What you will be able to do
Define a company's real market and say what its growth, risks and bargaining positions imply for its accounts.
Apply the four-step plan (wealth generation, investment, financing, returns) in the right order and state a verdict.
Choose between trend, peer and standard comparisons, and name the blind spot of each.
Turn statements into common-size form and read what they reveal.
Spot the usual red flags of flattering accounts, including cash that does not follow profit.
Explain what a score or a rating compresses, and why neither replaces understanding.
1The business explains the numbers
Before a single ratio, the analyst forms a view of the business. That view lets her predict the accounts, and a prediction that fails is the quickest way to learn something.
Start with the market the customer sees
A market is the set of products a customer treats as substitutes
Statistical classifications group companies by main activity. A buyer decides differently. Someone choosing how to commute compares a 3,000 euro electric bike with a second-hand one, a season ticket, a scooter. The market is whatever the customer treats as a substitute. The analyst defines it that way, then divides it into segments such as city, trekking and cargo bikes.
Growth must be split into volume, price and life-cycle stage
Sales rise because more units are sold, because each unit sells for more, or both. The two effects multiply, they do not add. Volume growth in a young market is relatively easy to buy. Price and mix growth suggest strength, or inflation.
The life-cycle stage tells the analyst how cash behaves. A growth phase eats cash: factories, stock and customer credit expand before profit arrives. A mature business releases cash.
Growth in a sector is not the same as profit in a company
Market risk adds four questions. Are buyers first-time (fashion and subsidy driven) or replacement (steadier)? How easily can a competitor enter? Could a cheaper substitute appear? Could regulation or subsidy change demand overnight?
Follow the product from supplier to customer
Bargaining power sits where the alternatives are few
Every product passes through links: suppliers, the maker, distributors, the user. Each link bargains with its neighbours, and the one with fewer alternatives keeps more of the margin. The analyst asks where a squeeze would show in the accounts.
| Link around Harmony | Who has the power | Where it would show |
|---|---|---|
| Battery cell and component suppliers | Few suppliers, so they have it | Purchase prices, gross margin swings, safety stock |
| Dealers (70% of sales) | Large dealer groups can threaten to switch brands | Payment terms, receivables, discounts |
| Online customers (30% of sales) | Low cost of switching, so they have it | Marketing spend, returns, promotions |
How a firm produces and sells decides how much cash it ties up
Harmony builds to a forecast ahead of a spring peak, so it must hold stock. Outsourcing assembly would lighten the balance sheet but hand margin to the contractor. Selling online replaces dealer credit with payment at order, and adds marketing cost and returns.
People and sustainability belong in the diagnosis
Who owns the firm and who runs it shapes the risks it takes
Founders hold 40% of Harmony, Northlight Ventures 35%, business angels and family 20%, employees 5%. A venture fund has a fixed life and wants to sell within a few years. Founders may build for decades. The horizons pull investment, dividends and listing timing apart.
Key people matter too: if Léa Moreau or Samir Haddad left, would Harmony keep its engineering edge?
Sustainability is now a cost, a risk and a condition of access to money
For Harmony the question is the battery: where cells are sourced, what happens to them at end of life, how much carbon the supply chain carries. These facts affect cost, regulation and the terms banks offer. Reporting rules (CSRD in the European Union, ISSB standards internationally) make such data public and comparable. CF45 covers the topic.
1.1 Two weak links
Harmony sells 70% of its bikes through dealers, and one dealer group buys 30.0 of the 120.0 sales. A single Asian supplier provides most of the battery cells. Which of Harmony's two positions is weaker? Name one line of the accounts that would show the weakness in each case.
The supplier link is weaker. Harmony can find other dealers more easily than another source of the right cells. It would show in purchase prices, the gross margin and the safety stock carried in inventories.
The dealer group is a real concern too: it buys 30.0 of 120.0, or 25% of sales. If it left or demanded longer payment terms, receivables and sales would both move. Watch receivable days and discounts.
Hugo, a carpenter, wants to lend 15,000 euro to a friend’s cycle workshop and plans to start with its balance sheet. Explain why you would first ask who its customers and suppliers are.
Hugo. A balance sheet shows the past. Customers and suppliers show whether the workshop can keep selling, and at what margin. One dominant customer or supplier makes it fragile whatever the balance sheet says.
Nadia. The competitor may simply be bigger. Divide profit by revenue, or by the money invested, to compare equals.
a drag-and-drop game with twelve short statements about Harmony, such as “one dealer group buys a quarter of our bikes” or “cell prices rose 9% this year”. The learner drops each into the box it belongs to (market, growth, value chain, production, distribution, people or sustainability) and sees which line of the accounts it will change. The learner discovers that nearly every number has a business cause.
