07.4 · High-growth & loss-making companies¶
Why this matters: since 2021 India has listed food delivery, payments, insurance marketplaces, beauty e-commerce, quick commerce and a string of SaaS and fintech companies — most of them loss-making at IPO, some now profitable, some worth a fraction of their listing price. None can be valued on P/E, and an EV/Sales multiple without a margin argument is a guess. The method is to value the business the company will become, and to be explicit about the path, the dilution, and the probability of getting there.
Learning objectives — after this lesson you can:
- Analyse a loss-making company through unit economics, contribution margins and cohorts.
- Build a path-to-profitability model with explicit fixed-cost and contribution-per-unit drivers.
- Sanity-check EV/Sales multiples against the terminal margin they imply, and use the Rule of 40.
- Account for ESOP dilution and "adjusted EBITDA" definitions.
- Set a terminal state for a company that has not reached it, and recognise the patterns in India's 2021 IPO cohort.
Prerequisites: 05.1 Business models & unit economics, 06.3 DCF step by step · Time: ~90 min
1. Three questions¶
- Is the business losing money by choice or by nature? A company with positive contribution per unit that loses money because it is spending on growth (marketing, new cities, engineering) can stop and be profitable. A company whose contribution is negative — each order loses money before overheads — has no such switch. Find the contribution line first.
- What does the profitable end-state look like, and who already looks like it? A food-delivery platform at scale in a mature market; a SaaS company at 25% operating margins; a lender at 3% RoA. If nobody anywhere has reached the end-state you are assuming, your terminal margin is a hypothesis with no base rate.
- How much dilution and how many years between here and there? Losses are funded by equity; every round and every ESOP grant reduces your share of the end-state.
2. Unit economics and contribution margin¶
The core disclosure is contribution margin: revenue per unit minus all variable costs of serving the unit (delivery cost, payment fees, discounts, customer support, sometimes marketing), before fixed overheads. Indian new-age companies define it differently — some include marketing, some not; some report it as % of GMV/GOV, others as % of revenue — so rebuild it from the P&L before comparing:
Cohorts turn contribution into value: for each acquisition month, plot retained customers and their spend over time. A cohort that flattens at 40% retention with rising order frequency is a business; one that decays to 10% is a marketing expense with a logo (05.1 §4.3). Companies that stop disclosing cohorts after the DRHP have usually seen them worsen.
3. A path-to-profitability model¶
Dhruv Delivery (fictional quick-commerce platform), base year: GMV ₹12,000 Cr, take rate 18% (revenue ₹2,160 Cr), average order value ₹450 (26.7 Cr orders a year), contribution per order −₹5 (it loses money on every order before overheads), fixed costs ₹900 Cr (technology, corporate, dark-store leases treated as fixed), EBITDA −₹1,033 Cr.
The model has three drivers: order growth (density lowers delivery cost per order), contribution per order (AOV up, delivery cost per order down, advertising income), and fixed-cost growth (8% a year).
| Y1 | Y2 | Y3 | Y4 | Y5 | |
|---|---|---|---|---|---|
| Order growth | 40% | 32% | 25% | 20% | 15% |
| Orders (Cr) | 37.3 | 49.3 | 61.6 | 73.9 | 85.0 |
| Revenue (₹ Cr) | 3,024 | 3,992 | 4,990 | 5,988 | 6,886 |
| Contribution / order (₹) | −5 | +3 | +10 | +16 | +22 |
| Contribution (₹ Cr) | (187) | 148 | 616 | 1,183 | 1,870 |
| Fixed costs (₹ Cr) | 900 | 972 | 1,050 | 1,134 | 1,224 |
| EBITDA (₹ Cr) | (1,087) | (824) | (434) | 49 | 646 |
| EBITDA margin | −35.9% | −20.6% | −8.7% | +0.8% | +9.4% |
| Cumulative EBITDA burn | (1,087) | (1,911) | (2,345) | (2,295) | (1,649) |
| Rule of 40 (growth % + margin %) | 4 | 11 | 16 | 21 | 24 |
g = [0.40, 0.32, 0.25, 0.20, 0.15]; cpo = [-5, 3, 10, 16, 22]
rev, orders, fixed = 2160.0, 26.67, 900.0
for t, (gg, c) in enumerate(zip(g, cpo), 1):
rev *= 1 + gg; orders *= 1 + gg; fixed *= 1.08 if t > 1 else 1
ebitda = c * orders - fixed
print(t, round(rev), round(ebitda), f"{ebitda / rev:.1%}")
What the table says:
- Break-even is in Y4 and depends entirely on contribution per order rising from −₹5 to +₹16: that is the thesis. Every quarter, the number to check is contribution per order, not GMV growth.
