In one worked reconciliation, expenses came in 17% under budget and the landlord still owed one tenant $100,000 while collecting $300,000 more from another. The arithmetic was fine. The setup was not. Here is where reconciliations actually break, and why a spreadsheet cannot tell you.
You budgeted electricity at $500,000 and spent $1,000,000. During the year you billed tenants estimates based on the budget. Now you recover the difference. That is the entire exercise: what you billed versus what you should have billed, tenant by tenant.
Two points follow that most guides skip. First, you reconcile against what was billed, not what was collected. Collections are a receivables problem; the reconciliation books what each tenant should have been charged. Second, the swing is two-sided and the two sides move independently. In Kardin’s worked example, actuals came in 17% below budget, and the landlord still owed one tenant $100,000 back while collecting an extra $300,000 from another. Occupancy had changed. Expense variance alone told you nothing.
That is the first thing to understand about CAM reconciliation errors: the number that looks wrong is usually correct for the inputs it was given. The inputs are where to look.
Every recoverable expense belongs to a pool, and every tenant who owes a share has to be assigned to it. When a tenant is missing, the pool total is still fully allocated, so everyone else’s pro rata share is a little too high, or the landlord absorbs the gap as leakage. Nothing flags it, because the math balances.
The most common way this happens is timing. A reconciliation needs two rent rolls, a January one and a December one, because only together do they capture every tenant who occupied space during the year. Reconcile off a single year-end rent roll and everyone who moved out mid-year disappears, along with the estimates you billed them, which still have to tie.
A related version: a tenant is renamed in place rather than expired and recreated, so the suite carries a stale ID and the new lease terms never attach. Or a new expense account is added mid-year without a category, so it never calculates at all.
Controllable expense caps are the single most expensive place to be wrong, because the error compounds year after year.
There are three cap methodologies, and the lease decides which applies: cumulative and compounded, where the cap grows on the prior year’s cap; cumulative but not compounded, where it grows by a fixed dollar amount; and non-cumulative, where it grows over the lesser of last year’s actual controllables or last year’s cap, and so ratchets down over time. When the lease is silent, cumulative and compounded is the default assumption. Most controllable expenses come in under the cap most years, so under the non-cumulative method the cap shrinks every year. Practitioners call this cap decay, and it is rarely modeled.
The error that costs the most is carrying the original lease cap through a base year reset. When a base year resets, the lease is saying that year’s rent covers that year’s expenses, and the controllable cap resets from that year as well. Carry the old cap forward and you under-recover on every subsequent reconciliation for the life of the lease. Kathy Lim, a CRE veteran of 30 years, put it plainly in a BOMA Power Hour session on reconciliation best practices:
“That’s wrong. You’re losing money for the landlord on that basis.”
Two more cap details that produce disputes: a cap is a ceiling, not a floor, so it is not an amount you are entitled to bill when expenses come in lower. And carryover language, which lets unrecovered dollars above the cap be recaptured in a later year with headroom, is not standardized. It reads differently depending on which lawyer, broker and asset manager wrote the lease. A template that assumes one form of carryover across a whole building is wrong on some of the leases.
Gross-ups adjust variable expenses to what they would have been at a stated occupancy, usually 95% occupied. The lease language matters: occupied, not leased. Done correctly, it is what the lease allows. Done with the wrong method, it overstates or understates a pool for every tenant in the building.
The most common mistake is using a simple gross-up, cost divided by occupied square feet extended to 95%, on utilities. Utilities have a fixed baseline load. Even at full vacancy the chillers, elevators, lobby and parking lot lighting run. A simple gross-up treats every dollar as variable, so electricity comes out overstated. The right approach applies a vacancy consumption rate, and that rate is a judgment call that varies by building: 30 to 50% for general office electricity, lower in an efficient building with occupancy sensors, higher with twenty-year-old HVAC in a hot climate. For water, the answer depends on what is using the water. An industrial building with a small office component and landscaped grounds has a vacancy consumption rate of 50 to 75%, because the water bill is an irrigation bill. A downtown office tower with no exterior landscaping might be 5 to 10%.
Two more that surface at reconciliation. Post-pandemic rent rolls broke the management fee gross-up in many buildings: when old leases are at $30 a foot and new ones at $50, or a large tenant is burning off free rent, a blended gross-up understates the fee. Free rent is an invisible leak. And some costs should never be grossed up at all. A day porter is a day porter whether the building is 60% or 95% occupied; you would not hire a second one as the building fills. The rule is simple: do not gross up a cost you would not actually add.
This is the highest-value mistake in Kardin’s entire training archive, and it is a budgeting error that only surfaces at reconciliation.
When a tenant renews, rent, tenant improvements and leasing commissions are carried forward from the renewal assumption. Expense reimbursements are not, unless someone sets them up again. The renewed tenant sits on the rent roll paying rent, absent from every recovery pool, and the reconciliation dutifully allocates their share to everyone else or to the landlord. Renewals get signed all year; the damage shows up the following spring, when the reconciliation is run.
