NFL salary cap carryover: the money that does not expire

Open any list of cap space in March and the figures look like they were produced by subtraction: one league cap number, thirty-two sets of commitments, thirty-two answers. They were not. A sizeable part of what separates the club at the top of that list from the club at the bottom is a letter, signed by an owner, filed in January, that most people watching football have never heard of.

NFL salary cap carryover is that letter. It is the mechanism by which room a club did not use in one season becomes room it may use in the next, and it is the single reason a league with a cap that is identical for every club produces thirty-two different working ceilings.

The agreement is blunt on the first half of that. Article 13 opens by saying the salary cap is the same amount for each club, then immediately that no club may have a team salary exceeding it. Two sentences, one ceiling, thirty-two identical numbers. Everything after that is accounting, and carryover is the largest single piece of it.

What NFL salary cap carryover actually is

The provision lives at Article 13, Section 6, and it is short enough to describe exactly.

A club may designate an amount of its current league year room to be credited to its team salary for the next league year, by giving the league written notice signed by the owner. The notice may name a specific dollar figure or a percentage of the room available. It must be received before four in the afternoon, New York time, on the day following that club's final regular season game. The agreement calls that moment the Carry Over Measuring Date. If a club designates more than it turns out to have, the notice is automatically amended down to the whole of its room rather than being thrown out.

Then the mechanism, which is the part worth slowing down for. The designated amount is charged against the current year's team salary and credited to the next year's, and the two happen together at the deadline.

Nothing physical moves. No money is set aside, no account is funded, no cheque is held. The club voluntarily spends its remaining room on nothing at all, in a year that is finishing, in exchange for the same quantity of room in a year that has not started. Carryover is not a savings account. It is a cap charge taken deliberately in order to buy a cap credit.

That framing explains two things that confuse people. It explains why carryover has no effect on what an owner has actually paid out, and it explains why a club with an enormous carryover credit can still be short of cash room to sign anybody, because the credit is permission rather than funding.

• 32Clubs that receive the identical salary cap figure
• 51Contracts counted in full before the regular season
• 100Share of unused room a club may carry, per cent
• 90Minimum team cash spending across a period, per cent

All four are set by the 2020 collective bargaining agreement. The spending figure is the minimum team cash spending requirement applying to the 2021 to 2023, 2024 to 2026 and 2027 to 2030 periods.

The deadline is not the same day for every club

The Carry Over Measuring Date is defined as the day following the club's final regular season game. Not the day after the season ends. Not a fixed date on the calendar. The day after your last game.

Regular season games are played across several days, so the measuring dates are staggered by whichever slot a club drew in the final week. A club finishing on a Saturday has an earlier deadline than one finishing on a Sunday, which has an earlier deadline than one finishing on a Monday night. The window is a day wide and it opens at a different hour for different clubs.

Two consequences follow, and both are more interesting than the trivia sounds.

The first is that the measurement happens before the postseason. Whatever a club does in January, whichever players it signs off the street for a playoff run, whichever elevations it makes, none of it touches the number that was measured the morning after the regular season ended. The carryover figure is a snapshot of a season that is over, taken while another one is still running.

The second is that the notice must be signed by the owner. Not the general manager, not the club's capologist, not the head of football administration. The agreement specifies the owner's signature on a document that has to arrive by a stated hour on a stated day. That is deliberate friction: it makes the election an act of the club rather than a routine filing, and it makes forgetting it expensive in a way that is nobody's fault but the club's.

It is worth stating the default plainly, because it is the opposite of what most people assume. Unused cap room does not roll forward on its own. If nothing is filed, nothing carries. The room simply ceases to have been.

