Swiggy — four quarters of chasing breakeven

The State of Play

Swiggy closed FY26 with ₹23,053 crore in revenue, up roughly 51% on the year, and a net loss of ₹4,154 crore — its largest since FY20. Those two numbers are the whole company in miniature: a food-delivery business that prints steady, profitable growth, bolted to a quick-commerce business (Instamart) that has spent the year burning money at a declining rate while management repeatedly promised the bleeding would stop “next quarter, give or take.” As of the May 2026 call, the promise stands at: Instamart reaches contribution-margin breakeven in the April–June 2026 quarter, on a full-quarter basis. April, management says, is already behind them and they are “pretty confident.”

The Company

Founded in 2014, Swiggy runs a hyperlocal commerce platform through one app. The FY25 annual report — its first as a listed company, after the November 2024 IPO — lists five reportable segments: Food Delivery (the original business: restaurant orders, monetised through restaurant fees, delivery charges, and advertising), Quick Commerce (Instamart — groceries and increasingly everything else from dark stores), Out-of-home consumption (DineOut restaurant dining, SteppinOut events), Supply Chain and Distribution (warehousing services for brands and retailers), and Platform Innovations (an incubator that birthed Swiggy Sports and the snacking app Snacc during FY25 — more on Snacc’s fate later).

Sriharsha Majety, a BITS Pilani and IIM Calcutta alum who co-founded the company, is MD and Group CEO. Anand Kripalu — the EPL Limited global CEO — chairs the board as an independent director. There is no promoter line in the shareholding table at all; Swiggy is a genuinely widely-held company, and the register has been churning fast. Public shareholding fell from 86% in December 2024 to 54% by March 2026, while institutions hoovered up the float: DIIs more than tripled from 7.75% to 25.45%, FIIs went from 6.19% to 14.59%. Combined institutional ownership went from roughly 14% to 40% in six quarters. This matters beyond sentiment — domestic ownership crossing 50% would let Swiggy convert to an “IOCC” structure and run an inventory-led model for Instamart, currently barred by FDI rules. By the February 2026 call it stood at about 47%.

The balance sheet was rebuilt twice over the period: reserves swung from deeply negative (−₹7,785 crore in March 2024) to +₹18,053 crore by March 2026, courtesy of the IPO and a QIP, while borrowings dropped from ₹16,437 crore to ₹2,551 crore. The cash funds an operation that still consumes it — FY26 operating cash flow was −₹2,898 crore, free cash flow −₹3,809 crore.

The Story So Far

The year’s plot is a single question asked four times: when does Instamart stop losing money? The answer moved. Watching how it moved is the most instructive part of the whole exercise.

Q1 FY26 (call: July 31, 2025) — the promise is a window

The quarter Swiggy reported in July 2025 was the hangover after a binge. The March 2025 quarter had seen 316 dark stores added — roughly 150 in March alone — and Q1 absorbed their full costs while they ramped. Instamart’s GOV still grew 108% year-on-year, average order value jumped 26%, and contribution margin improved 100 basis points despite the drag. But order growth slowed to about 4% sequentially — an eight-quarter low, as Ambit’s Ashwin Mehta pointed out — because Swiggy was deliberately consolidating baskets through its Maxxsaver programme and dropping low-value orders.

The central commitment, from CFO Rahul Bothra:

“Our guidance remains that we expect to get to contribution margin neutrality between December and June quarter of calendar year 2026.” — Rahul Bothra, Q1 FY26 call

Note the width of that window: six months of wiggle room. He also promised the next quarter’s improvement would beat the 100 basis points just delivered — “all of this improvement is not back ended.”

On money: cash stood at about ₹5,500 crore against a free-cash-flow burn of roughly ₹2,800 crore over the prior two quarters. Asked directly whether Swiggy needed to raise equity, Bothra was unambiguous: “we do not need to raise the equity.” File that one away.

Two side plots opened. Swiggy announced it was separating from Rapido — the bike-taxi firm it had invested in had entered food delivery, creating a conflict. And the company flagged that its 4.3 million square feet of dark-store space could theoretically support double the business without new stores, so capex would turn “measured.”

The quarterly P&L recorded the worst loss of the period: −₹1,197 crore on revenue of ₹4,961 crore.

