LIHTC Opportunity Finder

The three questions the locator separates

1. Where is housing need acute? — high cost burden, deep AMI gap, severe rent burden
2. Where is a deal executable? — basis-boost eligibility, civic readiness, population scale
3. Where is competition light? — funding-recency headroom, LIHTC saturation gap

The locator never collapses these into a single black-box score. You see why a jurisdiction rises and can re-weight the dimensions for your own deal strategy.

The universe — which jurisdictions get scored

Starts with Colorado's 482 incorporated places, towns, and CDPs (TIGER 2024). Each target deal type then filters this universe differently. Today's implementation ranks the 158 jurisdictions with QCT and/or DDA designation (basis-boost eligible). Target state adds preservation, workforce-resort, and Prop 123 deal types with relaxed designation requirements.

How each jurisdiction is scored

Every Colorado place gets five 0–100 scores. The scores answer five questions a developer asks about any deal site. They're blended into a composite using weights that change depending on what kind of deal you're trying to do (see the deal-type table below).

1. Housing Need — how stressed is the rental market here?
   Combines what share of renters and owners pay more than 30% of income on housing, what share pay more than 50% (severe burden), and how many low-income households can't find an affordable place at their income. Then ranked against every other Colorado jurisdiction, so a score of "87" means this place is the 87th-percentile most-stressed market in the state.
   Data: HUD's CHAS housing-cost dataset (2018–2022) + Census American Community Survey rent-burden tables. (technical detail: tenure-blended cost burden + severe burden + AMI gap, percentile-normalized; CHAS Table 7 + ACS DP04 GRAPI/SMOCAPI)
2. Recency — how long since the last LIHTC deal closed here?
   A community that hasn't received a CHFA award in 15+ years has more geographic-balance headroom under CHFA's framework. Score caps at 25 years (anything older, or never funded, scores the max). For 9% Competitive deals this is one of the larger contributors because CHFA's QAP gives explicit weight to geographic distribution.
   Data: CHFA's live property feed — 926 Colorado projects from 1987–2025. We read CHFA's "Award Year" rather than HUD's "Year Placed in Service" because an award reserved in 2024 represents committed CHFA capital in the community, even if construction has not yet started — and that reflects the same signal CHFA's own QAP weighs when balancing geographic distribution.
3. Basis Boost — does the IRS give this place a bigger budget?
   The federal LIHTC statute (IRC §42) gives a 30% bigger eligible-basis cap to deals in federally-designated low-income areas. That translates to roughly $3–5M of extra equity on a typical 60-unit project — a hard math line, not a soft preference. Score: 100 if the place is in both a Qualified Census Tract (QCT) and a Difficult Development Area (DDA); 60 if just one; 0 if neither.
   Data: HUD's 2026 QCT designation list (224 Colorado tracts) + HUD's 2026 DDA list (10 Colorado nonmetro counties).
4. Population — is the renter base big enough to fill the building?
   A 60–100 unit project needs to lease up to ~95% in 12–18 months. Score scales from 0 (places under 500 people) to 100 (15,000+). Most important for 4% bond deals which usually need 100–200 units absorbing quickly; less critical for 9% deals which can be smaller (40–60 units) in a tighter market.
   Data: CHAS households at or below 100% AMI × 2.5 average household size — a proxy until raw ACS B01003 population is wired in.
5. Civic Readiness — is the local government actually ready to support a deal?
   Seven yes/no checks against what makes a deal smooth: did the place file a Prop 123 commitment with DOLA, do they have a current Housing Needs Assessment, do they have a comprehensive plan, do they have an inclusionary-zoning ordinance, do they run their own affordable-housing fund, do they have a housing authority, and are there affordable-housing nonprofits active here? The more yeses, the higher the score (out of 100). A Prop 123 filing is referenced in CHFA QAP scoring criteria; the other six matter for permit timelines, letters of support, and soft-debt sources to stack with the LIHTC equity.
   Data: hand-curated policy scorecard (547 Colorado jurisdictions × 7 dimensions) + DOLA's Prop 123 commitment list (217 jurisdictions filed) + the local-resources directory.

Two more signals show up as columns but not in the composite score — Market Capture (the gap between FMR market rent and LIHTC 60% AMI max rent — tells you whether the deal can pencil at 60% AMI or needs a deeper mix) and Prop 123 status (filed / not filed / N/A for CDPs). Both surface in the table so you can screen on them; they're not bundled into the score because they're decision-critical thresholds rather than smooth gradients.

