The Bottom Line
A board that defers a capital repair has not saved money. It has borrowed money — from the building — at an interest rate nobody disclosed, on a note that appears on no balance sheet, and that no one ever voted to accept. The rate has two components: construction costs escalate every year you wait, and deterioration compounds, so the scope of the eventual job grows too. Together they routinely produce an implicit rate well above what the association could have borrowed at from a bank. The board avoided debt by taking on the most expensive debt available to it.
A deferral is a borrowing decision, and it is the only one a board makes without a vote
Every board understands that taking a loan is a financing decision. It requires a resolution, a lender, a rate, an amortization schedule, and usually a fight. What almost no board recognizes is that voting to postpone a scheduled capital repair is also a financing decision, and structurally identical: the association keeps cash this year and agrees to pay a larger sum later. The difference is not economic. It is that the loan from the bank is written down, and the loan from the building is not.
The economics are unremarkable. Any deferral of a cost into the future is a borrowing, and the price of moving money across time is an interest rate — the same discounting logic Case, Fair and Oster apply to public deficits, where postponing an obligation does not eliminate it but shifts it, with interest, onto a later period. What makes the association's version dangerous is not that the concept is subtle. It is that the note is invisible. There is no lender to send a statement, no covenant, no amortization table, and no line on the balance sheet. The obligation accrues anyway, and the only party who will ever see the payoff figure is the board sitting in the room three or four turnovers from now.
The useful exercise is to make the note visible, which means deriving the rate.
The first component of the rate: construction costs escalate, and they escalate faster than prices in general
The replacement gets more expensive every year you wait, and it gets more expensive faster than the general cost of living. This is not a forecast; it is a structural feature of how construction inputs are priced. Case, Fair and Oster describe the mechanism through producer prices — the prices of inputs at earlier stages of production, which move before and more sharply than the finished-goods prices consumers see. Construction inputs sit upstream and they are concentrated: a small number of commodities (steel, cement, asphalt, copper, resins, lumber) plus a skilled-trade labor market with a slow supply response. Upstream, concentrated, supply-inelastic inputs escalate faster than a broad consumer basket. That is why a producer price index read in October tells a board more about the roofing bid in March than any general inflation number will.
The worked example below uses a 5% annual escalation assumption. It is illustrative and unremarkable, the point of the exercise does not depend on it, and a board should run the arithmetic with whatever escalation rate its own reserve study carries.
The second component: the cost curve is convex
Escalation alone is the smaller half of the rate. The larger half is that the thing being deferred is not sitting still — it is deteriorating, and deterioration does not proceed in a straight line. It bends.
Consider what happens when a roof membrane fails and is left in place. In year one the job is a membrane replacement. Water then reaches the sheathing, and the job is a membrane plus sheathing. The sheathing stays wet and the framing goes, so structural carpentry enters the scope. Water reaches the ceiling below, adding drywall, insulation and finishes. It reaches an owner's contents, and now there is a claim. The claim reaches the carrier, and now there is an insurance posture — a loss history, a deductible conversation, possibly a non-renewal. Behind that sits a litigation posture.
The cost of the same repair does not rise in a straight line as you defer it, because it is not the same repair anymore. Every year of delay adds scope. That is what makes the cost curve convex, and convexity is why the arithmetic of deferral is so much worse than boards expect. A board that models deferral as "it will cost a bit more next year" is modeling escalation and ignoring convexity — and convexity is the part that ruins them.
Magnolia's pool, at years 1, 3, 5, and 10
Magnolia Recreational HOA carries pool resurfacing in its reserve plan at $84,500, scheduled for April FY2025. Suppose the board defers it to protect the operating surplus.
Magnolia Recreational HOA is a composite illustration built for the FOAM series. The community, its vendors, and its financial institutions are fictional; the structures and the arithmetic are real.
Two assumptions drive the model, both stated openly so a reader can substitute their own. Escalation runs at 5% per year. Added scope — the convexity — is a multiplier on the job: 1.00 in year one (the shell is still just a shell), 1.15 by year three (tile and coping have joined the scope), 1.40 by year five (shell repair and plumbing penetrations), and 2.10 by year ten (structural shell repair, deck replacement, equipment damaged by chronic leak-driven cycling).
| Deferred by | Cost if escalation alone | Added-scope multiplier | Actual cost of the job | Premium over doing it now | Implicit annual rate |
|---|---|---|---|---|---|
| Do it now | $84,500 | 1.00 | $84,500 | — | — |
| 1 year | $88,725 | 1.00 | $88,725 | $4,225 | 5.0% |
| 3 years | $97,819 | 1.15 | $112,492 | $27,992 | 10.0% |
| 5 years | $107,846 | 1.40 | $150,984 | $66,484 | 12.3% |
| 10 years | $137,642 | 2.10 | $289,047 | $204,547 | 13.1% |
The right-hand column is the number that should stop a board. It is not a metaphor. It is the compound annual rate that turns $84,500 into the eventual cost of the work, derived from nothing but the two assumptions stated above.
The implicit rate, measured against the rates the association could have chosen instead
The board's stated reason for deferring is almost always that it does not want to take on debt. That reasoning inverts the moment the rate is on the table.
