How to tell what a company is really worth

₹100 today is worth more than ₹100 promised next year, because today's ₹100 can be invested and grow.

Think of it likeWould you rather have ₹100 now or ₹100 in five years? Almost everyone says now. Valuation just puts a number on how much more.

Present value = Future amount ÷ (1 + r)ⁿ   (r = discount rate, n = years)

The return investors need to be paid for the risk of owning this company. Riskier companies get a higher rate, which makes their future cash worth less today.

Think of it likeA fixed deposit at a big bank pays less than a loan to a friend's new business. The more risk you take, the more return you demand.

Cost of equity = Risk-free rate + Beta × Equity risk premium
WACC = Cost of equity × (equity share) + Cost of debt × (1 − tax) × (debt share)

You can't forecast a company year by year forever, so after 5–10 years you assume it grows at a slow, steady rate for good and value all of that in one step.

Think of it likePlanning a career, you might map the next five years job by job, then just assume steady raises until retirement.

VALARTH ruleConfidence is marked down when more than 75% of the value comes after the forecast period.

Terminal value = Cash flow in the first year after the forecast ÷ (Discount rate − Forever growth)

A company is worth all the cash it will generate in the future, discounted back to today.

Think of it likeThink of buying a mango orchard. You don't pay for the trees, you pay for all the mangoes they will produce over the years, minus the cost of looking after them.

Value = Σ Cash flowₜ ÷ (1 + WACC)ᵗ + Terminal value ÷ (1 + WACC)ⁿ
Free cash flow = After-tax operating profit − (Capex − Depreciation) − Increase in working capital

For a company still losing money, forecast sales and ask what margin it will earn once it grows up. Then value that future company and allow for the chance it never gets there.

Think of it likeValuing a medical student: today they earn nothing, but you can estimate what a doctor earns later, and allow for the chance they drop out.

Value = (1 − Chance of failure) × DCF value + Chance of failure × Value if it fails

Like a DCF, but it counts only the cash left for shareholders after the company pays and borrows from lenders.

Think of it likeYour household budget after the EMI is paid: what's left is what the family can actually spend or save.

FCFE = Net profit − (Capex − Depreciation) − Increase in working capital + Net new borrowing

A share is worth all the dividends it will ever pay you, discounted back to today.

Think of it likeValuing a rental flat by the rent it will pay you every year.

Value = Next dividend ÷ (Required return − Growth)

A company is worth its book value plus whatever it earns above what shareholders need, for as long as it can keep doing so.

Think of it likeA shop you could sell for its stock and fittings (book value). If it earns more than a fair return on that, a buyer pays extra for the shop.

Value = Book value + Σ (ROE − Cost of equity) × Book value ÷ (1 + Cost of equity)ᵗ

A company only creates value when it earns more on its capital than that capital costs.

Think of it likeBorrowing at 12% to open a shop that makes 20% leaves you richer. Borrowing at 12% for a shop that makes 10% leaves you poorer, even though the shop 'makes a profit'.

Economic profit = (ROIC − WACC) × Invested capital

Value the business as if it had no debt, then add the tax savings debt brings and subtract the cost of possibly going bust.

Think of it likePricing a house on its own, then separately adding the tax break on the home loan and subtracting the risk of missing EMIs.

APV = Unlevered value + PV(tax shields) − PV(expected distress costs)

How many rupees investors pay for each rupee of yearly profit. Value a company by applying what similar companies trade at.

Think of it likeIf flats in your building sell for ₹12,000 per sq ft, a 1,000 sq ft flat is worth about ₹1.2 crore.

Value per share = EPS × Peer P/E

Adjusts P/E for growth. A high P/E can be reasonable if profits are growing fast.

Think of it likePaying more rent for a flat in an area where prices are rising quickly.

PEG = P/E ÷ Expected EPS growth (%)

Values the whole business, debt included, against its operating profit before depreciation. Useful for comparing companies with different amounts of debt.

Think of it likeComparing two shops by what they earn before the owner's loan payments, so the one with a big loan isn't unfairly penalised.

Equity value = EBITDA × Peer EV/EBITDA − Debt + Cash

Values a company by its sales, because it has no profits yet to measure.

