What you will be able to do
Explain why a start-up is financed with equity in stages, and why a bank will not do the same job.
Compute pre-money, post-money, price per share and dilution, and follow a cap table across rounds.
Value a young company with the venture capital method, and read a price as an implied return.
Compare term sheets by simulating exits: liquidation preference, anti-dilution, vesting.
Cost venture debt properly, and name the main French instruments for a start-up.
1Why start-ups run on equity, raised in stages
A start-up cannot borrow its way to success and cannot be valued like a mature firm. These two facts shape everything that follows.
Failure is the normal outcome, so only equity can carry the risk
Most of a start-up's value is a small chance of a very large result
Of a hundred young companies, many disappear within a few years, a few do well and one or two are worth a fortune. Nobody knows in advance which. Cash flow is negative for years, and the founders hold most of their wealth in one company, the opposite of the diversified investor of CF15.
Rounds are staged so that each cheque buys one answer
Money arrives in tranches because uncertainty falls in steps
Investors pay for one milestone at a time. If the answer is no, they stop and lose a small amount. If it is yes, the next round is priced higher. This is a real option (CF26): paying a little now for the right to put in more later. It disciplines the founders, because the next cheque depends on results. It helps them too, because each tranche is raised after the risk has dropped, so less of the company is given up per euro.
Business angels are wealthy individuals investing their own money. Venture capital funds take over for rounds in the millions. The exit comes through a listing (CF37), a sale (CF38) or a buyout (CF40).
A few outliers repay the whole fund, so investors demand a great deal
Runway is the clock that rules the founder's year
Runway is cash divided by monthly burn
Burn is the cash the company loses each month. Runway is how many months it survives at that rate.
Runway (months) = cash in the bank / monthly cash burn
Raising a round takes six to nine months, and a round is meant to cover twelve to twenty-four months. A founder who starts with six months left negotiates from weakness, because every investor can see the date on which the cash ends. After the seed round Harmony holds 1.0 million and burns 0.05 million a month: 1.0 / 0.05 = 20 months, so the process must begin by month 11 at the latest.
1.1 The deadline before the money runs out
In mid-2018 Harmony holds 0.72 million and burns 0.06 million a month. Compute the runway. The Series A will take six to nine months: by when must Léa and Samir start? What happens to that deadline if the burn rises to 0.08 million after two months of hiring?
Runway = 0.72 / 0.06 = 12 months. With a six to nine month process, the founders must start by month 3 (12 − 9). If the burn rises to 0.08 after two months, 0.72 − 0.12 = 0.60 remains, which lasts 7.5 months, so the cash ends at month 9.5 (2 + 7.5). The latest start was month 0.5 (9.5 − 9), which has already passed: they must start at once. Burn rises faster than founders plan, so the start date needs a margin.
Karim, your uncle, runs a successful plumbing company and has a good bank. He cannot understand why a young firm with an exciting product does not simply take a bank loan, and why investors accept that most of their bets will lose. Explain it to him in four or five sentences, without the words option or equity.
Karim. A bank gets its money back plus interest and nothing more if the company is a hit, but loses if it fails, so it cannot take that bet. Investors lose a little on many failures and win a lot on a few successes, so they can afford the risk the bank cannot.
Camille. Ask what the 60,000 let her do and what the remaining 70% is worth. If the money let the studio double in value, she owns 70% of something twice as big, which is 40% more than before. The sale is good when the growth it pays for exceeds the share she gave away.
1
Why will a bank not lend to Harmony in 2016 to fund its first bikes?
No such rule; banks lend against assets that keep their value.
Many founders would prefer to avoid dilution; the lender is the one who refuses.
The problem is the risk, not the rate.
Right. Only equity can carry the risk of the idea itself.
A fund's ten bets return 17.5 on 10.0. Without the single best deal they return 5.5. What does that imply?
Nobody knows in advance which one will win.
It made 1.75×.
Right. A few outliers repay the whole fund.
The average hides four total losses.
Harmony has 0.72 million and burns 0.06 a month. A raise takes six to nine months. By when must the founders start?
That is when the money runs out.
That allows only six months, the shortest case.
Right. Runway 12 months, less 9 months of process.
Too late for a nine-month process.
2Every round is arithmetic on the cap table
The cap table lists who owns what. Each round changes it by a few simple rules, and valuation without profits is disciplined by one method.
