Sum-of-the-parts Valuation Where Each Part Is

Figures converted from Indian rupees at historical FX rates — see data/company.json.fx_rates (the 2026 valuation cluster converts at 0.01059 USD/INR). Ratios, margins, and multiples are unitless and unchanged.

Sum-of-the-Parts — Each Part Valued on Its Own Terms

The rest of this report has already settled that Edelweiss must be valued as a sum of its parts. This page does the harder thing the phrase "valued properly" demands: it values each of the seven businesses with the lens that fits its economics, cuts every stake to Edelweiss's actual economic ownership, nets the holdco debt (not the consolidated debt), credits unallocated parent assets, applies a realistic conglomerate discount, and divides by shares — then stress-tests the two or three assumptions that actually move the answer.

The headline is sobering for a "parts-worth-more-than-the-whole" story. Summed gross, the attributable parts are worth meaningfully more than the $1,227 million market cap. But once you subtract $679 million of corporate net debt and apply the discount the market rationally attaches to an unlisted, promoter-controlled, related-party-dense holdco, the central estimate lands close to — not far above — today's price. The re-rating to ~2.5x book has largely priced the break-up. The upside that remains is conditional, not embedded.

The three BigValues above show the bear/base/bull per-share outputs of the full waterfall built on this page; the central column ($1.16) is the base case. Market price $1.30 (25 Jun 2026). Derived on this page from segment economics, third-party transaction marks, and the equity bridge; not a company figure.

How each part is valued — the lens matters more than the number

An alternatives manager, a mutual fund, an ARC, two lending books and two insurers cannot share one multiple. The discipline of a proper SOTP is matching the basis to the business and tagging how hard the evidence is. Three of the seven carry third-party cash marks set within the last year; the other four rest on book or embedded value and are softer.

Source: lens choice is analyst judgment per business type; mark-quality tags reflect whether a third party has set a cash price. Transaction marks per the FY2026 investor presentation (Apr 2026) and press reporting cited below.

Two principles run through the whole build, and both cut against the loose bull arithmetic that floats around this name:

A mark is not a price, and a 100% mark is not an attributable value. Edelweiss owns 100% of only three of these businesses. The EAAA and Nido marks especially must be cut to Edelweiss's economic share before they enter the sum — and for Nido that share has collapsed (see the correction below).

Subtract holdco debt, not consolidated debt. Each subsidiary's borrowings already sit inside its own equity value. The only debt that nets against the parts is the $679 million of corporate (holdco) net debt — not the $1,105 million consolidated figure. Using the consolidated number would double-count subsidiary leverage; using zero would ignore the parent's real obligations.

The single biggest correction: Nido is now a ~27% stake, not ~55%

The existing illustrative SOTPs in the report value the housing arm at roughly $268 million attributable, assuming Edelweiss keeps ~55% of Nido after Carlyle's 45% purchase, and quoting an implied equity of ~$487 million. Both inputs are wrong, and the error roughly triples the housing line.

Press reporting (Livemint, Business Standard, NDTV Profit, Feb 2026) is explicit: Carlyle (via CA Sardo) and Salisbury Investments — Aditya Puri's family office — will jointly hold ~73% of Nido, through a $64 million secondary purchase plus a $159 million primary infusion ($222 million total), pending RBI/NHB/CCI approval and a target close around 31 July 2026. That math implies a post-money equity value of $222 mn ÷ 0.73 ≈ $305 million — not $487 million. The $487 million figure double-counts the $159 million primary (new capital into Nido is not a purchase of Edelweiss's existing shares). And Edelweiss's residual is ~27% fully diluted, not ~55%.

Source: Carlyle/Salisbury 73% stake and $222 mn ($159 mn primary) per Carlyle press release (10 Feb 2026) and Livemint/Business Standard/NDTV Profit (Feb 2026); post-money equity and residual stake derived from the transaction structure. Pre-deal Nido book equity $90 mn per FY2026 presentation p.5.

The part-by-part build (base case)

Below is the auditable core: for every business, its valuation basis, the implied 100% value, Edelweiss's economic ownership, the attributable value, and the evidence tag. Ownership percentages are taken from the FY2025 annual report's group-structure table (EAAA and the mutual fund stepped down post their FY2026 placements; Nido is shown post-Carlyle).

Source: segment equity and PAT from FY2026 presentation p.5 & p.8; Life embedded value $250 mn p.47; EAAA placement p.18; ownership % from FY2025 Annual Report group-structure table (EAAA/MF/NBFC/Nido/Zuno 100%, EARC 60%, Life 80%) stepped for the FY2026 EAAA 4.4%, MF 15% and Carlyle/Nido transactions. Multiples and the Zuno figure are analyst estimates.

Source: attributable column of the table above. EAAA alone is $861 mn — roughly 47% of the gross parts and ~70% of the market cap — the concentration that defines the whole thesis.

EAAA is the SOTP. At a $900 million platform mark, Edelweiss's ~95.6% stake ($861 mn) is seven-tenths of the entire market cap and 47% of the gross parts. Every conclusion on this page is, first and foremost, a bet on that one mark. The mutual fund ($270 mn) is a distant second; the two insurers and two lending books together ($613 mn gross, but much of it un-attributable or low-ROE) are the long tail.

