Explainer
NBA trade rules explained: salary out buys salary in
The NBA trade rules explained in detail: the salary matching bands, what each apron actually stops, trade exceptions, the Stepien rule and the deadline.
By CricketTaken EditorialPublished Explainer19 min read
Type two rosters into a trade machine and the reason it refuses the deal is almost never talent. It is that the two salary columns do not agree with each other. Somebody has proposed a fair basketball exchange and the league office would reject it inside a minute, because a trade in this sport is not a swap of players. It is a purchase, and the currency is outgoing salary.
That single idea carries most of what follows. The NBA trade rules explained properly are a sequence of arithmetic tests applied in a fixed order, and once you know the order you can look at any rumoured deal and work out in about thirty seconds whether it can legally exist. The tests have changed since 2023, several of them significantly, and a great deal of what circulates about them describes the previous agreement or an imagined version of the current one.
The soft cap underneath all of this, the exceptions, the tax bands and the way team salary is assembled, is set out in the piece on how the NBA salary cap actually works. This one is about what happens when two clubs try to move contracts across the line between them.
- 4Ways an over-the-cap team may match salary
- 7Transactions tied to the first apron
- 4Further transactions tied to the second apron
- 2Months a newly acquired contract cannot be combined with another
Counts taken from Article VII of the collective bargaining agreement. The restricted transaction list has eleven rows in total; seven are tied to the first apron and four to the second.
Outgoing salary is the currency, not the player
A club under the salary cap has it easy. It absorbs the incoming contracts into its room, plus a $250,000 allowance on top, and nothing else needs to happen. It does not have to send anybody back. This is the entire reason a rebuilding team with space is useful to everybody else: it is the only kind of club that can take a contract without giving one up.
Above the cap, that route closes. An over-the-cap team acquiring a player must point to a specific written permission that allows it, and for trades that permission is the traded player exception. The size of the exception is set by what the team is sending out. Send out nothing, acquire nothing.
Two consequences follow that people find counter-intuitive.
The first is that the test is applied to each side separately. A trade is not checked as a whole. Each acquiring team's post-trade salary is examined on its own terms, which means a deal can be perfectly legal for one club and impossible for the other. Most collapsed rumours die here: the star's team can do the deal, the team receiving the star cannot.
The second is that a player's basketball value is irrelevant to the arithmetic. A twenty-year-old on a rookie deal and a declining veteran on the same figure are identical inputs. This is why the third team appears. A club with room, or with a large contract nobody wants, gets paid in draft capital simply for standing in the middle of the transaction and making the columns balance.
There is a further wrinkle that matters once the aprons come into it. The number tested against a cap or a tax line is team salary. The number tested against an apron is apron team salary, which is calculated differently: it adds in bonuses that team salary excludes, adds the value of an outstanding qualifying offer to a restricted free agent, and takes out the cap holds for unsigned free agents and the charges for empty roster spots that team salary carries. Two clubs with identical payrolls on a public database can sit on opposite sides of an apron.
The four ways an NBA trade can be made to balance
The agreement gives an acquiring team exactly four routes, and they do not stack. Picking one closes the others for that transaction.
Cap room. A team under the cap takes back up to its room plus $250,000. It may not combine this with any of the exceptions below in the same trade, which sounds like a technicality and is not. A club with $8m of room and a $12m trade exception cannot add them together to absorb a $20m contract. It picks one.
The standard traded player exception. One contract goes out, one or more come back, and the incoming total may not exceed one hundred per cent of the outgoing player's pre-trade salary plus $250,000. This is the plain one-for-one route, and it is the only route that remains open to every team in the league regardless of payroll.
The aggregated standard exception. Two or more contracts go out simultaneously, and the incoming total may not exceed one hundred per cent of their combined salaries plus $250,000. Combining salaries is what the agreement calls aggregation, and it is treated as a separate and more restricted permission than sending out one contract, which is the change most people have not absorbed.