1
Harmony sold 6.7% more bikes at a 7.1% higher average price. Sales grew by…
Subtracting makes no sense here.
Right. 1.067 × 1.071 − 1.
Volume grew too.
The two effects multiply, they do not add; the sum also understates.
How should an analyst define a company's market?
Harmony sells across Europe.
Classifications group sellers, not buyer choices.
Right. An e-bike competes with a transport pass and a scooter.
Customers do not care how the product is made.
One supplier provides most of Harmony's cells; one dealer group buys 25% of its bikes. Which position is weaker?
It reaches margins, receivables and inventories.
The one with fewer alternatives is weaker.
Right. It would show in purchase prices, gross margin and safety stock.
It is a real concern, but dealers are easier to replace.
2Four questions, always in the same order
The accounts are read with a plan. The order is deliberate: each question needs the answer to the one before.
Wealth, investment, financing, returns
The first question is whether the business generates wealth
Look at sales growth, EBITDA, the EBITDA margin and operating profit. If the firm earns little on its sales, nothing downstream can be good.
The second question is what that wealth costs in investment
Growth has a bill: fixed assets, working capital, capex against depreciation, capital needed per euro of sales. A firm can have a high margin and absorb every euro in new machines and stock.
The third question is who paid, and whether they can be repaid
Only after seeing cash generated and consumed can the analyst judge the funding: net debt, interest cover, the debt and equity mix, liquidity.
The fourth question is whether the money was well used
Returns compare operating profit and net income with the capital employed and equity invested, then with the return investors require. This is the verdict step.
| Step | Question | Typical measures | Module |
|---|---|---|---|
| 1 Wealth generation | Does the business earn a margin and grow? | Sales growth, EBITDA margin, EBIT | CF10 |
| 2 Investment | What capital does growth require? | Capex / depreciation, working capital, capital turnover | CF11 |
| 3 Financing | Can the funders be repaid? | Net debt / EBITDA, EBITDA / interest, gearing | CF12 |
| 4 Returns | Does it earn more than it costs? | ROCE and ROE against required returns | CF13 |
Why the plan looks like this
The plan is the product of history. When lending meant claiming collateral, analysts valued assets for liquidation. Later came liquidity ratios, then self-financing capacity and working capital needs. From the 1990s shareholders pushed analysis towards returns and free cash flow. After 2008 liquidity and solvency returned to the front. Each era fixed on the risk it had just met, and the four steps keep all of them.
Harmony run through the plan
Harmony passes step by step, with a thin margin of safety at the end
Growth has a bill that profit may not pay
2.1 Cyclora in four steps
Run the four steps on Cyclora, in € million: sales 60.0, EBITDA 4.8, depreciation 3.0, net interest 1.6, capex 2.5, capital employed 60.0, net debt 26.4, working capital 16.8. Compute the EBITDA margin, the operating margin, capex over depreciation, net debt over EBITDA, EBITDA over interest and ROCE after tax at 25%. Which step breaks first, and what would you ask the lenders?
EBITDA margin 4.8 / 60.0 = 8.0%. Operating profit 1.8, a margin of 3.0%. Capex over depreciation 2.5 / 3.0 = 0.83. Net debt over EBITDA 26.4 / 4.8 = 5.5. EBITDA over interest 4.8 / 1.6 = 3.0. ROCE after tax 1.8 × 0.75 / 60.0 = 2.25%.
Step 1 fails first: the margin is thin. Capex below depreciation suggests under-investment. Leverage of 5.5 with cover of 3.0 is tight, and a return of 2.25% is below any plausible cost of capital. Ask lenders about debt maturities, covenants and the plan to restore the margin.
2
In what order should the four questions be asked?
Returns are the verdict, so they come last.
Funding can only be judged once cash generated and consumed are known.
The order is deliberate.
Right. Each question needs the answer to the one before.
Harmony's NOPAT is 9.0, capital employed 84.0 and WACC 9%. The wealth created beyond investors' requirement is…
1.7 is the gap in points between ROCE and WACC, not an amount.
That ignores the cost of the capital used.
That is the capital charge.
Right. 9.0 − 7.56.
Cyclora: sales 60.0, EBITDA 4.8, depreciation 3.0, capital employed 60.0. Its ROCE after tax at 25% is…
That is the operating margin, before tax and on sales.