Pre-money plus new money equals post-money, and the stake is money over post-money
Pre-money value is what the company is worth before the new money arrives. Post-money value is the value afterwards. The cheque is simply added, and the investor's share follows.
Post-money = pre-money + new money
Investor's stake = new money / post-money
| Round | New money (m) | Pre-money (m) | Post-money (m) | New investors' stake | Price per share |
|---|---|---|---|---|---|
| Seed 2017 | 1.0 | 4.0 | 5.0 | 20.0% | 1.00 |
| Series A 2019 | 6.0 | 18.0 | 24.0 | 25.0% | 3.60 |
| Series B 2022 | 17.0 | 43.0 | 60.0 | 28.3% | 6.00 |
Harmony's three rounds. Investors typically buy between a fifth and a third of the company per round.
Dilution is not a loss when the pie grows
Founders watch their percentage fall from 100% to 40% and feel robbed. The right question is what the stake is worth: percentage × post-money value.
Camille, a designer you know, sold 30% of her small studio to an investor for 60,000. She says she has lost a third of her business. Explain how she can decide whether the sale was a good one, using the idea of what her remaining 70% is worth.
Without profits, value is a bet on the exit, discounted at a very high rate
A discounted cash flow (CF27) needs forecasts nobody can make, and multiples (CF28) need profits that do not exist. Investors use the venture capital method: forecast the value at exit in about five years, choose the return needed, and work out the stake that delivers it.
Stake needed today = investment × (1 + target)ⁿ / exit value / share kept after later dilution
The target is high, 30% to 60% a year for early rounds and about 20% before a listing. It covers the chance of failure, which no plan shows, the founders' optimism and the difficulty of selling shares quickly.
The reverse use is the most useful: any price implies a return if the plan comes true. If that return looks too low for the risk, the price is too high.
a cap table simulator with sliders for money raised, pre-money value and pool size, and a toggle for pool before or after the money. Bars show each holder's percentage and the founders' stake in euros after each round. The learner discovers that the pool shuffle can move the founders' result as much as a visible change in headline valuation, and that a bigger cheque at a lower valuation can leave founders with less value.
2.2 A fund invests 3.0 million for 15%. What are the pre-money and post-money values?
Post-money = 3.0 / 0.15 = 20.0 million. Pre-money = 17.0 million.
2.3 Why is a venture capitalist's 40% target return no evidence that the average investment earns 40%?
The target is what investors ask on each deal so that winners cover losers. It is a price of risk. The actual return depends on how many investments fail, and averages much lower.
2.4 Two offers raise 2.0 million: 10.0 million pre-money with a 10% post-money pool inside it, or 9.0 million pre-money with no pool. Which is better for the founders?
Option 1: post-money 12.0, pool 10% × 12.0 = 1.2, so the founders' effective pre-money is 12.0 − 2.0 − 1.2 = 8.8 million. Option 2: 9.0 million. Option 2 is better, although its headline is lower.
2
Northlight invests 6.0 million at a pre-money value of 18.0 million. What price per share does it pay, with 5.0 million shares before the round?
That is 6.0 / 5.0.
That is the Series B price.
Right. 18.0 / 5.0 = 3.60; it receives 1.667 million new shares.
That divides post-money by the old share count.
Series B: 17.0 at 43.0 pre-money, with a 5% pool created before the money. What is the effective pre-money for the old owners?
Right. 60.0 − 17.0 − 3.0: the pool is paid by the existing shareholders.
That is the post-money.
That is the headline, which includes the pool.
The pool is subtracted, not added.
The founders fall from 100% to 40% over three rounds. What happened to the value of their stake?
Each round raised the price per share.
Right. Dilution is not a loss when the pie grows.
Percentage and value are different things.
Each round's price gives a value.
3The term sheet decides who is paid first when the outcome disappoints
Price is half of a venture deal. The other half is a set of rules that matter little when things go well and decide everything when they do not.
A liquidation preference lets the investor take their money back first
A liquidation preference gives investors the right, in a sale or wind-up, to receive a multiple of their money before other shareholders receive anything. The usual multiple is 1×. With a non-participating preference, the investor takes the preference or converts to ordinary shares, whichever pays more. With a participating preference it takes the preference and then also shares the rest, so it is paid twice.
