Research · Social Housing

Research Addendum

Procurement Amendment and Q&A Analysis — What the New HUD Materials Change, Validate, and Add

21 March 2026 · Greg Williams · steko.co.nz/thinking

This addendum is the fourth document in an analytical sequence that began with the Strategic Analysis delivered on 28 February 2026. The Strategic Analysis diagnosed the procurement landscape. The HUD Model Practitioner Notes examined the financial model architecture at formula level. The Layer 2 Detailed Report applied a simulation engine to a synthetic programme-scale scenario. Nineteen days later, HUD has issued its first Notice of Amendment, published 44 Q&A responses, and a tranche of OIA material. The analytical question is straightforward: do any of these new materials change, validate, or add nuance to the prior findings?

A note on analytical method: This document draws inferences from HUD’s published materials. Inferences are distinguished from direct findings throughout. Where HUD says something explicitly, the addendum states so. Where the author is reading between the lines, the text identifies the reasoning as inference. An inference may be correct, but it is not evidence, and it should not be presented to a board as though it were.

Five findings dominate this addendum

AInsurance obligation expanded. Amendment 1 redefines “Required Insurance” from financier requirements to full replacement value under the General Terms and Conditions. Material cost increase for weather-exposed or high-value locations. The CPIDI index lag now operates against a higher insured value floor. The cost consequence compounds for 25 years.
BNo renegotiation mechanism. HUD has confirmed this explicitly. Q&A 41: outside the Qualified Change in Law clause, there is no provision to renegotiate the Agreed Amount. Set at contract execution, indexed by formula, cannot be reopened.
CContingency governance lever validated by HUD. Q&A 10 states four purposes for contingency — including absorbing the difference between index-adjusted and actual costs. Q&A 35 confirms it is paid regardless of actual costs. Zero exhaustion at Medium level is consistent with HUD’s own description.
DHUD financial model contained a formula error for 19 days. Amendment 7 corrects cells F32, G32, H32. The original summed per-typology figures instead of calculating a weighted average. The Layer 2 simulation used an independent engine and is unaffected.
ENegative inflation index reduces the Agreed Amount. Q&A 40: if the relevant index is negative, the payment shrinks while fixed costs do not. The contingency reserve absorbs this — the same reserve covering index lag, capital replacements, and legislative change.
None of the new information suggests the prior findings were wrong. Several are now confirmed by HUD’s own published statements. The direction of travel since 28 February is consistent: the risk allocation favours the Crown, and the contract architecture has been tightened, not loosened.
Signal 1 — Strategic PartnersStrategic Partners receive a reduced evidence burden on the highest-weighted criterion (Q&A 12). HUD’s existing Strategic Partners are not required to submit evidence for Criterion 1 (delivery partner capability — 20% of marks). Whether a CHP holds Strategic Partner status materially affects the Criterion 1 narrative strategy.
Signal 2 — Wraparound Services SilenceForty-four questions across three weekly releases. Not one addresses the structural asymmetry the Strategic Analysis identified as the centrepiece governance finding. The recommended Q&A question should be submitted before the 10 April deadline regardless of this silence.
34 days remain to 24 April. The analytical infrastructure is validated. The four inputs requested in the Layer 2 report remain the critical path. None of this counsels against applying. All of it counsels against applying without full knowledge of what the contract requires.
Five findings that dominate this addendum FINDINGA Insurance obligation expanded. Amendment 1 redefines “Required Insurance” to full replacement value. Material cost increase for weather-exposed locations. The CPIDI index lag now operates against a higher insured value floor. The cost consequence compounds for 25 years. FINDINGB No renegotiation mechanism. HUD has confirmed this explicitly. Q&A 41: outside the Qualified Change in Law clause, there is no provision to renegotiate the Agreed Amount. Set at contract execution, indexed by formula, cannot be reopened. The single most important confirmation in the Q&A sheet. FINDINGC Contingency governance lever validated by HUD. Q&A 10 states four purposes for contingency — including absorbing the difference between index-adjusted and actual costs. Q&A 35 confirms it is paid regardless of actual costs. Zero contingency exhaustion at Medium level is consistent with HUD’s own description. FINDINGD HUD financial model contained a formula error for 19 days. Amendment 7 corrects cells F32, G32, and H32 — the affordable rental per-home summary row. The original summed per-typology figures instead of calculating a weighted average. Any CHP that populated the model before 18 March was working with a grossly inflated metric. FINDINGE Negative inflation index reduces the Agreed Amount. Q&A 40: if the relevant index is negative, the payment shrinks while fixed costs — debt servicing, staff, insurance — do not. The contingency reserve absorbs this gap. The same reserve covering index lag, capital replacements, and legislative change.