That uses EBITDA instead of EBIT.
That is the EBITDA margin.
Right. 1.8 × 0.75 / 60.0, below any plausible cost of capital.
3Numbers mean something only against a comparison
A margin of 15% is neither good nor bad. It becomes a statement once it is set against something. There are three yardsticks, and each has a blind spot.
Three yardsticks, three blind spots
| Yardstick | What it answers | Blind spot |
|---|---|---|
| Trend (the firm over time) | Is it improving? | A firm can improve and still be weak |
| Peers (similar firms now) | Is it better than those facing the same market? | Peers may differ in size, mix, accounting or year end |
| Standards (rules of thumb) | Is it inside the zone lenders accept? | Rules are sector specific and age badly |
Trend shows direction, and two lines moving apart carry the message
| € million | N-2 | N-1 | N | Index N (N-2 = 100) |
|---|---|---|---|---|
| Sales | 90.0 | 105.0 | 120.0 | 133.3 |
| EBITDA | 11.7 | 14.7 | 18.0 | 153.8 |
| EBITDA margin | 13% | 14% | 15% | |
| Sales growth | 16.7% | 14.3% | ||
| EBITDA growth | 25.6% | 22.4% |
Sales grew by a third over two years, EBITDA by more than half. Profit outgrows sales when fixed costs are spread over more volume, prices rise faster than costs, or the mix improves. The analyst tests each (CF10).
Peers show rank, and the choice of peers is a judgement that can be gamed
| Year N | Harmony | Velora | Cyclora |
|---|---|---|---|
| Sales (€ million) | 120 | 1,200 | 60 |
| EBITDA margin | 15.0% | 14.0% | 8.0% |
| Net debt / EBITDA | 1.67 | 1.5 | 5.5 |
| ROCE after tax | 10.7% | 12.0% | 2.3% |
| Working capital / sales | 20% | 14% | 28% |
Harmony sits between a large, efficient listed rival and a small, struggling maker. A comparison is a source of questions. Velora's EBITDA margin is a point below Harmony's, yet its operating margin is the same 10% and its return is higher: it outsources more, so it needs less capital, and much less working capital, for each euro of sales. Peers must be restated on the same basis, for example for leases (CF8). And whoever picks the peer group picks the answer: a weak list flatters any firm.
Standards give a threshold, and thresholds belong to a sector and an era
Lenders use rules of thumb. A net debt to EBITDA ratio below about 2 to 3 is often considered comfortable for an industrial firm. Interest cover well above 4 is usually treated as safe. Such figures help to triage but do not fit a stable utility or a cyclical miner. Since 2022, when borrowing costs rose, interest cover has again become a binding constraint.
3.1 Growth that does not pay
Cyclora’s sales rose from 56.0 to 60.0 million, while its EBITDA fell from 5.6 to 4.8. Compute sales growth, EBITDA growth and the EBITDA margin in each year. What kind of growth is this, and what two explanations would you test?
Sales growth: 60.0 / 56.0 − 1 = 7.1%. EBITDA growth: 4.8 / 5.6 − 1 = −14.3%. The margin falls from 5.6 / 56.0 = 10.0% to 8.0%.
This is growth bought at the expense of margin. Test price cuts and unrecovered cost rises.
Common-size statements make firms of different sizes comparable
Divide every line by one number and the size disappears
A common-size income statement expresses each line as a percentage of sales. A common-size balance sheet divides each line by total assets or, in the economic view, capital employed. Harmony and Velora become comparable at once, and the method shows where in the statement two firms diverge.
| % of sales, year N | Harmony | Cyclora |
|---|---|---|
| Sales | 100.00 | 100.00 |
| EBITDA | 15.00 | 8.00 |
| Depreciation | 5.00 | 5.00 |
| Operating profit | 10.00 | 3.00 |
| Net interest | 1.67 | 2.67 |
| Profit before tax | 8.33 | 0.33 |
| Tax at 25% | 2.08 | 0.08 |
| Net income | 6.25 | 0.25 |
The two firms spend the same 5% of sales on depreciation. The gap opens at EBITDA, 15% against 8%. Below it, interest takes 2.67% of Cyclora’s sales, which is 89% of its operating profit (1.6 / 1.8). Cyclora works almost entirely for its lenders. On the balance sheet, Harmony’s capital employed of 84.0 is 71% fixed assets and 29% working capital, funded 64% by equity and 36% by net debt.
Nadia runs a nursery school and hears that a competitor “makes more profit”. Explain why comparing profit in euros is unfair, and what to compute instead.
