Support Us
Harness Architecture
Essay ·
October 14, 2026

Financial Metrics

Author
Clifton Harness
Date
October 14, 2026
Address
/harness-architecture/
financial-metrics

One deal runs through these studies, in the order a developer meets them: acquire, fund, operate, value, distribute. Buy a block of eight parcels for $3M. Build 100 rental homes for $10M all in, land included. Borrow $6M and raise $4M from the owners. Lease 95 of the homes (five sit empty) for $600,000 a year more than the building costs to run, then sell in year five. From the assemblage on, each study opens on that deal and then changes one assumption at a time; drag any handle to test your own. It is a teaching model, not a forecast: the homes are built and fully running from the first year, income never grows, and taxes and selling costs are zero.

Acquire · Real Property

Before the price, settle what is being bought. The deal buys its eight parcels in fee simple: the whole property, with no end date. It will then grant leases, which give a tenant possession of one home for a term and no ownership. A condominium unit is a third interest: one particular unit plus a share of the common areas. A tenancy-in-common share is a fourth: a percentage of the whole that names no particular room. The investors who fund this deal will hold something similar, a share of the whole venture rather than particular apartments. The drawing uses a small 12-unit building so the four interests stay legible; notice that the undivided share hatches the whole building however large it grows.

Illustrative 12-unit building: every unit is 900 ft² and common space totals 2,400 ft². Condominium common interests are assumed equal, one twelfth per unit. Tenancy in common is an undivided interest in the whole property; a lease provides possession for its term. The deed, declaration, lease and applicable law determine the actual rights.

Acquire · Assemblage

Assemblage means buying adjoining parcels to use as one site. The deal’s block is eight 80 × 100 ft parcels at a $300,000 base price each, $2.4M. Bought separately, each parcel keeps a 10 ft setback, a required open strip, along every lot line. Joined, only the outer setback remains, recovering 15,600 ft² of buildable area. Sellers can see that too, so the deal pays a 25% premium over base price: $3M, still 11.1% less per buildable square foot. Past the 40.6% break-even premium, the price has eaten the geometry; a holdout, the last owner who will not sell cheaply, can push it there.

Eight illustrative 80 × 100 ft parcels cost $300,000 each at base price. Separate parcels each have a 10 ft setback on all sides. The joined case assumes internal lot lines and their setbacks can be removed while the outer setback remains. This geometric assumption requires applicable approvals. Costs include acquisition only; joining does not by itself establish allowed floor area or project value.

Fund · Sources & Uses

Every dollar the project uses needs a source. The deal’s uses are $3M of land, $5.6M of hard costs (the construction itself), $1.12M of soft costs (design, engineering and permits, modeled as 20% of hard cost) and a $0.28M contingency held for surprises: $10M. Lenders fund 60% loan-to-cost, the debt as a share of total cost, or $6M. Equity, the owners’ own cash, covers the remaining $4M. Because the loan is sized as a share of cost, an overrun splits the same way. Add $2M of hard cost and soft cost and contingency rise with it, so uses grow $2.5M, to $12.5M: lenders fund $1.5M of the increase and the owners owe $1M more. Equity takes the whole overrun only if the loan stays at $6M; then the owners owe all $2.5M.

Land and hard cost are editable dollar amounts. Soft cost is fixed at 20% of hard cost; contingency is fixed at 5%. These are illustrative budget assumptions. Debt = total uses × loan-to-cost; equity is the exact remainder. The bars share a fixed $24M drawing scale. No interest, fees or profit is included.

Fund · Capital Stack

The capital stack ranks the same $10M by who is repaid first. Senior debt, $4M, is first in line. Mezzanine debt, $2M, is repaid after the senior loan but before any owner. Equity is last: $3M from investors and $1M from the developer. Drag the proceeds handle to preview a sale. At the $12M value the deal earns below, all $10M of principal, the amount each party originally put in, comes back with $2M left over; at $6M, both lenders are whole and the owners recover nothing. Who receives that $2M depends on a waterfall, the agreement that pays cash out tier by tier. This drawing only orders the return of principal.

This illustration recovers principal sequentially: senior, mezzanine, investor equity, developer equity. The equity split stays 3:1. This is an explicit teaching assumption, not a universal equity waterfall. It omits interest, fees, preferred returns and negotiated profit sharing. Proceeds above initial funding are unallocated.

I have been asked by several architects how Real Estate Development "works", and since we are not a verbal species, I am going to do a series of infographics to explain some high-level concepts.