Plain 1× non-participating preferences are common in European venture deals. Participating ones, especially without a cap, move value from founders to investors and are a warning sign.
Anti-dilution protects against a down round, at the founders' expense
A down round is priced below the previous round. Anti-dilution clauses compensate earlier investors by lowering their conversion price, which gives them more shares at no cost. The full ratchet resets the old price to the new one. The broad-based weighted average, the market standard, adjusts in proportion to the share of the company sold at the lower price.
New conversion price = old price × (A + B) / (A + C)
A is the shares outstanding before the round, B the shares the new money would buy at the old price, C the shares it actually buys.
3.1 A larger down round
Same situation, but the bridge is 6.0 million at 4.80. Compute the founders' stake with no protection, with the weighted average and with the full ratchet. Then say why full ratchets are rare even though investors would like them.
New shares: 6.0 / 4.80 = 1.25 million. No protection: 4.0 / 11.25 = 35.56%. Weighted average: B = 1.0, new price = 6.00 × 11.0 / 11.25 = 5.867, Series B converts into 2.898 million shares, 0.064 more, founders 35.35%. Full ratchet: 0.708 million more shares, founders 4.0 / 11.958 = 33.45%.
Full ratchets are rare because each cheaper round hands the founders' shares to the earlier investor, and later investors refuse to join a company where one class can take so much.
Vesting and control clauses keep the team together and the investors informed
Vesting means shares become the founder's own only gradually, typically over four years with a one-year cliff before anything vests. A founder who leaves early sells back the unvested part at nominal value. If Samir holds 1.6 million shares and leaves after 18 of 48 months, 18 / 48 = 37.5% have vested: 0.6 million stay his and 1.0 million return to the company.
Other clauses decide who can block or force a sale. Investors receive monthly accounts and a veto on a new round, a sale or large borrowing. A drag-along forces minority holders to sell if a large majority does, and a tag-along lets them sell alongside a major holder. At the earliest stage, a convertible note or SAFE (simple agreement for future equity) postpones the price: the money turns into shares at the next round, at a discount or a valuation cap.
an exit waterfall. A slider sets the sale price from 0 to 200 million. Toggles switch between plain shares, 1× non-participating, 1× participating and participating with a cap. A stacked bar shows what founders, Northlight, other investors and the pool receive. The learner discovers that preferences decide the split below a threshold price and vanish above it, while participation shifts value even at high prices.
3.2 A company is sold for exactly its last post-money value. Why does a non-participating investor not care whether it converts?
At that price the investor's share as an ordinary shareholder equals the 17.0 million it invested, so preference and conversion pay the same.
3.3 Why does a participating preference cost founders more than it appears?
It is paid before the founders at every price and is not diluted by growth. At a 60.0 sale it gives the investor 29.2 instead of 17.0, a transfer of 12.2 million, larger than a visible valuation concession and far less obvious.
3.4 Why does a cliff protect Léa as much as it protects Northlight?
It stops Samir leaving within months with a full stake and leaving Léa to do the work, which protects the company that both she and the investor own.
3
Series B holds a 1× non-participating preference on 17.0 (28.3%). The company is sold for 120.0. What does Series B receive?
That adds the preference to the converted share.
Converting pays more: 28.3% × 120.0.
Right. Above 60.0 the preference is irrelevant.
That is the participating version.
Why are full ratchets rare even though investors would like them?
They protect the earlier investor at the founders' expense.
Right. In the 6.0 bridge case the founders fall to 33.45%.
They can be written in, but are seldom accepted.
They apply to down rounds.
Why does a one-year cliff protect Léa as much as Northlight?
Vesting delays ownership, it does not speed sales.
It has no direct effect on price.
Right. It protects the company both of them own.
Unvested shares return to the company, not to Léa.
4Debt, public money and the French toolbox
Equity is the expensive part of the funding. Other sources, and the tools French law offers, make it go further.
Venture debt extends the runway, but its all-in cost exceeds the coupon
Venture debt is a loan to a start-up already backed by venture investors. The lender looks at the shareholders and the next round, not at profits. It charges a high rate, an arrangement fee and often warrants: rights to buy a few shares at a fixed price. It stretches a round by some months. It cannot finance a business with no path to the next round.