1. 1The club plays its final regular season gameThe measuring date is the day after that game, whichever day of the week it happens to be. Anything the club does in the postseason arrives too late to affect the calculation.
2. 2Room is measuredRoom is the gap between the club's team salary and the cap at that moment, built up from a season of injuries, releases, unclaimed game cheques and bonuses that were charged but never earned.
3. 3The owner signs the noticeNothing carries over by default. The club must file written notice, signed by the owner, naming either a dollar figure or a percentage of the room it has available.
4. 4The notice must arrive before four in the afternoonNew York time, on the measuring date. A designation larger than the room actually available is automatically amended down to the whole of the room rather than being refused.
5. 5The charge and the credit happen togetherThe designated amount is charged against this year's team salary and credited to next year's. No cash is set aside; only the permission to spend moves.
6. 6The incentive settlement runs on what is leftUnearned likely bonuses net against earned unlikely ones. A club that carried everything has no room left to absorb an overage, so the overage lands on next year's books instead.
7. 7The credit shows up in MarchWhen the new league year opens, the club's working ceiling is the identical league cap figure plus its credit, less any charge. Thirty-two clubs, one cap, thirty-two different ceilings.

The sequence is the same for every club, but the first step falls on a different day depending on when each club played its final regular season game.

Where the leftover room comes from in the first place

Nobody sets out to reach January with room to spare. Almost every dollar of carryover is the residue of four distinct processes, and only one of them is deliberate.

Room preserved on purpose. A club that knows next spring is the one where it has to sign a quarterback, or that its own good young players hit free agency in eighteen months, will decline to spend now. This is the version that gets written about, and it is usually the smallest of the four.

The top 51 rule, and what it hides. Between the start of the league year and the first day of the regular season, only the 51 highest valued contracts on a club's books count in full. Players outside that group are not free, but they count only for amounts above a threshold: for an undrafted rookie, the part of his salary exceeding the minimum active or inactive list salary, and for everyone else, the part exceeding twice that minimum. Rosters through the summer run far beyond 51, so a large number of players are contributing very little or nothing to the club's team salary while the offseason ceiling is being reported.

In-season attrition. Players get released. Players get injured and replaced by cheaper players. Practice squad members are signed and returned. Every one of those transactions changes the running total, almost always downwards, and the accumulated effect across seventeen weeks is generally larger than anything a club planned.

Bonuses that were charged and never earned. Incentives classified as likely to be earned are charged against the cap during the season whether or not the player achieves them. When he does not, that money was never spent, and it is sitting in the room at the end of the year.

46%33%21%

• Savings from in-season releases and injury replacements11
• Room preserved on purpose for next spring8
• Likely to be earned bonuses that were charged but not earned5

Invented figures in index points, used to show the shape rather than any club's actual position. The largest slice is usually the one nobody chose.

Show the numbers

Worked example: where one club's leftover room came from

Item	Value
Savings from in-season releases and injury replacements	11
Room preserved on purpose for next spring	8
Likely to be earned bonuses that were charged but not earned	5

The top 51 rule deserves one more sentence because of what it does in the other direction. When the regular season begins, the accounting flips and every contract on the roster counts. Injured reserve counts. The practice squad counts. A club whose books looked comfortable in July can find itself tight in the second week of September without having signed anybody, purely because the counting rule changed underneath it. That is also why a club's carryover is measured in January rather than in August: the January figure is the honest one, taken while the strict counting rule is in force.

Room is a residual, not a balance

The word "room" invites the wrong mental picture. It sounds like a balance in an account, drawn down by purchases. It is a subtraction, and what is subtracted is a long list, most of which was fixed long before the season began.

A club's team salary is the sum of everything the agreement makes it count. The paragraph 5 salaries of the players on its roster. The prorated slice of every signing bonus and option bonus it has ever paid, including bonuses paid to players who left years ago. Roster bonuses, reporting bonuses and workout bonuses. Incentives classified as likely to be earned. Franchise and transition tenders it has applied. Offer sheets it has signed. Termination pay owed to players it released.

Look at what is on that list and how little of it moves after March. Proration is fixed by contracts already signed. Tenders are set in the spring. Dead money from past releases is a schedule, not a decision. Roster bonuses are dated. The likely-to-be-earned classification is made before the season starts and does not change during it.