Q2 FY26 (call: October 30, 2025) — delivery, and a raise that wasn’t needed

The promise made in July was kept, with room to spare. Instamart’s contribution margin improved 200 basis points — double the prior quarter’s gain — moving from −4.6% to −2.6%. GOV grew over 100% year-on-year for the third straight quarter. Overheads grew just 5% sequentially against roughly 25% GOV growth: the operating leverage management kept invoking was visibly showing up. Only about 40 stores were added; the network now stood near 1,100 dark stores, up from roughly 600 a year before.

The breakeven window narrowed — to its far end. The Dec ‘25–Jun ‘26 range from July quietly became a single date:

“We’ve also guided and reiterated our guidance of being able to demonstrate contribution margin profitability by June 2026 quarter.” — Rahul Bothra, Q2 FY26 call

Pushed by Citi and Kotak on whether breakeven could arrive earlier — December or March — management declined to pull it forward. Bothra went further: by June the entire store network would be contribution-positive. At the time only about 25% of stores were, with the negative 75% sitting at −5.2% margin; his defence was that the bulk of the network had been added between December and March and the standard 6–12 month maturation hadn’t elapsed.

Then came the wrinkle. Three months after “we do not need to raise the equity,” a QIP was going to the board. The framing: growth capital, strategic reserves, innovation capital — emphatically not because cash was short, and “we don’t expect the need to raise any further capital if and when we were to raise this additional QIP.” The Rapido stake sale proceeds were due the following quarter. Long-term targets got their fullest articulation here: Instamart at ~7% contribution margin and ~4% EBITDA margin in steady state, quick-commerce ad revenue reaching 6–7% of GOV, food delivery at a 5% medium-term EBITDA margin.

The quarter’s net loss narrowed to −₹1,092 crore on revenue of ₹5,561 crore.

Q3 FY26 (call: January 29, 2026) — the peak, and the sharpest question of the year

This is where the narrative got complicated. The festive quarter brought strong top-line optics — food delivery GOV up 20.5%, monthly transacting users up 22% — but Instamart’s reported contribution margin didn’t improve. Management’s explanation: the underlying, structural improvement was about 100 basis points, but roughly 90 of those were reinvested into a “No Fee” campaign (free delivery above roughly ₹299) that cost an estimated ₹70–90 crore and, by management’s own admission, saw “limited adoption and retention.” Contribution loss per order worsened to ₹19. Absolute numbers stayed ugly: contribution still around −₹200 crore for the quarter, adjusted EBITDA loss in the ₹800–900 crore range.

The June 2026 promise survived, restated by Instamart CEO Amitesh Jha as “contribution margin zero in the quarter of AMJ’26,” with 250 basis points of improvement to come over two quarters. And Bothra added a new claim with a clear expiry date:

“We are calling out that this would be the peak of the investments. And from here on, gradually, you will see that reduction happening.” — Rahul Bothra, Q3 FY26 call

The sharpest exchange of the entire year came from Invesco’s Manish Poddar, who pointed out that with no public market-share or customer-quality metric, investors had no way to verify management’s distinction between “good growth” it was keeping and “bad growth” it was nobly declining — and that “in the last 6 quarters since the listing, there have been a couple of times where you have earlier said that you’ll breakeven… they were goalposts which got delayed.” Bothra’s answer leaned on the largest city already being contribution-positive, a quarter of the network in the black, and an invitation: “Do hold us accountable to that.”

The competitive backdrop, in management’s telling, was “irrational” — six-odd players discounting hard, a large traditional retailer adding 600+ dark stores. Swiggy’s stance hardened into a slogan: “we are not going to throw good money at bad growth.” Orders per day were dismissed as a “vanity metric.” Cash, post-QIP, stood near $2 billion.

Net loss: −₹1,065 crore on revenue of ₹6,148 crore. Marginally better, still heavy.