The expanded traded player exception. This is the wide band, and it is available only to a team that will finish the trade below the first apron. It allows the greater of two calculations: either the lesser of two hundred per cent of outgoing salary plus $250,000, or outgoing salary plus a fixed sum of $7.5 million indexed to the cap; or one hundred and twenty-five per cent of outgoing salary plus $250,000.
That last sentence is a mess to read, which is why almost nobody quotes it correctly. It resolves into three clean bands.
The three bands, and where they break
Work the formula through and the "greater of the lesser of" language collapses into something a person can hold in their head.
For a small outgoing salary, doubling it is the most generous answer, so the rule pays out two hundred per cent plus $250,000. For a large outgoing salary, a flat $7.5m of headroom is worth less than a percentage, so the rule pays out one hundred and twenty-five per cent plus $250,000. In the middle, the flat sum wins, and the team may take back the outgoing salary plus that indexed $7.5 million.
The two breakpoints fall out of the arithmetic rather than being written down. The first sits $250,000 below the indexed sum. The second sits at four times that gap. In the agreement's base season, where the indexing fraction equals one, that puts the crossovers at $7.25m and $29m of outgoing salary, and both rise with the cap thereafter.
Notice the shape. The bands are deliberately generous at the bottom and deliberately tight at the top. A team can turn $5m of salary into $10m, which is how contenders convert a spare rotation contract into a real player. A team cannot turn $40m into anything like $80m, which is how the agreement stops the largest contracts being used as blank cheques.
The choice between the standard exception and the expanded one is not automatic, and this is a genuinely subtle piece of front office work. The expanded exception sits on the restricted transaction list. Using it caps the team at the first apron for the rest of the league year. A club that expects to spend into the tax later may deliberately take the narrower hundred per cent match instead, and pay for the privilege in incoming salary, to keep the apron ceiling off its own books.
The $250,000 that disappears when you cross a line
Buried in the same section is a rule that is short, easy to miss, and does more day-to-day work than most of the apron provisions people argue about.
If a team's apron team salary immediately after the trade would exceed the first apron, the $250,000 allowance in every one of those routes is reduced to zero.
So for any club finishing above the first apron, matching is not "roughly a hundred per cent". It is one hundred per cent, to the dollar, and a hundred per cent is a ceiling rather than a target. Take back a single dollar more than you send out and the deal is void. Front offices deal with this by building trades around contracts whose salaries are already known to line up, or by adding a minimum-salary player to the outgoing side purely as ballast.
It also explains a pattern that looks like coincidence in reporting. Deals involving expensive teams tend to arrive with an odd extra name attached, someone nobody was discussing, on a small contract. That player is not a sweetener. He is arithmetic.
- Split the trade in two and test each acquiring club separatelyThe rules are applied to each side on its own numbers. A deal can be entirely legal for the club sending the star and impossible for the club receiving him, which is where most rumoured trades die.
- Is the acquiring club under the salary cap afterwards?If so, the incoming contracts are absorbed into its room plus $250,000, and it need not send anybody back. Room cannot be combined with any trade exception in the same deal, so the club picks one route and loses the other.
- If not, it must buy the incoming salary with outgoing salaryAbove the cap there is no absorption and no cheque that fixes it. The size of the permission is set entirely by what goes the other way.
- How many contracts is it sending out?One is the standard exception and is open to every club in the league. Two or more is aggregation, which the agreement treats as a separate and more restricted permission tied to the second apron.
- Will its apron team salary finish below the first apron?If so, the wide bands are available: two hundred per cent plus $250,000 on small contracts, outgoing salary plus an indexed $7.5m in the middle, one hundred and twenty-five per cent plus $250,000 at the top. Using them caps the club at the first apron for the rest of the year.
- Will it finish above the first apron?The $250,000 allowance falls to zero and the wide bands close. Matching becomes one hundred per cent of outgoing salary exactly, and a single dollar over voids the transaction.