Think of it likeValuing a new restaurant by its footfall and bills, before it has paid off its fit-out.

EV = Revenue × Peer EV/Sales

Compares the share price with the company's net worth on its books. The standard multiple for banks and lenders.

Think of it likeA bank is mostly a pile of loans. The question is whether that pile is worth more or less than what's written in the books.

Value per share = Book value per share × Peer P/B

Works out what P/E a company deserves from its own growth, returns and risk, instead of copying what peers trade at.

Think of it likeInstead of pricing your flat by what the neighbour got, working out the rent it can earn and what that rent is worth.

Fair P/E = Payout × (1 + g) ÷ (Cost of equity − g),   Payout = 1 − g ÷ ROE
Fair P/B = (ROE − g) ÷ (Cost of equity − g)

Turns the DCF around. Instead of forecasting and getting a value, start from today's price and work out what growth the market must be expecting.

Think of it likeSeeing a flat listed at ₹3 crore and asking what rent it must earn for that price to make sense. Then checking what rent flats there actually earn.

Solve for growth g such that DCF(g) = Market price

When one company runs very different businesses, value each one with its own best method and add them up.

Think of it likeValuing a joint family's assets: the farm, the shop and the rented flat are each priced differently, then added.

VALARTH ruleThe default holding-company discount is 20%, adjusted from 10% to 40% by how much cash the group passes up and its governance. Real market discounts vary widely.

Value = Σ Value of each business − Holding costs − Holding-company discount

Value what the company owns at today's market prices, minus everything it owes.

Think of it likeYour net worth: add up your flat, gold and savings at today's prices, subtract your loans.

NAV = Market value of assets − All liabilities

What would it cost to build this company's factories from scratch today? If the market values it for less, you're buying capacity cheaply.

Think of it likeIf a new car costs ₹10 lakh and a year-old one sells for ₹6 lakh, the used one is cheap for what it does.

Compare EV with Capacity × Cost of new capacity per unit

What shareholders would get if the company shut down today, sold everything and paid off everyone it owes.

Think of it likeClosing a shop: stock sells at a discount, equipment for scrap, and the supplier gets paid before you do.

Liquidation value = Σ Asset × Recovery rate − Liabilities

For cyclical companies, value them on their average profit through a full up-and-down cycle, not this year's.

Think of it likeA farmer's income swings with the monsoon. You'd judge the farm by an average year, not a drought or a bumper crop.

Normalised profit = Current capacity or sales × Average margin over the cycle

A drug still in trials might be worth a lot, or nothing. Value it if it succeeds, then multiply by the chance it gets approved.

Think of it likeA film still in production: you'd count its box-office potential, but only after allowing for the chance it's shelved.

rNPV = Σ P(reach stage t) × Cash flowₜ ÷ (1 + r)ᵗ − Σ P(reach stage t) × Trial costₜ ÷ (1 + r)ᵗ

A life insurer is worth its own money, plus profits still to come from policies it has already sold, plus the value of policies it will sell in future.

Think of it likeAn LIC agent's book of business: commissions still due on policies already sold, plus what they'll earn from new customers.

Value = Embedded value + VNB × Multiple

Some companies earn a return fixed by a regulator. Value them by that fixed return on the capital they're allowed to invest.

Think of it likeA toll road: the government sets the toll, so income is predictable for the length of the contract.

Value ≈ Regulated equity × Allowed return ÷ Cost of equity

Some assets aren't worth using today but could be later. That possibility has a value, like an option contract.

Think of it likeOwning a plot outside the city with no road yet. It earns nothing, but if the highway comes, it's worth a lot.

Black–Scholes with a cost of waiting y:
Value = S·e^(−yT)·N(d₁) − K·e^(−rT)·N(d₂)
d₁ = [ln(S/K) + (r − y + σ²/2)T] ÷ (σ√T),   d₂ = d₁ − σ√T
A common rule of thumb for y is 1 ÷ years left on the licence.

The most a cautious investor should pay, based on today's profits and book value.

Think of it likeA strict elder's rule: never pay more than a set multiple of what you can see today.

