Guaranteed contracts in sport, and why the NFL is different

A player signs for four years and a number with eight digits in it, and within a day the number has been repeated everywhere as though it described his bank balance. It does not. It describes the most he could ever be paid if everything goes well, and in one of the major North American leagues that ceiling is a long way above the floor. Guaranteed contracts are the reason the same headline figure means completely different things in different sports, and they are the largest single difference between the labour markets of the world's richest leagues.

Get the guarantee wrong and every comparison you make afterwards is wrong. A basketball player and an American footballer who sign for the same reported total have not signed for the same money, have not taken on the same risk, and in most cases have not even agreed to the same kind of document.

This is how the guarantee works, what the different flavours of it actually cover, why one sport resists it, and what a player gets in return for going without.

What a guaranteed contract actually promises

Strip away the sport and a guarantee is a simple promise: this money is owed regardless of what happens next. Not owed if the player performs. Not owed if he stays fit. Owed.

That sounds like the ordinary condition of a contract, and in most industries it broadly is. Employment law in much of the world makes an employer who ends a fixed-term contract early liable for what remains, or at least for a substantial slice of it. Sport is the odd case, because sport wrote its own rules through collective bargaining and one set of those rules went a different way.

Three separate questions hide inside the word.

What triggers the payment? A guarantee is written against specific events. It might cover being released because a better player arrived. It might cover being released because the player got hurt. It might cover being released because the club ran out of cap room. Those are different promises, and a contract can contain one, two or all three.

When does the promise attach? Money guaranteed at signing is certain from the day the pen moves. Money guaranteed on a future date is not: it becomes guaranteed only if the player is still under contract when that date arrives, which gives the club a decision point and the player an anxious spring.

Is anyone required to set the money aside? A promise from a wealthy club is not the same as a promise backed by cash in an escrow account. Some leagues require the money to be funded in advance, and that single administrative rule turns out to shape behaviour more than most fans would guess.

Answer those three questions about any deal and you know what it is worth. Answer none of them and you are repeating a press release.

Fully guaranteed, injury guaranteed and guaranteed at signing are three different things

American football has the most developed vocabulary here, because it is the sport where the distinctions do the most work. The categories are worth learning even if you never watch a snap, because they are the clearest map of what a guarantee can and cannot be.

Guaranteed for injury is the narrowest of the useful protections. If the player suffers an injury playing the sport, cannot pass the club doctor's physical, and is released in that condition, the salary is still owed. That is real protection and it is the protection players want most, because the injury is the risk they cannot manage. It also has an edge that catches people out: a player who recovers, passes the physical, and is then released has not triggered the injury guarantee at all. The protection ended the moment he got healthy.

Guaranteed for skill covers being outperformed. If the club releases the player because a rookie is better, or a free agent is cheaper, or the scheme changed, the money is still owed. This is the guarantee clubs resist hardest, because it is the one that removes their ability to correct a mistake.

Guaranteed for cap covers being released for roster accounting reasons rather than football ones. On its own it is close to worthless, since a club that wants to release a player can almost always find a football reason.

Fully guaranteed is not a fourth category. It is all three at once. A salary that carries injury, skill and cap guarantees together is owed whatever the reason for the release, which is exactly what fans assume every contract in every sport already means.

Sitting across those categories is the timing question, and it is where most of the misunderstanding lives.

A signing bonus is guaranteed by the simplest mechanism available: it has already been paid. The money is in the player's account, usually within weeks of signing, and no subsequent release can retrieve it. That is why the signing bonus is the number an agent fights hardest for, and why a deal with a large bonus is a genuinely better deal than one with the same total and a small one.

Guaranteed at signing is the phrase that separates a real commitment from a conditional one. Money guaranteed at signing is certain. Money that becomes guaranteed on a date in the second or third league year is a promise the club can walk away from at any point before that date, and clubs do exactly that. The vesting date is not a formality. It is a scheduled opportunity to reconsider, and a player whose second-year salary vests in March spends February being valued by people with a calendar in front of them.

Then there is the number nobody puts in the headline: the practical guarantee. That is the amount a player will actually collect given how clubs behave, which is usually the guaranteed-at-signing money plus whatever the club has effectively committed itself to by the structure of the deal. If cutting a player in year three costs more in accelerated charges than keeping him, he is functionally guaranteed for year three whether or not the word appears in the document.

Finally, offset language. Where a released player is owed guaranteed money and then signs elsewhere, offset language lets the original club reduce what it pays by what the new club pays. Without offset, the player collects twice. Agents fight over this clause with an intensity that surprises people, because for a well-regarded veteran it is worth a substantial share of the guarantee.