- Will it finish above the second apron?Salaries may no longer be combined at all, and no cash may be sent. One contract out buys one or more contracts in, at one hundred per cent or less, and that is the whole toolkit.
- Now test every player against the calendarRecently signed, recently extended, recently acquired or recently given a raise above twenty per cent: each carries its own waiting period, and one frozen player invalidates the entire structure.
- Last, count the first-round picksIf the selling club would be left without its own first-rounder in two consecutive future drafts, the league by-laws refuse the deal whatever the two front offices have agreed.
The sequence follows Article VII of the agreement. Every dollar amount named here is a fixed allowance written into the rules rather than an example, except the $7.5m, which is the base figure and rises with the salary cap.
The aprons are not hard caps until you make them one
Here is the part that is most often described wrongly, including by people who should know better.
Neither apron is a hard cap in the way the NFL's ceiling is a hard cap. There is no rule saying a team's payroll may not exceed either line. A club above the second apron can re-sign its own free agents for any amount its Bird rights allow and can sign anyone to a minimum contract, and its payroll can climb without limit while it does so.
What the aprons actually do is govern methods, in two directions at once.
The first direction is a gate. A team may not carry out a listed transaction if, immediately afterwards, its apron team salary would exceed the line that transaction is tied to. That is a straightforward prohibition, tested at the moment of the deal.
The second direction is a latch, and it is the one that catches people. A team that carries out a listed transaction may not, for the remainder of that league year, have an apron team salary above the corresponding line. The team has hard-capped itself, by its own choice, by using a particular tool. Nobody imposed the ceiling on it in July. It picked the tool up in August and the ceiling came attached.
There is a third rule that operates in the offseason and almost never gets described at all. Between the end of the regular season and the end of the league year, several of the listed transactions are tested against the following season's apron level as well as the current one, with options assumed exercised and a set of other conservative assumptions applied. A club cannot use a tool in June on the basis that it will get under the line by October.
And one more, which is a straightforward trap: a team that signs anybody using the taxpayer mid-level exception is barred for the rest of that year from the whole first-apron block of transactions, including the wide salary matching bands. One signing in July can quietly remove the club's best trade route in February.
What each apron takes off the table
The restricted list runs to eleven rows. Seven are tied to the first apron and four to the second, and the second-apron restrictions sit on top of the first rather than replacing them.
Tied to the first apron: the bi-annual exception, the non-taxpayer mid-level exception, acquiring a player through a sign-and-trade, signing a player during the season whose terminated contract had paid more than the non-taxpayer mid-level, using the expanded matching bands, using a trade exception left over from an earlier season, and a transitional exception that applied for a single year only.
Tied to the second apron: aggregating two or more salaries in a trade, sending cash to another team, using a trade exception that was created by a signed-and-traded contract, and signing anybody with the taxpayer mid-level exception.
Read that second list again and the design becomes obvious. Above the second apron a team keeps exactly one way to acquire a player it did not draft or already employ: send out one contract, take back no more than that contract was worth. Every method of building a bigger salary out of smaller ones is gone. So is the ability to pay another club to take a problem off its hands. The roster can still be expensive. It just cannot be reshaped.
The draft consequences of finishing above the second apron, including the pick that is frozen years out and what repeat seasons do to it, are covered in the piece on the line that takes tools rather than money, and they belong to the by-laws rather than to this arithmetic.
- Outgoing contract A14m
- Outgoing contract B9m
- Outgoing contract C6m
Constructed round numbers, not a real trade. A club sending out three contracts and taking back one is aggregating, which is a second-apron transaction, so it must finish under that line and stay under it all year. Above the first apron the $250,000 allowance is zero, so the incoming contract may not exceed $29m.
Show the numbers
| Item | Value |
|---|---|
| Outgoing contract A | 14m |
| Outgoing contract B | 9m |
| Outgoing contract C | 6m |
A trade exception is a receipt, not a wallet
When a team sends out more salary than it takes back, the difference does not evaporate. It becomes a traded player exception: a written permission to acquire, later and without sending anything back, up to that amount of salary plus the usual allowance.
