Picking a credit program and competition bucket — Texas

What you are actually choosing

9% credits are rationed by the same fixed federal per-capita ceiling every state gets, allocated in a single annual round. 4% credits are officially non-competitive — TDHCA's own process manual calls it "the non-competitive tax credit program" because the credit itself is as-of-right once a project clears volume-cap availability and underwriting. What's actually competed for on the 4% side isn't the credit, it's the private activity bond volume cap that has to come with it, and that's run by a separate state agency using a lottery, not a score.

One vintage warning before citing anything here: Texas repeals and re-adopts its entire Qualified Allocation Plan every calendar year as a brand-new rule, not an amendment to a standing one. By statute the Board must submit a proposed QAP to the Governor by November 15, and the Governor must approve, reject, or modify and approve it by December 1 (Tex. Gov't Code Section 2306.6724(b)) — and this isn't a formality: the 2026 QAP's own cover page documents the Governor approving it on December 1, 2025, with a modification to Section 11.202(1)(O). A tool that treats any year's point values, deadlines, or set-aside shares as a stable constant is wrong on its face here in a way California's amendment-only rulemaking doesn't force.

The 2026 federal break, and why Texas didn't react the way California did

The same federal change hit both states. The One Big Beautiful Bill Act (P.L. 119-21) Section 70422(b) amended 26 U.S.C. Section 42(h)(4)(B) to add a second path that waives the 9%-competitive-allocation requirement for 4% credits: either 50% of a building's aggregate basis (building plus land) is bond-financed, or 25% is bond-financed and at least one bond issue dated after December 31, 2025 itself finances at least 5% of that basis. This is federal law, identical in both states.

California's regulator reacted by immediately capping every bond award at 30% of aggregate basis (40% by exception) specifically to prevent gaming the new lower floor. TDHCA did not follow suit. Its own December 4, 2025 guidance memo states the position directly: "TDHCA does not currently have a requirement that limits the 25% test further." The only Texas-specific basis-percentage ceiling that exists is much older and unrelated to this change — a conditional 55% cap under Tex. Gov't Code Section 1372.037(b) that only switches on in years where statewide residential-rental bond requests exceed 55.75% of the state ceiling as of October 20.

Whether that trigger is actually active for the 2026 program year could not be confirmed from anything fetched this session. BRB's own 2026 program-year page states requests "represented less than 55.75 percent of the state ceiling as of October 20, 2024" — citing 2024 on a page describing the 2026 program year, which reads like a stale, uncorrected carryover rather than a confirmed current figure. Do not port California's 30%/40% cap into a Texas model: it doesn't exist there. And don't assume the 55% test is definitely off, either — resolve the October 20, 2025 residential-rental request total against 55.75% of the $4,280,825,835 ceiling ($2,386,560,403) before relying on either answer for a live deal.

The 9% side: points are a gate, geography is the actual competition

Texas has no single published maximum score the way California's regulation sums cleanly to 109. Summing every category that feeds the QAP's own "Self Score Total" gives a self-computed ceiling of 142; adding the categories TDHCA scores separately (Local Government Support, Quantifiable Community Participation, State Representative input, and either Community Organization input or the Concerted Revitalization Plan/Opportunity Zone item) gives a self-computed Total Score ceiling of 177. Neither number is published by TDHCA — both were derived from the regulation text this session, and 2026's real data ran one and two points short of each: the observed maximum Self Score Total was 141, and the observed maximum Total Score was 175, out of 106 active applications.

That near-half-field tie is resolved almost entirely by geography, not score. Outside of USDA-financed Rehabilitation deals (tied on Self Score and Total Score alike, resolved by which property is oldest), the real tiebreaker sums the straight-line distance from the site to the nearest of four amenity types — a public park, an ISD school campus, a full-service grocery store, and a public library meeting minimum weekly-hours thresholds — and lowest sum wins. If still tied, lowest HTC request per Low-Income Unit wins; if still tied after that, greatest distance from the nearest other same-Target-Population award in the prior 15 years wins.

The 4% side isn't scored at all — priority tiers and a lottery

TDHCA's own 2026 process manual is explicit: 4% applications "do not compete regionally against one another in terms of achieving a score high enough to secure an allocation." Instead, an applicant self-selects one of six Priority tiers on the bond application, which sets nothing but queue position relative to other 4% applicants.

Volume cap is granted by one annual lottery (filing window October 6–20, drawing November 12, first reservations issuing January 2), then strictly by priority tier, then lottery number, then date received. Miss the lottery window and the priority election stops mattering entirely: Tex. Gov't Code Section 1372.0231(j) puts every post-lottery application behind every application submitted before the deadline, regardless of tier — you go to the back of the whole year's line, not a parallel lower-priority lane.

For Priority 0, 1, or 2 applicants there's a sequencing trap with a hard clock: Tex. Gov't Code Section 1372.0321(e) requires BRB to have evidence that an HTC application was submitted to TDHCA before it will reserve volume cap. Mechanically, once BRB is ready to issue a reservation it emails that the applicant is next in line — there are 3 business days to submit the complete HTC application to TDHCA, and missing that window gets the slot cancelled and handed to the next applicant, with no appeal path described in any source read this session.

Bucket election: 26 subregions, 3 set-asides, and a six-step waterfall

Where California sorts 9% applications into eleven geographic apportionments and four set-asides, Texas sorts them into 13 Uniform State Service Regions, each split into an Urban and a Rural subregion — 26 buckets total — plus three set-asides layered differently than California's: Nonprofit (at least 10% of the ceiling, and a qualifying ownership structure is automatically enrolled unless affirmatively waived), USDA (5%, with a routing quirk — USDA-financed Rehabilitation draws from the At-Risk pot, while USDA-financed New Construction draws from the applicant's ordinary regional pot, not a separate carve-out), and At-Risk (at least 15% — more than triple California's 5% At-Risk share).