Q4 FY26 (call: May 8, 2026) — closer, narrower, and one promise quietly retired

The fourth quarter showed the “peak investment” claim roughly holding. Instamart’s contribution margin averaged −180 basis points for the quarter but exited March at −110 — a 5.5-percentage-point improvement over the year. The No Fee experiment was killed in the third week of January, accounting for about half the quarter’s improvement in the NOV-to-GOV ratio; the mix of low-value orders had halved over the year. The top city reached roughly +3% contribution margin and EBITDA breakeven at the city level. No new dark stores were added at all — and net customer additions fell to 0.5 million for the quarter, from a peak of 3 million, which management characterised as deliberate churn of low-value, discount-seeking users. Morgan Stanley’s Gaurav Rateria made sure the contrast was on the record.

The breakeven promise was reiterated one more time, now at its most precise:

“…along with the reiteration of our guidance of achieving breakeven in the current quarter… This is for the entire quarter and not just on an exit basis… Now that April is behind us, we are pretty confident of being able to achieve that.” — Rahul Bothra, Q4 FY26 call

But something else slipped out of the frame. ICICI’s Abhisek Banerjee asked whether the earlier indication — EBITDA breakeven four quarters after contribution breakeven — still held. It was not reaffirmed:

“…we have carefully decided not to speculate on when EBITDA profitability will come through… EBITDA profitability will be a choice that will be made at a later point in time.” — Rahul Bothra, Q4 FY26 call

In its place came a much grander, much vaguer ambition: ₹1 lakh crore of quick-commerce net order value in the medium term, at a 35–50% CAGR over 3.5–5 years. HSBC’s Prateek Maheshwari did the arithmetic on the call — at current frequency that implies 45–50 million users, against 0.5 million quarterly adds — and management conceded the churn headwind would last another two quarters before customer numbers moved “healthily” again. Strategy talk shifted to differentiation: a clean-label private brand called Noice (high-protein eggs, low-preservative bread), a cookware line called Triply. Snacc, the 10-minute snacking app launched with some fanfare in FY25, was shut down during the quarter; a new low-frequency food app called Toing is in “green shoots.” Quarterly overheads in quick commerce sat around ₹700 crore — mostly marketing — and Macquarie pegged annualised free cash flow at roughly −$400 million, undisputed.

The quarter’s net loss narrowed to −₹800 crore on revenue of ₹6,383 crore — the best print of the year, though flattered by an unusual ₹266 crore of other income against the ₹59–121 crore of recent quarters.

The ledger: said vs. delivered

Across the four calls, the scorecard reads roughly so. Kept: the Q2 contribution improvement would beat Q1’s (promised +100bps-plus, delivered +200); store additions would slow sharply (316 in a quarter became 40, then zero); operating margin improved every single quarter, from −19% of sales in Q1 to −11% in Q4; net loss narrowed three quarters running. Moved: “breakeven between December and June” became “June,” which became “the AMJ quarter in full” — always inside the original window, but always at its far edge, exactly the pattern Poddar called out. Reversed: “we do not need to raise the equity” (July) preceded a QIP (October) by one quarter. Withdrawn: the EBITDA-breakeven-four-quarters-later framing, replaced by “a choice that will be made at a later point in time.” Meanwhile the absolute annual loss still widened — ₹3,117 crore in FY25 to ₹4,154 crore in FY26 — because depreciation (₹612 → ₹1,217 crore) and interest (₹101 → ₹200 crore) from all that dark-store and warehouse building landed below the operating line, swamping the percentage-margin gains above it.

Where Things Stand

The business enters FY27 with two very different engines. Food delivery is the dependable one: GOV growth of 18.8%, then ~20%, then 20.5% across the year, contribution margin around 7.3%, advertising above 4% of GMV and already exceeding the segment’s entire EBITDA, guidance held at 18–20% growth with a 5% steady-state EBITDA margin “in line of sight.” Nobody on any of the four calls spent much time worrying about it.

Instamart is the bet. The trajectory above the contribution line is genuinely, measurably better — from −5.5% at peak burn to −1.1% at the March exit — and the company has stopped paying for growth it considers low-quality, at the visible cost of decelerating user adds while competitors discount. Everything now rests on the AMJ 2026 quarter delivering the full-quarter breakeven that has been promised, in progressively firmer language, on four consecutive calls. Below the contribution line, ₹700 crore a quarter of overheads (mostly marketing, hostage to a competitive environment management itself calls irrational) stands between contribution-zero and actual profitability — and the company has explicitly declined to say when that gap closes. Capex should fall meaningfully, with warehousing build-out “now behind us.” The cash pile ($2 billion as of the January call) buys plenty of time; the −$400 million annualised FCF run-rate spends it. The next shareholders’ letter answers the only question that matters this year.