Four things about them are routinely got wrong.
They are not money. A trade exception cannot be used to sign a free agent, cannot be split across two separate trades in most useful ways, and cannot be combined with another exception or with cap room to absorb one bigger contract. It buys contracts, one transaction at a time, up to its size.
They expire in one year. The window runs twelve months from the date of the original trade, not to the end of the season and not to the following deadline. A large exception created in a February deal is dead the following February, and the second half of its life is the useful half only if the team is still contending.
They do not shrink. From 10 January most unused exceptions are reduced a little every day for the rest of the regular season, on a pro-rata basis. Trade exceptions are specifically excluded from that decay, along with minimum-salary and disabled-player exceptions. A trade exception is worth on its last day exactly what it was worth on its first.
And a team with meaningful cap room cannot hold one at all. The agreement only lets a club use these exceptions if its team salary is at or above the cap, or below it by less than the exception is worth. Go far enough under the cap and the exception is extinguished, absorbed into the room it duplicates. This is the trade-off every rebuilding club has to price: use the space, or keep the exceptions, never both.
The first apron then adds one more limit. Using a trade exception that arose in an earlier season is a first-apron transaction. An expensive team can bank a large exception in June and find it unusable in December.
Draft picks, the Stepien rule and trading in alternate years
Nothing in the labour agreement governs the trading of draft picks. That sits in the league's by-laws, and the operative restriction is named after the owner whose spending spree in the early 1980s produced it.
A club cannot trade away its own first-round picks in consecutive future drafts. It has to be left holding a first-rounder in at least every other year. The practical shape of this is immediately visible in any list of a team's outgoing obligations: picks go out in alternating years, and the gaps are not accidents of negotiation but the by-law drawing a line.
Three habits grow out of it.
Pick swaps exist because a swap is not a trade of the pick itself. A club that has already committed the picks it can commit can still give away the right to exchange positions in a draft, which transfers most of the downside risk without transferring the selection. In a deal for an established player, swap rights are often the last thing on the table because they are the last thing still available.
Protections exist because a conditional obligation is easier to sell internally than an absolute one. A pick protected in the top four conveys only if it falls outside that range, and the trade document specifies what happens if it never conveys at all, usually turning into second-rounders after a defined number of years. This is also why where a team lands in the lottery can determine whether an obligation agreed three years earlier is discharged or rolls forward again.
And the rule binds the seller as much as the buyer. A rebuilding club cannot cash in every first-rounder it owns at once even if it wants to. The by-law forces it to keep a floor under itself, which is the entire point: it exists to stop a team mortgaging a decade in a single summer and leaving the league with an unwatchable franchise it cannot fix.
The poison pill nobody notices until February
A player who signs an extension of his rookie-scale contract creates an asymmetry that has killed more midseason deals than any apron.
Until the extension actually starts, he is still being paid his rookie-scale salary, which may be modest. When his team trades him, its outgoing salary is that modest figure, because outgoing salary is his current salary.
For the acquiring team, the agreement deems his salary to be the average of his final rookie-scale year and every year of the extended term. If the extension is a large one, that average can be several times what the trading team is sending out. The trading club is spending pocket change and the acquiring club is being charged a fortune, on the same player, in the same transaction.
Nothing about this is a penalty. It is a consistency rule: it stops a club from signing a player to a back-loaded extension purely so that another club can absorb him cheaply, and it stops the extension itself being used to disguise the real cost. But the practical effect is that a recently extended young player is close to untradeable for several months, unless the acquiring team happens to have enormous cap room, and that window closes on the first day of the following league year when the extension takes effect and the two numbers become the same again.
Bird rights survive a trade, except when they do not
The right to re-sign your own player over the cap is the most valuable thing a team owns that does not appear on any balance sheet, and the trade rules treat it carefully.
