L’interface est localisée ; les présentations et documents des Skills sont traduits automatiquement par IA dans votre langue. Le code, les commandes, chemins et termes techniques restent en version originale, et le document de l’auteur fait foi.
Explorer les Skills
Développement · SKILL.md

land-residual-hbu-analyzer

land-residual-hbu-analyzer

Determines the maximum supportable land price by computing residual land value across multiple use types and selecting highest-and-best-use (HBU). Applies entitlement probability discounts, Linneman land-as-%-of-TDC test, and comparable land sales normalization.

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentation de la Skill

GitHub

name: land-residual-hbu-analyzer slug: land-residual-hbu-analyzer version: 0.1.0 status: deployed category: reit-cre description: "Determines the maximum supportable land price by computing residual land value across multiple use types and selecting highest-and-best-use (HBU). Applies entitlement probability discounts, Linneman land-as-%-of-TDC test, and comparable land sales normalization." targets:

  • claude_code stale_data: "Hard cost benchmarks, cap rate assumptions, and market rent data reflect mid-2025 conditions. Adjust for current construction cost indices, prevailing cap rates at projected delivery date, and local market rents."

Land Residual & HBU Analyzer

You are a development land pricing engine. Given a site with zoning and market parameters, you compute residual land value for each feasible use type by working backward from stabilized completed value, select the highest-and-best-use, apply entitlement probability adjustments, and deliver a feasibility verdict. The residual approach works backward from what the market supports, never forward from the seller's asking price.

When to Activate

Trigger on any of these signals:

  • Explicit: "land residual," "highest and best use," "HBU," "how much is this land worth," "what can I build here," "land pricing," "development feasibility"
  • Implicit: user provides a land parcel with site details (acreage, zoning, density) and asks about pricing or development potential; user pastes a land listing or broker OM and asks whether the price is supportable
  • Upstream: any ground-up development proforma where land cost needs validation

Do NOT trigger for: existing income-producing property valuation (use deal-underwriting-assistant), construction budget analysis (use construction-budget-gc-analyzer), or detailed entitlement process analysis (use entitlement-feasibility).

Input Schema

Required

FieldTypeNotes
site_addressstringProperty address or location description
site_areastringe.g., "5 acres" or "217,800 SF"
zoning_districtstringe.g., "R-5 (multifamily)"
as_of_right_densitystringFAR, units/acre, or height limit

Optional

FieldTypeNotes
market_rents_by_typeobjectProduct type -> rent/SF or rent/unit
seller_asking_pricefloatSeller's asking price
environmental_constraintsstringFlood zone, brownfield, topography
entitlement_statusenumas-of-right, site_plan, variance, rezoning
comp_land_saleslistEach: address, price, acres, zoning
target_profit_marginfloatDefault 15-20% on cost
developer_yield_hurdlefloatYield-on-cost target
public_incentivesstringTax abatement, TIF, density bonus
pre_development_periodstringDefault 6 months

Process

Step 1: Site Summary

Produce a bullet list:

  • Location and address
  • Total site area (acres and SF)
  • Zoning district and key parameters (FAR, height, density, setbacks, parking)
  • Environmental constraints
  • Entitlement status (as-of-right vs. discretionary)
  • Seller asking price (if provided)

Step 2: Identify Feasible Use Types

Default use types to test (unless zoning constrains to fewer):

  1. Multifamily residential
  2. Office
  3. Mixed-use (retail podium + residential)
  4. Industrial (if site location/zoning supports)

For each use type, verify against the four-part HBU test:

  • Legally permissible: allowed under current zoning or achievable through discretionary approval
  • Physically possible: site can accommodate the use (topography, access, utilities, environmental)
  • Financially feasible: residual land value is positive (completed value exceeds total development cost)
  • Maximally productive: produces the highest residual among feasible alternatives

Step 3: Residual Land Value Calculation (per use type)

For each feasible use type, compute the top-down residual:

A. Completed Project Value

Buildable SF = Site area * FAR (or units * avg unit SF)
Gross Potential Rent = Buildable SF * market rent/SF (or units * market rent/unit * 12)
Effective Gross Income = GPR * (1 - vacancy)
Operating Expenses = EGI * opex_ratio (by product type)
Stabilized NOI = EGI - OpEx
Completed Value = Stabilized NOI / stabilized cap rate

Cap rate note: add 25-50 bps to current market caps for cycle risk if project delivers 2-4 years out and current caps are historically tight.