- Cumulative burn is ~₹2,350 Cr before the turn. The company needs that much capital (net of what it has) — which means equity raises and dilution, or debt it cannot service. If it has ₹1,500 Cr of cash, there is a funding round in Y2–Y3 at whatever price the market then offers.
- The Rule of 40 (revenue growth % + EBITDA margin % ≥ 40 for a healthy software/platform company) is failed throughout: growth is high but margins are deeply negative. The rule is a heuristic from US SaaS; apply it as a screen, not a valuation.
3.1 The terminal state and what EV/Sales implies¶
Suppose Dhruv trades at 6x revenue (EV ₹12,960 Cr). Is that expensive? Convert it into what it assumes: continue the model to Y8 at ~10% growth with EBITDA margin reaching a 15% steady state (a plausible mature quick-commerce margin — nobody has demonstrated it in India yet; verify against the latest Eternal/Blinkit and Swiggy Instamart disclosures): Y8 revenue ≈ ₹9,165 Cr, EBITDA ≈ ₹1,375 Cr. Today's EV is 9.4x that Y8 EBITDA — but Y8 is eight years away; discounted at 14%, the PV factor is 0.35, so the EV is ~27x the present value of year-8 EBITDA, before the dilution needed to get there. That is the honest translation of "6x sales": a growth-stock multiple on profits eight years out, in a business model whose steady-state margin is unproven.
The general recipe:
If the implied multiple is far above what mature comparables trade at, the price needs either a higher terminal margin, faster growth, or a lower cost of capital — and you should say which you believe.
4. Dilution: ESOPs and rounds¶
Loss-making companies pay in shares. Two channels:
- ESOPs: Ind AS 102 expenses the grant-date fair value; Indian new-age companies have run ESOP charges of 2–10% of revenue and present "adjusted EBITDA excluding ESOP cost". Treat the expense as real (it is the cost of the people) and count the shares: 2% annual dilution for five years is a 9.4% haircut to your per-share value (\(1.02^5 = 1.104\)).
- Funding rounds: model the cumulative burn, the cash on hand, and the price at which the gap is filled. A ₹1,000 Cr raise at half today's price dilutes existing shareholders far more than the same raise at today's price — the path matters, and a bear market during the burn phase is the main way these investments fail even when the business succeeds.
Per-share value = (terminal equity value) ÷ (today's shares × cumulative dilution factor), discounted. Always.
5. Patterns from India's 2021–24 cohort¶
Facts to verify before citing; the patterns are the lesson:
| Pattern | Examples (verify specifics) | Lesson |
|---|---|---|
| Contribution turned positive, then EBITDA, then the multiple re-rated | Zomato (now Eternal): food delivery reached adjusted-EBITDA break-even in FY23–24; the stock re-rated strongly, then Blinkit's quick-commerce losses reopened the burn debate | The market pays for the turn in contribution; it then re-prices when a new burn segment is added |
| Regulatory jump risk | Paytm: RBI's Jan-2024 restrictions on Paytm Payments Bank removed a core business overnight | Regulated platforms carry a discrete, unhedgeable risk — scenario it explicitly (case I8) |
| Marketplace with take-rate pressure | Nykaa (beauty e-commerce), PB Fintech (insurance/loan marketplace) | Take rates compress as competition and regulation act; model take rate as a driver, not a constant |
| Duopoly capital war | Swiggy vs Zomato/Eternal in quick commerce (2024–26) | Both burn until one blinks; the capital cycle (05.4) applies to dark stores as it does to steel mills |
| Profitable at IPO, still de-rated | Several 2024–25 listings priced for growth that slowed | Profitability is necessary, not sufficient; the growth-multiple compression of 06.5 applies |
India notes
- SEBI's ICDR rules require new-age issuers to disclose Key Performance Indicators (GMV, contribution margin, cohorts, CAC) in the DRHP with auditor certification, and to explain the basis of the issue price against them (the 2022 "BIP" rules) — save the DRHP; later disclosures are thinner.
- Indian new-age companies typically report adjusted EBITDA excluding ESOP cost and sometimes excluding "one-time" items every quarter; rebuild reported EBITDA and free cash flow yourself.
- Ind AS 116 makes dark-store and warehouse leases part of D&A and interest — EBITDA flatters lease-heavy quick-commerce models; look at EBIT or pre-Ind AS 116 EBITDA and free cash flow after lease payments.
