Product
Platform AI agents Security & governance
Who it's for
Asset management Acquisitions Finance & IR
Company
Resources PricingTrust Center Read the thesis Book a demo →
Analysis

The real cost of a 10-day close

For roughly half of institutional real estate teams, the books take six to ten days to close. The cost everyone counts is analyst overtime. The cost that actually matters is the one nobody puts on a slide: every decision made during those ten days is made on numbers that are not yet trusted.

Built AI · March 2026 · ~7 min read
Built AI
Analysis · March 2026 · 7 min read
CloseFinance
Key takeaways
  • A ten-day close is a recurring blackout: for a third of every month the firm operates without a trusted picture of its own performance.
  • The close drags because reconciliation is manual, serial, and dependent on systems that hold different versions of the same truth.
  • The real bill has four lines, labor, stale decisions, eroded investor trust, and burnout, and only the first shows up in a budget.
  • Continuous reconciliation moves the work off the critical path, turning a ten-day build into a roughly three-day review.

A ten-day close is not a bookkeeping inconvenience. It is a ten-day window, every single month, in which your firm operates without a reliable picture of its own performance. For the first third of the quarter, the asset managers, the controller, and the CIO are all working from a prior period that has gone cold and a current period that has not yet set. The business keeps moving. The numbers do not catch up until the close lands, and by then the decisions have already been made.

The reason this persists is that the close looks like a finance problem, so it gets measured as one: days elapsed, hours logged, headcount assigned. Measured that way, a ten-day close is merely expensive. Measured by what it costs the rest of the firm, it is something worse. It is a recurring blackout.

Why the close drags

Walk the actual sequence and the drag is obvious. Nothing in it is hard. All of it is manual, serial, and dependent on the step before it being finished and correct.

The general ledger has to be reconciled against the property management system, because the two disagree more often than anyone admits: a charge posted in one and not the other, a payment applied to the wrong period, a recovery accrued in the PMS that the GL has not seen. The bank has to be reconciled against the GL. Intercompany balances between entities and funds have to net to zero, and when they do not, someone has to find the missing leg. Accruals have to be booked for the invoices that have not arrived yet but belong to the period. And then, only then, can anyone write the variance commentary that explains why actuals diverged from the plan, which is the part the CIO and the investors actually read.

The close sequence
Five manual steps, each waiting on the one before it
1
GL ↔ PMS recon
Reconcile the general ledger against the property management system where the two disagree.
2
Bank recon
Reconcile the bank against the GL once the ledger is settled.
3
Intercompany
Net intercompany balances between entities and funds to zero, finding any missing leg.
4
Accruals
Book accruals for invoices that have not arrived but belong to the period.
5
Variance commentary
Only then write the commentary that explains why actuals diverged from plan.
Each step is serial and dependent on the prior one being finished and correct. A single unexplained variance late in the chain can send the whole close back two days.

Each of these steps waits on the one before it. Each is owned by a person copying figures between systems that were never designed to agree. A single unexplained variance late in the chain can send the whole close back two days. The process is not slow because the people are slow. It is slow because it is a manual reconciliation of systems that hold different versions of the same truth, performed under deadline, once a month, forever.

A ten-day close means that for a third of every month, the firm is steering by a rear-view mirror that has not finished developing the photo.

The costs nobody books

The visible line item is labor: experienced controllers and analysts spending the first third of every month on reconciliation rather than analysis. That is real money, and it is the smallest part of the bill.

Stale decisions

The expensive cost is decisional. Capital calls, distribution timing, covenant headroom, hold-or-sell judgments, leasing concessions, all of these get made on a continuous basis, not on the day the close finishes. When the close runs ten days, every decision made in that window is made on numbers that are either stale or provisional. Usually it is fine. Occasionally a distribution goes out that the final numbers would have sized differently, or a covenant drifts closer to a test than the provisional figures showed. The risk is invisible precisely because it only bites sometimes, which is exactly why it is tolerated.

