Chapter 34 — Key Takeaways

A one-page reference card. Every ratio threshold below is an approximate industry heuristic, not a rule — sureties and banks each set their own, and they move with the market and with your sector. Accounting treatment varies by jurisdiction and by the size and type of the reporting entity; your CPA is the person who knows which version applies to your company this year.

Key Takeaways

  • Your cost-to-complete forecast is not a project document. It is an input to your company's financial statements, its bonding capacity, and its ability to bid the next job. Somebody you will never meet makes a large decision using a number you typed on a Thursday afternoon.
  • The WIP schedule has four inputs and seven consequences. Contract amount, total estimated cost at completion, cost incurred to date, and amount billed to date. Everything else is arithmetic. Three of the four are verifiable against the ledger and the contract file — total estimated cost is the only soft number on the page, and it is yours.
  • Percent complete is cost incurred ÷ total estimated cost at completion. The denominator is not the original budget and not the contract value. It is the current forecast.
  • Over-billing is a liability whose repayment currency is work. Billing ahead is normal, often healthy, and is the owner funding your job. It is also a dollar of cost you will incur later with no matching billing.
  • Under-billing is usually a symptom. Four causes: timing (benign), poor billing discipline, unbilled change work, and an unrecognized cost overrun. A hidden overrun disguises itself as an under-billing — the stale denominator inflates revenue earned while billings, driven by verified field progress, do not move.
  • Over- and under-billings never net on the balance sheet. Over-billed jobs sum to a current liability; under-billed jobs sum to a current asset. Only the WIP schedule shows the net.
  • Fade measures the forecast, not the outcome. A job can fade $400,000 and still finish profitably. What matters is when fade appears: early fade means the team found it and re-based honestly; late fade in a lump means either nobody knew or nobody said, and both are serious findings.
  • Billings do not appear anywhere in the profit calculation. Two jobs at the same percent complete have earned the same profit no matter how differently they are billed. One of them just has more of somebody else's cash.
  • A job parked at the same percent complete for months is not stable — it is being held there by a denominator rising as fast as the numerator. Every dollar it spends goes straight to gross profit reduction.
  • An anticipated contract loss is recognized in full, immediately. The estimated loss is charged in the period it becomes known and the profit already recognized reverses, so the charge to income is far larger than the loss itself.
  • A bonding program is not an asset you own. It is a credit line the surety can withdraw, and its withdrawal correlates with your need for it. Underwriters look at Capital, Capacity, and Character — and Character is not on any statement.
  • The audit is the control, not the expense. It is the procedure that tests total estimated cost on every significant contract. It does not just cost more; it removes the place the problem was hiding.
  • Growth is a use of cash, not a source of it. Every new job funds 45–90 days of cost before the first payment and parks retention permanently. Forty million dollars of new revenue can demand $6,000,000 of liquidity and return $1,000,000 of profit.
  • The failure cascade is legible on the WIP months in advance: growth outruns working capital → staff dilution degrades forecasting → fade appears and is absorbed → over-billing covers the gap → the jobs end and the over-billing reverses → the surety cuts the program → working capital goes negative and the covenant trips. Steps 1 through 5 are all on the schedule.
  • A company that punishes the first honest fade report is manufacturing its own surprises. Humiliate a $200,000 fade at 30 percent complete and you have not prevented it — you have moved it to 80 percent complete, at four times the size, when nothing can be done.

Action Items — this week, on your job

  1. Ask for your job's WIP row. Not the cost report — the row: contract amount, total estimated cost, percent complete, revenue earned, gross profit earned, billed, over/(under). Most project managers have never seen it. Ask anyway.
  2. Compute your net cash to finish. Remaining to bill = contract − billed against remaining cost = total estimated cost − cost to date. If remaining cost exceeds remaining billings, your job will consume cash for the rest of its life. Say so out loud, in writing, this month.
  3. Compute your fade since last period, name the cause in one specific sentence, and write it down whether or not anyone asked for it. Not "productivity" — the event, the dollars, and the date.
  4. Inventory your unapproved change work. Dollar value, notice status, and date on each item. Every dollar of it is cash you are lending the owner without a note.
  5. Check your billing against your production. Walk the building. If the WIP says 60 percent and the superintendent says one third, the WIP is wrong in the direction that overstates revenue.
  6. Find your retention step-down provision and the date it triggers. Then find out whether anyone is tracking it. On Northgate, retention drops from 10 to 5 percent at 50 percent completion and the full effect is roughly $2,400,000 of cash held by Kestrel rather than by the owner.
  7. Compare your subcontract payment terms to your owner payment terms. If the owner pays in 30 days and you pay subs in 21, you have volunteered to be the bank on every subcontract on the job.
  8. Sit down with your CFO for twenty minutes and ask the five questions in §34.9. You will walk out a different kind of project manager.

