The price requires future owner earnings
Reference ₹363.65; current statutory earnings are small relative to market value.
Watch: Watch cash conversion.
Home / Research / Eternal Limited
The price requires future owner earnings
Eternal Q2 FY27 shareholder letter and signed results, 21 Oct 2026
Eternal Q2 FY27 shareholder letter and signed results, 21 Oct 2026
Eternal Q2 FY27 shareholder letter and signed results, 21 Oct 2026
Eternal Q2 FY27 shareholder letter and signed results, 21 Oct 2026
Reference ₹363.65; current statutory earnings are small relative to market value.
Watch: Watch cash conversion.
Q2 statutory revenue rises 77.6% YoY; LFL adjusted growth 58%.
Watch: Use comparable NOV and contribution.
Attributable profit ₹125 crore versus ₹65 crore.
Watch: Do not extrapolate current P/E.
Calculated PBT less other income plus finance and D&A.
Watch: Not adjusted EBITDA.
16 November 2026 | Q2 FY27 ended September 30 | Reference close ₹363.65, November 13. Source-verified institutional longform; full independent QA pending.
WATCH over 18–36 months. Eternal has established two unusually valuable consumer habits: ordering meals and replenishing household goods quickly. The investment debate has moved beyond whether these services have demand. It is now about how much capital is needed to serve incremental demand, how much of the customer economics survives competition, and whether the parent can convert operating profit into cash after store fit-outs, inventory and leases. Q2 statutory consolidated revenue from operations ₹24,138 crore, up 77.6% from ₹13,590 crore, is a striking headline. But the shift to first-party inventory in Blinkit changes the revenue denominator. The shareholder letter's like-for-like consolidated adjusted revenue growth of 58% is a more useful growth comparison, though still a non-GAAP measure. Neither figure is equivalent to growth in cash earnings.
Key takeaways
The counter-thesis is that attractive early cohorts reflect low competition and favourable site selection; the next 2,249 stores may be less productive, and competing networks can force higher discounts and delivery costs.
The falsifier is not a single quarter of lower margins during expansion. It is a sustained failure of mature cohorts to generate enough cash to finance new cohorts, alongside deterioration in retention or order density.
Management now plans around 5,000 stores by March 2028 versus 2,751 today. The stock price ₹363.65 implies approximately ₹334,922 crore market capitalisation using the reported 921 crore paid-up shares.
This is a substantial valuation for a business with H1 statutory attributable profit of ₹217 crore and negative cash after physical investment and lease payments. The market is paying for future earnings power, not current earnings.
Our model therefore asks what future owner cash is required to justify that price rather than extrapolating the current P/E.
Food delivery is a marketplace with real-time execution obligations. Restaurant supply, discovery, delivery partner density, payment, customer support and dispatch are coordinated within minutes. A larger order pool can improve delivery utilisation and restaurant choice, but the network effect is not absolute: customers and restaurants can use multiple platforms. The economically important variables are order frequency, delivery cost per order, restaurant commission, platform fees, advertising, discounts and repeat-customer acquisition cost. An additional order from an existing customer can be more valuable than a subsidised first order if retention and delivery economics are favourable. Management's decision to reinvest incremental surplus in growth should be judged on repeat contribution after incentives, not gross order growth alone.
Key takeaways
Management says approximately 90–95% of SKUs are first-party and expects no further working-capital change solely from that mix shift. That does not mean absolute working capital stops growing. A larger store network and broader assortment can require materially more inventory even at a constant first-party share. Supplier credit can finance part of this, but payment terms are not permanent equity capital.
A useful unit model is orders per day × net average order value × contribution margin, less store fixed costs and allocated central costs. For illustration, 1,000 orders/day at an assumed ₹600 basket yields ₹6 lakh daily NOV or ₹21.9 crore annual NOV at 365 days.
A 5% mature-city adjusted EBITDA/NOV ratio would imply ₹1.095 crore annual adjusted EBITDA on that illustrative volume. This is not an observed store profit or a company forecast: the actual basket, operating days, lease costs, central allocations and initial capex vary.
The point is sensitivity. At 800 orders/day and a 3% margin the same basket generates only ₹0.526 crore annually. A 20% density miss and 200 basis-point margin miss together more than halve earnings. That is why cohort payback, not store count, is the critical expansion KPI.
Hyperpure can improve the group supply chain without requiring a consumer private label. It serves around 85,000 restaurants versus 62,000 a year earlier, with 25% of orders arriving within two hours. A dependable procurement service can reduce restaurants' own inventory and food waste, improving retention.
However, low-margin wholesale turnover can inflate consolidated revenue without creating equivalent value. The correct measures are gross profit per delivery, receivable days, spoilage, repeat procurement and return on working capital.