3.2 Name the blind spot of the trend yardstick.
A firm can improve and still remain weak, because a trend says nothing about the level against others or against a required standard.
3.3 Why does a common-size statement help to compare Harmony and Velora?
Dividing by sales removes the effect of size, so the structure of the two statements can be compared line by line.
3.4 What must an analyst do to peer data before comparing?
Restate the figures on a common basis: same accounting treatment for items such as leases, same definitions of EBITDA and net debt, and where possible the same year end.
a table of Harmony, Velora and Cyclora with a switch for the yardstick (own trend, peer average, lender standard) and a selector for the ratio (net debt over EBITDA, EBITDA margin, ROCE). The verdict line, “better”, “in line” or “worse”, changes as the learner switches. The learner discovers that the same ratio earns different verdicts under different yardsticks.
3
What is the blind spot of the trend yardstick?
Right. A trend says nothing about the level.
It can be computed for any firm with history.
That is the blind spot of the peer yardstick.
That is the blind spot of standards.
Cyclora's sales rose from 56.0 to 60.0 while EBITDA fell from 5.6 to 4.8. This is…
EBITDA fell 14.3% while sales grew 7.1%.
Right. Margin fell from 10.0% to 8.0%.
Leverage would make profit grow faster than sales, not fall.
Nothing suggests that; test prices and costs first.
Why do common-size statements help compare Harmony with Velora?
Tax stays as a line, as a share of sales.
Right. Structure can then be compared line by line.
They only change the unit.
Peers still need restating on the same basis.
4Trust the accounts only after trying to break them
Accounts are prepared by managers who prefer good news. The final layer of the method looks for where flattering choices usually hide, and then asks how much weight a score or a rating can bear.
Cash that does not follow profit is the first alarm
A short list of warning signs catches most problems
Receivables grow much faster than sales: sales brought forward, or customers who struggle to pay.
Inventories grow faster than sales: unsold goods or a coming write-down.
Profit rises while operating cash flow falls.
One-off charges that reappear every year.
Changes in depreciation lives, provisions or revenue policy that help the year’s result.
Costs moved onto the balance sheet, especially development costs.
Adjusted measures with long add-back lists and no clear reconciliation.
Funding that hides debt, such as supplier finance or factoring (CF8).
The first three signs share one idea: profit is an opinion, cash is a fact. An analyst who sees the two drift apart for more than a year or two should ask why before doing anything else.
Capitalising a cost improves every headline ratio and changes no cash
Adjusted figures deserve one rule: demand the bridge. A measure such as “adjusted EBITDA” is legitimate when reconciled to reported profit and applied consistently. IFRS 18, applying from 2027, requires such measures to be disclosed and reconciled in the notes. Count how many years the same “exceptional” cost has appeared.
A clean audit opinion is a filter, not a guarantee of health
4.1 Cyclora’s receivables
Cyclora’s receivables were 9.0 on sales of 56.0 a year ago and are 14.0 on sales of 60.0 today. Compute the growth of sales and of receivables and the days of sales in receivables each year. Give two explanations and say what you would ask for.
Sales grew by 60.0 / 56.0 − 1 = 7.1%. Receivables grew by 14.0 / 9.0 − 1 = 55.6%. Days of sales: 9.0 / 56.0 × 365 = 58.7 a year ago, 14.0 / 60.0 × 365 = 85.2 today.
Customers may pay slowly, or sales may have been pushed with generous terms. Ask for the ageing of receivables and post year-end credit notes.
a simulator on the Harmony table where the learner drags a slider for the amount of cost capitalised, from 0 to 4 million. EBITDA, margin and leverage move with it. A fixed line shows EBITDA minus capex, which does not move. The learner discovers that policy choices change headline ratios, never the cash.
Scores and ratings compress a diagnosis into one judgement
A score is a weighted sum of ratios fitted to past failures
A scoring model picks a few ratios, gives each a weight and adds them up. The weights are chosen to separate firms that failed from firms that did not.
The best known real model is the Altman Z-score, first published in 1968 for listed manufacturers. CF13 shows how it is built. Scores have limits. They are fitted to a sample from one period and one industry, so they age. When lenders rely on them, a low score raises the cost of credit and can help cause the failure it predicts. Managers can also adjust ratios to move a score.