Findings A and B carry the highest strategic weight. Finding C validates the prior analysis. D and E introduce new risks.

Notice of Amendment — 18 March 2026

HUD issued its first Notice of Amendment on 18 March 2026, amending the Flexible Fund Stage One Application Pack. Seven amendments are declared. A revised Commercial Term Sheet (V2) and a corrected Financial Model Response Form (V2) were issued alongside the notice. The amendments are analysed below in order of materiality.

Amendment 1 — Insurance Definition Expansion

Materiality: High

The definition of “Required Insurance” in Item 8 (Termination if a Home becomes Uninsurable) has been expanded. The original scoped Required Insurance to whatever the External Financier requires. The amended definition adds an obligation to meet the General Terms and Conditions — which will require full replacement value coverage for all homes.

Full replacement value is a higher cost floor than financier-required cover. A financier’s requirement protects the outstanding loan balance, which declines over time. Full replacement value protects the physical asset at current reconstruction cost, which typically increases with construction cost inflation.

The Strategic Analysis identified the CPIDI index lag as a structural risk (Observation 1). This amendment increases the insured value floor against which that lag operates.

Inference (flagged)This amendment may have been prompted by HUD’s own risk analysis. If Required Insurance was scoped only to financier requirements, a CHP could maintain minimal insurance and face an Uninsurable Home termination event with inadequate cover. The expanded definition closes that gap — but the cost sits with the CHP, not the Crown.

SME Consideration

On a programme-scale deployment, full replacement value coverage will be materially more expensive than financier-minimum cover. The Layer 2 insurance assumption of $2,800 per unit per annum (Assumption 10) should be reviewed against full replacement value quotations. For locations with weather event exposure — particularly the Far North (Cyclone Gabrielle, 2023) — the premium loading may be substantial.

Amendment 7 — Financial Model Formula Error

Materiality: High

Cells F32, G32, and H32 in the Input sheet contained incorrect formulae. The original was a simple sum of four per-typology per-home figures — mathematically meaningless. The corrected formula calculates a weighted average by number of homes in each typology.

Gap analysis completed. A full cell-by-cell comparison of all 18 worksheets confirms the three declared corrections are the only substantive changes. No undeclared modifications found.

Layer 2 simulation impact: None. The simulation engine was built independently and does not reference the HUD model’s summary row.

Amendments 2 & 3 — “Insurance” to “Dwelling Insurance”

Materiality: Medium

“Dwelling Insurance” narrows the Agreed Amount component to the physical dwelling.

Inference (flagged)Non-dwelling insurance costs — liability, public liability, D&O — are the CHP’s cost and not recoverable through the Agreed Amount. The insurance architecture is now precise: the Agreed Amount covers dwelling insurance only, but at full replacement value. Everything else is the CHP’s exposure.

Amendment 3 — Affordable Rental Apportionment Basis

Materiality: Medium

Changed from market-value-based to a simple Apportionment Percentage. For a mixed-value affordable rental portfolio, this may redistribute internal funding away from the most expensive properties.

Amendment 5 — Land Covenant Terminology

Materiality: Low–Medium

The covenant now correctly states “social affordable housing” rather than “social housing.” Tenure-appropriate, marginally better for balance sheet flexibility on the affordable rental tranche.

Amendments 4 & 6 — Housing Option Terminology

Materiality: Low

The same “social” to “social affordable” correction. No change to commercial substance.

Commercial Term Sheet V2

A full gap analysis has not been performed. The seven amendments are taken on good faith, supported by the confirmed integrity of the financial model disclosure. The legal advisor should independently verify the full V2 term sheet.

Q&A Intelligence — 20 March 2026

44 questions across three weekly releases. Organised by strategic significance.

Benchmark and Cost Assessment

Q7 — Benchmarking will not be released. The $46,000 figure is confirmed unchanged in V2. CHPs are submitting into a benchmark they cannot see.

