NBA Cap, Tax, and Apron Outlook 2026-27 through 2030-31

League-wide reference for the BOS-WAS-UTA trade analyses

Date: 2026-06-10 Status: Current. Supersedes nothing; CONFIRMS the threshold table embedded in data/okc_luxury_tax_projection_2025_31.csv and data/bos_tatum_trade_cap_2026_31.csv. Companion data: data/trade_prep_20260610/cap_thresholds_2025_31.csv, data/trade_prep_20260610/cba_trade_rules_reference.csv

---

1. Conclusion first

The repo's threshold table is confirmed, not superseded. But the consensus "10 percent every year" growth story is already broken: the first cap year under the new national TV deal (2026-27) came in at +6.69 percent, dragged down by local media revenue. Every 2027-28+ row in this table is a C-tier ceiling, not a central estimate. Plan trades against the 2026-27 numbers (B tier, league memo) and treat out-year apron distances as soft by roughly 5 to 7 million per year of sub-10-percent growth.

For the BOS-WAS-UTA shape specifically: none of the three teams projects as an apron team in 2026-27, so the binding constraints are the matching formulas and the hard caps a trade ATTACHES, not pre-existing apron restrictions. A trade executed before July 1 is also tested against next season's aprons under the offseason lookahead rule (measured, CBA Art VII Sec 2(e)(2)(ii)).

---

2. Threshold table 2025-26 through 2030-31

Evidence:

• 2025-26: official NBA setting of 2025-06-30. cba-knowledge/sources/cba-2023-changes.md Section 1; independently confirmed by SalarySwish (WebFetch 2026-06-10). Measured.
• 2026-27: NBA memo to teams, 2026-03-24, reported by Shams Charania. Cap 165M, floor 149M, tax 201M, apron1 209M, apron2 222M. Captures: data/raw/cap_outlook_20260610/yahoo_cap_2026_27_memo_20260610.html, data/raw/cap_outlook_20260610/basketnews_cap_projection_20260610.html. Spotrac tracker carries the same 165M max (data/raw/league_cap_2026.html, line 5731). Measured at B; the league does not officially set 2026-27 until late June 2026 after the audit.
• 2027-28 onward: mechanical 10 percent compounding from the 2026-27 memo base. Inferred (documented formula). Cross-checked against SalarySwish full projection table (WebFetch 2026-06-10), which compounds from 165,472,000 and therefore runs about +0.3 percent above every row here. Divergence is base-choice only, not method.

Timeline note on 2026-27: the projection moved 165M (mid-2025) up to 166M (late 2025) back down to 165M (March 2026 memo, local media revenue dip). cba-knowledge/sources/cba-2023-changes.md Section 1 records the direction of the March revision backwards ("revised up from 165M"); the March memo was a revision DOWN from 166M. One secondary writeup (ClutchPoints) lists the floor as 147M; the memo figure is 149M and 90 percent of 165M is 148.5M, so 147M is wrong. Flagged, not corrected in place per append-only discipline.

Relationship to prior repo table: identical in every cell to the table used in data/okc_luxury_tax_projection_2025_31.csv and data/bos_tatum_trade_cap_2026_31.csv. CONFIRMED.

---

3. Growth assumptions and sensitivity

The consensus expectation of repeated 10 percent years rests on the 76 billion dollar, 11-year ESPN/NBC/Amazon national deal that began 2025-26, with the CBA capping annual cap growth at 10 percent (cba-knowledge/sources/cba-2023-changes.md Section 11; awfulannouncing.com coverage, tier D). The cap-growth ceiling itself is measured (CBA smoothing); the expectation that the ceiling binds every year is opinion-tier consensus.

The uncomfortable read: 2026-27, the second year of the new national deal, grew only 6.69 percent (154.647M to 165M) because local/RSN revenue fell. SalarySwish models 2026-27 as a 7.00 percent year and then reverts to 10 percent for 2027-28+. If local media keeps dragging, the 10 percent rows above are ceilings.

Sensitivity: 2027-28 at 7 percent instead of 10 percent (all thresholds scale near-proportionally with the cap; aprons are defined as 2023-24 base times cap ratio, CBA Art VII Sec 2(e)(3)):

Implication: any team plan that "clears the 2027-28 apron by under 7M" on the 10 percent path does not actually clear it. Per data/bos_tatum_trade_cap_2026_31.csv, BOS clears every 2027-28 threshold by 33M+ in both scenarios, so this sensitivity does not bind BOS; it binds whichever counterparty carries an apron-adjacent book into 2027-28.