How much rests on hard marks versus estimates

"Valued properly" also means being honest about evidence quality. Of the $1,829 million gross attributable base, about two-thirds sits on third-party cash transactions and one-third on book/EV estimates — but even the "hard" two-thirds carries an asterisk: the EAAA and MF marks come from small strategic placements (4.4% and 15%), not arm's-length majority sales. The one true majority-control price in the stack is Carlyle's Nido deal — and that line is tiny after the ownership correction.

Source: derived by summing the attributable base-case values from the part-by-part table — EAAA + MF + Nido (marks) vs EARC + NBFC + Life + Zuno (estimates).

Are the marks rich or conservative? Cross-checking the two that matter

A mark is only useful if it survives a fundamentals check. The two big lines — EAAA and the mutual fund — both pass, sitting within or below relevant benchmarks. That is reassuring for the base case, but it also means there is little hidden conservatism to harvest.

Source: EAAA/MF PAT and AUM from FY2026 presentation p.5/p.8/p.34; WestBridge 57x and 30–60x benchmark per Business Standard/ET (Dec 2025); Nippon AMC ~44x P/E and 360 ONE market cap $4,685 mn on ~$62 bn AUM per Competition tab and market data; Indian life-insurer P/EV is a general market range.

The net read: the EAAA mark is full, not cheap (a 32x trailing platform inside a market that pays 27–44x for scaled listed AMCs — and EAAA must list against 360 ONE, eight times its AUM). The mutual fund looks cheap on forward earnings (35x FY26 vs a 30–60x benchmark) — the one place the base case may understate. Life at 1.0x EV is deliberately conservative given the losses, leaving option value if it reaches breakeven. There is no large pocket of buried conservatism that rescues the SOTP — which is exactly why the base case clears so little above debt.

The equity bridge: from gross parts to per share

This is where loose "parts-worth-more" arithmetic goes to die. The gross attributable parts ($1,829 mn base) look comfortably above the $1,227 mn cap — but the bridge to per-share equity erodes most of that cushion.

Source: gross parts from the attributable table; holdco net debt $679 mn (flat YoY) per FY2026 presentation p.14; unallocated holdco assets (~$212 mn owned property, with the ~$106 mn fund investments treated cautiously to avoid double-counting parent cash) per Q4 FY2026 transcript p.4; 18% discount is analyst judgment.

Source: pre-discount equity $1,362 mn and post-discount $1,117 mn ÷ 946 million shares; market price per NSE close 25 Jun 2026.

The bridge tells the whole story in three numbers. Without any holdco discount, the parts are worth ~$1.44/share — about 11% above the price. Apply the conglomerate discount the market rationally attaches to this structure (~18%) and you get ~$1.18 — about 9% below the price. In other words, most of the "gap to intrinsic" people cite is simply the discount the market is already, correctly, applying. Whether that discount is "anomalous" (bull) or "warranted" (bear) is the real argument — and given a flat-at-$679 mn holdco debt still partly funded by ~10% retail NCDs, an unlisted promoter-controlled structure, and a regulator that has policed the credit subsidiaries once, the discount is not obviously unjustified.

Bear / base / bull — and what actually swings it

The range below carries the same three scenarios through the whole bridge — different per-part values, different unallocated credit, and different discounts. It brackets the existing Bull $1.75 / Bear $0.90 anchors, but with a base case ($1.16) that is more sober because it carries the corrected Nido stake and a realistic discount.

Source: each scenario runs gross attributable parts − $679 mn holdco debt + unallocated assets, then the stated discount, ÷ 946 mn shares. Bear cuts EAAA to $741 mn and applies 25%; bull lifts EAAA to $1,112 mn (listing re-rate) and applies 10%. Market price $1.30.

Source: derived on this page. The two columns that move the most between scenarios are the EAAA mark and the holdco discount — confirmed by the sensitivity grid below.

The answer to "which assumptions matter" is unambiguous: the EAAA value and the holdco discount. The grid below holds everything else at base and flexes only those two. Note that you need both a high EAAA mark and a compressed discount to get a number with real upside — at the base EAAA mark, even a 10% discount only reaches $1.29 (spot). That is the mathematical statement of the Watchlist verdict.

Source: per-share output holding all parts except EAAA at base, flexing EAAA's 100% value and the holdco discount; ÷ 946 mn shares. Spot $1.30 sits inside the grid, between the base/10% and base/18% cells.

What the parts are worth versus what you'll actually get — and when

A proper SOTP separates intrinsic value from realisable value. As of 29 June 2026, the two marks that carry the SOTP are both still unrealised:

EAAA IPO — SEBI approved the DRHP on 23 April 2026; the issue is a 100% offer-for-sale (~$159 mn) by the Edelweiss promoter entity, targeted for a Jul/Aug-2026 window. It has not priced or listed. The IPO has slipped from a Q4-FY2024 plan — roughly two years. A list below the $900 mn private mark is the single largest downside risk.

Carlyle/Nido — announced 10 Feb 2026, pending RBI, NHB and CCI approvals, targeted close ~31 July 2026. The ~$64 mn secondary cash to the holdco is not yet banked.

Until those print, the base case is not a price you can capture — it is a mark you are waiting on. A modest realisation-timing haircut is already embedded in the discount; a further slip would push the realisable value toward the bear case even if the underlying marks hold. This is why the SOTP, valued properly, supports a Watchlist rather than a Buy: the parts are real, the marks are defensible, but net of debt, ownership and discount they do not offer a margin of safety at $1.30 — they offer a fairly-priced option on two pending events.