The default is generous. A player who changes teams by trade keeps the continuous service that generates those rights. Three consecutive seasons without signing elsewhere as a free agent produces full rights, and a trade does not interrupt the count. His new club inherits the ability to exceed the cap to keep him. A waiver claim is treated less kindly, and only preserves the count if it happened in the first of the three seasons.
Three complications sit on top of that.
A player on a one-year contract who would hold full or early rights at the end of it has a veto. He cannot be traded without his consent, unless he agreed at signing to give that right up. And if he does consent, the agreement then treats him, for the purposes of his next contract with his new team, as though he had arrived as a free agent rather than by trade. His rights are gone. The veto exists precisely because exercising it costs him something real, and clubs increasingly ask for the waiver at signing rather than negotiate for it in February.
A sign-and-trade has its own list of conditions, all of which have to be satisfied at once. The player must have finished the previous season on that roster. The contract must run at least three seasons and no more than four. Its first season must be fully protected for lack of skill. It has to be signed before the regular season begins. It cannot be signed using the non-taxpayer or room mid-level exceptions. And the acquiring club must have room for the first-year salary, which caps the whole manoeuvre at the first apron. The mechanics of the rights being exercised on the other side of that deal are in the guide to how free agency actually works.
And there is an anti-inflation rule aimed squarely at sign-and-trades. Where an over-the-cap team re-signs its own qualifying free agent and immediately trades the new contract, its outgoing salary for matching purposes is deemed to be the greater of his previous salary or half of his new first-year salary. Giving a player a large raise on the way out of the door does not create matching capacity out of thin air. This is the modern replacement for what used to be called base-year compensation, and it is narrower and cleaner than the old version.
The calendar is part of the trade rules
A trade that is legal in January can be illegal in November, on identical numbers, because a set of waiting periods sits underneath everything else.
A drafted rookie who has signed, or any player who has signed a two-way contract, cannot be traded for thirty days.
A free agent who signs a standard contract cannot be traded until the later of three months from signing or 15 December. This is the reason the trade market genuinely opens in mid-December rather than in October, and the reason a club that has just spent its exceptions in July has nothing to move for five months.
A player re-signing with his own team, above the cap, on a raise of more than twenty per cent over his previous salary, is frozen until the later of three months or 15 January. Minimum contracts are exempt.
A player who signs a long veteran extension or renegotiates his deal cannot be traded for six months, and the rule runs in both directions: a club that acquires a player cannot give him that kind of extension for six months either. A designated veteran extension locks the player in place for a full year.
Then there is the aggregation clock, which is the one that decides what a deadline day actually looks like. A contract acquired in a trade cannot be combined with another contract for two months. So a club that makes a December deal has, by default, disarmed its own aggregation ability until February. The agreement then carves that back out: if the contract was acquired on or before 16 December, the restriction lifts from the day before the deadline. December acquisitions are deliberately kept tradeable in February, and January acquisitions are deliberately not.
One more, easy to miss and occasionally decisive. Where a team is combining three or more outgoing contracts and taking back fewer players than it is sending, no more than one of those outgoing contracts may be a minimum-salary player. That restriction is switched off between 15 December and the deadline, which is exactly when consolidating three role players into one rotation player is a thing clubs want to do.
What is actually being counted at the deadline
The league fixes the deadline date each season and it falls in February, weeks before the regular season ends. After it, no player may be traded until the following league year. The market does not slow down, it stops.
Three things behave differently as it approaches.
Unprotected salary counts for less. For matching purposes, a traded player's salary is reduced by whatever part of his remaining base compensation is not fully protected against injury and lack of skill. A player on a large but unguaranteed deal is worth far less as outgoing salary than his headline number, which is why some contracts that look like perfect matching pieces are useless. From 8 January this reverses: salary is deemed fully protected for the rest of the regular season regardless of what the contract says, and those players suddenly become tradeable at their full value. The deadline market and that date are not unconnected.