Trader's lens
A loss-making growth stock is a long-dated call with a funding-dependent path: the strike is the cumulative burn, the expiry is the runway, and the vol is enormous. Two things a vol trader will recognise: theta is real (every quarter of burn moves value to the next funding round's investors), and the position's delta is not to revenue but to contribution per unit — the variable that decides whether the option finishes in the money. Trade the contribution line; the GMV line is noise with a marketing budget.
Common mistakes
- Valuing on EV/Sales without translating it into an implied terminal margin and multiple.
- Accepting management's contribution-margin definition.
- Ignoring cumulative burn and the funding path (and the price at which it will be funded).
- Treating ESOP cost as non-cash and therefore ignorable — then also ignoring the share count.
- Assuming a terminal margin no company in the model has ever achieved.
- Extrapolating GMV growth when the drivers are discounting and marketing spend.
- Ignoring regulatory jump risk for platforms in payments, lending, insurance distribution or gaming.
Key terms¶
| Term | Meaning |
|---|---|
| Contribution margin | Revenue minus variable costs of serving a unit, before fixed overheads |
| Contribution per order / unit | Contribution ÷ units; the variable that decides break-even |
| GMV / GOV | Gross merchandise / order value flowing through a platform |
| Take rate | Platform revenue ÷ GMV |
| Cohort analysis | Retention and spend of customers grouped by acquisition period |
| CAC / LTV | Customer acquisition cost / lifetime value |
| Path to profitability | A driver-based forecast of when and how contribution overtakes fixed costs |
| Burn / runway | Cash consumed per period / periods of cash remaining |
| Rule of 40 | Growth % + margin % ≥ 40 as a health screen for software/platform businesses |
| Adjusted EBITDA | Company-defined EBITDA excluding ESOP and "one-time" costs; rebuild before use |
| ESOP dilution | Increase in share count from employee options; a real cost to per-share value |
| Terminal state | The mature economics (growth, margin, capital intensity) the valuation assumes the company reaches |
| Implied terminal multiple | Today's EV grown at the discount rate ÷ terminal-year EBITDA |
Check your understanding¶
- Dhruv's contribution per order stalls at +₹10 from Y3 onward. When does it break even, if ever?
Answer
Contribution at ₹10 × orders: Y4 = 739 vs fixed 1,134; Y5 = 850 vs 1,224 — the gap narrows only if orders keep growing faster than fixed costs (8%); at 15% order growth it is still −₹89 Cr in Y10, with cumulative burn above ₹4,200 Cr. The investment fails on time and dilution even if "it eventually works".- Rebuild the implied terminal multiple if Dhruv's terminal margin is 10% instead of 15%.
Answer
Y8 EBITDA = 9,165 × 10% = ₹917 Cr; EV 12,960 / 917 = 14.1x nominal; ÷ 0.351 PV factor = ~40x the present value of year-8 EBITDA — before dilution. A 5-pp change in an unproven terminal margin moves the implied multiple by 50%.- A company reports "adjusted EBITDA margin of +3%" with an ESOP charge of 6% of revenue. What is the reported EBITDA margin, and how should you treat the difference?
Answer
−3%. The 6 pp is compensation paid in shares: real cost, recurring while the company pays people that way. Use the −3% for profitability, and add the share issuance to the dilution schedule.- Why is contribution per order a better leading indicator than GMV growth for a quick-commerce company?
Answer
GMV can be bought with discounts and marketing (which sit in variable or fixed costs); contribution per order captures whether each incremental order adds or destroys money after those costs. Break-even is contribution × orders − fixed costs; only the first term's sign tells you if growth helps.- List three ways an Indian platform's take rate could fall.
Answer
Competition (a rival subsidises merchants/customers); regulation (caps on commissions, e.g. proposals on food-delivery or app-store fees; RBI rules on payment charges); mix shift toward lower-commission categories or larger merchants with negotiating power.Go deeper¶
- Aswath Damodaran, The Dark Side of Valuation — young and growth companies; and his periodic blog valuations of Zomato, Paytm and Swiggy at IPO (useful models of method, whatever one thinks of the numbers).
- Bill Gurley, "All Revenue Is Not Created Equal: The Keys to the 10X Revenue Club" (2011) — what a high EV/Sales multiple requires.
- Any 2021–24 Indian new-age DRHP, "Key Performance Indicators" and "Basis for Offer Price" sections — SEBI-mandated unit-economics disclosures.
- Case I8 Zomato & Paytm IPOs and case G5 Amazon — the divergent outcomes and the one that worked.
← Previous: 07.3 Cyclicals & commodities · Module index · Next: 07.5 Holdcos, conglomerates, PSUs & MNC subsidiaries →