How to Read a Construction WIP Schedule

If you are the CFO or controller of a construction company, you do not need this schedule explained. You build it. What is worth talking about is the review layer — the questions to run against it before it leaves your office, because everyone downstream is going to run their own.

Your surety recomputes your percentages. Your bank tests covenants off the balances it produces. Your auditor selects jobs from it and works backward. Most WIP problems are not arithmetic failures; they are two judgment inputs that nobody stress-tested.

The Engine, Briefly

For each open contract:

1. Percent complete = costs to date ÷ estimated total cost
2. Revenue earned to date = percent complete × contract price
3. Gross profit earned to date = revenue earned − costs to date
4. Billing position = revenue earned − billings to date, positive is underbilled

Cost-to-cost is the dominant input method in construction, but it is a choice, not a requirement. ASC 606 requires a measure of progress that depicts the transfer of control, and cost-to-cost qualifies only where cost incurrence tracks performance. Where it does not — a job with substantial uninstalled materials, or one where a large early mobilization payment distorts the cost curve — the method needs an adjustment or a different measure.

Two inputs carry nearly all the risk: estimated total cost and contract price. Everything else on the schedule is arithmetic downstream of those two.

A Sample Schedule

Five open jobs at year-end. Each one is here to illustrate a different review issue.

Two footing checks that should run before anything else. Revenue earned less costs to date must equal gross profit earned: $8,515,000 − $7,687,400 = $827,600. And billings less revenue earned must equal net overbillings: $8,650,000 − $8,515,000 = $135,000, which is $420,000 less $285,000.

Backlog is $5,385,000 with $520,400 of profit in it — a 9.7% margin on remaining work, which is among the first figures an underwriter derives.

Estimate to Complete: Where the Real Control Lives

The ETC is prepared by people who are evaluated on the outcome they are estimating. That is not a character flaw, it is a structural conflict, and it needs a control rather than a caution.

Four questions worth answering in your own close process:

Who challenges the ETC, and on what basis? A PM revising their own estimate without a second set of eyes is not a control. The challenge does not need to be adversarial — comparing committed costs to the remaining estimate, or units installed to units remaining, is usually enough to surface an optimistic number.

Which jobs had cost activity but no change in the ETC? Run that comparison every month. A job that burned $400,000 while its estimated total cost held perfectly flat is either genuinely on plan or nobody looked. The report takes ten minutes and finds more problems than any other single procedure.

Are committed costs captured? Executed subcontracts and issued purchase orders represent cost you are going to incur. An ETC built only from historical burn rates will systematically understate.

Is the contingency documented or buried? Estimating contingency inside the ETC is defensible. Contingency that quietly releases into margin as the job progresses is how a portfolio produces steady margin gain, and it means your early-stage jobs are understated right now.

Cost Cutoff: The Understatement Nobody Looks For

Percent complete is only as good as the cost accrual. Costs incurred but unrecorded at period end understate the numerator, which understates percent complete, revenue, and gross profit — and creates paper underbillings that do not exist.

The recurring sources: subcontractor invoices received and sitting unposted, work performed by subs who have not invoiced, material delivered with the invoice in transit, unposted equipment burden, and payroll accrual through the cutoff date. Every one of these is ordinary, and every one moves the schedule.

The related question is consistency between numerator and denominator. If indirect costs and equipment burden are allocated into costs incurred, the ETC has to include them on the same basis. Mixing a burdened numerator with an unburdened denominator produces a percentage that means nothing, and it is a surprisingly common defect.

Margin Fade Analysis

This is the section that tells you how the company is actually running. Compare bid margin to current margin, job by job, every period.

A portfolio bid at 14.4% now carries 9.7%. Two jobs account for the entire $660,000, which is the more useful finding — concentrated fade points at specific estimating or execution failures, while fade spread evenly across a portfolio points at a systemic issue in how work is priced or how change order work gets performed.

Track the trend by job across periods, not just against bid. A job that fades two points a quarter for three quarters is telling you something a single-period comparison will not.

And watch the inverse. Margin that improves as jobs progress usually means original estimates carried padding, which means revenue is being deferred into later periods and every early-stage job on the schedule is understated today.

Loss Jobs

Harbor Point expects a $120,000 loss. Percentage of completion has recognized $78,000 of it at 65% complete. That is not the answer.

When a contract is expected to lose money, the full estimated loss is recognized as soon as it is determinable, not ratably. Harbor Point needs an additional $42,000 accrued so cumulative gross profit on the job equals the full negative $120,000, with the offset to a liability — typically captioned provision for loss on uncompleted contracts. The loss provision requirement for construction-type contracts was retained when ASC 606 superseded the rest of the old construction guidance.

Two practical points. The loss is measured on the total contract, so a job that is profitable overall but has a losing phase does not trigger it. And the accrual unwinds as the remaining work is performed, which means it needs to be recomputed each period rather than left standing.

Missing loss accruals is among the most common defects in contractor-prepared schedules, and it always errs toward flattering the current period.

Contract Price: The Other Judgment Input

Approved change orders belong in the contract price. Beyond that it gets harder.