Expiring contracts stop moving. A team cannot trade a player after the deadline that falls in the last season of his contract, or in any season that could be his last depending on an option. It is one of the few provisions in this area with no exception attached.
And the buyout market has its own clock. A player waived after 11:59pm eastern on 1 March cannot play in the play-in or the playoffs for anyone that season, save for a narrow injury provision. So the useful buyouts happen in the days after the deadline and before that date, and a contending club's last chance to add anybody at all is a two-week window it does not control.
Cash sits alongside all of this as a quiet constraint. A club may pay or receive a limited aggregate amount across all of its trades in a year, set at 5.15 per cent of the salary cap, and money paid and money received both consume the same allowance rather than cancelling out. A team that has spent its cash budget buying second-round picks in June cannot be paid to absorb a contract in February.
Four checks to run on any trade you read about
Next time a deal is reported, or proposed to you as obvious, run these in order.
Take each acquiring team separately and ask where its apron team salary lands afterwards. Below the first apron, the wide bands are available and the deal probably works on salary. Above it, matching is exactly one hundred per cent. Above the second apron, ask whether the club is combining contracts, because if it is, the trade cannot happen at any price.
Count the outgoing contracts on each side. One is unrestricted. Two or more is aggregation, which is a second-apron transaction and carries a ceiling for the rest of the year.
Check the dates on every player in the deal. Recently signed, recently extended, recently acquired, recently given a big raise: each has its own waiting period, and one frozen player invalidates the whole structure.
Then look at the picks. Count forward through the first-rounders the selling club still controls, and if two consecutive years would be gone, the by-law says no, whatever the two general managers have agreed.
Most of what circulates fails on the first check. A good deal of the rest fails on the third, which is the one nobody makes. And if you want to see how differently the same problem gets solved when a league has no aprons at all but a much smaller pool of money, the comparison with basketball's other salary system is worth an hour. The rest of the archive's work on the sport sits under basketball, including the pieces on what a maximum contract can actually be worth and on the contracts that do not count against the cap at all.
Common questions
How does salary matching work in an NBA trade?
A team already over the salary cap cannot simply absorb an incoming contract. It has to send salary out, and the amount it sends out determines the amount it may take back. A team that will finish below the first apron gets the widest bands, roughly double the outgoing salary on small contracts and a quarter more on large ones. A team that will finish above the first apron may take back no more than one hundred per cent of what it sends, exactly.
What is an NBA trade exception?
A trade exception is the unused difference left over when a team sends out more salary than it takes back in a single deal. It is not money and it cannot be spent on a free agent. It is a permission to absorb one or more contracts later without sending anything else out, and it lasts one year from the date of the original trade. Teams above the first apron cannot use one created in an earlier season.
What is the Stepien rule in the NBA?
The Stepien rule is a league by-law that prevents a club from trading away its own first-round draft picks in consecutive future drafts. In practice it means a team can deal a first-rounder only in alternate years, so that it never faces two drafts in a row with nothing of its own to select with. It is the reason so many deals are built out of pick swaps and protections rather than clean picks.
Can a team above the second apron make trades?
Yes, but only simple ones. A second-apron team may send out one contract and take back one or more, provided the incoming salary does not exceed the outgoing salary. What it cannot do is combine two or more of its own contracts to match a single larger one, send cash in a deal, or use an exception created by a sign-and-trade. That is why big second-apron rosters tend to break up in one direction only.
When is the NBA trade deadline and what happens after it?
The league fixes the date each season and it falls in February, several weeks before the end of the regular season. After it passes no team may trade any player until the following league year, so the only route to a new player is the waiver market. A player waived after 11:59pm eastern on 1 March is ineligible for that season's play-in and playoffs with any team, which is what makes the days before that date busier than they look.
Filed under Basketball·nba · trades · salary cap · contracts · roster building · cba