Common Mistakes — and the fix

Mistake What it costs The fix
Holding a known scope gap out of the forecast "to see if it can be absorbed" Revenue recognized that was not earned; the gap returns larger and later Book it the month you believe it. Absorption is a plan with a name and a date, not a hope
Treating cost-to-date against budget as a forecast A rear-view mirror sold as a windshield; fade arrives in a lump at 85 percent Cost-to-complete, monthly, with the superintendent in the room
Reading percent complete without reading estimated gross profit A job frozen at 98 percent charges every dollar it spends straight to profit and nothing looks wrong Read the estimated-gross-profit column across periods, always. Flag any job at the same percent complete twice
Mistaking over-billing for cash the company owns Spending an advance; a cash crisis in the last 15 percent of every job Track the reversal schedule. Over-billing is a forecast of when cash gets tight
Calling an under-billing "timing" without a date A hidden overrun sitting on the balance sheet as an asset "Timing" requires a date and a document. No date, no timing
Confusing the fade with the charge to income Underestimating how badly a job flipping to a loss hurts the quarter Charge = profit already recognized + the full remaining loss
Front-loading a schedule of values to fund a hole rather than a job The last 15 percent funds itself out of nothing; and it is a misrepresentation question, not a cash-flow tactic Value each line to what it costs, with a disclosed allocation of mobilization and general conditions
Job borrow — moving cost between jobs so a bad one looks acceptable Two corrupted forecasts, a corrupted unit-cost database, and a misstatement Never. Code cost to the job that incurred it, including shared crews, and fix the reporting problem instead
Booking a change order as approved revenue before it is approved Revenue recognized on work you may not be paid for Approved means approved. Unapproved change work is a cash and entitlement problem, not a revenue line
Carrying a $200,000,000 bonding program on review-level statements Nobody independently challenges the forecast; the fade surfaces years late and all at once Upgrade the assurance level as the program grows, not after
Growing revenue without growing working capital Bonding ratios deteriorate; the surety reduces the program at the worst moment Retain earnings, and treat the working-capital turnover ratio as a hard governor on volume
Punishing the first honest fade report Every project manager in the room learns to wait; the next fade is four times larger Early fade is a management item; late fade is a performance item — in writing, and applied

Decision Framework

Build a WIP row — six steps, in order

1. Contract amount        = original contract + APPROVED change orders
2. Estimated gross profit = contract amount − total estimated cost at completion
3. Percent complete       = cost incurred to date ÷ total estimated cost at completion
4. Revenue earned         = contract amount × percent complete
5. Gross profit earned    = revenue earned − cost incurred
                          ... and it MUST equal estimated gross profit × percent complete.
                          If it does not, stop. You have an arithmetic error.
6. Over/(under) billing   = amount billed to date − revenue earned

Then foot the schedule. A WIP schedule that does not foot is a WIP schedule somebody edited by hand, and that is a finding all by itself.

Read a WIP schedule — the seven-point check

  1. High percent complete with high remaining margin → a fade waiting to happen. Closeout is where cost hides.
  2. A late gain → name the event that produced it, or it did not happen.
  3. A large under-billing → which of the four causes? Do not accept "timing" without a date.
  4. The same percent complete for multiple periods → a dispute, a hidden cost, or an earlier overstatement quietly catching up.
  5. Percent complete by cost ≠ percent complete in the field → walk the job. An accountant cannot do this from an office.
  6. A job materially below the company's average margin → it consumes the same supervision and the same bonding capacity for less return.
  7. The aggregate billing position and its direction → net over-billed and rising is usually fine; flat while revenue grows means cash is about to get harder; net under-billed at any scale is this month's problem.

The company numbers, and roughly where comfort lives

Number Formula Rough comfort zone
Working capital Current assets − current liabilities Enough to fund 45–90 days of cost on all jobs at once
Current ratio CA ÷ CL Generally 1.3–2.0+; below about 1.2 draws attention
Quick ratio (Cash + receivables) ÷ CL Generally 1.0+
Debt to net worth Total liabilities ÷ equity Often comfortable below about 2.0–2.5
Working-capital turnover Revenue ÷ working capital Uneasy above roughly 15–20×
Months of backlog Backlog ÷ (revenue ÷ 12) Often 9–18 months
Aggregate bonding program Against working capital Often lands around 10–20×
Single-project limit Against the aggregate Commonly about one third to one half

Where the line is on a forecast

You are wrong when your honest best estimate turns out badly. That happens constantly; that is what a forecast is.

You are misstating when you:

  • believe the remaining cost is one number and report a different, better one;
  • hold a known scope gap out of the forecast to see if it can be absorbed;
  • book an unapproved change order as approved revenue;
  • recognize a contract claim as revenue when its resolution is genuinely uncertain (there are legitimate treatments; they are narrow, they require specific conditions, and your CPA decides — not you);
  • move cost between jobs so a bad job looks acceptable.

The difference is reliance. Being optimistic in your head makes you wrong. Being optimistic in a document a surety uses to guarantee $60,000,000 of your obligations, or a bank uses to lend you $15,000,000, induces somebody to take a risk they would not otherwise have taken, on information only you had. What the specific legal consequences are — what constitutes a misrepresentation, what standard of knowledge applies, what remedies exist, what changes on federally funded work — varies by jurisdiction and is a question for counsel. The principle does not vary.

The one sentence

Your monthly cost report is the raw material of next year's bid and a line in this year's financial statements at the same time. Report the number you believe, in the month you believe it, with the reason written down.