[DTrim99]

Summary

Closes #9679.

Stacked on #9585 (merge that first) — the diff includes #9585's commits until it merges.

New Mexico's net capital gains deduction now gives each married-filing-separately spouse one-half of the deduction computed on the couple's combined net capital gain. Previously, each separate filer's own gains determined the deduction, which was then halved.

Law

• NMSA 7-2-34(B), as amended by 2019 HB6, section 14(B): "Married individuals who file separate returns for a taxable year in which they could have filed a joint return may each claim only one-half of the deduction provided by this section that would have been allowed on the joint return."
• 2024 HB252, section 8(B) keeps the same subsection (B) from 2025.
• PIT-ADJ line 16 instructions (2023, 2024, 2025): "A taxpayer and spouse who file separate returns ... may each claim only one-half of the net capital gains deduction allowed on a joint return."

Changes

• nm_net_capital_gains_deduction: each spouse filing separately heads their own tax unit, so the formula reaches the other spouse's unit through the marital unit. It pools the two units' long-term gains (including capital gain distributions) and short-term gains only when the marital unit holds exactly two people who both head a separate-filing unit. This matches the guard in ar_medical_expense_deduction_floor, since without explicit marital units everyone shares one default unit. The IRC 1222(11) netting and the joint-return cap/floor then apply to the combined amount, and each spouse receives the separate-filer share. Other filing statuses are unchanged. A separate filer with no identifiable spouse still uses its own gains, halved, as before.
• New parameter gov.states.nm.tax.income.deductions.net_capital_gains.separate_filer_share (0.5) replaces the hardcoded where(separate, 2, 1) divisor.
• Removes the limitation comment Fix AR, MT and NM income tax gaps found in TAXSIM comparison sweep #9585 added.

Before and after (MFS couples)

Case	Year	Before (spouse 1 / spouse 2)	After (each)
$2,000 long-term gain each	2023	$500 / $500	$800 (combined $4,000; max(40% x 4,000, 1,000) = 1,600; halved)
+$20,000 and -$15,000 long-term	2023	$4,000 / $0	$1,000 (combined $5,000; 40% = 2,000; halved)
$10,000 gain on one spouse only	2023	$2,000 / $0	$2,000 (joint 4,000; halved)
$3,000 and $1,000 long-term	2025	$1,250 / $500	$1,250 (combined $4,000; joint cap $2,500; halved)

Tests

Nine cases added at the bottom of nm_net_capital_gains_deduction.yaml: gains on one spouse only, gains split between spouses, a long-term loss offsetting the other spouse's gain, a short-term loss offsetting the other spouse's long-term gain, combined gains at the 2023 $1,000 cap and above the 2025 $2,500 cap, a couple with a dependent and explicit marital units, and joint and single regressions.

• nm_net_capital_gains_deduction.yaml: 30 passed
• NM deductions folder: 93 passed
• NM integration test: 4 passed

🤖 Generated with Claude Code

https://claude.ai/code/session_01LqCGPgxDukji8Bhs57c5rH

[hua7450]

Waiting for #9585 to merge first, then I’ll approve this one.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (909176a) to head (fec9b03).
⚠️ Report is 457 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##              main     #9689   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            4         1    -3     
  Lines           76        30   -46     
  Branches         2         0    -2     
=========================================
- Hits            76        30   -46     

Flag	Coverage Δ
unittests	100.00% <100.00%> (ø)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.