Two structural caveats on the out-years (both measured): the 2023 CBA expires after 2029-30 with a mutual opt-out after 2028-29 (cba-knowledge/sources/cba-2023-changes.md Section 17), so the 2030-31 row assumes successor-CBA continuity. And minimums, rookie scale, MLEs, BAE, tax brackets, and the expanded-TPE 7.5M leg all scale with the cap (CBA Art VII: 9.12 pct of cap NT-MLE line 9219, 5.678 pct room MLE line 9320, 5M x cap-ratio taxpayer MLE lines 9258-9268, 5M x cap-ratio tax bracket lines 7196-7200, 7.5M x cap-ratio matching leg lines 9408-9411 of cba-knowledge/sources/cba-2023-full-text.txt), so every exception value inherits the same growth uncertainty.

---

4. Apron restriction matrix

All rows measured, tier A, from the official 2023 CBA text (cba-knowledge/sources/cba-2023-full-text.txt; the Transaction Restrictions Table is Art VII Sec 2(e)(4), lines 7475-7542). The CBA frames these as hard-cap triggers: doing the transaction caps your Apron Team Salary at the stated level for the rest of the cap year; equivalently, a team already above the level cannot do the transaction.

Precision notes the secondary sources blur (all tier A):

• "No cash at the first apron" is wrong. Cash payment is a SECOND apron trigger (row I).
• "First apron teams cannot sign-and-trade" is imprecise. RECEIVING an S&T player is the first-apron trigger (row C). Signing-and-trading your own free agent out is legal at any level; only USING the resulting TPE is restricted (row J, second apron).
• "Second apron teams cannot use any TPE" operates through row F: a standard TPE used in a later season caps you at the FIRST apron, which a second-apron team fails by definition. Same-trade simultaneous 100 percent matching remains legal above the second apron, one outgoing contract at a time, with the +250,000 zeroed (Sec 6(j)(3), lines 9432-9436).
• Offseason lookahead (Sec 2(e)(2)(ii), lines 7403-7426): from the day after the regular season through June 30, rows E through J are also tested against the FOLLOWING season's apron levels, with all options assumed exercised and current thresholds assumed (Sec 2(e)(3), lines 7444-7472). A trade signed today, 2026-06-10, must pass both the 2025-26 test and the 2026-27 lookahead.
• Second apron draft penalties (Art VII Sec 2(f), lines 7699-7767): finish a cap year above the second apron (status as of last regular-season game) and your first-round pick in the draft after the seventh following season is frozen (untradeable). Be a second-apron team in 2 or more of the next 4 years and that frozen pick is moved to the last pick of the first round. Fewer than 2, it unfreezes after the third clean year.

---

5. Trade salary matching, with worked 2026-27 dollars

Formulas, tier A, CBA Art VII Sec 6(j) (lines 9358-9436). The 7.5M leg scales by cap ratio to 2023-24: 7,500,000 x (165,000,000 / 136,021,000) = 9,097,860 in 2026-27.

For a team over the cap and below the first apron post-trade (expanded TPE, simultaneous): max incoming = greater of [lesser of (200 pct of outgoing + 250,000) or (outgoing + 9,097,860)] or (125 pct of outgoing + 250,000)

2026-27 breakpoints (inferred, computed from the A-tier formula at the B-tier cap):

Worked examples at 2026-27 numbers:

1. Below-tax team sends 5,000,000: may take back 10,250,000.
2. Below-tax team sends 10,000,000: may take back 19,097,860.
3. Below-tax team sends 40,000,000: may take back 50,250,000.
4. First-apron team (post-trade above 209M): sends 25,000,000, takes back at most 25,000,000. The 250,000 cushion drops to zero (Sec 6(j)(3)) and anything above 100 percent is row E, which it cannot trigger.
5. Second-apron team (post-trade above 222M): sends one 30,000,000 contract, takes back at most 30,000,000. It cannot aggregate a 15M + 15M package (row H), cannot add cash to sweeten (row I), and cannot route the return through a stockpiled TPE (row F).
6. Room team with 20,000,000 in space: absorbs up to 20,250,000 with no outgoing salary at all (Sec 6(j)(1)(v)); absorption is not on the restrictions table and attaches no hard cap.