The same family of error includes mid-year acquisitions, where the months before ownership have to be zeroed and occupancy overridden, or the unowned period falls into vacancy leakage as phantom underbilling. And mid-year changes to a tenant’s recovery method, which need two rent roll rows, because the lease dates drive the recovery language.
Two things have to tie before a reconciliation means anything. The starting expense pool has to equal the income statement. And the estimates billed, as keyed into the reconciliation, have to equal the accrual recoveries line on the income statement. If either is off, every tenant statement is off by a share of the difference.
The second one fails more often than it should, because invoicing is not billing. Producing a tenant statement does not tell accounting what to book. The handoff between the reconciliation and the accounting system is a step, not a side effect, and in a spreadsheet workflow it is the step most often skipped.
A smaller version of the same problem: rounding. A calculation runs on 26.404 and the statement shows 26.4, so the statement and the calculation disagree by cents, and tenants do not like rounding. Display precision and calculation precision have to be managed separately.
None of the five errors above is arithmetic. A spreadsheet will compute a pro rata share on the wrong denominator perfectly. The failure is structural.
A CRE budget has three components in a fixed order: rental revenue first, then expenses and capital, then recoveries, which depend on both. Kardin co-founder Karen Schwartz describes it this way:
“Once the rent roll is in place and all of your expenses are in place, then you can set up all of the recoveries.”
Recoveries are downstream of everything else by construction. A spreadsheet does not enforce that dependency. When occupancy changes, or a tenant renews, or a base year resets, the recovery tab does not know unless someone tells it, and in a fifty-tab workbook maintained by three people across a year, someone usually does not. It is one of the clearest signs a team has outgrown its budgeting spreadsheet.
A purpose-built tool fails differently. It does not guess the cap methodology for you, and it should not, because it cannot read the lease. What it does is keep the recovery logic attached to the tenant and the pool, so that when the rent roll changes, the exposure recalculates, and the roll-forward from budget to reforecast to reconciliation carries the caps, exclusions and adjustments with it instead of asking someone to rebuild them.
Three habits, all from practitioners rather than software vendors.
Reconcile against an audit report, not a total. In Kardin the recoverable operating expense leakage report exists for this. Its two heuristics are worth stealing whatever tool you use: unexpectedly low CAM recovery usually means schedules are not set up, and an unexpectedly high cap adjustment usually means a keying error, an extra zero somewhere.
Trace any wrong number the same three hops. Start at the recoveries calculation and spot the anomaly. Go to the tenant’s recovery detail to learn the pro rata share, the reimbursement method and, critically, which schedule the tenant is assigned to. Then go to the cap calculation or the expense account to see what is actually driving it. The middle hop is the one most people skip.
Count periods, not payment dates. Twelve monthly bills, four quarterlies. What you are looking for are anomalies, an eleventh month or a fifth quarter, not which year a check cleared.
A tool does not know what your lease says. It will not grow a cap for you, decide a vacancy consumption rate, or tell you whether a renewal carried its reimbursements. What it can do is make those decisions visible and keep them attached to the tenant from budget through reforecast to reconciliation, so that a decision made once in August is still in force the following March. That is the difference between a reconciliation that is wrong in an auditable way and one that is wrong in a way nobody finds until a tenant does.
If your reconciliation process still starts from a blank spreadsheet each spring, see how Kardin handles recoveries, or ask us to walk through your setup. Kardin works alongside Yardi, MRI, RealPage and other accounting systems rather than replacing them; the reconciliation exports back to your system for billing.
Budget with clarity.
Mostly setup rather than arithmetic: tenants missing from an expense pool, controllable caps carried through a base year reset, gross-ups that apply a simple method to utilities with a fixed baseline load, reimbursements that dropped off when a tenant renewed, and estimates that were never tied to the accounting system. The calculation reproduces whatever setup it is given.
Because the reconciliation is tenant-level and driven by occupancy and lease terms, not by the expense total. If a large tenant moved out and a smaller one moved in, the pro rata shares changed regardless of what happened to expenses. In Kardin’s worked example, a 17% expense saving coexisted with a $100,000 refund to one tenant and a $300,000 collection from another.
The reconciliation depends on two rent rolls, year-end financials and the billed estimates from the accounting system, and each has to tie. In a spreadsheet, every one of those is a manual import, and every correction is a re-import that can overwrite other work. The handoff back to accounting, booking what each tenant should have been billed, is a separate step that a spreadsheet does not produce. Each of those is a place the close waits.
By keeping the recovery logic attached to the tenant and the pool rather than to a cell. Caps, exclusions, gross-up methods and adjustments roll forward from budget to reforecast to reconciliation, so they are set once rather than rebuilt each cycle. Audit reports flag leakage by source: vacancy, base years, caps and free reimbursements. The tool does not read the lease for you; it makes sure what you decided is still in force.
The gap between recoverable operating expenses and what is actually recovered. Four named sources: vacancy, base years, caps and free reimbursements. Some leakage is contractual and expected. In Kardin’s demo portfolio roughly half of recoverable expenses went unrecovered, which is why the leakage report is treated as an audit tool: an unexpected number usually means a setup problem, not a market one.