40%20%40%

• Signing bonus paid up front24
• Base salary guaranteed at signing12
• Base salary carrying no guarantee24

A constructed contract, not a real one. The same total splits into money already paid, money the club owes whatever happens, and money the club may never pay at all.

Show the numbers

Invented example, where a four-year $60m deal actually sits

Item	Value
Signing bonus paid up front	24
Base salary guaranteed at signing	12
Base salary carrying no guarantee	24

Why the NFL is the outlier on guaranteed contracts

Every explanation you will read attributes it to injuries. The injury rate is real and it matters, but it is not the reason, because plenty of hazardous jobs come with guaranteed pay. The reason is history, and specifically the order in which two things happened.

Basketball and ice hockey got their guarantees from competition between leagues. When a rival competition appears and starts signing established players, the incumbent clubs stop negotiating from strength. A player with two bidders can ask for the money to be certain, and in the 1960s and 1970s players in both sports did exactly that. The rival leagues did not last, but the contract norms they forced into existence did, because a norm once established is very hard to claw back.

Baseball got there by a different route. Its players broke the reserve system and won free agency in the mid 1970s, and the guarantee arrived with it: once a player could choose his employer, the employer had to offer terms that were worth choosing, and certainty was the cheapest thing to offer a man who had spent his career unable to leave. The language that made a salary the club's obligation regardless of the player's ability to perform became standard within a few seasons.

American football's timing was worse in two ways. It faced its serious rival competition before free agency existed, so the leverage arrived in the wrong order. And when its players finally won free agency in the early 1990s, they won it in a package that included a salary cap. That is the crucial detail. Free agency without a cap creates an auction. Free agency with a hard cap creates an auction with a budget, and a club negotiating against a budget will pay a higher headline number in exchange for keeping the right to stop paying.

Two structural features then locked it in.

The first is squad size. A club carrying more than fifty players under contract has more than fifty chances to be wrong, and it will be wrong about a great many of them. Guaranteeing everyone is a different proposition from guaranteeing a squad of fifteen.

• 53Players on an NFL active roster
• 26Players on an MLB active roster
• 23Players on an NHL roster
• 15Players on an NBA standard roster

Roster limits set by each league's collective agreement. The league with the largest squad is the league that guarantees the least, and the relationship is not a coincidence.

The second is the funding rule, which almost nobody outside the industry has heard of and which shapes behaviour more than the cap does. Guaranteed money owed beyond the current league year has to be funded, meaning placed into an escrow account, rather than simply promised. A club offering a large multi-year guarantee is therefore not making a commitment on paper. It is writing a very large cheque now, years before the money is due. For an owner with limited liquidity, that is the binding constraint, and it explains why the clubs most willing to guarantee are the ones with the deepest balance sheets rather than the ones with the most cap room.

Something recent belongs here. The players' association brought a grievance arguing that clubs had colluded to resist fully guaranteed veteran deals after one such contract was signed in 2022. An arbitrator dismissed it in January 2025 while finding that the league's management council had encouraged clubs to reduce guarantees at an owners' meeting that year. The association appealed, and in April 2026 a three-member panel upheld the dismissal, again affirming that clubs had been invited to limit guarantees while holding that the evidence did not establish that they had actually agreed to. The outcome is a strange one to explain: the players lost, and the written findings say the invitation happened. It is the clearest public account yet of why the guaranteed contract has not spread in that sport, and it did not come from the bargaining table.

What happens when a club releases a player, step by step

The mechanics of a release are where the abstraction becomes money, and they are worth following once in detail.

1. 1The club decides the player is not worth his next salaryNothing about this decision is contractual. A general manager compares what the player will cost against what a replacement would cost, and the comparison is made against the non-guaranteed portion, not the headline total.
2. 2The calendar decides whenSalary that vests as guaranteed on a future date sets the deadline. A club that wants out must act before the vesting date, which is why so many veterans are released in a narrow window early in the league year rather than when their form dipped.
3. 3The player is released, and the route depends on service timeA player with enough accrued seasons becomes a free agent immediately outside the season. A younger player passes through waivers first, where another club can claim him and take on the contract as written.
4. 4Non-guaranteed salary simply disappearsThe base salaries for future years the player will not play were never an obligation. They vanish from both the club's books and the player's expectations, which is why the headline total was never a real number.
5. 5Guaranteed money still owed acceleratesAnything guaranteed that has not yet been paid becomes due, and the whole remaining amount lands on the current year's salary cap rather than being spread across the seasons it was written for.
6. 6Unamortised signing bonus accelerates with itA signing bonus is spread across the contract for cap purposes even though it was paid in one lump. Release the player early and every remaining slice of that proration is pulled into the current year.
7. 7A post-June 1 designation can split the chargeClubs may designate a limited number of releases so the acceleration falls across two seasons instead of one. It buys cap room this year at the cost of carrying a charge into the next.
8. 8Offset language decides who pays the balanceIf the player signs elsewhere and the contract contained offset language, the original club deducts the new salary from what it owes. Without offset, he collects from both.