The board deferred the work to avoid taking on debt, and in doing so took on the most expensive debt available to it. It could have funded the work from reserves, at a true cost of the 3.5% the fund would otherwise have earned — the blended yield on Magnolia's certificate ladder. It could have borrowed from a bank near 7.5%, with a schedule, a disclosed rate, and a vote. Instead it borrowed from the building at 12.3%, without recording the transaction anywhere.
Least-cost decisions: the right way to compare patch against replace
Boards resist this arithmetic because they suspect it is rigged — that "replace on schedule" is being compared against a straw man in which nothing is done at all. It is not. The honest comparison is against what boards actually do, which is patch and defer, and there is a standard tool for making it rigorously.
Managerial accounting has an explicitly labeled category for exactly this. Garrison calls them least-cost decisions: capital choices in which no alternative produces revenue on either side, so there is no return to compute and no profit to compare. The criterion collapses to something simpler — discount every outflow in each alternative back to today, and choose the one with the least-negative net present value. The machinery survives intact. Only the sign changes. This matters more for associations than for any firm, because an association is a permanent least-cost environment: it never sells anything, and nearly every capital decision it makes is a choice between two ways of spending.
Apply it to Magnolia's pool. Option A resurfaces on schedule: one outflow of $84,500 today. Option B patches and defers for five years: $9,000 in each of years one and two, $14,000 in each of years three and four as the failures get harder to hold, then the full job in year five at the $150,984 established above. Routine pool service is identical under both options, so it does not differ between the alternatives and is properly excluded. Discount at 3.5% — the association's opportunity cost of funds, being what the money would have earned had it stayed in the reserve.
The option that was chosen because it was cheaper has a present value of cost of $169,050, against $84,500 for the option that was rejected because it was expensive. The board spent $46,000 on patches in order to avoid an $84,500 job, and then did the $84,500 job anyway — except that by the time it did, the job cost $150,984.
Payback is the reason boards keep choosing the roof that fails first
One more mechanism is worth naming, because it operates quietly. When a board asks "what is the payback period on this roof?" it has already guaranteed the wrong answer.
The payback rule measures how long an investment takes to return its cost, and then stops looking. Garrison is direct about the consequence: payback is blind to everything after the payback point. It cannot see longevity, residual value, or the failure that arrives in year twelve. A rule that ignores everything past the cut-off will systematically prefer the option that returns cash soonest — which, in a capital-components context, means the cheaper, shorter-lived option. A board that demands a payback period for a roof will always conclude the roof is a bad investment, and that conclusion is a property of the tool, not a fact about the roof.
Where the arithmetic stops
Everything above is arithmetic, and arithmetic is something a board can do for itself. What it cannot do is establish the inputs. Whether a component is deteriorating, how much useful life it has left, what it will cost to replace, and what a community's reserve contribution ought to be are not questions a board should answer from the boardroom. They are established by a reserve study prepared by a qualified professional, and by engineers who have examined the actual component. This article explains what deferral costs once someone competent has told you the repair is due. It cannot tell you that yours is.
The division of labor is the point. The professional supplies the condition, the life, and the cost. The board supplies the decision. What has been missing is a way for the board to price its own decision, and that is the gap this arithmetic fills.
What to do with this
- Price every deferral before you vote on it. Require the same three numbers you would require of a loan: the amount, the term, and the rate. The rate is the compound rate that grows today's cost into the cost on the day the work actually gets done. If nobody can produce it, the board is voting on a financing transaction it has not priced.
- Put the escalation assumption in writing. Ask your reserve professional what escalation rate the study carries, minute it, and use that number — not a general inflation figure — in every deferral discussion for the year.
- Ask for the added-scope estimate separately. Escalation is the easy half. Ask the professional: what enters the scope of work if we wait three years, and what enters it if we wait five? The answer is the convexity — the half boards never model.
- Compare alternatives on present value of cost, never on this year's cash. Discount every outflow in each option back to today at the rate your reserve fund earns, and take the smaller number. "Patch and defer" and "replace on schedule" are both spending plans and must be compared as such.
- Ban payback as a decision criterion, in writing. Keep it as a liquidity screen if you find it useful, but never let it rank two capital alternatives. Put the distinction in the board's financial policy so it survives the next turnover.
- When you do defer, minute it as a financing decision. Record the amount deferred, the estimated cost on the deferred date, and the implicit rate. The next board is entitled to know what note it inherited.
Sources and further reading.
- Garrison, Noreen & Brewer, Managerial Accounting, 16th ed. (McGraw-Hill, 2018), ch. 13 — least-cost decisions (the criterion for capital choices that generate no revenue on either side, where the correct rule is the least-negative net present value), and the structural blindness of the payback method to all cash flows beyond the payback point.
- Case, Fair & Oster, Principles of Macroeconomics, 13th ed. (Pearson, 2020), ch. 7 — producer prices and the mechanism by which upstream input costs move earlier and more sharply than general consumer prices; ch. 14 — the discounting logic of deferring an obligation into a later period. Cited for mechanism only; the text predates the 2021–23 construction cost surge and no rate in this article is drawn from it.
- Quiry, Dallocchio, Le Fur & Salvi, Corporate Finance: Theory and Practice, 4th ed. (Wiley, 2014), ch. 34 — the expropriation effect, in which value is transferred from one class of claim-holders to another with no exchange of flows. The deferral note in this article is drawn on owners who have not moved in yet.