A rating adds judgement to the arithmetic
A rating is an opinion on the likelihood that a borrower will meet its obligations. Credit rating agencies publish them for large issuers, banks compute internal ones, credit insurers and some central banks assess firms too. A rating combines a financial profile, which is the arithmetic of the four steps, and a business profile, which is the analysis of Part 1. The market divides ratings into investment grade and speculative grade, and the borrowing cost changes sharply across that line. Ratings lag events, so an analyst reads the outlook as well as the letter. CF33 shows how ratings shape debt choices.
Lenders now also use machine-learning models with hundreds of inputs. They predict default among thousands of small borrowers better than a simple score, but they offer patterns without explanation, a poor basis for a large corporate loan.
4.2 Why do receivables growing faster than sales deserve a question?
It can mean sales were brought forward with generous terms, customers are not paying, or the firm is relaxing credit to win sales. Each reading points to a risk.
4.3 A manager capitalises costs. Which ratios improve and which number stays the same?
EBITDA, the EBITDA margin and net debt over EBITDA improve. EBITDA minus capex, and the cash itself, stay the same.
4.4 Name two limits of a scoring model.
Any two of: fitted to one period and industry so they age; can be self-fulfilling if lenders rely on them; can be gamed by managers; fit poorly to businesses with intangible value.
4
Harmony capitalises 2.0 of development spending. What does not change?
It rises to 11.0.
It falls to 1.50.
Right. Cash is identical; only the presentation moves.
It rises to 20.0.
Cyclora's receivables grew 55.6% while sales grew 7.1%. What should the analyst ask for?
They did not here: days went from 58.7 to 85.2.
Right. Slow payers or sales pushed with generous terms.
Irrelevant to collection risk.
The question is about Cyclora's own customers.
Which statement about scoring models is right?
They compress it, they cannot replace it.
Right. Useful for triage, never a verdict.
They are fitted to past failures.
An audit says nothing about failure risk.
Summary and cheat sheet
Les formules du module
Touchez un terme pour voir ce qu’il veut dire et sa valeur chez Harmony.
Volume et prix se multiplient
1 + = (1 + ) × (1 + )
ROCE après impôt
= × (1 − ) /
Contre un CMPC de 9,0 % : 1,7 point de création de valeur.
Richesse créée au-delà de l’exigence des investisseurs
= −
Jours de ventes en créances
= / × 365
| Idea | What to remember |
|---|---|
| Business first | Define the market as the customer sees it, split growth into volume and price, find who holds bargaining power. |
| Predict the accounts | Write down what the business implies for capex, stock and credit terms, then check. A failed prediction is a question. |
| Four-step plan | Wealth generation, investment, financing, returns. Always in that order. |
| Three yardsticks | Trend shows direction, peers show rank, standards show tolerance. Each has a blind spot. |
| Common-size | Divide by sales or capital employed so firms of different sizes can be compared. |
| Cash against profit | If profit rises and cash does not follow, ask why before anything else. |
| Capitalised costs | Raise EBITDA, margin and leverage ratios without moving cash. Use EBITDA minus capex as a check. |
| Audit opinion | A clean opinion means the accounts are fair. It does not mean the firm is safe. |
| Scores and ratings | A weighted sum of ratios, or an expert opinion. Useful for triage, never a verdict. |
| Harmony | Margin 15%, net debt / EBITDA 1.67, ROCE 10.7% against WACC 9.0%: thin but real value creation. |
The one-sentence version
Understand the business first, put the accounts through four questions in order (wealth, investment, financing, returns), judge every figure against a trend, a peer group and a standard, and try to break the accounts before you trust them.
Key terms
| Term | Plain-language meaning |
|---|---|
| Common-size statement | A statement where each line is shown as a percentage of one base, such as sales. |
| Bargaining power | The ability of one link in the chain to keep more of the margin because its partners have fewer alternatives. |
| Capitalising a cost | Recording spending as an asset instead of an expense, so it reduces profit over several years instead of at once. |
| Alternative performance measure | A profit figure defined by management, such as adjusted EBITDA, that is not a standard accounting subtotal. |
| Scoring model | A formula that adds weighted ratios to estimate how likely a firm is to fail. |
| Investment grade | A credit rating in the upper part of the scale, where default risk is considered relatively low. |
Going further
Vernimmen, Quiry, Le Fur — Finance d'entreprise (Dalloz), chapter on the introduction to financial diagnosis and the online chapter on the history of financial analysis.
Porter (1980) — Competitive Strategy, on the structure of industries and bargaining power.
Altman (1968) — Financial ratios, discriminant analysis and the prediction of corporate bankruptcy, Journal of Finance.
Independent educational material. Not investment, legal or tax advice. All companies and figures in the running case are fictional.