The sequence described is the American football case, because it is the one with a genuine release mechanism. The equivalent in a fully guaranteed league stops at step four, because the salary does not disappear.

The result of that sequence is dead money: cap space consumed by somebody who is not in the building. Dead money is not a penalty and it is not a mistake, whatever the phrase suggests. It is simply the accounting catching up with cash that was already spent. A club that pays a bonus in year one and releases the player in year three has had the benefit of spreading the charge and now has to settle up. The detail of how that spreading works is set out in the piece on how the American football cap is actually accounted for, and the mechanism explains most of what looks like reckless spending.

Two identical headlines, two completely different deals

Here is a constructed comparison, invented for the arithmetic rather than drawn from any real contract.

Two players sign four-year deals reported at $60m.

Player A receives a $24m signing bonus, paid immediately. His base salaries are $4m, $8m, $12m and $12m. The first two base salaries are fully guaranteed at signing. The last two carry no guarantee.

Player B receives an $8m signing bonus. His base salaries are $6m, $10m, $16m and $20m. The first is fully guaranteed. The second is guaranteed for injury only. The last two carry no guarantee.

Both totals are $60m. Nothing else about them is comparable.

Player A knows on the day he signs that he will collect $36m: the bonus, plus two guaranteed base salaries. If he is released after two seasons, that is what he ends up with. Player B knows he will collect $14m, and one more year is protected only if he happens to be injured when the club wants rid of him. If he is healthy and released after one season, he has taken $14m from a deal reported at $60m. He would have been better off with a two-year deal at a much lower headline number and a bigger bonus, and any competent agent would tell him so.

• Player A
• Player B

Headline total60m60m

Guaranteed at signing36m14m

Collected if released after two years36m24m

Collected if all four years are played60m60m

Constructed contracts, not real ones. Figures in millions of dollars. The two deals are identical in every report and nothing alike in what they oblige anyone to pay.

Show the numbers

Invented example, the same $60m headline under two structures

Item	Player A	Player B
Headline total	60m	60m
Guaranteed at signing	36m	14m
Collected if released after two years	36m	24m
Collected if all four years are played	60m	60m

The lesson is not that Player B was badly advised. It is that the headline is a function of the structure, and a player who accepts a thin guarantee can extract a larger headline in return. Clubs are happy to inflate a number they may never pay.

This is also why contract comparisons within a sport are so frequently nonsense. Two deals with the same average annual value can differ by tens of millions in what will actually change hands, and the average annual value is the number that gets used for every list of the highest paid athletes ever compiled.

How to read a headline contract figure

Four numbers matter, in this order.

Guaranteed at signing. The floor. What the player would collect if it all went wrong tomorrow.

Cash flow in the first two years. The most reliable proxy for what a club really thinks, because it is the money the club has committed to before it has any new information. A deal with a big year-three number and a small year-one number is a deal the club expects to restructure or escape.

The out. The first point at which the club can walk away for a manageable cost, calculated as remaining guarantees plus unamortised bonus. Every long contract has one and it is rarely at the end.

New money against total value. Extensions are routinely reported as though the whole document were new. A player with two years remaining who signs a three-year extension is often reported as having signed for five years and the combined total, which flatters the deal by including money he had already been promised.

Only after those four does the total mean anything, and by then you will have noticed that it usually does not.

What a player extracts for giving up the guarantee

Non-guaranteed money is not charity. A player who accepts contract risk is being compensated for it, and reading the compensation is how you tell a good deal from a bad one in a league without guarantees.

The most common exchange is a larger headline in return for a shorter certain period. A club will pay a higher annual rate for the right to stop paying, because the option to exit has value and the club is buying it. That premium is real money in the first two years.

The second is structure. Signing bonus over base salary, roster bonuses payable early in the league year rather than late, per-game active roster bonuses that pay for availability rather than performance, and the conversion of base salary into bonus at the moment of signing. Every one of those moves cash forwards, and cash in hand is the only guarantee that cannot be renegotiated.