Management's decision not to pursue broad private labels is economically coherent: owning brands would add product development, marketing and inventory risk and could conflict with advertising customers. The opportunity cost is foregone private-label gross margin, which should be weighed against incremental brand-building costs rather than assumed free.
District is an option on going-out commerce, with different competitive protection. Restaurant reservations and non-exclusive tickets can be accessed through alternative channels, so the platform has less control over underlying fulfilment than food delivery or quick commerce.
Exclusive inventory, venue relationships and physical experiences can improve defensibility. Management's FY30 goal of ₹28,000 crore NOV and ₹1,400 crore adjusted EBITDA implies a 5% margin. That is guidance, not realised profitability.
The segment should earn its own capital allocation rather than being valued as a guaranteed extension of the food delivery network.
All amounts are ₹ crore unless specified. The signed consolidated unaudited results appended to the October 21 shareholder letter show Q2 revenue 24,138 versus 20,211 in Q1 and 13,590 a year ago. Growth is 19.43% sequential and 77.62% annual. Total income 24,489 includes other income 351. Expenses 24,166 reconcile to PBT 323. Current tax 175 and implied deferred tax 23 reconcile to attributable profit 125. Q1 PBT 272 less tax 180 gives profit 92; prior Q2 PBT 129 less net tax 64 gives 65. Attributable profit grows 92.3% YoY, but from a small base. Q2 PBT/revenue is 1.34%, versus 1.35% in Q1 and 0.95% a year earlier. The statutory operating margin is not expanding at the same pace as revenue.
Key takeaways
Management defines adjusted EBITDA by adding share-based payment expense and subtracting lease rentals. These adjustments are economically important: employee equity compensation dilutes owners, and rent is a real cost. Do not compare this proxy directly with segment adjusted EBITDA/NOV margins.
The expense bridge reveals the cost of the growth strategy. Purchases of stock-in-trade rise to 15,746 from 8,795, reflecting the inventory model and scale. Advertising and sales promotion 1,138 versus 806 rises 41.2%; delivery and related charges 3,581 versus 2,213 rises 61.8%; depreciation 625 versus 376 rises 66.2%.
As a share of reported revenue these costs can appear to fall because gross merchandise revenue is now included. That is not proof of lower cost per order. Like-for-like revenue and NOV, orders and unit contribution are needed to assess true efficiency.
H1 revenue 44,349, PBT 595 and attributable profit 217 reconcile to the two quarters. H1 operating cash flow 278 versus 38 in the prior half is an improvement, but physical and intangible investment cash outflow 1,656 versus 788 has more than doubled.
Operating cash less capex is minus 1,378, versus minus 750. Lease principal 361 and lease interest 281 are financing cash flows, so a lease-adjusted cash proxy is minus 2,020, before other interest and investment income.
The prior comparable is minus 1,092 (38 − 788 − 195 − 147). This is not a standard free cash flow definition; it is an explicit cash-after-investment-and-lease measure. It shows that expansion is still consuming cash despite positive statutory earnings.
Working capital explains much of the difference. H1 operating profit before working-capital changes is 1,955, but cash generated from operations is 322. Inventory absorbs 2,078 and trade payables provide 2,861.
Other financial assets absorb 1,011 and other assets 908. The net working-capital and other operating adjustments reduce pre-working-capital cash by 1,633. Management attributes the quarterly cash decline partly to festive inventory and the restaurant settlement cycle.
That is plausible and testable: if seasonal inventory unwinds and settlement normalises, the next quarter should show improved cash conversion. If the cash deficit persists, the structural cost of growth is higher than the seasonal explanation suggests.
The shareholder letter reports cash balance 16,475 at September end, down 1,813 in the quarter, implying 18,288 at June end. This is a broad treasury measure, not statutory cash equivalents. The consolidated balance sheet shows cash equivalents 467, bank balances 80, current investments 2,591 and non-current investments 12,071.
These sum to 15,209, not the management cash balance; classification and other instruments must be reconciled before treating the entire 16,475 as freely distributable. Lease liabilities total 5,757 (4,860 non-current plus 897 current), up from 4,592 in March.
Equity attributable to owners 31,275 versus 30,980. Goodwill 5,737 is unchanged; right-of-use assets rise to 5,360 from 4,294 and PPE to 3,040 from 2,033. The business is becoming more asset-intensive, even if it retains a technology platform's customer interface.
The accounting quality question is not whether the numbers are legitimate. It is whether the metrics chosen for valuation match the economic claim. Reported revenue is affected by principal-versus-agent accounting; adjusted EBITDA excludes share-based compensation and includes rent; statutory profit includes treasury income; operating cash includes working-capital timing; and lease cash is classified in financing.
