How the numbers are worked out

Given the component costs and remaining lives your association or its reserve professional supplied, we show the contributions and trade-offs that avoid a cash shortfall.

We do not perform a reserve study, determine compliance with any statute, or decide whether a condominium project is eligible for any form of financing. Several states require a study prepared or verified by a qualified professional; nothing here substitutes for that.

Every amount is stored and computed in whole cents as an integer. Nothing in the model touches a decimal dollar, because binary floating point cannot represent one exactly and the errors accumulate across thirty years of compounding.

Money rounding goes through one function, which rounds half away from zero — so a credit of fifty cents and a charge of fifty cents round to the same magnitude. The usual Math.round is asymmetric about zero and would not. Whole-year and whole-dollar rounding elsewhere (the contribution solver searches in dollars; per-unit monthly figures are rounded to the cent) is ordinary and stated where it happens.

Any amount split between units goes through the largest-remainder method. The shares always sum to exactly the total, and the split is deterministic — the same inputs always produce the same table. A per-unit column that does not add up reads as broken even when the total is right.

A component is scheduled in assessed year + remaining useful life. If that year is already past, it is scheduled in the first year of the plan and flagged as deferred maintenance rather than quietly ignored.

Components marked as recurring are then rescheduled every full useful life until the horizon ends. A component whose first replacement falls beyond the horizon still accrues toward the fully funded balance — the association is still using it up.

A cost you enter is a cost in today’s dollars. To spend it in a future year we compound it:

cost in year Y = cost today × (1 + rate)^{Y − base year}

One general rate applies unless you override it per category. Roofing and paving genuinely do not inflate alike, but a single rate is the right default — the dominant uncertainties in any small association’s plan are the replacement cost and the remaining life, not whether roofing runs half a point above asphalt.

Interest is credited on the balance available to earn it. Which balance that is depends on a timing convention you choose. Our default assumes contributions arrive and projects are paid through the year, so on average about half of the year’s net cash flow earns a return:

interest basis = opening balance + (net cash flow ÷ 2)

This is our modelling choice, not an industry standard. The National Reserve Study Standards do not prescribe an interest-timing convention. You can select start-of-year or end-of-year instead, and whichever is in force is printed in the packet.

Interest earned by an association is generally taxable, so we apply an effective rate you control. A residential association filing Form 1120-H is generally taxed at 30% on homeowners-association taxable income — but that is computed after directly connected deductions and a $100 specific deduction, and the election is annual. A regular Form 1120 generally uses 21%. State tax may add a further layer.

We therefore model a single editable effective rate, default 30%, and we do not choose the form for you or account for those deductions. For most small associations this assumption matters far less than the replacement costs and remaining lives.

Interest is never credited on a negative balance. A negative reserve balance is a modelling signal that you are short, not a real overdraft, and compounding a phantom borrowing cost onto it would overstate the problem and corrupt the shortfall figure.

The fully funded balance is what the fund would hold if every component had been saved for exactly on schedule. It is the sum, across components, of the portion of each one’s life already used up:

fully funded balance = Σ (replacement cost × effective age ÷ useful life)

Effective age is useful life − remaining useful life, which is the reserve-study definition. It is not necessarily how old the thing is: a roof inspected and given more life left gets a lower effective age than its installation date implies, and that is the intent.

Percent funded is then simply your balance divided by that figure.

Both the headline figure and every row of the projection are measured at the start of the fiscal year, against that year’s opening balance, in that year’s dollars. One convention throughout, which is also how a reserve study presents percent funded.

Two details follow from it and are worth stating, because they are the places a hand-built spreadsheet usually differs:

• A component due this year is fully accrued this year. The association should have saved all of it by the time the year opens.
• Costs are converted to the year being valued, including the headline figure. A cost you entered in 2020 dollars is not a 2026 replacement cost, and treating it as one would understate the liability by every year of inflation in between.

An earlier version measured the projection at year end and the headline at year start. That produced a visible discontinuity, and a real defect: a component with a one-year life replaced annually had an effective age of zero at every year end, so the fund read as permanently fully funded while spending money every year.

The conventional bands — under 30% weak, 30 to 70% fair, 70% and above strong — are an industry convention for gauging special-assessment risk. They are not a rule, a grade, or a compliance state, and they are not part of the reserve-study standards. A well-funded association with a large near-term project can be in worse shape than a less-funded one with well-timed contributions, which is why we always show the projected low balance beside them.