Eroded investor trust

Investors notice cadence. A firm that can answer a position question in an afternoon reads as in control. A firm that responds with "we will have that once the quarter closes" reads as one that does not have its own numbers at its fingertips. Over a relationship, the firms that report fast and report live accumulate trust, and trust is the currency of the next fund. The close speed is not just an internal metric. It is visible to the people deciding whether to re-up.

Burnout and retention

There is also a human cost that compounds. The monthly grind of reconciliation is the work analysts least want to do and most resent, because it is rote, high-pressure, and invisible when it goes right. It is also the work that drives the best people out, the ones who joined to underwrite and analyze, not to chase a charge across two systems at 9pm on close day. A ten-day close is a recurring retention tax on exactly the talent you can least afford to lose.

What the close really costs

Count it honestly and the bill has four lines, not one: the labor of the people doing it, the decisions made on stale numbers during the window, the investor trust lost to slow answers, and the analysts who leave because reconciliation is the worst part of the job. Only the first line shows up in a budget.

How a continuous close compresses it

The way to fix a ten-day close is not to do the same ten days faster. It is to stop treating the close as a once-a-month event and start treating reconciliation as a continuous capability that runs in the background, so that close day is a review, not a build.

That requires the underlying data to be unified continuously rather than assembled monthly. When the GL, the PMS, and the bank feeds are normalized into one model that reconciles itself as data arrives, the conflicts surface the day they occur, not on day six of the close. A charge that posts in the PMS and not the GL is flagged in near real time, while the context is fresh and the fix is small, instead of being discovered three weeks later as one of forty mysteries to resolve under deadline.

On top of that continuously reconciled foundation, the rote drafting can be done before the human ever opens the period. A close agent can reconcile GL, PMS, and bank, draft the accruals that belong to the period, and write a first pass of the variance commentary, every figure traced back to its source, then pause and wait. The controller does not start the close from a blank page. They open a drafted, reconciled, sourced close and spend their time on the judgment calls that actually need a controller: which accrual estimate is right, how to characterize a genuine variance, what to escalate. Nothing posts and nothing is final until they sign off. The machine does the reconciliation; the human owns the close.

The effect is to move the bulk of the work off the critical path. A close that took ten days because ten days of manual reconciliation had to happen in sequence becomes a close of about three days, because most of the reconciliation already happened continuously and the rest arrives pre-drafted for review.

Time to close
From a ten-day build to a three-day review
Today
10 days
With Built AI
3 days
Most of the reconciliation already happened continuously, so close day is a review of drafted, sourced figures rather than a build from a blank page.
You do not get a faster close by working the close harder. You get it by making reconciliation a thing that happens every day instead of a thing that happens for ten days a month.

What a three-day close unlocks

Compress the close and the blackout window shrinks with it, which changes more than the finance team's month.

  • Decisions on current numbers. Capital calls, distributions, and covenant judgments get made against a picture that is days old, not weeks. The provisional-number risk does not disappear entirely, but the window in which it lives drops from a third of the month to a few days.
  • Faster, more credible investor answers. When the books are nearly always close to current, a position question gets an answer in an afternoon rather than after the quarter. That cadence is what institutional investors increasingly expect, and what differentiates the firms they trust.
  • Analysts on analysis. The team that spent the first third of every month reconciling spends it interpreting instead. The work gets more interesting and the people stay longer.
  • A continuous, not quarterly, capability. The same reconciled foundation that compresses the close is what makes live investor reporting and continuous variance analysis possible at all. The close is the gateway capability.

A ten-day close is the most expensive cheap problem in institutional real estate: small on the budget, large everywhere else. Built AI compresses it by unifying the GL, PMS, and bank into one continuously reconciled model, and by letting a close agent draft the reconciliations, accruals, and variance commentary, every number cited to source, for a human to review and approve rather than build from scratch. The machine reconciles; the controller decides; nothing is final without sign-off. To see your own close walked from ten days toward three, see how it works for finance and IR or book a walkthrough.