How to value a loss-making company

The Dark Side problem

A young company, pre-profit and maybe pre-revenue-model, gives the standard machine nothing to hold. Its earnings are negative, so there's nothing to extrapolate. It has no history to average and few or no profitable comparables to put a multiple on. And there's a real probability it won't exist in ten years, which breaks the survival assumption. Damodaran's Dark Side of Valuation answer is to keep valuing and reverse the direction of estimation. Instead of growing the present forward, you describe the mature destination and discount it back, then haircut for the chance the company never gets there.

The six-step architecture

1. Size the TAM without flattering it. Start with the total addressable market in the year of maturity, say year 10. Today's market is the wrong number, and so is the promoter's conference-slide TAM. Build it two ways and take the lower: top-down from category spend, bottom-up from users × frequency × ticket. Then narrow it from TAM to serviceable market to the segment this firm's model actually monetizes.
2. Earn the market share instead of assuming it. What share can this competitor plausibly hold at maturity against the competition, given its right to win? Base rates help. In most two-player Indian internet categories the leader has settled near 45 to 60% share and #2 near 30 to 40%. "We only need 1% of a huge market" is the oldest lie in fundraising. Share is built from capacity: cities, riders, engineers. You can't conjure it out of the TAM.
3. Draw the margin path. What operating margin does the mature business earn? Anchor to the best structural comparable that already exists. A mature food-delivery platform's economics look like a marketplace's, at 4 to 5% EBIT as a % of gross order value. A mature SaaS firm earns 25 to 30% of revenue. Draw the path from today's losses to that endpoint. It matters because the interim losses are negative cash flows you have to finance and discount.
4. Pay for growth through sales-to-capital: reinvestment = Δrevenue / sales-to-capital ratio. Young platforms look capital-light, but they burn capital as operating losses and intangibles. Use the ratio to keep growth paid for, and check that the implied ROIC at maturity is believable. If your model says 60% ROIC forever, competition would like a word.
5. Let the discount rate age with the company. Young firms carry top-decile risk, so start the cost of capital high: 13 to 15% for a listed Indian cash-burner, 10 to 12% in the US (as of mid-2026 anchors). Glide it down to a mature-company rate by year 10 as margins and survival firm up.
6. Apply the survival haircut, the step everyone skips.

Value = P(survival) × going-concern DCF + (1 − P(survival)) × distress value

How do you estimate the probability of survival?

The DCF you just built is the value if the story gets to play out. Some of the time it doesn't. Funding markets close, the model never reaches contribution profit, a price war outlasts the balance sheet. You estimate P(survival) from two things. The first is base rates. In US BLS/Census data roughly half of new firms are gone within 5 to 7 years, though a listed, funded company stands far above that average. The second is the specific balance sheet. Work out runway = cash ÷ annual burn. Check access to capital. Then ask whether the business is contribution-positive, able to slow spending to survive, or structurally negative, dying without outside feeding. Distress value is what a failure scenario actually returns to equity: remaining cash plus the fire-sale value of assets and brand. That's often 10 to 25% of book, and sometimes zero after preferences. For a funded, listed, contribution-positive platform, a P(survival) of 85 to 95% is defensible. For a pre-revenue cash furnace, think 40 to 60%.

Mini-example: a Swiggy-shaped platform (illustrative, modeled on ~FY25 scale: pull the actual filings)

Food delivery plus quick commerce. At FY25 scale the whole Indian category has gross order value (GOV) of ≈ ₹1.6 lakh crore, growing fast. The company has revenue of ≈ ₹15,000 crore and is still EBITDA-negative. It holds cash of ≈ ₹8,000 crore from its IPO, against a market cap of ≈ ₹80,000 crore (mid-2026).

Now build a conservative base case, one step at a time. Category GOV compounds ~14% to ≈ ₹6 lakh crore by year 10. The company holds 35% share, so its GOV is ₹2.1 lakh crore. A 22% take-rate turns that into revenue of ≈ ₹46,000 crore. Mature EBIT at ≈ 5% of GOV is ≈ ₹10,500 crore, which leaves NOPAT of ≈ ₹7,900 crore. The continuing value at year 10 (g 5%, ROIC 25%, WACC 11%) is ≈ ₹1.09 lakh crore. Discount it along a rate glide from 14% to 11% and it's worth ≈ ₹34,000 crore today. Add the interim burn-then-build cash flows (≈ −₹1,500 crore PV) and the cash. Then haircut at P(survival) = 90%, with ₹8,000 crore of distress value. You land on an expected value ≈ ₹38,000 to 42,000 crore, roughly half the market price.

Don't read that as "short it." Read it as a sign that the price is buying a bull story, one that needs roughly 45%+ share and 6%+ EBIT/GOV and near-certain survival. Your job, with the narrative-and-numbers loop, is to write that implied story down, test it for plausibility, and name what would falsify it. Hold the opposite caution too. In 2022–23, the same arithmetic run pessimistically on Zomato near its lows would have missed the fastest profitability inflection in Indian internet history. The method's output is a map of what must be true at each price, and it should come without a sneer.

The young-company checklist (run before trusting any Dark Side model: yours or anyone's)

1. Is the smallest unit (order, ride, account) contribution-positive after all variable costs including discounts and delivery? If not, scale multiplies losses.
2. Is the runway, cash ÷ current burn, at least as long as the time to contribution breakeven? Who funds the gap, and at what dilution?
3. Is the claimed TAM the market the firm actually monetizes (its take-rate slice), or the whole category's GMV?
4. Is share backed by capacity and retention (cohorts, frequency), or only by marketing spend that stops working when it stops?
5. Does the margin endpoint have a living comparable, or does it require economics no one in the category has ever earned?
6. Who else is funded to fight for the same share, and what does the category look like if everyone executes their deck?
7. What's the dilution schedule? ESOPs and pending raises matter, because value per share needs the future share count. Today's won't do.
8. Governance: founder control, dual-class, related-party leakage (your forensic lens), audit quality.
9. Regulatory kill-switches: take-rate caps, labor reclassification, data rules. Young models often monetize a regulatory gap.
10. What did insiders do at the last three prices: raise, sell, pledge?

A "no" on items 1 and 2 is a survival-probability event. Don't try to absorb it with an input adjustment.