Financial Statement Analysis & Valuation: Reading the Past to Price the Future

Two companies in the same industry both earn a 20% return on the owners' money. A new analyst calls them equals. A graduate asks three questions first. How much of that 20% comes from the business, and how much from borrowing? Is the profit backed by cash, or by accruals that may reverse? And does the growth that justifies the price earn more than it costs to fund?

Those questions are the course. Financial statement analysis takes the statements the previous subjects taught you to read and rebuilds them as economics: operations on one side, financing on the other, profit tested against cash. Valuation then asks what that economic machine is worth, and, in its most honest form, what a market price must assume to be reasonable.

The craft is not a formula that outputs a price. It is a way of locating the few assumptions on which a value rests and looking at those hard. A graduate rarely says "the value is X." They say "the value is about X, if the return on new investment stays above Y, and half of it comes from after year five."

This explains how analysts think. It is not investment, tax or legal advice, and gives no view on any real security. The companies, prices and peers below are invented.

Separate the business from how it is funded

The oldest tool in the course is the DuPont identity, developed at the DuPont company in the early twentieth century. It takes return on equity and breaks it into three pieces that are easy to compare across firms:

The same 20% can come from very different recipes. A supermarket makes a thin margin but turns its assets over four times a year. A luxury house makes a fat margin on assets that turn slowly. A bank has a tiny margin on assets and an enormous multiplier, which is why comparing bank ROE with grocer ROE tells you little. The identity does not say which is better. It says where each return comes from and what could change it.

Without JavaScript: a grocer with a 2% margin, turnover of 4.0 and multiplier of 2.5 earns 20% on equity; a luxury house with 25%, 0.5 and 1.6 also earns 20%; a stylised bank with 14%, 0.07 and 10.2 earns about 10%.

The second step is to strip out the financing. Analysts reorganise the statements into operating assets and liabilities and financing items (debt, cash, equity). Profit before interest, taxed as if the firm had no debt, is NOPAT (net operating profit after tax). The money tied up in running the business is invested capital. Their ratio, ROIC, measures the business whatever its funding. Debt then changes the owners' return without changing the business:

Without JavaScript: with no debt return on equity is 13.5%; with $400 of debt it is 18.5% and interest cover is 5.6 times; with $800 it is 43.5% and cover is 2.8 times. ROIC stays 13.5%.

This is the picture to hold: leverage is an amplifier, not an engine. It lifts ROE in good years and cuts it in bad ones. Leases matter here too: since IFRS 16 (see Financial Accounting & Reporting (36/45)) they are debt-like, so a careful analyst adds lease liabilities to invested capital and adjusts EBITDA before comparing firms across rulebooks.

Earnings quality: do the profits turn into cash?

Profit is an opinion built on estimates. Earnings quality asks how much of a profit figure is cash-backed and repeatable. A simple test is the gap between profit and operating cash flow, scaled by assets. This accrual ratio is the part of earnings that rests on estimates rather than money received. In 1996 Richard Sloan, in The Accounting Review, reported that firms whose profits came more from accruals and less from cash were followed by weaker profits and weaker stock returns, as if the market treated both kinds of earnings as equally lasting. That finding drew decades of work. Later studies found the pattern weaker in more recent years, and researchers still argue whether it reflects mispricing, risk or the way accruals are measured.

In 1999 Messod Beneish built the best-known screening formula for earnings manipulation, a weighted mix of eight ratios. Each compares this year with last: receivables relative to sales, gross margin, the share of assets that are not in plain operating categories, sales growth, depreciation, selling and admin costs, leverage, and total accruals relative to assets. A score above about −1.78 (papers use thresholds in this region) marks a company as resembling past manipulators.

Without JavaScript: a steady firm (all indices near 1, accruals zero) scores about −2.4; a firm whose receivables grew 40% faster than sales, with margin slipping, 30% growth and accruals at 5% of assets scores about −1.56, above the usual threshold.

Screens have false positives (a fast-growing firm looks suspicious) and false negatives (a careful fraud does not). Analysts use them to decide where to read closely. The usual list of things to read: receivables or inventory growing faster than sales; capitalised costs rising; a gap between profit and cash that widens year after year; "one-off" charges that recur; "adjusted" earnings that add back ordinary costs; and shifts in estimates that conveniently move profit across a target. Auditing & Assurance (36/47) is the profession on the other side of the same list.

Growth creates value only when returns beat the cost of capital

Growth sounds good. But growth has to be paid for: more stock, more equipment, more customers on credit. The cash a firm can pay out is its NOPAT minus what it reinvests. If a firm reinvests a share of NOPAT and earns ROIC on that new money, its growth rate is the reinvestment share times ROIC. Rearrange, and value depends on three numbers:

Read it as a lesson. When ROIC equals WACC, growth adds nothing: the extra cash flow is worth exactly the money put in, and value equals NOPAT ÷ WACC whatever g is. When ROIC is above WACC, growth is worth having. When ROIC is below WACC, every dollar invested in growth destroys value, and the faster the growth the worse it gets. Take NOPAT of 100 and WACC of 9%. With no growth the firm is worth 1,111. At 3% growth and ROIC 15% it is worth 1,333; at 3% growth and ROIC 6% it is worth only 833.

Without JavaScript: at NOPAT 100, WACC 9% and growth 3%, value is 1,333 if ROIC is 15%, 1,111 if ROIC is 9% (the same as no growth), and 833 if ROIC is 6%.

This is why economists and strategists care about competitive advantage. A firm can only keep ROIC above WACC while rivals cannot copy it: through brands, networks, scale, patents or cost position. Practitioners differ on how fast excess returns fade. Many valuations assume ROIC drifts to WACC over a decade or so; others argue that a few firms sustain returns for decades. That disagreement is a parameter, not a flaw in the method.