The third is term. A shorter contract puts the player back on the market while he still has leverage, which is worth more than a long deal he will never see the end of. This is the reasoning behind the two-year and three-year deals that look modest in a table and are frequently the smartest contracts in a sport.

The fourth is the collection of clauses that limit what the club can do: no-trade protection, guarantees that vest earlier than the club would like, injury protection extending beyond the contract, and the absence of offset language. None of these appear in the headline and all of them are worth negotiating over.

There is a fifth item that is rarely framed as compensation but functions as it. In a league without guarantees, the mechanism that lets a club retain a player it does not want to lose becomes a form of protection in its own right, though a bitterly resented one. The franchise designation used in American football pays a fully guaranteed one-year salary set by formula, which is why players who hate being tagged still take the money.

Guarantees, the hard cap and dead money

A guarantee and a cap pull against each other, and the tension explains most of the accounting oddities in capped leagues.

A cap is an annual limit. A guarantee is a multi-year promise. Committing guaranteed money to future seasons is committing cap space you do not yet have, which is why capped leagues all build a mechanism to force the accounting to reconcile when a deal ends early.

In American football, that mechanism is acceleration: everything owed lands now. In basketball, contracts are guaranteed as a norm, but the guarantee is not universal in the way people assume. Deals routinely contain partially guaranteed and non-guaranteed seasons, and there is a date early in the calendar year after which a season's salary becomes fully guaranteed. Waive a player and his guaranteed salary stays on the cap. The stretch provision lets a club spread that charge over a longer period, calculated as twice the remaining years plus one, which reduces the annual hit and extends the pain. The interaction of guaranteed money with the tightest spending restrictions in the sport is covered in the piece on what the second apron actually restricts, and it is the clearest current example of guarantees limiting what a club can do years later.

Baseball has no cap, so a guarantee there is a balance sheet problem rather than a roster problem. A released player is still owed his salary, less whatever a new club pays him, and the money still counts towards the competitive balance tax at its average annual value. The club gets no relief for having admitted the mistake, which is exactly the design intent set out in the explanation of how baseball's tax on payroll works.

Ice hockey took a third route. Contracts are guaranteed, and the buyout formula lets a club terminate one by paying a defined fraction of the remaining salary, spread over twice the remaining years, with the cap charge following a formula rather than the cash. It is the most explicit acknowledgement anywhere in sport that a guarantee is a price rather than an absolute.

Compare all four and the pattern is clear: guaranteed money never disappears, it only moves. The differences between the leagues are differences in where it is allowed to land, and the full comparison sits in the survey of how the major leagues set their spending limits.

Guaranteed contracts in football, and the transfer market as the release valve

European football has guaranteed contracts and almost nobody describes them that way, because the guarantee is so complete that it needs no name.

A football contract is a fixed-term employment contract. There is no release mechanism. A club cannot cut a player, cannot waive him, and cannot stop paying him because a better one arrived. Terminating without just cause exposes the club to compensation under FIFA's transfer regulations, and the framework exists precisely to stop clubs treating contracts as optional. A player under contract to the end of the decade will be paid to the end of the decade by somebody.

That creates a problem the American leagues solve with a release. Football solves it by selling.

The transfer fee is best understood as the price of ending a contract early with the selling club's consent. A club that wants a player out does not terminate the deal; it finds someone to take the contract on and charges for the privilege, or pays for the relief. That is why a club will accept a fee far below a player's perceived worth, and why a club sometimes pays another club to take a player: the wage liability it sheds is worth more than the fee it forgoes. The mechanics of when this can happen, and why the window compresses so much activity into so few days, are set out in the explanation of why the transfer window exists at all.

Three consequences follow, and all three are visible every season.

The squad that cannot be reduced. A player who will not accept a move cannot be removed. He can be excluded from the squad list, trained apart, and left out of every matchday, and he still collects every penny. The register of players eligible for a competition becomes the substitute for a roster cut, and it is a poor one, since the wage keeps flowing.

The settlement. Clubs buy their way out by agreeing a lump sum to tear the contract up, usually at a discount to the remaining wages because the player wants to go somewhere he will play. This is a negotiated release in everything but name.

The subsidised loan. A club sends a player out and pays a portion of his wages to make him affordable. Nobody calls this dead money and it is exactly that: cash leaving the building for a player who is not in it.