The genuinely variable component is small: the salaries of players who arrive and leave, and the difference between what a departing player was costing and what his replacement costs.

Which means a January carryover figure is mostly a report on decisions taken in March and August, plus whatever the season did to the roster. It is not an achievement, and it is not a plan. It is a residual, and treating it as evidence of financial discipline is reading the wrong end of the process.

What a club can and cannot do in the last month

The idea that a club manages its carryover in December collapses on contact with how salary is actually charged.

Paragraph 5 salary is earned week by week across the regular season, in equal instalments, and it is charged as it is earned. A player released with three weeks remaining saves the club three weeks of his salary and nothing else. Fourteen weeks of it have already been paid and already counted. On top of that, any unamortised signing bonus he is carrying accelerates immediately, which frequently costs more than the three weeks saved.

So a late-season release is one of the least efficient ways to create room in professional football, and a late-season signing is one of the cheapest ways to add a player. The same weekly accrual that makes releasing somebody in December pointless makes signing somebody in December almost free, because the club is buying a fraction of a salary.

That asymmetry is why the useful work on a carryover number happens in the spring. A club that wants room in January builds it in March, by declining to commit it, and then watches the season either preserve it or eat it. What December offers is the ability to spend room the club has decided it does not want to carry, and one narrow decision about incentives, which is where the next section starts.

The incentive settlement runs through the same window, and it can go the other way

This is the part of carryover that almost nothing written about the subject covers, and it turns the election into a genuine decision rather than a formality.

The same section of the agreement that handles carryover also handles the annual reconciliation of performance bonuses. Incentives are classified before the season as likely to be earned or not likely to be earned, based on whether the player hit the equivalent target the previous year. Likely ones are charged during the season. Unlikely ones are not.

At the end of the season the two are netted against each other. If bonuses that were included in team salary but not actually earned exceed bonuses that were earned but never included, the difference is credited to the club's team salary for the next league year. If the reverse happens, and earned bonuses resulted in the club paying salary in excess of the cap, the amount by which it exceeded is charged to next year's team salary.

So there are two adjustments travelling into the next league year, and they point in opposite directions. The carryover credit, which a club chooses. And the incentive settlement, which a club does not, and which can be negative.

Now the interaction, which is the reason cap staff care about this in December rather than in March. A carryover designation reduces the room a club is treated as having at the end of the season for the purposes of that netting. Carry 100 per cent of the room and the club's end-of-season room is zero for the settlement. Any overage from earned unlikely bonuses therefore has nothing left to absorb it, and it lands on next year as a charge.

Which means the last dollar a club carries forward is precisely the dollar that would have soaked up an incentive overage. Carrying everything is not free; it converts a problem that could have been settled this year into a debit on next year's books. A club with players sitting on live unlikely-to-be-earned incentives going into the final week has a real calculation to run, and it has to run it before the owner signs.

Carryover is cap, and the floor is cash

Here is the constraint that stops carryover being a strategy on its own, and the agreement is explicit that carryover does not touch it.

Article 12 sets a Minimum Team Cash Spending requirement: a guaranteed floor, measured in cash actually paid to players, of 90 per cent of the salary caps across each multi-year period. The periods are defined in the agreement as 2021 to 2023, 2024 to 2026, and 2027 to 2030, with the earlier 2017 to 2020 period set at 89 per cent. Any shortfall at the end of a period is paid, on or before the following 15 September, by the club with the shortfall, directly to the players who were on its roster at any time during the applicable seasons, allocated as the players' association instructs.

Read the two rules side by side and the tension is obvious. Carryover is a cap mechanism, measured in room. The floor is a cash mechanism, measured in dollars leaving the building. A club that underspends to bank room has, in the same act, moved itself closer to a cash shortfall it will have to settle at the end of the period. And the settlement is not a fine paid to the league, which a club might treat as a cost of business. It is a payment to the specific players it declined to pay at the time.