A French start-up has a toolbox of its own
The structure exists to organise the exit
A venture fund has a life of about ten years and must sell. Preferred shares normally convert to ordinary shares at a listing, so the preferences stop. Harmony's listing at 12.00 a share is twice the 6.00 of Series B, so Northlight's 3.5 million shares are worth 42.0 million on paper, 2.47× its money.
Valuations peaked in 2021 and fell sharply in 2022 and 2023, so down rounds and tougher terms became common. A founder raising in 2026 should assume that preferences and ratchets matter most when valuations are fragile.
4
Harmony borrows 3.0 million of venture debt at 10%, with a 0.06 fee and warrants worth 0.25 at the end. What is the all-in cost?
That ignores the fee and the warrants.
Adding the fee as two points is not how a rate works.
That counts the warrants as a yearly cost.
Right. The rate that equates 2.94 received with 0.30, 0.30 and 3.55 paid.
Harmony spends 0.6 million on eligible research in 2017. What does the CIR bring, and when?
Small firms get a cash refund.
The credit is 30% of eligible spending.
Small and young companies are refunded the year after.
Right. About 3.6 months of burn.
Harmony lists at 12.00. What happens to Northlight's preferred shares?
Conversion goes to ordinary shares.
Right. That is 2.47 times its 17.0.
Preferred shares normally convert at a listing.
The fund sells or keeps shares; it is not repaid.
Summary and cheat sheet
Les formules du module
Touchez un terme pour voir ce qu’il veut dire et sa valeur chez Harmony.
Pré-money, post-money et part de l’investisseur
= / ( + M)
Post-money 24,0 ; 1,667 million d’actions nouvelles à 3,60.
Pré-money effective quand le pool est dans la pré-money
= − M −
Série B : 60,0 − 17,0 − 3,0 = 40,0, sous les 43,0 affichés.
Méthode du capital-risque
= × (1 + )ⁿ / /
Anti-dilution par moyenne pondérée
= × ( + ) / (A + )
Pont de 3,0 à 4,80 : la série B reçoit 0,034 million d’actions de plus.
| Idea | What to remember |
|---|---|
| Why equity | Start-up risk is the risk of the idea itself. Lenders gain nothing from success, so only equity can carry it. |
| Staged rounds | Each round pays for one answer. Investors can stop; founders sell less per euro as risk falls. |
| Power law | A few winners repay a fund, so investors push for high returns, control and protection. |
| Runway | Cash divided by monthly burn. Start raising 9 to 12 months before the money ends. |
| Pre- and post-money | Post = pre + new money. Investor's stake = new money / post-money. |
| Option pool | Reserved before the money arrives, it lowers the effective pre-money of existing owners. |
| Dilution | Judge a round by the value of the stake, not its percentage. |
| VC method | Work back from the exit at a high target return; use it in reverse to read the return a price implies. |
| Liquidation preference | Protects investors at low exits and is irrelevant at high ones. Participating versions are costly. |
| Anti-dilution | Weighted average is the norm; a full ratchet moves large value to investors in a down round. |
| Venture debt | All-in cost includes fees and warrants, so it exceeds the coupon. |
| France | SAS for flexibility, BSPCE and BSA for people, BSA AIR for speed, Bpifrance and CIR for public money. |
The one-sentence version
A start-up sells equity in stages to investors who accept many failures for a few big winners, and the price, the ownership table and the small print together decide how any outcome is shared.
Key terms
| Term | Plain-language meaning |
|---|---|
| Cap table | The list of owners with their shares and percentages. |
| Pre-money value | What the company is worth before new money arrives. |
| Post-money value | Pre-money value plus the new money. |
| Dilution | The fall in an owner's percentage when new shares are issued. |
| Burn and runway | Monthly cash loss, and the months the cash lasts at that rate. |
| Liquidation preference | Investors' right to be repaid before other shareholders in a sale. |
| Vesting | Shares earned gradually by staying with the company. |
| Down round | A round priced below the previous one. |
| Option pool | Shares reserved for future employees. |
Going further
Vernimmen, Quiry, Le Fur — Finance d'entreprise (Dalloz), chapter on business creation and start-up financing.
Metrick and Yasuda — Venture Capital and the Finance of Innovation (Wiley).
Feld and Mendelson — Venture Deals (Wiley).
Independent educational material. Not investment, legal or tax advice. All companies and figures in the running case are fictional.