A DCF is a sensitivity table, not a number

Corporate Finance (34/51) showed how to discount a forecast: project free cash flows for some years, add a terminal value, discount at WACC. This piece adds the discipline that makes it honest. A DCF's output depends on two inputs that nobody knows: the discount rate and the growth after the forecast ends. Both live in the denominator of the terminal value, which is typically most of the total. A one-point change in either moves the answer by roughly a tenth to a fifth, and moving both at once moves it much more.

So practitioners never present a single value. They present a grid. Below, an invented company starts with 100 of free cash flow, grows it at a chosen rate for five years, then grows at a lower steady rate for ever. The grid varies WACC and terminal growth around your central choices.

Without JavaScript: with 6% growth for five years, a 9% WACC and 2.5% terminal growth, enterprise value is about 1,832, of which about 75% is the terminal value; moving WACC to 8%–10% and terminal growth to 1.5%–3.5% gives a range of about 1,440 to 2,568.

Where do the inputs come from? The WACC needs an equity risk premium, and estimates of it differ by a point or two among respected sources. Terminal growth is capped by sense: a firm cannot grow faster than the economy for ever, or it would eventually be the economy. Damodaran, whose free datasets many analysts use, makes the point that the discipline of a valuation lies in telling a story and attaching numbers to it, not in the arithmetic.

Multiples and the enterprise value bridge

The quick alternative to a DCF is to price a company by what the market pays for similar ones: a multiple such as price to earnings, or enterprise value to EBITDA. The rule that separates careful analysts is consistency: whatever is in the numerator must be paid for by the same people as the denominator. Price per share to earnings per share both belong to shareholders. But enterprise value (EV) is the value of the whole operating business, which belongs to everyone who funds it, so it pairs with an operating measure such as EBITDA, EBIT or sales.

Leases show why the care matters. Under IFRS 16, lease costs leave EBITDA, so the lease liability must be in EV. Under US GAAP, operating-lease cost stays above EBITDA, so it should not be. Mix the two and the multiple is wrong. Comparing an IFRS firm with a US GAAP firm needs an adjustment either way.

Without JavaScript: at $15 a share, EV is $1,900m with leases as debt and EBITDA of $245m, a multiple of 7.8; on the lease-as-rent basis EV is $1,780m, EBITDA is $220m and the multiple is 8.1. Applying the peers' median 8.5 times to $245m implies about $16.83 a share.

Multiples are quick and market-based, and they inherit the market's mood: if the whole sector is expensive, so is your answer. The choice of peers is the main judgement: same industry, similar growth, similar risk and similar accounting.

Reverse the model: what does the price assume?

The last habit is to run the DCF backwards. Instead of choosing assumptions and getting a value, take a market price, fix everything else you believe, and solve for the assumption that makes the model equal the price. With constant growth the algebra is short: EV = FCF × (1 + g) ÷ (WACC − g), so g = (EV × WACC − FCF) ÷ (EV + FCF). Alfred Rappaport and Michael Mauboussin popularised the approach as expectations investing: the price is a bet about what the future holds, so ask what future.

The point is not that the implied number is right or wrong. It is that you now know what you would have to believe. If the price implies growth forever above the economy's, doubt the price or the model. If it implies shrinkage, ask what the market sees.

Without JavaScript: at $15 a share the enterprise value is $1,780m; at a 9% WACC and $100m of free cash flow the price implies growth of about 3.2% a year for ever.

That 3.2% is plausible for a mature firm in a growing economy and heroic in a shrinking one. The honest conclusion is a question, and the question is what a student of valuation learns to ask first.

What a graduate just sees

• The same ROE can be built from margin, turnover or leverage; only the first two are the business.
• Leverage amplifies return and risk together; ROIC does not depend on funding.
• Profit that outruns operating cash flow, year after year, deserves a long look.
• Growth adds value only when new investment earns more than the cost of capital; at equality it adds nothing.
• A DCF is a grid of values for a grid of beliefs, and the terminal value usually carries most of it.
• Match numerator to denominator: enterprise value with operating measures, equity value with earnings.
• Leases, minorities and cash belong in the bridge, and rulebooks differ on how.
• A price is a bet; reverse the model to see the bet.

Where it shows up next

• Investments & Portfolio Theory (36/49): the cost of equity in every WACC, and the question of whether prices can be beaten.
• Financial Markets & Institutions (36/51): how credit analysts and rating agencies read the same statements for lenders.
• Taxation (36/48): the tax rate inside NOPAT and the effects of interest deductibility.
• Entrepreneurship (34/56): the same value drivers for a firm with no profit yet.
• In the world: an equity research note, a bank's credit committee, a takeover bid, a private-equity model and a founder's fundraising pitch.

If you want to go deeper

• 📘 Textbook. Tim Koller, Marc Goedhart and David Wessels, Valuation: Measuring and Managing the Value of Companies (Wiley, 8th edition, 2025), the standard practitioner text and the home of the value-driver formula; or Krishna Palepu and Paul Healy, Business Analysis and Valuation (Cengage), a classic course text. Check for the current edition.
• 📘 Gentler. Aswath Damodaran, The Little Book of Valuation (Wiley), a short, plain-language route; and Howard Schilit and Jeremy Perler, Financial Shenanigans (McGraw Hill), for earnings-quality red flags.
• 🌐 Free text. Aswath Damodaran's site at pages.stern.nyu.edu/~adamodar, with free datasets, spreadsheets and lecture material on valuation.
• 🌐 Free course. MIT OpenCourseWare 15.535 Business Analysis Using Financial Statements (Spring 2003), and 15.401 Finance Theory I (Fall 2008) for the valuation theory beneath it.