Awards flow through a six-step waterfall: USDA selection statewide by score, then At-Risk selection statewide by score, then initial selection within each of the 26 subregions (subject to embedded priorities — an Elderly-Development cap in large-population regions, a mandatory award to the top concerted-revitalization-plan project in counties over 1.7 million, and Supportive-Housing award caps), then Rural Collapse (unused rural pools flow statewide until at least 20% of total credits go to Rural applications), then Statewide Collapse (remaining credit flows to the highest scorer in whichever subregion is proportionally most underserved), and finally a Contingent Nonprofit Set-aside Step that repeats the prior steps after force-selecting the highest-scoring Nonprofit application if the 10% floor still isn't met — which can bump a higher-scoring non-nonprofit application out of an award it would otherwise have received.

What actually binds on the 4% side is subceiling exhaustion, not a percentage cap

Texas dedicates only about 26.25% of its private-activity-bond ceiling to multifamily rental housing — a full third goes to single-family mortgage revenue bonds, and nearly another third to "all other" categories (solid waste, exempt facilities, and similar) that barely register in California's system, which dedicates roughly 90% of its ceiling to multifamily. A Texas multifamily developer is competing for a much smaller slice of a somewhat larger total pie.

Demand is running well ahead of that supply. As of November 15, 2025, the multifamily subceiling had $3,020,372,882 available against $5,134,616,422 requested — 170% oversubscribed, the most oversubscribed category in the entire PAB program that year. By September 3, 2026, the local-HFC slice alone had $1,598,048,078 requested against $786,601,746 available (more than 2x oversubscribed), while TDHCA's own 20% share ran comparatively light — and only $17,220,829 of the full $4,280,825,835 ceiling remained available statewide. Treat this as one data point, not a trend, the same caution due any single-year subscription snapshot.

The queue narrows on a fixed schedule regardless of subscription: local-HFC regional sub-pools collapse into one statewide pool on March 1, a priority boost for lower-income-area Priority 2/3 projects activates June 1, and on August 17 all remaining subceilings collapse into a single pool where lottery number and receipt order alone decide. The live 2026 spreadsheet shows real churn from the March 1 collapse alone: of 30 applications tracked afterward, 16 were reserved, 6 withdrawn, 5 converted to carryforward, and 3 closed — a 20% outright-withdrawal rate among deals that made it that far.

Calendar, cost of entry, and the doors that don't reopen

Program election isn't the only irrevocable choice in this phase. Local Government Support resolutions, State Representative letters, and Neighborhood Organization statements all become permanent the instant they're submitted to TDHCA — the QAP says so explicitly for each — and a State Representative's opposition letter is worth −8 points, the only item in the entire scoring system that subtracts. A scoring-notice appeal has its own hard clock: 7 calendar days to file (Tex. Gov't Code Section 2306.6715), 14 days for an Executive Director response, no new evidence permitted at any stage, and the Board's decision is final. TDHCA's own 2026 records name five granted appeals — but winning doesn't guarantee funding: one of the five, Trails Apartments, never appeared anywhere in the July award list at all, because an appeal can restore eligibility or score without changing the underlying regional-allocation math. Separately, Readiness-to-Proceed is worth only 1 point, but the Board "cannot and will not waive" its site-acquisition-and-permit deadline, and a miss costs a 2-point penalty on every application submitted, for up to two subsequent Application Rounds (Tex. Gov't Code Section 2306.6710(b)(2)). And a Previous Participation Review finding against any Applicant, Developer, or Guarantor is a categorical eligibility gate, not a point deduction — it disqualifies outright, and it follows that party across every deal, not just the one that triggered it.

The state credit, hybrid deals, and what the sources do not say

Texas does have its own state Housing Tax Credit — House Bill 1058 (88th Legislature, 2023), first claimable in 2026, with a real calendar (Intent to apply by February 27, Request Form by September 4, Allocation Certificates in December) and a $3,000,000 minimum request. It is a genuinely different animal from California's: optional and non-gating, not a mandatory dependency — TDHCA does not withhold a bond allocation or an HTC award for failing to win a piece of it, unlike CDLAC's rule that voids a bond allocation outright if requested state credits aren't scheduled to be awarded. A total annual cap of $25,000,000, split evenly between 4% and 9% deals, is widely repeated across secondary summaries of the program, but that specific dollar split was not found stated anywhere in the QAP's own text this session — treat it as reported, not independently confirmed, until traced to the Texas Tax Code or Insurance Code provisions that actually authorize it.

Hybrid 9%/4% structuring gets a real mention in California's own materials — a size-factor boost from splitting a large project into phases. Nothing comparable turned up for Texas in the sources read for this phase: no worked hybrid structure, no combined-developer-fee arithmetic, and no funding-line data broken out for hybrid deals. What is knowable is that the 9%/4% choice is structurally more binary in Texas than in California in the first place, because the 4% side has no scoring or tiebreaker gradient to hybridize against — a Texas hybrid decision reduces to sequencing two entirely separate agency processes (TDHCA's single annual scored round, and BRB's rolling lottery) rather than blending one continuous formula, and nothing in the research read for this phase describes that combination being done in practice. Treat the mechanics as a structuring question for bond counsel rather than something to model from the regulations alone.