We solve for the contribution that achieves each of the three goals current industry guidance recognises. The search is over whole dollars, so the number you are shown is one that was actually run through the model and checked. Where an increase is needed, it is the smallest whole-dollar amount that works. Where your current contribution already satisfies the goal, we report it unchanged rather than searching for something lower.

Baseline funding — cheapest, highest risk

The balance never goes below zero

Threshold funding — you choose the floor

The balance never goes below a dollar floor, or percent funded below a floor after the first year

Full funding — our default deadline is 10 years, which is not a standard

Reach 100% funded within N years and stay there

The ten-year deadline is our default, not an industry standard — the standards describe full funding as attaining and maintaining reserves at or near 100% and name no period, and requiring exactly 100% is stricter than “at or near”. You can set your own. But some deadline is load-bearing, and it is worth explaining why. Without one, “be fully funded” is satisfied by almost any plan that limps along for two decades and then overshoots, since the tail of any escalating contribution is naturally flush. On our own example building, the baseline plan reaches and holds 100% funded from 2046 — while passing through 0% funded in 2033. Calling that full funding would be indefensible. Requiring 100% immediately is the opposite error: it means funding the whole accrued liability at once, which for a six-unit building can be several times its annual budget.

The goals are not a ladder. Holding a floor in every year is most expensive in the early years, when the accrued liability is largest and nothing has been replaced yet, while full funding is measured at a deadline. So a 30% floor held throughout can genuinely cost more than reaching full funding. We show all three and never rank them.

There is no “statutory” funding goal, though older reserve-study material named one. State requirements are not a single algorithm — Hawaii prescribes named funding methods, Florida’s structural-integrity regime prescribes covered components and professional involvement, Illinois sets a reasonableness standard. Collapsing those into one button would produce a number that looks statutory and is not. Where you have a requirement to meet, enter it as a threshold.

Each cash event is allocated on its own basis. Dues can be split by ownership percentage while a deck replacement is shared equally between only the four units with decks and a loan is serviced by square footage — all in the same plan. Allocating the yearly total instead would silently blend those bases together.

Where percentages in your declaration do not total 100%, we split proportionally and say so rather than refusing to calculate. A declaration that does not add up is common, and it is yours to reconcile.

Reviewing a condominium project, a lender often looks at what share of the budget goes to reserves. That share is itself part of the income it is measured against, so the contribution needed is not simply a percentage of the operating budget:

contribution ≥ operating budget × share ÷ (1 − share)

For a 10% share that is the operating budget divided by nine, not by ten — a difference of about 11% of the answer, and a common hand-calculation error.

The threshold itself is not a constant, so we never treat it as one. It is a table keyed by the date a loan application is dated, and every figure we print carries the date the guidance was read:

CurrentlyGuidance read July 30, 2026

10%

Applications dated August 3, 2026 or laterGuidance read July 30, 2026

10%

Applications dated January 4, 2027 or laterGuidance read July 30, 2026

15%

We will tell you where your figures sit relative to the cited guideline. We will never tell you a project is eligible, approved or warrantable. That determination belongs to a lender and depends on insurance, litigation, delinquency rates, owner-occupancy, commercial space, needed repairs and project type — none of which are in a reserve plan. Many small projects are also exempt from full project review entirely.

The calculation engine is pure: no clocks, no randomness, no network. The same plan always produces identical numbers, so a forecast regenerated in a year matches the original figure for figure. (A regenerated packet differs in one place — the preparation date printed on it.) That determinism is what makes “check our arithmetic” a real offer rather than a slogan.

Every figure in the forecast opens on screen to show its inputs and its arithmetic. Beyond that, you can export a spreadsheet whose cash flow is built from live formulas — not a dump of results — and recompute it independently: change a cost or a rate and the balances move. The fully funded balance and percent funded columns are exported as values rather than formulas, and the workbook says so on its own first sheet. Your full data is exportable on every plan, including the free one.

Before public launch the model is reviewed by a reserve professional or association accountant. Until that review is published, treat this page as our own description of our own work.

Everything on this page is computed live from the worked example, so it cannot drift from what the product actually does.

Opening balance

$41,500

Fully funded balance, start of first year

$242,545

Lowest balance in the projection

($326,289) in 2054

First year the fund runs out

2029