Graham number = √(22.5 × EPS × Book value per share)

What the company is worth if it never grows again. The gap to the full value shows how much you're paying for growth.

Think of it likeValuing a job by today's salary alone, with no future raises.

EPV = Normalised operating profit × (1 − tax) ÷ WACC

The cash an owner could take out each year without hurting the business.

Think of it likeWhat a shop owner can take home after keeping shelves stocked and fixing the fridge.

Owner earnings = Net profit + Depreciation − Maintenance capex − Increase in working capital

A fairly priced company's P/E is roughly equal to its growth rate plus dividend yield.

Think of it likeA quick rule a shopkeeper might use: the price multiple shouldn't run ahead of how fast the business grows.

Lynch ratio = (EPS growth % + Dividend yield %) ÷ P/E

Rank every company on two things, how cheap it is and how good a business it is, then favour those high on both.

Think of it likePicking a restaurant that is both highly rated and reasonably priced.

Score = Rank(EBIT ÷ EV) + Rank(EBIT ÷ (Working capital + Net fixed assets))

Nine yes-or-no tests of profitability, debt and efficiency. More yeses means a financially healthier company.

Think of it likeA nine-point health check-up. You don't need every test normal, but 2 out of 9 is a worry.

F-score = count of 9 binary signals

Combines four balance-sheet ratios into one score that predicts the risk of going bust. VALARTH uses Altman's later version built for non-manufacturers and emerging markets.

Think of it likeA credit score for companies.

Z″ = 6.56·(Working capital ÷ Total assets) + 3.26·(Retained earnings ÷ Total assets) + 6.72·(EBIT ÷ Total assets) + 1.05·(Book equity ÷ Total liabilities)

Eight ratios that tend to move in a suspicious way when a company is dressing up its profits.

Think of it likeA shopkeeper whose sales are 'up' but whose unpaid bills are piling up faster. Something doesn't add up.

M = −4.84 + 0.92·DSRI + 0.528·GMI + 0.404·AQI + 0.892·SGI + 0.115·DEPI − 0.172·SGAI + 4.679·TATA − 0.327·LVGI

Instead of one forecast, build three believable stories, each with its own growth, margins and value, and give each a probability.

Think of it likePlanning a trip for rain, normal weather and sunshine, rather than assuming one forecast is right.

VALARTH ruleDefault chances are 25% bear, 50% base, 25% bull, adjusted by company type and checked against past results.

Range = Bear value to Bull value

Shows how the value changes when you nudge the two assumptions that matter most.

Think of it likeChecking how your EMI changes if the interest rate moves half a percent.

Value(WACC, g) across a grid

Run the valuation thousands of times with inputs drawn at random from realistic ranges, and look at the spread of results.

Think of it likeInstead of predicting one cricket score, simulating the match 10,000 times and seeing how often your team wins.

Distribution of Value over 10,000 random draws

No single method is trusted alone. VALARTH weights the ones that suit the company, and says how confident it is based on how much they agree.

Think of it likeAsking three doctors. If all agree, you trust the diagnosis. If they disagree, you know it's a hard case.

VALARTH ruleThe weights for each company type, the agreement bands (under 15% spread is good, over 35% is poor) and the five parts of the confidence score are VALARTH's own choices. They are published and don't change stock by stock.

Value = Σ Weightᵢ × Methodᵢ   ·   Confidence 0–100

Guess what the startup could sell for in a few years, then discount at the high return investors need for the risk.

Think of it likeBuying a sapling: you guess the price of the grown tree, then pay much less today because many saplings die.

Post-money = Exit value ÷ (1 + Target return)ʸᵉᵃʳˢ,   Pre-money = Post-money − Investment,   Stake needed = Investment ÷ Post-money

Value the startup in three outcomes, success, muddling along, and failure, and weight each by its chance.

Think of it likeBetting on three possible results of a match with different odds.

Value = Σ Probability × Scenario value, discounted

Startup investors usually get their money back first when the company is sold. So a headline valuation overstates what ordinary shares are worth.

Think of it likeIn a family property sale, the relative who lent money gets repaid before the rest is split.

Payout waterfall by seniority and preference terms