The Four Checks

1. Quality and moat. Two businesses, two answers. Food delivery is one half of an effective duopoly — network effects between restaurants, riders, and users that took a decade and billions to build, with advertising revenue above 4% of GMV already exceeding the segment’s entire EBITDA. That half has a real moat, though Swiggy is the smaller player in it. Instamart, where most of the capital is going, has none yet: six-odd competitors discounting “irrationally” in management’s own words, a frequency-and-habit game where Swiggy is deliberately shedding users (net adds down from 3 million a quarter to 0.5 million) while rivals spend. The differentiation bets — Noice eggs, Triply cookware — are early and unquantified by management’s own admission. Call it half a moat: durable in the legacy engine, entirely unproven in the growth engine.

2. Returns on incremental capital and runway. On the reported numbers, reinvestment is still destroying value: ROCE of −24.1% and ROE of −29.0% per the snapshot, FY26 operating cash flow of −₹2,898 crore, free cash flow of −₹3,809 crore. The trend is the only positive — operating margin has improved from −110% of sales in FY20 to −14% in FY26, and quarterly from −19% to −11% across the year, while Instamart’s contribution margin went from −5.5% at peak burn to −1.1% at the March exit. The runway is genuinely long (management’s ₹1 lakh crore quick-commerce NOV target implies 35–50% growth for years), and food delivery now grows 18–20% on little incremental capital. But a long runway at returns that are currently negative — and a steady-state Instamart EBITDA ambition of only ~4% — is a thin engine. Everything depends on margins that do not yet exist.

3. Capital allocation for the stage. Mixed, with the mix improving. For the stage — a land-grab with a funded balance sheet — heavy reinvestment is the right design, and the operational discipline has been real: dark-store additions went 316 → 40 → zero across the year, the No Fee experiment was killed within a quarter of underperforming, Snacc was shut down, the Rapido conflict was exited, capex is guided to fall with warehousing “now behind us.” Against that: “we do not need to raise the equity” in July preceded a QIP in October — funding the burn through dilution one quarter after denying the need (equity capital went from ₹3 crore to ₹229 crore over two years, with ₹9,397 crore of financing inflows in FY26 alone) — and the EBITDA-breakeven timeline was quietly withdrawn just as the contribution target came due. No dividends, no buybacks — correctly, for a business burning roughly $400 million a year. Rational in deed, slippery in word.

4. Price. As of the June 2026 snapshot, the stock trades at ₹248 — near its 52-week low of ₹238, down from a ₹474 high — for a market cap of ₹68,500 crore, with no P/E (there are no earnings), 3.82 times book, and roughly 3 times FY26 revenue of ₹23,053 crore that grew 51%. The halving of the price has taken the perfection out of it: the market is no longer paying up for the breakeven story, and the food-delivery duopoly half alone carries serious value against that figure. But a company with a −₹4,154 crore annual loss, negative free cash flow, and a withdrawn EBITDA timeline cannot be called cheap on its economics — the price is a wager that the AMJ 2026 promise lands and that the quick-commerce war rationalises. Less demanding than it was; still speculative on what it must become rather than fair on what it is.

Sources

• Concall transcripts (4): Q1 FY26 (call July 31, 2025), Q2 FY26 (October 30, 2025), Q3 FY26 (January 29, 2026), Q4 FY26 + full-year (May 8, 2026) — all BSE filings, converted to markdown. The October 2025 entry on screener had a presentation only, no transcript.
• Annual report (1): FY25 (Swiggy’s first as a listed company), high-signal sections. Note: the trimmed extract was light on hard financials; financial data above leans on the screener snapshot and calls.
• Screener.in snapshot: consolidated quarterly and annual tables, ratios, shareholding — fetched 2026-06-05 (logged-out session; document list may omit anything very recent).
• Research files: vault/Sources/Earnings/Swiggy Ltd/ — raw transcripts, AR sections, snapshot, and per-document digests (not published).