Here is number 1: The Capital Stack

From the original infographic post

Operate · NOI

The 100 homes rent at $1,250 a month: $1.5M a year if every one is leased. Vacancy, five empty homes, costs $75,000. Operating expenses, including property taxes, insurance and running the building, take $825,000 and do not shrink when homes sit empty. What remains is net operating income, NOI: $600,000 a year, earned by the property before any loan is paid. Every later number in this deal multiplies or divides it. Drag vacancy to 25% and NOI halves.

Illustrative 100-home property with uniform monthly rent. Effective rental income = potential annual rent × (1 − vacancy). NOI = collected rent − annual operating expenses. Fixed expenses include assumed property taxes and insurance; they do not automatically fall when homes are empty. Other income and credit losses are zero here. Debt service, capital spending and income taxes are excluded.

Value · Cap Rates

A capitalization rate, or cap rate, is NOI divided by price: the annual income yield a buyer accepts. Rearranged, it prices income. The deal’s $600,000 at an assumed 5% indicates $12M. The deal cost $10M, a 6% yield on cost (NOI divided by total cost), and the one-point gap between what the building yields and what a buyer accepts is the $2M of value the development created. The 5% is an assumption, not a market quote. Drag it to 10% and the same income is worth $6M: exactly the debt.

Capitalization rate = annual NOI ÷ property value; rearranging gives indicated value = NOI ÷ cap rate. This is a one-year, unlevered income comparison. It does not model income growth, sale costs, financing or a full investment return. The displayed 3%, 5% and 10% rates are illustrative comparisons, not current market quotations.

Distribute · Leverage

Leverage, paying for part of an asset with borrowed money, decides how much of the NOI reaches the owners. Count the deal’s first-year asset return as NOI divided by cost, 6%, and charge one assumed 5% interest rate on all $6M of debt; the model does not price the mezzanine separately. Interest takes $300,000 and leaves $300,000, a 7.5% return on $4M of equity. Borrowing helped because the asset earns more than the debt costs. Lenders also test the debt service coverage ratio, DSCR: NOI divided by the year’s loan payments, which are interest only here. The DSCR card shows this deal covering its interest 2.0 times. Step to the 25% vacancy example: NOI halves to $300,000, a 3% asset return on the $10M cost, the same $300,000 of interest takes all of it, DSCR falls to 1.0× and the owners receive nothing. The debt handle is loan-to-cost, the same 60% as in Sources & Uses.

A single-year, interest-only illustration: equity profit = asset profit − interest; return on equity = equity profit ÷ initial equity. Debt is assumed repaid at its original principal; fees, taxes, amortization and exit terms are omitted. Assuming full debt repayment makes calculated equity shortfalls below −100% possible. Debt magnifies positive or negative results according to the spread between the asset return and borrowing cost.

Distribute · IRR

The internal rate of return, IRR, is the annual rate at which the dated cash going in and coming out nets to zero, so it counts when money arrives as well as how much. The deal’s owners pay $4M at year 0, receive $300,000 a year, and sell in year five at the $12M value; repaying the $6M of debt leaves them $6M. That is $7.5M back, a money multiple (cash returned divided by cash invested) of 1.875, and a 14.93% IRR. Hold three more years and $8.4M comes back, yet IRR falls to 11.62%. The sale supplies most of the return: without it, even twelve years of distributions, $3.6M, fall short of the $4M.

Illustrative equity cash flows: one initial payment at year 0, equal nonnegative distributions after operating and financing costs at each year-end, and an additional net exit payment in the final year. IRR solves the discount rate that makes their present value zero. Exit cash is assumed already net of selling costs and loan repayment. Money multiple = total cash returned ÷ initial cash invested. These examples contain one sign change, so any sequence with a positive inflow has a unique IRR; no positive inflow means no finite IRR.

Thanks Andrew Kastanos for teaching me this back in 2014!

From the original infographic post

Now push one assumption through the whole chain. Raise vacancy to 25% in the NOI study and NOI falls to $300,000. That just pays the $300,000 of interest, a 1.0× DSCR (the 25% vacancy example in the Leverage study), so the owners receive nothing each year. At the same 5% cap rate the building is worth $6M, which repays both lenders and none of the $4M of equity. One empty home in four wipes out the owners first, because they stand last in the stack.

That is the architectural point of the series. Debt and cap rates only scale an income; the income they scale comes from the plan: buildable area recovered by the assemblage, homes that lease at the assumed rent, and a hard cost that holds. A $2M hard-cost overrun becomes $2.5M once soft cost and contingency follow it; at 60% loan-to-cost the lenders fund $1.5M of it and the owners $1M. The same $600,000 of NOI is now a 4.8% yield on $12.5M. At a 5% cap rate that building is worth $12M, less than it cost. The drawing sets the deal before the spreadsheet sees it.

End of essay ·
Clifton Harness
·
October 14, 2026
Original on onlyplans.io ›