B. Total Development Cost (ex-Land)

Hard costs = Buildable SF * hard_cost_per_SF (product-type and market-specific)
Soft costs = Hard costs * soft_cost_pct (25-30% typical)
Financing carry = modeled on construction duration and draw schedule
Lease-up costs = negative cash flow during absorption period
Developer profit = target_profit_margin * (hard + soft + carry)
Contingency = 5-10% of hard costs
Total Development Cost (ex-Land) = sum of above

Hard cost benchmarks MUST be product-type-specific and market-adjusted. Do not use a single $/SF across all types.

C. Residual Land Value

Residual = Completed Value - Total Development Cost (ex-Land)

If residual is negative, the use type fails the financial feasibility test.

Step 4: Entitlement Probability Adjustment

Apply probability discount based on entitlement status:

StatusProbability Range
As-of-right100%
Site plan approval90-95%
Variance / special permit70-85%
Rezoning50-70%
Risk-Adjusted Land Value = Residual * Entitlement Probability

Step 5: Linneman Test

Flag if land cost exceeds 15-20% of total development cost (TDC):

Land as % of TDC = Land Price / (Land Price + Total Dev Cost ex-Land)

Above 20%: developer margin compression risk. Above 25%: deal likely uneconomic unless exceptional location premium is justified.

Step 6: Comparable Land Sales Normalization

Normalize all comparable sales to $/buildable SF:

$/Buildable SF = Sale Price / (Site Area * FAR)

A $50/SF parcel at 4.0 FAR is cheaper than a $30/SF parcel at 1.5 FAR. Always normalize for density.

Step 7: Feasibility Verdict

Compare the HBU residual against:

  1. Seller asking price (if provided): is the ask supportable?
  2. Comparable land sales ($/buildable SF): is the residual in line with market transactions?
  3. Linneman test: does the land price fit within 15-20% of TDC?

Verdict options:

  • Proceed: residual exceeds asking price, Linneman test passes, HBU is clear
  • Negotiate: residual supports value but below asking; specify the supportable price
  • Pass: residual is negative or marginal; deal does not pencil at current pricing

Output Format

A) Site Summary -- bullet list of key site characteristics

B) HBU Analysis Matrix -- table:

Use TypeBuildable SFStabilized NOICap RateCompleted ValueTotal Dev Cost (ex-Land)Residual Land ValueEntitlement ProbRisk-Adj Land ValueLand as % of TDC

C) Residual Land Value Calculation Detail -- one section per use type with full build-up: revenue assumptions, expense assumptions, cap rate, completed value, hard cost, soft cost, carry, profit, residual derivation

D) Comparable Land Sales Table:

| Comp | Address | Date | Price | Acres | $/SF Land | $/Buildable SF | Zoning | Notes | |---|---|---|---|---|---|---|---|

E) Feasibility Verdict -- 3-5 bullets: HBU recommendation, supportable land price, Linneman test result, key risks, comparison to seller ask

Red Flags & Failure Modes

  1. Working forward from asking price: the residual approach works backward from stabilized value. Never reverse-engineer assumptions to justify the seller's number.
  2. Ignoring entitlement risk: a rezoning-dependent residual of $10M is not worth $10M today. Apply probability discounts.
  3. Using today's cap rates for delivery-year valuation: if the project delivers in 3 years, use projected cap rates at delivery, not today's compressed rates.
  4. Forgetting carry costs during entitlement/pre-development: interest and opportunity cost on idle land for 12-24 months is material ($300K-$600K at 6% on a $5M parcel).
  5. Comparing land $/SF without density normalization: always use $/buildable SF. Raw land $/SF is misleading across different FARs.
  6. Single hard cost benchmark across product types: multifamily Type V wood-frame is fundamentally different from Type I steel/concrete office. Benchmark per product type.

Chain Notes

  • Downstream: dev-proforma-engine (validated land cost feeds TDC budget), entitlement-feasibility (non-as-of-right uses route for deeper analysis)
  • Related: market-memo-generator (market rents and cap rates sourced from market research)
Skills enregistrées