Per-deal-type weighting (re-tuned live by your target selection)

Different deal types succeed on different signals. The locator doesn't use one fixed weighting — it re-tunes the formula based on which round you're trying to compete in. Here's what each deal type is actually optimizing for, in plain English:

9% Competitive — geographic-gap is the differentiator

9% deals are scarce — Colorado gets roughly 30–40 awards a year for hundreds of applicants. CHFA's Qualified Allocation Plan explicitly rewards geographic distribution and underserved-market scoring categories, so the single biggest predictor of success is "this jurisdiction hasn't been funded recently." Need still matters (CHFA scores deeper income-targeting), but a high-need market that just got two awards last cycle is harder than a moderate-need market that hasn't seen LIHTC in 15 years. Population and civic readiness matter less than for other deal types because 9% awards go through CHFA's most rigorous diligence — they evaluate readiness during application rather than relying on pre-screening.
Heaviest: Need (30%) + Recency (22%) + Civic (18%)

4% Bond — population scale is the differentiator

4% bond deals are non-competitive (you "buy your way in" via private-activity bonds), so saturation matters less. What kills 4% deals is not enough scale to absorb the transaction. Bonds carry fixed costs (issuance fees, trustees, legal, ongoing compliance) that only pencil at 100+ units; smaller deals can't carry that overhead. A 60-unit 4% bond deal in a 1,500-person town typically doesn't work; the same project on 9% credits often does. Population is therefore the dominant factor. Civic readiness matters more than for 9% because the local jurisdiction has to issue the private-activity bonds — they need an active housing finance authority or a willing partner.
Heaviest: Population (30%) + Need (25%) + Civic (18%)

Preservation — basis-boost stacking is the differentiator

Preservation deals refinance and re-syndicate existing LIHTC properties whose initial 15-year compliance period is ending. The economics depend almost entirely on the subsidy stack: you're combining 4% acquisition/rehab credits with expiring LIHTC, possibly Rental Assistance Demonstration (RAD), maybe HOME or CDBG, and a refi of the existing permanent loan. Basis-boost eligibility (QCT/DDA) is the biggest single lever. Recency matters less because preservation deals deliberately target a sub-market with proven existing absorption — the LIHTC units already exist and are already leased. Population matters less because you're working with an existing project's footprint. Civic readiness matters for the refi entitlement (some jurisdictions require new public hearings on preserved deals).
Heaviest: Basis/Subsidy (35%) + Civic (20%) + Need (20%)

Workforce / Resort — scale + civic capacity (resort towns punch above their weight)

Resort and workforce-housing markets (Aspen, Steamboat, Vail, Crested Butte, Durango) have severe need but tiny populations. They've adapted by building strong civic housing strategies — dedicated funds, employer partnerships, strong housing authorities. A typical workforce deal involves bond financing (so scale matters), a local employer or resort backing the offer, and a sophisticated public-private entitlement path. Basis-boost is less of a differentiator because resort markets often don't qualify for QCT and only some are in DDA counties. The lock is execution capacity: can this town actually deliver a 100-unit bond deal in 24 months?
Heaviest: Need (25%) + Population (25%) + Civic (20%)

Prop 123 Local — civic readiness IS the gate

Prop 123 (HB22-1304) created two state-funded affordable-housing programs that require local jurisdictions to commit to a 3% annual baseline expansion of affordable housing stock before they can access state dollars. Without the commitment, no Prop 123 deal happens — full stop. So the differentiator isn't need or market size, it's "has this jurisdiction done the local-policy work to unlock the state funding stack?" Prop 123 ✓, comp plan ✓, housing lead ✓, housing authority ✓, local funding ✓, IZ ordinance ✓ — those seven civic-capacity flags decide whether a deal is even legally possible. Basis-boost matters less because Prop 123 dollars don't require federal QCT/DDA designation. Need still matters (the funds prioritize high-need markets), but readiness gates everything.
Heaviest: Civic (30%) + Need (25%) + Basis/Subsidy (20%)

Balanced (Any) — exploratory, no specific lens

For exploratory work when you haven't picked a deal type yet. Roughly equal distribution across all five dimensions surfaces jurisdictions that score well on multiple dimensions rather than one. Use this when scouting CO for new deal opportunities and want a broad list to narrow from.
Balanced: ~20% each

The numerical table below summarizes these weights. The highlighted cells in each column show the heaviest dimension(s) for that deal type.