The agreement closes the door on any argument about this by saying, inside the carryover provision itself, that nothing in it affects the guaranteed league-wide cash spending or the minimum team cash spending obligations. The two systems were written to be independent, and they were written that way because a carryover rule without a cash floor is an invitation to bank room permanently and field the cheapest roster the sport allows.

The multi-year measuring window is what makes the floor bearable. A club is not required to spend 90 per cent every season. It is required to average it across the period, which leaves room for a rebuilding year followed by an aggressive one. The bill arrives at the end of the period, not at the end of the year, which is exactly the rhythm a carryover strategy runs on.

Why a hard cap needs a carryover rule at all

Strip the mechanism away and ask what a hard cap without carryover would produce.

It would produce a use-it-or-lose-it incentive, and use-it-or-lose-it incentives generate the same behaviour in every organisation that has ever had one. A club approaching the end of a season with room remaining would have a reason to spend it on something, anything, because the alternative is watching it evaporate. December would become a month of pointless signings, of guaranteeing salary to players nobody needs, of restructuring contracts for no reason other than to burn room before it disappeared.

Worse, it would punish exactly the behaviour the system wants. A club that identified a strong free agent class one year away and declined to spend into a weak one would be penalised for the foresight.

Carryover removes both problems with one provision. Room retains its value across the boundary, so there is no deadline pressure to waste it and no penalty for patience. What remains is a club making a straightforward allocation decision between this season and the next, which is the decision it ought to be making.

The design also explains why the election is opt-in rather than automatic. Making it a filed, signed, dated act keeps a record of the amount, fixes it at a moment nobody can argue about afterwards, and lets the incentive settlement be calculated against a known figure. Automatic rollover would work perfectly well for the simple case and produce arguments in every complicated one.

The same cap, three different ceilings

The practical effect is that "cap space" in a March report is almost never the league cap minus a club's commitments. It is the adjusted figure, and the adjustment is what makes clubs different.

• League cap
• The club's adjusted ceiling

Carried 14 forward in January100index114index

Carried nothing forward100index100index

Owes 6 from an incentive settlement100index94index

Index numbers with the league salary cap set to 100. The three clubs are invented. Every club in the league receives the identical figure in the first column; the second column is what each one may actually spend to.

Show the numbers

Worked example: one league cap, three working ceilings

Item	League cap	The club's adjusted ceiling
Carried 14 forward in January	100index	114index
Carried nothing forward	100index	100index
Owes 6 from an incentive settlement	100index	94index

Twenty index points separate the top club from the bottom one before either has made a single decision about a player. That gap is not revenue, and it is not an owner's willingness to spend. It is bookkeeping from a season that finished, and it is the closest thing the league has to a legal advantage in cap room.

The gap is also self-limiting in a way the equivalent gaps in other sports are not. A club can only carry what it did not spend, and what it did not spend is dragging it towards the cash floor. A club can only carry it forward one year at a time, since the credit becomes part of next year's room and is then subject to the same election again. And the credit is a one-off while contracts are recurring, which is the trap the next section is about.

A large carryover is not the same as a good cap position

The most common misreading of these figures treats a carryover credit as evidence of health. Often it is. Sometimes it is the opposite, and telling the two apart takes one question.

A credit is a stock. It exists once, it can be spent once, and when it has been spent it is gone. A cap charge is a flow. It recurs every year for as long as the contract runs, and a contract signed today generates charges for four or five years.

Spending a stock on a flow is how clubs get into trouble. A club with a large credit can sign a player whose first-year charge fits comfortably inside it, and then discover that the second year of that contract has to be paid out of ordinary room that was already committed. This is the same arithmetic that makes converting salary into bonus so seductive and so dangerous: both moves create space now by borrowing from a year that will have its own problems.

The honest use of a carryover credit is to pay for something that is itself a one-off. A dead money spike from a release the club has already decided to make. A signing bonus paid in a year when the club has room and taken as proration in years when it will not. The cost of carrying an extra injury replacement through a season. Those are stocks meeting stocks, and the arithmetic closes.