There is an accounting layer under all of this that behaves very differently from the American model. A transfer fee is capitalised and written off across the contract term, while wages are expensed as they are paid, which means the guaranteed wage bill hits the accounts in a completely different shape from the fee. The consequences for what a club can afford are worked through in the piece on how transfer fees are spread across a contract. Where a league then imposes a spending control on top, as with the Premier League's profitability rules, guaranteed wages become the least flexible line in the accounts, because they cannot be cut in a bad year.

Who insures the risk a guarantee creates

Somebody is carrying the risk that a guaranteed player stops being able to play, and it is worth knowing who.

Where contracts are guaranteed, the club carries it, and the club generally reinsures. Disability cover comes in two shapes. Permanent total disablement pays out when a career ends, typically as a lump sum against the remaining contract value. Temporary total disablement pays a share of the wage during a long absence, usually after a deductible measured in weeks rather than in money, which is why cover is worth very little for a hamstring and a great deal for a knee reconstruction. Premiums scale with age, position, wage and medical history, and the exclusions do most of the work: pre-existing conditions, recurrences of a known problem, and anything arising outside the sport are routinely carved out.

Where contracts are not guaranteed, the risk sits with the player, and the market serves him worse. He can buy personal cover, and the two products that matter are career-ending disability and loss of value. Loss-of-value cover is the interesting one: it pays out when an injury before a contract is signed reduces what the player subsequently earns, which is the exact risk a young player without guarantees faces. It is expensive, it is written with a threshold that must be crossed before anything pays, and the disputes over whether a decline in earnings was caused by the injury or by the player's form are as ugly as you would expect.

The asymmetry is the point. A club insuring fifty contracts is pooling risk across a portfolio and buying at institutional rates. A player insuring one career is buying a bespoke policy on a single asset he cannot diversify. Moving the risk from the club to the player does not make it disappear. It makes it more expensive to carry, and somebody pays for that inefficiency.

What a players' association has to trade to win guarantees

Guarantees are bought. No league has ever handed them over, and no union has ever won them without giving up something a member could feel.

The currency is finite. A collective agreement is a package, and every clause is priced against every other. A union pushing for guaranteed contracts is pushing against a fixed pot, so the guarantee comes out of the share of revenue, or the cap ceiling, or the minimum salary, or the pension, or the number of games, or the discipline procedure. Owners will trade almost anything for cost certainty and they know exactly what certainty is worth to them.

The harder problem is internal. A union represents its median member, and the median member is not the star. Guarantees overwhelmingly benefit players signing large multi-year deals, who are a minority. A player on a short contract at the bottom of the roster benefits from churn, because churn is how he got his job in the first place, and a league in which every contract is guaranteed is a league with fewer openings. Ask a union to spend its whole negotiating budget on protecting the top of the market and a lot of hands do not go up.

That is why unions in guarantee-poor leagues tend to bargain for things the middle can use: minimum salaries, benefit accrual, injury protection extending past the end of a contract, guaranteed pay during rehabilitation, and rules limiting how easily a club can void a deal for conduct. Those are guarantees by another name, distributed towards the people who need them.

The other lever is the alternative to bargaining. Antitrust litigation, a grievance, decertification, a strike, a lockout weathered. All of them cost the players money now for a structural gain later, and all of them require a membership with short careers to accept present pain for a benefit that will arrive after most of them have retired. That is an unusually hard sell, and it is the reason a union of athletes is structurally weaker than its members' scarcity would suggest.

Ice hockey supplies the strangest twist on the whole subject. Contracts there are guaranteed, and the collective agreement also requires players to hold back a share of their pay in escrow so the agreed split of revenue between players and clubs holds once the season's revenue is known. A guaranteed contract whose final value depends on how much money the league made is a guarantee with a footnote, and it is a reminder that the word describes a range of promises rather than a single one.

The five things to check on any contract

Next time a signing is announced, ignore the total and find these.

The guaranteed-at-signing figure, which is the only number in the report that is certain to be paid. If it is not published, assume the reason is that it is embarrassing.

The signing bonus, because it is money that has already moved and cannot be recovered.

The vesting dates, which tell you when the club has to decide and therefore when the player is genuinely at risk.

The offset language, which determines whether a released player is paid once or twice, and which is the clause an agent will have fought hardest over.

The first realistic out, calculated as the remaining guarantees plus unamortised bonus in each future year. Find the season where that number drops below what a replacement would cost, and you have found the real end of the contract, whatever the term says.

Do that for two or three deals and the reported totals stop being interesting, which is the correct response to a number designed to be repeated rather than examined. The rest of the archive on how sport actually pays for itself sits in the multi-sport section, where the same argument about who carries the risk turns up in every league with a different answer.