Each is useful for a different purpose. A disciplined valuation must bridge them rather than choose the largest growth number.
The October 21 call makes clear that food delivery margins are expected to remain broadly flat over the next one to two years as incremental surplus is reinvested. This is a strategic choice, not a promise of immediate margin expansion. The next evidence should be cohort retention and frequency after promotional spending. If higher acquisition cost buys durable repeat orders, flat margins can create value. If the same customers require repeated incentives, the spending is maintenance rather than growth investment.
Key takeaways
The following year requires about 1,500 additions, again around 375 per quarter. The absolute pace is demanding but not a discontinuous jump from the revised near-term target. The harder issue is maintaining site quality, warehouse capacity and supplier availability as expansion moves beyond the best urban catchments.
Management's evidence for small cities is encouraging: similar ramp to 1,000 orders/day, 85–90% metro basket and comparable retention. These are management observations, not independently audited cohort tables.
We would request median and lower-quartile payback by opening cohort, including inventory, fit-out, lease deposit, warehouse allocation and closure costs. Averages can hide failed stores. The issuer also says cities representing 96% of NOV are contribution positive and cities representing about 70% are adjusted EBITDA positive; mature cities exceed 5% adjusted EBITDA margin.
This suggests broadening profitability but does not imply 96% of stores are profitable. The denominator is city NOV, not store count. A city with a large mature core can mask losses at recently opened stores.
The 60% three-year NOV CAGR guidance implies a 4.096x increase over three years. Q2 growth of about 79% is ahead of that rate, but management has not revised the long-term guidance. The expansion plan should be tested against actual orders, basket and density rather than assumed to guarantee the CAGR.
A 5,000-store network is 1.82x the current footprint; achieving 4.1x NOV over three years requires additional stores beyond that milestone, greater productivity, higher basket values or some combination. This is not a contradiction because the horizons differ; it is a useful decomposition of the growth hurdle.
District management reiterates the January 2026 break-even guidance of four to six quarters and FY30 goals. Three quarters into that window, the next one to three quarters are the relevant proof period. The appropriate test is adjusted EBITDA break-even without a collapse in NOV growth, followed by positive cash after event-related investment. Hyperpure's restaurant count growth of 37.1% is encouraging but should translate into repeat procurement and working-capital returns.
The new dark.ai initiative is explicitly inside Eternal and currently unmonetised. Management says costs are in Others and does not expect material capex unless demand requires infrastructure expansion.
We assign zero standalone value in the base case until there is evidence of paying customers, retention, gross margin after inference costs and a verified privacy/security architecture. The founder's product claims are not independent security certification.
The risk is not that experimentation is inherently bad; it is that a capital-intensive AI initiative could compete with the core commerce expansion for resources. A separate investment hurdle and transparent cost disclosure would protect the existing shareholder thesis.
Reference close ₹363.65 and 921 crore paid-up shares at ₹1 face value imply ₹334,922 crore approximate market capitalisation. Paid-up shares are a point-in-time basic count, not a fully diluted forecast. The reported Q2 diluted EPS is ₹0.13 versus basic ₹0.14; rounding prevents exact inference of diluted share count. H1 diluted EPS ₹0.23 is not annualised guidance. We use 950 crore illustrative FY30 diluted shares, explicitly an assumption allowing approximately 3.15% dilution from the current basic count. No observed peer multiple is claimed.
Key takeaways
Terminal equity values are ₹125,000/315,000/630,000 crore, or ₹131.58/331.58/663.16 per assumed diluted share. Discounting 3.38 years at 12% gives approximately ₹90/226/452 per share (discount factor about 1.466). The exact date convention is illustrative from November 2026 to March 2030. No dividends are assumed; treasury assets are not separately added because earnings include the return on capital and the scenario is an equity earnings model. This avoids double counting cash and treasury income.
The reverse hurdle at the current price is more instructive. At 950 crore future shares and 12% required return over 3.38 years, the required terminal equity value is approximately ₹506,000 crore. At 35x terminal P/E this requires roughly ₹14,460 crore FY30 attributable earnings.
At 25x it requires about ₹20,240 crore; at 45x about ₹11,240 crore. These earnings are many times H1 annualised attributable profit of ₹434 crore. The comparison is deliberately stark: today's earnings are suppressed by investment, so the market must be underwriting a very large future margin and scale improvement.
A premium terminal multiple alone cannot rescue a weak cash conversion outcome.
To connect the earnings hurdle to operations, consider an illustrative FY30 group revenue of ₹250,000 crore on a future consistent accounting basis. ₹14,460 crore attributable profit requires a 5.78% net margin; at ₹200,000 crore revenue it requires 7.23%.