Multi-team mechanics (measured): each team in a three-team trade is evaluated independently, its own aggregate outgoing versus incoming under its own exception and its own apron status. There is no trade-wide matching test. Cash counts per paying-and-receiving team against the 5.15 percent annual limit each direction.

---

6. Luxury tax and repeater rates (2025-26 onward)

Tier A, CBA Art VII Sec 2(d)(2), lines 7196-7292 of cba-knowledge/sources/cba-2023-full-text.txt. Repeater = taxed in 3 or more of the 4 preceding cap years. Bracket size = 5,000,000 x cap ratio (5,585,000 in 2025-26; about 6,065,240 in 2026-27).

Note: cba-knowledge/sources/cba-2023-changes.md Section 5 lists the pre-2025-26 rates as if current and approximates the repeater schedule; the table above is the controlling 2025-26+ schedule from the CBA text itself. The repeater penalty is now brutal at depth: 15M over the line costs a 2026-27 repeater about 51.6M (the CBA's own worked example, lines 7307-7312).

---

7. Key dates

Trades executed before 2026-06-30 sit in the 2025-26 cap year: matching runs on 2025-26 salaries and thresholds, plus the offseason lookahead against 2026-27 aprons described in Section 4.

---

8. What this means for a BOS-WAS-UTA three-teamer

Each leg is evaluated alone. The relevant archetypes:

BOS, over-cap below-tax team. Measured at B: BOS 2026-27 committed salary 178.1M pre-trade against the 201M tax line, 30.9M below the first apron (data/bos_tatum_trade_cap_2026_31.csv, row 2). BOS has the full expanded TPE: for any outgoing package between 8.85M and 35.4M it can take back outgoing + 9,097,860. Using that expanded matching attaches a first-apron hard cap at 209M, which costs BOS nothing at a 178M book. BOS can aggregate, can send or receive up to 8,497,500 cash, and can receive an S&T player (also apron-1 hard cap). BOS's flexibility is the engine of this trade shape.

Room-team archetype (the rebuilding side, WAS/UTA profile). A team with cap room absorbs salary up to room + 250,000 with nothing outgoing and no hard cap attached. This is the only archetype that can take back salary with no matching test at all, which is what makes a third team valuable: it converts an unmatchable two-team frame into two legal legs. Constraint to respect: the 90 percent floor (149M per the 2026-27 memo) must be met by season start, so absorbing bad salary is usually a feature for these teams, not a cost. If WAS or UTA is instead over the cap but far below the tax, it falls back to the same expanded TPE as BOS with enormous headroom.

Apron-team archetype. Not present in this trade on current data, and that is the point: no leg of BOS-WAS-UTA should be sacrificing assets to dodge apron rules. If a fourth team enters above an apron, it matches at 100 percent flat, cannot aggregate or pay cash above the second apron, and during this June window is tested against 2026-27 aprons with options assumed exercised.

The three most binding rules for this trade shape:

1. Per-team independent matching with the 2026-27 expanded TPE middle tier of outgoing + 9,097,860. This sets how much salary BOS can legally take back per dollar sent out, and it is the number every scenario CSV should use.
2. Offseason lookahead (Art VII Sec 2(e)(2)(ii)): any leg executed between now and June 30 must clear NEXT season's apron levels with all 2026-27 options assumed exercised. A team that looks safe on current-year numbers can fail the lookahead if option pickups push its 2026-27 apron salary over 209M or 222M.
3. Cash and aggregation hard caps at the second apron (rows H and I) plus the 8,497,500 per-direction annual cash limit. In a three-teamer cash is the usual lubricant for the room team taking back unwanted salary; the limit is per team per cap year and is not netted across trades.

Unverifiable or weak points, flagged:

• 2026-27 figures remain a league projection memo (B) until the official late-June setting. History says the official number can move by about 1M either way.
• WAS and UTA 2026-27 books are not yet compiled in this repo at the same evidence standard as BOS. The archetype claims above are structural, not team-measured. Compile both before the per-team analyses cite this section.
• Whether the 2023 CBA contains a minimum annual cap growth floor was not verified from the CBA text in this pass; no claim above depends on it.
• SalarySwish raw HTML could not be captured (connection refused from this host on both curl and Invoke-WebRequest); its figures are recorded from a WebFetch extraction dated 2026-06-10 and cross-checked against the mechanical path.
• The 2027-02-11 deadline date rests on a RealGM Wiretap report plus the bylaw formula; confirm when the league publishes the 2026-27 calendar.

---

Evidence file index