Q35 — Contingency paid regardless of actual costs. Confirms the mechanism the SA identified and Layer 2 modelled.

Q36 — Contingency must be justified. A CHP selecting higher contingency without articulating the risk basis will score less favourably.

Q39 — No ceiling, but cost assessment applies. The trade-off between contingency as 25-year insurance and contingency as benchmark penalty is confirmed by HUD.

Contract Architecture and Risk

Q10 — Contingency purpose defined explicitly. The most analytically valuable answer. Four purposes stated. HUD also lists five mitigating factors.

SME Consideration

Factor (b) — indexation as protection — is precisely what the SA identified as structurally inadequate. Factor (d) — diminishing interest rate risk — is valid but only material after significant debt repayment in the second half. Factor (e) — property ownership — does not generate cashflow during the term. Assess at programme scale, not at face value.

Q41 — No renegotiation outside Change in Law. The definitive answer. The Agreed Amount is a 25-year commitment with no periodic review.

Legal Advisor Brief

The brief must proceed on the basis that the Agreed Amount is fixed at contract execution and cannot be reopened. Every cost assumption is a commitment the organisation will live with for 25 years.

Q40 — Negative index reduces the Agreed Amount. Asymmetric downside not previously identified.

Inference (flagged)Sustained deflation is unlikely currently. Over 25 years, sector-specific sub-index deflation is not implausible. CPIDI could decline following benign weather; CPILARP in a severe housing correction. Low probability but material consequence — warrants modelling.

Q33 — Three Waters rates treatment. A genuine windfall/loss variable across 25 years.

Q38 — Tenancy Management LCI acknowledged as imperfect. Non-wage costs drift from the LCI. The CHP absorbs the drift.

Q28 — Payments start at tenanting, not at title. The gap between capital commitment and first payment must be self-funded.

Programme Structure and Competitive Landscape

Q9 — Leasing options detailed. Option 2 bypasses development costs, financing, and contingency entirely. Strategically significant.

SME Consideration

For a mixed ownership/lease portfolio, Option 2 brings blended cost closer to benchmark. But the loss of early termination compensation on the leased tranche means the CHP carries exit risk without protection. A mixed-tenure model was not tested and should be considered.

Q12 — Strategic Partners receive reduced evidence burden. Structural competitive advantage.

Inference (flagged)Whether the organisation holds Strategic Partner status materially affects the Criterion 1 narrative strategy. Confirm before the narrative is drafted.

Q22 — Re-pooling of uncommitted allocations. Stage 1 success does not guarantee Stage 2 funding.

Q29 — Programme-level assessment, not site-specific. The Layer 2 per-location modelling is a competitive advantage.

Q16 — Challenging development economics defined. Upfront funding pathway not previously modelled.

Wraparound Services — The Silence

The SA devoted an entire section to the funding gap. The Layer 2 report quantified it at $552,000–$840,000 per year, with 25-year cumulative exposure of $17.7M–$26.9M. Forty-four questions published. Not one addresses the asymmetry.

Inference (flagged)Two possible readings. Either no CHP has asked — suggesting the sector has not identified the asymmetry. Or HUD chose not to publish — suggesting policy sensitivity. Neither diminishes the materiality. The recommended Q&A question should be submitted before 10 April regardless.

Cross-Reference: Impact on Prior Findings

Strategic Analysis — Four Findings

SA Finding 1 — The paradigm shift is permanentReinforced

Q&A 23 confirms the operating supplement model is gone. Q&A 13 confirms no impact on existing contracts.

SA Finding 2 — The financial modelling riskReinforced & Sharpened

Q&A 7 confirms benchmarks will not be released. Amendment 7 reveals a formula error. Q&A 10 and Q&A 35 confirm contingency is the designed absorption mechanism. Better understood but not reduced.

SA Finding 3 — The wraparound services funding gapUnchanged — Silence Noted

No Q&A response addresses this finding. The structural asymmetry remains the most significant unaddressed governance risk.

SA Finding 4 — The contract is asymmetricReinforced

Q&A 41 (no renegotiation), Q&A 40 (negative index), Q&A 22 (sole discretion re-pooling), Amendment 1 (expanded insurance). Risk allocation favours the Crown; contract terms tightened.