These weights are live in the composite today — all five dimensions including Civic Readiness, all six deal types. The F9 rebalance (2026-05-26) lifted civic from 10–15% to 18–20% on 9% / 4% / Workforce-Resort because local Prop 123 commitment, IZ ordinance, housing-authority infrastructure, and soft-debt match proved to matter more than the original 4-dimension model credited. CDPs (unincorporated places) get a −8 composite adjustment on incorporation-sensitive targets (9% / 4% / Workforce-Resort / Balanced) because they lack independent permitting authority. See methodology doc §4 for full derivation.

Confidence — surface uncertainty, don't hide it

Every score carries a confidence rating reflecting input quality. Confidence does not change the composite — it changes how the result is presented.

• ★★★ High — all five dimensions from direct data; civic ≥ 5 of 7 dims filled; no major fallback used
• ★★ Medium — at least one dimension uses a fallback OR civic 3–4 of 7 OR population is a proxy
• ★ Low — multiple fallbacks OR civic <3 of 7 OR major dimension missing entirely

Confidence pills next to every metric are sprint-1 work; today the locator surfaces this as table column hover-tooltips and detail-panel inline notes.

How to use the result — a developer's decision flow

1. Filter. Pick a target deal type. Locator re-weights + re-filters.
2. Read top 10. Sort by composite. Skim reasons (does the rise make sense?) and risks (does any flag disqualify?).
3. Cross-check civic. Strong scores but no Prop 123 + no housing lead + no comp plan = harder to execute. Flag for early conversation.
4. Funnel forward to HNA. Click "📋 Open HNA" — verify locator signal matches deeper needs picture.
5. Funnel forward to PMA. Click "🗺️ Run market analysis" — verify rental demand, comparable rents, amenities, no infrastructure or environmental risk.
6. Build a concept. Click "💵 Build deal concept" — test 9% (30–60 units, 30/50/60 AMI) vs 4% (100–200 units, 50/60/80 AMI). See which pencils.
7. Export memo. Take it to your pipeline meeting. (planned)

Output bands

These bands describe where each community sits in a public-data readiness lens — not a public ranking of communities. Every community can move between bands as plans, partnerships, and funding postures evolve.

What this locator does NOT do

Be honest about scope. The locator is a jurisdiction-level screen, not site-screening or underwriting.

Caveats & known data quirks

1. HUD YR_PIS = 8888 placeholders (in-pipeline projects) are excluded from recency.
2. Project-to-jurisdiction matching uses PROJ_CTY string equality — a misspelled PROJ_CTY ("Ft Collins" vs "Fort Collins") misses match. Fuzzy-match is a backlog item.
3. Population is approximate — derived from CHAS HHs ≤100% AMI × 2.5 avg HH size, not direct ACS B01003. Resort areas with high second-home stock are actually closer to the renter-base truth via HH proxy than via B01003.
4. QCT-place membership requires tract's share of place > 5% OR place's share of tract > 20%. Sliver overlaps don't claim a QCT.
5. DDA designation in CO is county-based (all 10 CO DDAs are nonmetro counties); all jurisdictions in DDA counties inherit the designation.
6. Civic-capacity coverage is sparse. ~70% of jurisdictions have populated scorecard records; ~30% rely on county-level inheritance.
7. Recency uses YR_PIS (placed in service) which lags awards by 2–3y. A jurisdiction with 2024 awards but no PIS yet looks "stale." CHFA-pipeline integration is a backlog item.

Data sources

• HUD QCT 2025 — 224 CO tracts (data/qct-colorado.json)
• HUD DDA 2025 — 10 CO nonmetro counties (data/dda-colorado.json)
• CHFA HousingTaxCreditProperties_view (live) — 926 CO projects 1987–2025 (data/chfa-lihtc.json; HUD LIHTCDB fallback data/market/hud_lihtc_co.geojson with 716 projects through 2020)
• HUD CHAS 2018–2022 — county-level (64) + tract-level (1,447 records)
• TIGER 2024 — place-tract spatial join (data/hna/place-tract-membership.json)
• DOLA Prop 123 commitments — 217 filings (data/policy/prop123_jurisdictions.json)
• Housing policy scorecard — 547 jurisdictions × 7 civic dimensions (data/policy/housing-policy-scorecard.json)
• Local resources — housing leads, authorities, advocacy URLs (data/hna/local-resources.json)