The dishonest use is to treat it as a raised ceiling and sign recurring commitments against it. A club that does that is in the same position, one year later, minus the credit, plus the contracts. The full picture of how charges accumulate and accelerate is the context this decision sits in, and carryover is the smallest of the levers available in it, not the largest.

What carryover is actually good for

Four uses justify the mechanism, and they are the ones front offices genuinely plan around.

Timing a free agency push. The most valuable thing about carryover is that it lets a club match its spending to the market rather than to the calendar. Free agent classes are uneven. A club that banks room in a thin year and spends it in a deep one is buying the same money at a better price.

Paying for a decision already taken. A club that knows it is releasing an expensive player next March knows the dead money charge in advance, to the dollar. Carrying room forward to cover it is the cleanest possible use of the mechanism, because the liability is fixed, dated and certain.

Absorbing the September accounting change. Room measured under the top 51 rule overstates a club's comfort. Carryover credit is one of the things that stops the transition to full counting being painful, and it is why clubs with thin rosters and heavy injured reserve usage value it more than the headline suggests.

In-season flexibility. Room during the season buys replacements, elevations from the practice squad, and the ability to claim a useful player off waivers rather than watch him go somewhere else. It is unglamorous and it decides a surprising number of Novembers.

There is a fifth use that is really a defensive one. A club going into a season with players on rookie contracts approaching the end of their fixed terms knows the price of keeping them is about to jump. Room banked in advance of that is room that does not have to be found by cutting somebody. The fixed four-year pricing of drafted players makes the timing of that jump unusually predictable, which makes it unusually easy to save for.

What the mechanism does not do

Three limits are worth stating precisely, because each of them gets asserted the wrong way round somewhere every March.

It does not create money. A carryover credit raises the ceiling. It does nothing about whether an owner is willing to write the cheque underneath it. Cap room and cash are different currencies and clubs differ in both, independently.

It does not remove a commitment. Room carried forward sits alongside every charge already scheduled for next season. A club with a large credit and a larger set of committed charges is not in a better position than a club with neither, and the credit will be consumed by the difference before free agency opens.

It does not survive being spent twice. The credit becomes part of next year's room and is then subject to the same election, so in principle room can travel forward again and again with no expiry written into the provision. In practice it does not accumulate for long, because the cash floor is pulling the other way and because unused room is an asset earning nothing.

The comparison with other leagues is instructive here without needing any arithmetic. Basketball's system does not need a carryover rule because its cap is soft and full of exceptions, so a club's ability to spend is governed by which exceptions it holds rather than by a ceiling it might undershoot. A hard cap is the structure that creates the use-it-or-lose-it problem in the first place, and carryover is the patch. The mechanism exists because the cap is genuinely hard, not in spite of it.

Reading a carryover figure properly

Four questions turn a reported number into information.

How much of next year is already committed. A credit is only meaningful against the charges it has to cover. A club carrying a large credit into a year where most of its cap is already spoken for has less usable room than a club carrying nothing into a clean year.

Whether the credit is a plan or a symptom. Room left over because a club deliberately declined to spend is a strategy. Room left over because a club could not sign the players it wanted, or lost half its roster to injury, is a report on a season that went wrong.

Where the club sits in the spending period. The cash floor is measured across defined multi-year blocks, and a club that has underspent through the early years of one is under real pressure in the final year of it. That pressure is invisible in a cap number and completely visible in a cash one.

Whether there is a charge travelling the other way. The incentive settlement can arrive as a debit. A club that carried its entire room forward and had players earning bonuses nobody expected has both a credit and a charge landing in the same March, and only one of them makes the headlines.

None of those four requires access to anything private. The rules are in the agreement, the periods are defined in it, and the club-by-club figures are published every spring. What they will not tell you is the thing everyone wants to know, which is whether a club is about to spend. Carryover is permission, not intention, and a great many clubs have arrived in March with an enormous ceiling and no plans whatsoever.

More on how the money works, in this sport and others, is collected in the American football archive.