Layer 2 Detailed Report — Five Findings

Layer 2 Finding 1 — Benchmark gap is structuralUnchanged

$46,000 benchmark confirmed. Layer 2 figures unaffected by the formula error.

Layer 2 Finding 2 — Contingency is a governance leverValidated by HUD

Q&A 10 and Q&A 35 confirm the mechanism. The $8,700 per unit difference is now supported by HUD’s own published description.

Layer 2 Finding 3 — Capital allocation is tightUnchanged

Q&A 16 opens a potential upfront funding pathway not previously modelled.

Layer 2 Finding 4 — Wraparound gap is structuralUnchanged

See Wraparound Services — The Silence above.

Layer 2 Finding 5 — Four inputs convert to submissionUnchanged

The four inputs remain the critical path. The analytical infrastructure is validated.

Sources Not Analysed

OIA Responses — 16 March 2026. Substantial tranche. May contain intelligence on programme design history, benchmark methodology, and risk rationale. Analysis can be produced as a supplementary document if required.

Commercial Term Sheet V1 vs V2. Full gap analysis not performed. The legal advisor should independently verify.

HUD Model Practitioner Notes. The V2 correction does not affect the formulae analysed. The Practitioner Notes remain current.

Revised Recommendations

The SA’s five decisions and sprint plan remain the governing framework.

Legal Advisor Brief — Updated

Update with: Q&A 41 (no renegotiation); Amendment 1 (full replacement value insurance); V2 Commercial Term Sheet (verify independently against V1).

CFO Actions — Updated

Review insurance against full replacement value (Amendment 1). Note apportionment change (Amendment 3). Note Three Waters (Q&A 33). Consider leasing Option 2 (Q&A 9). Note upfront funding threshold (Q&A 16).

Application Narrative — Updated

Confirm Strategic Partner status (Q&A 12). Contingency justification must be evidence-based (Q&A 36). Submit the wraparound Q&A question before 10 April.

Sprint Plan

No dates changed. 24 April noon deadline confirmed unchanged.

34 days remain. The four inputs from the Layer 2 report remain the critical path.

Request the full paper

The full paper includes complete legal notices, IP reservation, and full analytical method note. Available on request.

Request the full paper →

Sources & Provenance

HUD (2026a). Budget 2025 Flexible Fund — Notice of Amendment. 18 March 2026.

HUD (2026b). Budget 2025 Flexible Fund — Questions and Answers. Consolidated 6, 13, 20 March 2026.

HUD (2026c). Budget 2025 Flexible Fund — Commercial Term Sheet V2.

HUD (2026d). Budget 2025 Flexible Fund — Financial Model Response Form V2.

Williams, G. (2026a). Strategic Analysis. 28 February 2026. Steko Consulting Limited.

Williams, G. (2026b). HUD Model Practitioner Notes. 1 March 2026. Steko Consulting Limited.

Williams, G. (2026c). Programme-Scale Financial Simulation. 12 March 2026. Steko Consulting Limited.

Colophon

Edition: 21 March 2026 (web publication March 2026)

This addendum was originally produced on 21 March 2026 as an analytical update to the Social Housing Financial Architecture research series. The analytical content is identical to the original publication. Organisation-specific references have been removed under the IP reservation terms.

How this article was produced

Produced under RPP-001 v0.1.0. Every external deliverable passes through evidence verification, adversarial review, and practitioner sign-off.

What the practitioner brought: Source document analysis, cross-reference mapping, inference/finding distinction discipline, all strategic recommendations. Editorial direction and publication approval.

What the production engine brought: Structured report production, cross-reference formatting, substance traceability verification, ANON-001 treatment, web publication at BWC template standard.

Powered by Claude Opus 4.6 · RPP-001 v0.1.0

Source verificationAll Q&A references cite specific question numbers. All amendments cite specific numbers. Financial model gap analysis: 18 worksheets, three corrections confirmed, no undeclared changes.
Substance traceability40/40 claims traced to source (100%). Four inference flags. Zero unattributed assertions.
Register compliancePASS — lens-not-subject, psychological register, brand consistency, ANON-001.

If you believe any claim requires correction, we welcome feedback at [email protected].

← Back to Social Housing Research
© 2026 Steko Consulting Limited · Originally produced 21 March 2026 · steko.co.nz