Unapproved change orders and claims are variable consideration. Including them requires a conclusion that it is probable a significant revenue reversal will not occur, and that conclusion needs support — the owner's written acknowledgment, a documented history of approval on similar work, or a schedule position that makes payment near-certain. "We always get paid eventually" is a belief, not evidence. This is worth its own analysis and it is where I see the most aggressive positions taken.

Retainage does not affect the contract price. Billed retainage is billed, and it counts in billings to date. Whether it presents as a receivable or a contract asset depends on whether the right to payment is conditional on anything more than the passage of time.

Booking It

Most contractor systems record revenue as billings during the period, then adjust to percentage of completion at close. Using the sample totals, and assuming zero opening balances:

Revenue comes down $135,000, from $8,650,000 of billings to $8,515,000 earned.

Then the loss provision:

In periods after the first, the entry adjusts the contract asset and contract liability accounts to the computed balances rather than posting the gross amounts. Booking the gross figures against non-zero opening balances double-counts, and it is a mistake that survives a long time because the schedule still foots internally.

Reconciling the WIP to the General Ledger

The schedule and the trial balance have to agree, and the reconciliation is worth documenting rather than eyeballing:

• Contract assets and contract liabilities per the schedule equal the GL balances after the adjusting entry. No plugs.
• Costs to date for open jobs plus costs on jobs closed during the period tie to cost of revenue in the GL.
• Change in cumulative revenue earned equals revenue recognized for the period. Period revenue is a difference between two cumulative figures, never a column on the job cost report.
• Billings to date tie to the billing register and to AR activity.
• Closed jobs carry zero contract asset and zero contract liability. Any residual is an unresolved closeout, and it usually means a final billing or a final cost accrual is missing.

Monthly Review Checklist

Run these before the schedule goes anywhere:

1. Any job over 100% complete. Costs exceeding the ETC means the estimate is stale — there is no 104% complete.
2. Any job with material cost activity and an unchanged ETC.
3. Identical gross profit percentages across unrelated jobs, which suggests margins derived from a target rather than from the field.
4. One job with a margin well above the portfolio. Sometimes real, sometimes early recognition.
5. Large underbillings on jobs near completion — unbilled change order work, billing lag, or overstated revenue.
6. Closed jobs still carrying over- or underbillings.
7. Loss jobs without an accrual, and standing accruals that were not recomputed.
8. Overbillings growing faster than revenue, which can mean operations are being funded from billings rather than profit.
9. Uninstalled materials sitting in costs incurred and inflating percent complete.
10. Cost coding transfers between jobs during the period, which move margin between contracts and distort fade analysis in both directions.

What Your Auditor Will Test

Knowing the procedures makes the audit shorter and the schedule better. On most construction engagements, expect:

Selection of jobs from the WIP by size, margin outlier, fade, and loss position — then agreement of contract price to the executed contract and approved change orders, and costs to date to the job cost detail and supporting invoices.

Challenge of the ETC, usually by comparing prior-period estimates to actual outcomes on completed jobs. A history of ETCs that consistently proved low is evidence about the reliability of the current ones, and it drives how much work gets done in this area.

Post-period cost testing, looking for costs recorded after year-end that belong before it — the single most effective test of whether percent complete is understated.

Recomputation of percent complete, revenue, gross profit, and the contract asset and liability balances for selected jobs, plus the schedule's mathematical accuracy in total.

Evaluation of unapproved change orders and claims included in contract price, and of the support for the variable consideration conclusion.

Confirmation or inspection of contract terms and, on some engagements, direct communication with owners or architects about completion status.

If your file already contains the ETC support, the cutoff analysis, the fade explanation by job, and the loss computations, most of this is a document request rather than an investigation.

Presenting It to Your Surety and Bank

Three things underwriters do that are worth designing for:

They recompute. If they cannot reproduce your percent complete from your own cost columns, the schedule reads as unreliable — which does more damage than a bad job does. Anything excluded from the calculation, uninstalled materials in particular, needs its own visible column rather than a footnote.

They keep last year's schedule. Fade on individual jobs is visible to them across periods even when portfolio totals look stable. Explain it before they find it.

They adjust working capital. Many underwriters remove some or all of the overbilled position on the theory that it is not a real asset. A company that looks liquid on billings can look thin after that adjustment, and it is better to know your post-adjustment position than to learn it during the annual review.

A short cover memo with the schedule — fade by job with causes, any loss accruals, the treatment of unapproved change orders and stored materials — costs an hour and changes the tone of the conversation.

The Short Version

The arithmetic is four steps. The risk lives in the estimate to complete, the cost cutoff, and what went into the contract price. Everything a reviewer will find comes out of one of those three.

Run the ETC comparison monthly, document the fade by job with causes, reconcile the schedule to the ledger rather than to itself, and put an explanation in front of anything a stranger would find odd. That is most of the job.

Mario Castro, CPA is the Managing Member of Castro CPA, PLLC in Celina, Texas. The firm performs audits, reviews, and compilations for construction contractors across Texas.

If you are building out a WIP review process — or want a second set of eyes on the schedule before it goes to your surety — that is a conversation worth having before year-end rather than during fieldwork. Learn more about construction CPA services for contractors, or read what your surety underwriter actually does with your financial statements.