Stage 4 Research Report: Housing Cost Model vs Industry Best Practice
Document under review: _posts/2023-05-17-house-cost.md
Date: 2026-02-23
A. Well-Known Rent-vs-Buy Calculators
NYT Rent vs Buy Calculator
The most widely referenced calculator. Runs year-by-year cost simulations for buying and renting, plots cumulative costs over time, and finds the break-even point. Accepts inputs for home price, down payment, mortgage rate, property taxes, maintenance, home appreciation, rent growth, investment return rate, and tax parameters. Updated post-2017 to reflect Tax Cuts and Jobs Act changes to mortgage interest deduction caps.
- https://www.nytimes.com/interactive/2014/upshot/buy-rent-calculator.html
Ben Felix / PWL Capital “5% Rule”
A simplified heuristic: multiply the home’s value by 5%, divide by 12. If comparable monthly rent is below that figure, renting is cheaper. The 5% decomposes into three unrecoverable cost categories:
- 1% property taxes
- 1% maintenance
- 3% cost of capital (opportunity cost of equity in the home vs invested in equities)
The key insight is that only unrecoverable costs matter — mortgage principal repayment is excluded because it builds equity. This aligns directly with the reviewed model’s “cost of ownership” vs “cash flow” distinction.
- https://pwlcapital.com/rent-or-own-your-home-5-rule/
Homo Economicus Calculator (Independent Institute, 2026)
Models a fully rational economic agent. Includes 13 input parameters: home price, down payment %, mortgage rate, loan term, property tax rate, PMI rate, maintenance rate, appreciation rate, initial rent, rent growth rate, investment return rate, inflation rate, time horizon, and closing costs. Explicitly models the renter investing the difference between owner costs and rent, compounded monthly.
- https://www.independent.org/article/2026/02/02/homo-economicus-rent-vs-buy-calculator/
Consumer Calculators (NerdWallet, Bankrate, Zillow, SmartAsset)
Generally compute monthly owner costs (PITI + maintenance + PMI) vs monthly rent, apply growth rates, discount by inflation, and model investment of the renter’s savings. They vary in how many assumptions they expose vs hard-code.
- https://www.nerdwallet.com/mortgages/calculators/rent-vs-buy-calculator
B. Academic Frameworks
Himmelberg, Mayer & Sinai (2005) — “User Cost of Housing”
The foundational academic framework. Published in the Journal of Economic Perspectives. Defines the annual user cost of owning as:
User Cost = P × (r_mortgage + property_tax + maintenance + risk_premium − expected_appreciation)
In equilibrium, user cost equals annual rent. Deviations suggest mispricing. This is the basis for the Fed’s price-to-rent analysis.
- https://www.aeaweb.org/articles?id=10.1257%2F089533005775196769
Cleveland Fed (Rappaport, 2010–2011)
Extends the user cost framework with explicit treatment of uncertainty and timing risk. Argues that conventional metrics (price growth, price-to-rent, price-to-income) are misleading because they fail to account for interest rate variation and local appreciation rate differences.
- https://www.clevelandfed.org/publications/economic-commentary/2011/ec-201106-buy-a-home-or-rent-a-better-way-to-choose
Tabner (2016) — NPV of Homeownership
Full Net Present Value framework. Finds that households typically need a 5–10 year holding period to achieve breakeven NPV. Inflation, the spread between imputed rent and opportunity cost, and holding period are the key determinants.
- https://www.sciencedirect.com/science/article/abs/pii/S1057521916301545
Price-to-Rent Ratio (Simple Heuristic)
Home Price / (Monthly Rent × 12). Below 15 favours buying; 15–20 is neutral; above 20 favours renting. Limitation: excludes interest rates, taxes, maintenance, and all dynamic factors.
C. Comparison: Reviewed Model vs Industry Standard
What the model does well
| Feature | Model | Industry standard | Assessment |
|---|---|---|---|
| Cash flow vs cost of ownership distinction | Yes | Ben Felix 5% Rule, Himmelberg user cost | Core differentiator, well-aligned |
| Opportunity cost of deposit | Yes | NYT, Himmelberg, Cleveland Fed | Included by better calculators, often omitted by simple ones |
| Tax effect on opportunity cost | Yes | Rarely included in consumer tools | Above average |
| Holding/carrying cost breakdown | Yes (interest, insurance, rates, fees, maintenance) | Standard | Comprehensive |
| Double-entry bookkeeping constraint | Yes | Unusual | Adds rigour; most calculators don’t expose accounting identity |
Gaps relative to industry standard
| Factor | Model | Industry standard | Impact |
|---|---|---|---|
| Capital appreciation | Excluded (acknowledged) | Critical variable in all serious models | High — single most influential input alongside mortgage rate |
| Selling transaction costs | Not mentioned | 5–8% of sale price (agent fees, transfer taxes) | Medium — significant unrecoverable cost |
| PMI | Not mentioned | Standard for deposits < 20% | Low — examples use 20% deposit |
| Rent growth over time | Not modelled (single-year snapshot) | Projected annually | Medium — limits decision usefulness over time |
| Mortgage interest deduction | Not mentioned | Included but jurisdiction-dependent | Low for Australian context (no equivalent deduction) |
| Closing costs at purchase | Not mentioned (stamp duty is included) | Standard | Partially covered via stamp duty |
| Investment of renter’s savings | Not modelled | Homo Economicus, NYT, NerdWallet all model this | Medium — completes the opportunity cost picture |
Unique strengths
- Explicit cash flow separation: “Can you afford this?” is a different question from “Is this a good deal?” The model answers both. Most calculators conflate them into a single break-even number.
- Accounting identity: The double-entry constraint means the model is internally consistent by construction. This is rare.
- Simplicity as a feature: The single-year snapshot avoids compounding speculative assumptions about future appreciation and rent growth. This is a limitation but also a form of honesty — the model doesn’t pretend to predict the future.
D. Common Criticisms of Rent-vs-Buy Models
These apply to the field generally, not just the reviewed model:
- Ownership bias (NMHC): Most calculators are framed as “should I buy?” rather than “should I sell and rent?” This creates framing bias.
- Sensitivity to unknowable assumptions: Results are extremely sensitive to assumed appreciation rate, investment return rate, and rent growth. Small changes flip the recommendation.
- Tax benefit miscalculation: Post-TCJA, far fewer US homeowners itemise. Many calculators overstate tax benefits.
- “Forced savings” confound: Models assume the renter invests the cost differential. Behavioural evidence suggests many don’t.
- Single-asset concentration risk: A home is a leveraged, undiversified, illiquid bet on a single asset in a single geographic market.
Sources:
Summary
The model’s core design — separating cash flow from cost of ownership, including opportunity cost with tax adjustment, and enforcing a double-entry constraint — is sound and aligns with the better academic and consumer frameworks. The main gaps (no appreciation modelling, no selling costs, no rent growth projection) are acknowledged limitations of the single-year approach. The model sits between a simple heuristic (price-to-rent ratio) and a full multi-year simulation (NYT, Homo Economicus): more rigorous than a rule of thumb, deliberately simpler than a forecasting tool.