's integration and cost delivery are lagging (EBITDA/t down 13% YoY) and the holding-company structure adds complexity; Shree Cement is the most efficient north Indian producer but trades at 50x+ on a self-imposed growth cap; JK Cement and Ramco are regional stories at leader-like multiples. Among pipes (in our Industrials bucket) we own Astral over Supreme (PMS) because the CPVC-led mix and adhesives optionality offer more growth for a similar price.

We own no defence (HAL, BEL, BDL: 40–60x for government-monopsony order books), no railways (RVNL, IRFC, Titagarh) and no renewables (Suzlon, Inox Wind, Waaree) — great businesses for an extended period, but the valuations and investor faith are extreme. Among capital-goods bellwethers, ABB, Siemens and Cummins at 50–70x price a private-capex boom that machinery data (26% of GFCF) does not yet confirm. Bharat Forge is a good company at 40x with a defence premium; we prefer RK Forgings (Autos) for the same end-markets at a cyclical trough. Supreme Industries is owned only in AIF I; in the PMS Astral is the pipes expression.

SpiceJet is a going-concern question with 1.6% share; Air India is unlisted. There is no second airline worth owning in India. The undisclosed "Others" (5.0%) include the portfolio's Real Estate (0.6%) and Textiles (0.5%) exposure — see the "Other sectors" page.

Maruti is the PV leader but at 30x prices a recovery that CNG/hybrid mix has already delivered, and it is the most exposed to small-car weakness; Tata Motors is a JLR story; M&M is the best-run OEM but at 30x+ after a 3x move we prefer Bajaj's export optionality and net cash. Hero MotoCorp is losing share to scooters and EVs; TVS and Eicher are great franchises at 40–45x. Among components, Bharat Forge carries a defence multiple and Motherson a European-auto cycle we do not want.

Reliance Industries gives Jio only as one-third of a conglomerate with refining and retail cycles we do not want to own for this thesis; a Jio Platforms listing would be the way to revisit. Vodafone Idea is an equity option on government forbearance, not a business. Indus Towers (held in the AIFs as Value) is a tenancy annuity whose growth is now capped by consolidation.

Reliance Retail is unlisted; Vishal Mega Mart is a good value-retail story at a similar multiple with less proven unit economics; Zepto is the IPO we are watching, not buying. Relaxo is losing volume in mass footwear, Bata has not grown in five years and Metro Brands is premium-priced for premium footwear. Apparel is expressed through Trent (in FMCG).

Hindalco is 60% Novelis (US rolling, tariff and auto-cycle exposure) and trades at a similar multiple; NALCO is the cheapest but is a PSU with an alumina-price, not metal, earnings driver and a capacity-addition record that is slow. We own no steel: the China export overhang and the domestic capacity race make it the wrong metal for a deficit thesis.

There is no listed multiplex peer of scale. Studio and music stocks (Saregama, Tips, Zee) are content bets with weaker economics; OTT platforms are unlisted or foreign. Broadcasters are structurally challenged by streaming.

We own no specialty-chemical compounder (PI, SRF, Navin, Deepak) at 40–60x — the China-plus-one story has been priced for three years while Chinese dumping and agrochemical destocking hit earnings. Coromandel is the quality fertiliser name but at 25x prices its phosphates diversification. Tata Chemicals is a soda-ash cycle we do not want.

We own no large-cap developer (DLF, Lodha, Godrej Properties at 40–60x on pre-sales narratives) and no textile exporter at scale; these are option-sized positions.