These are sensitivity denominators, not forecasts. The current Q2 statutory net margin is 0.52%. First-party accounting makes revenue-based margin comparisons imperfect, so a full model should instead build food delivery NOV × contribution, Blinkit NOV × mature margin minus new-cohort losses, District and Hyperpure economics, central costs, depreciation, tax and share-based compensation.
The scenario is a valuation hurdle, not a claim that these operating inputs have been independently forecast.
Sensitivity: at base ₹9,000 crore earnings, each five turns of terminal P/E changes terminal equity by ₹45,000 crore, or ₹47.37 per future share before discounting. A 10% earnings miss reduces base terminal value by ₹31,500 crore.
Raising future diluted shares from 950 to 1,000 reduces per-share value by 5% at unchanged equity value. A higher required return also reduces present value materially because the payoff is distant. The bear scenario is not a floor, and the bull scenario is not a target.
The base present value below the reference price argues for patience unless the operating evidence supports the upper end of the earnings range.
The most important near-term catalyst is the post-festive working-capital unwind. H1 cash after capex and lease payments is negative ₹2,020 crore. The next report should show whether inventory normalises, whether trade payable growth remains sustainable and whether cash generation improves despite accelerated store openings. A decline in broad cash balance is not automatically negative if new cohorts earn high returns; persistent negative cash after investment without cohort disclosure is a different matter.
Key takeaways
Third, food delivery growth spending should be evaluated through repeat frequency and contribution after incentives. Management's flat-margin outlook makes a temporary margin dip unsurprising, but a prolonged decline with no retention improvement would challenge the reinvestment thesis.
Fourth, District break-even is a near-term accountability point. The original four-to-six-quarter window is approaching its end. Verify actual adjusted EBITDA, cash investment and exclusivity economics rather than assuming the FY30 target is achieved. Fifth, the AI initiative should remain ring-fenced until monetisation and infrastructure requirements are disclosed. The core commerce thesis does not require assigning speculative value to dark.ai.
Regulatory risks are real. The signed results disclose contested GST orders of ₹420 crore and ₹14 crore and show-cause notices of ₹119 crore on delivery charges, before interest and penalties. The combined disclosed principal exposure is ₹553 crore; this is not a booked liability or a prediction of loss.
Management contests the matters and the auditor draws attention without modifying the review conclusion. Changes to gig-worker social security and state rules could alter delivery costs; the current disclosed assessment is not a guarantee of future immateriality.
Model a higher cost per order and test whether fees or density can absorb it. Competition from other quick-commerce and food-delivery networks, supplier bargaining, inventory spoilage, lease commitments and technology reliability are additional risks.
Alternative thesis: a premium valuation may be justified if Eternal becomes the default high-frequency commerce infrastructure across Indian cities, with mature cohorts funding expansion and new categories increasing basket size. The strongest evidence is the observed small-city ramp, broadening city profitability and management's confidence in payback.
The strongest contrary evidence is the current cash deficit, rising lease liabilities and the large reverse earnings hurdle. We would change the stance on verified sustained positive cash after investment, strong mature-cohort returns and a credible path to the implied FY30 owner earnings, not merely a higher store target.
S1: Eternal shareholder letter dated October 21 2026, Q2 FY27, exchange filing https://nsearchives.nseindia.com/corporate/ZOMATO_21102026152512_SHLsigned.pdf . Pages 2, 8–15: definitions, store guidance, small-city evidence, District, Hyperpure and cash commentary; appended signed consolidated unaudited results, balance sheet and cash flows: statutory figures and auditor review. S2: October 21 earnings call, filed October 27, https://www.bseindia.com/xml-data/corpfiling/AttachLive/64f36e3b-87e9-426e-be3e-515dca4525d2.pdf . Pages 2–4: food delivery reinvestment, small-city expansion, first-party mix and AI initiative. Price reference: Sage canonical price history November 11–13 2026. Both source texts are accepted canonical documents with matching substantive audit hashes and full-window scores. The statutory income and cash-flow arithmetic above is independently recalculated from extracted signed results, but original-page visual parity, complete segment NOV/adjusted EBITDA tables, treasury classification, fully diluted share schedule and independent peer valuation remain open. The shareholder letter's image-only annexure tables are not reconstructed from guesses. The report is a source-verified longform, not a full institutional QA certificate. All scenarios and unit examples are illustrative analyst assumptions, not guidance or observed store economics.
Eternal Q2 FY27 shareholder letter and signed results, 21 Oct 2026
Eternal Q2 FY27 shareholder letter and signed results, 21 Oct 2026
Analyst illustrative calculation
Eternal Q2 FY27 shareholder letter and signed results, 21 Oct 2026
This is dated research, not a live market quote. Scenarios and historical performance are not guarantees.