SAFE-style warrants or convertible bonds: which seed instrument

Quick answer. The SAFE-style warrant (in France, the BSA AIR) and the convertible bond both let an investor put money in at seed stage without fixing the valuation: the investor enters the capital at a future round, with a discount and often a cap. The decisive difference: the warrant is not debt, the convertible bond is, with interest and repayment if conversion does not happen.

Raising funds at seed stage almost always stumbles on valuation: the company is too young to fix it without guesswork, yet the investor wants in now. Rather than negotiating a value neither side can defend, two instruments defer the question to the next round: the warrant and the convertible bond. Both rely on securities giving access to capital, framed by the French Commercial Code (notably article L228-91). But their legal nature, their balance-sheet treatment and their consequences for founders differ deeply. Here is how to decide.

Both instruments share a logic: invest today, value later. At seed stage, fixing a share price is difficult and risky for both parties. The warrant and the convertible bond let the investor put in funds immediately, then convert their contribution into shares at a future round, on preferential terms.

These terms rest on two levers found in almost every contract. The discount reduces the entry price compared with the next round, rewarding the early risk-taking: an investor entering with a 20% discount pays 80% of the series A price for their shares. The valuation cap sets a maximum value used for their conversion, protecting them if the company gains a lot of value between the two rounds. These mechanisms structure a first round well before a firm valuation is possible, as we set out for the seed to series A raise and for how to structure a first round of love money and business angels.

It must be said plainly: deferring is not removing. The valuation will come back, and both the discount and the cap will translate into real dilution for the founders at conversion. That is the most common mistake: signing a simple instrument thinking you avoid dilution, then discovering its extent at the next round.

The fundamental distinction lies in the legal nature of the instrument, and it commands everything else.

The warrant is not debt. The investor pays funds and receives subscription warrants, with no repayment claim or interest. If the next round happens, the warrants give the right to shares; their risk is that of a shareholder. If it does not happen, they have no right to repayment: they share the fate of the company. The convertible bond is debt. It bears interest and must be repaid if conversion does not occur. The investor stays a creditor until conversion, which protects them more, but this debt sits in liabilities and weighs on the financial analysis of the company.

Two practical consequences flow from this table. First, the convertible bond gives a deadline: at maturity, it either converts or is repaid. This deadline protects the investor but creates a cash constraint for the company if the round is delayed. Second, the convertible bond increases liabilities, which can hinder a future bank financing request or complicate the reading of the balance sheet by a series A investor.

The choice depends on the balance of power, the maturity of the project and the desired effect on the balance sheet. Rather than one instrument being superior in absolute terms, reason by situation.

The more mature the project and demanding the investor, the more the convertible bond imposes itself. The earlier the round and the stronger the trust, the more the warrant suits. In practice, both instruments often coexist on the same round, each investor entering with the one that matches their risk profile. This choice also combines with the taxation of the subscription, which we address below and detail in our article on SME subscription relief and the JEI schemes.

The choice of instrument also has a tax side, for the investor and for the company. Subscribing to the capital of an SME can give a right to the IR-PME income tax reduction (article 199 terdecies-0 A of the French General Tax Code), at the standard rate of 18%, raised to 25% for subscriptions to approved social-utility solidarity enterprises (ESUS) and certain solidarity property companies. Payments are taken into account up to 50,000 euros for a single person and 100,000 euros for a couple taxed jointly, with an obligation to keep the shares for five years.

Precision matters: this advantage requires a genuine subscription to capital. The warrant, which results in shares being issued at the next round, and the convertible bond, which gives a right to shares only at conversion, are not treated the same way over time. It is the operation that actually leads to entry into the capital that must be examined against the IR-PME conditions, not the initial intention. On the gain side, the later disposal of the shares is in principle subject to the flat tax, whose overall rate reaches 31.4% in 2026 for securities capital gains, after the rise in social levies. This point does not decide the choice of instrument, but it must be built into the investor's simulation. For arrangements with an interposed company, it combines with holding company taxation.

Our view: the instrument is chosen, the contract is calibrated#

Neither the warrant nor the convertible bond is better in absolute terms: they answer different needs. The warrant favours simplicity and the absence of debt, the convertible bond the protection of the investor and a firm deadline. Our conviction, formed file after file, is that the debate often misses the point: it is not the name of the instrument that creates or destroys value for founders, it is the calibration of the discount, the cap and the maturity.

A valuation cap set too low can heavily dilute the founders if the company grows fast, sometimes far more than a classic capital increase would. Conversely, a carefully negotiated cap and a reasonable discount preserve the balance of the cap table. Our role is to simulate the conversion under several next-round valuation scenarios, before signing, so the founders see precisely what they are giving up. A poorly calibrated instrument cannot be undone: conversion applies mechanically when the day comes.

The underestimated risk: stacking caps across several investors#

The trap we see most often is not the choice between warrant and convertible bond, but the accumulation of instruments with different caps issued over the months. Each business angel negotiates their own cap and their own discount, and the company ends up with a patchwork of conversion terms. At the series A round, all these conversions add up, and the cumulative dilution of the founders far exceeds what they had anticipated instrument by instrument. The cap table must be kept up to date and simulated globally, not contract by contract.

Founders at seed stage wanted to raise around 300,000 euros from several business angels, without fixing a valuation they considered premature. A first group of investors, close to the project and confident, subscribed in warrants: fast cash-in, no debt in liabilities, a 20% discount and a valuation cap negotiated for the future round. A more cautious investor, who wanted to keep a creditor status in case the series A was delayed, preferred an interest-bearing convertible bond with an eighteen-month maturity.

Simulating the conversion clarified the picture. With a series A round valued well above the cap, the warrant subscribers converted at the cap value, more favourable for them and therefore more dilutive for the founders than the discount alone. On the convertible bond side, the accrued interest was added to the converted principal, slightly increasing the number of shares issued. The cumulative dilution of the two instruments, once added together, exceeded the founders' initial estimate made contract by contract. The consolidated reading of the cap table, done before signing, made it possible to renegotiate a cap and allocate the envelope differently, rather than suffering the dilution at conversion.

In practice: securing the choice of instrument#

• Start with the investor's need: do they want a creditor status and a firm deadline (convertible bond) or do they accept shareholder risk with no debt (warrant)?
• Simulate the conversion under at least three next-round valuation scenarios, to measure the real effect of the discount and the cap on founder dilution.
• Keep a consolidated cap table from the first euro raised, and update it with every new instrument signed.
• Check the balance-sheet impact: a convertible bond increases liabilities, to anticipate if a bank loan is planned.
• Validate the tax treatment of the subscription against the IR-PME relief and its holding conditions before promising anything to the investor.
• Have the cap and maturity clauses reviewed: these decide the dilution, not the name of the instrument.

• The warrant offers no right to repayment: if the next round never happens, the investor loses, which can strain the relationship. Explain it clearly upfront.
• The convertible bond has a firm maturity: if the series A is delayed, the company must be able to repay or renegotiate, or risk default. This is a real cash constraint.
• A valuation cap set too low is the costliest mistake: it can dilute founders beyond what a classic capital increase would have done.
• Stacking instruments with heterogeneous terms complicates conversion: reason on cumulative dilution, not instrument by instrument.
• The IR-PME relief requires a genuine subscription to capital and compliance with the holding period: do not present it as acquired without checking the conditions.
• The convertible bond sits in liabilities: it can hinder a bank financing request or the reading of the balance sheet by a future investor.
• The legal documentation remains essential: these instruments are simple in principle, but their conversion clauses durably commit the split of capital. A review by your adviser and, where relevant, a lawyer, is necessary.