Chapter 34 — Construction Finance and Accounting: WIP Schedules, Over/Under Billing, Bonding Capacity, and Why Profitable Contractors Go Broke

The Hook: The Row I Had Never Seen

It was a Tuesday in early February of Year 2, and I was in Owen Baptiste's office because I wanted two more people on Northgate and I had been told no twice.

Owen is Kestrel's chief financial officer. He has a window and a whiteboard nobody is allowed to erase. I had known him for nine years and had spoken to him, in total, for maybe four hours, most of it about whether a subcontractor's insurance certificate was acceptable.

He turned his monitor toward me. On it was a spreadsheet I had never seen. Fourteen rows. About a dozen columns. Small type.

"Do you know what this is?"

"It looks like a job list."

"It's the work-in-progress schedule. Everybody calls it the WIP." He said the letters like a word — whip. "This is the company. Find your job."

I found it. Row one, because it was the biggest: Northgate Outpatient Pavilion. And then I read across, and I stopped, because I did not recognize my own project.

I knew my contract value. I knew my cost to date to the dollar, because Lorena Vasquez and I had closed the January cost report eight days earlier. I knew my cost-to-complete forecast, because I had spent two afternoons building it with Margo Deacon and Wei Chen. Those numbers were all there, in the first few columns, exactly as I had submitted them.

Then came columns I had never computed in my life. Revenue earned to date. Gross profit earned to date. Billings in excess of costs and estimated earnings.

Twenty-two million six hundred eighteen thousand two hundred eighty dollars of revenue earned. Eight hundred fifty-nine thousand five hundred thirty-six dollars of gross profit earned. Seven hundred twenty-one thousand seven hundred twenty dollars of over-billing.

"Where did those come from?" I asked.

"From you," Owen said. "Every one of them is your cost-to-complete forecast run through a formula. You hand me a number in the middle of the month and I turn it into revenue, profit, and a line on the balance sheet. Then that balance sheet goes to a surety and a bank."

I sat there trying to decide whether that was a compliment or a warning.

Then Owen scrolled down three rows and put his finger on the screen.

Rivermont Elementary School #12. Curtis Boone's job. A twenty-two-million-dollar public school on the other side of the metro. I had never set foot on it. I had never read a document from it. I knew about it the way you know about weather in a state you do not live in.

"That one," Owen said, "is why you didn't get two more people in March."

I said something like what?

"In February of last year, that row gave back a hundred and eighty thousand dollars of estimated gross profit. Then a hundred fifteen more. Then a hundred fifteen more. It has given back four hundred ten thousand dollars and Curtis still tells Nadia it's a timing issue." He tapped the screen twice. "Four hundred ten thousand dollars of fade on a job like that is a meaningful share of this company's entire net income for the quarter it landed in. When it landed, Nadia froze incremental staffing on every job that wasn't already committed. That was March. You asked for two people in March."

I have been a project manager for twenty-two years. I had run schedules, priced changes, argued about float, defended cost reports, and closed out buildings. And I had somehow gone the whole way without understanding that my job is a row on a page, my forecast is an input to a financial statement, and somebody else's bad row can reach across the company and take my staffing.

That is the chapter. Here is what it gives you.

You are going to learn to read and build a work-in-progress schedule, which is the single most important document in construction finance and the one almost no textbook written for project people bothers to teach. You will learn what over- and under-billing actually mean, why one of them is normal and the other is usually a symptom. You will learn what fade is and why a surety underwriter looks at it before anything else. You will learn how bonding capacity gets decided and what your job does to it. And you will learn — mechanically, with numbers, step by step — the specific sequence by which a contractor that reports a profit every single quarter runs out of money and stops existing.

🏃 Fast Track: If you already build WIP schedules, skim §34.1 and §34.2, then go to §34.4 (fade), §34.7 (bonding), and §34.8 (the failure cascade). The 📋 Try it in §34.5 is worth twenty minutes regardless — it is the diagnostic a CFO would run on you in an interview.

🔬 Deep Dive: The forecasting discipline this chapter depends on is Chapter 28. The cash mechanics underneath it are Chapter 32. The company-level context — backlog, bonding, margins, why this industry is shaped the way it is — is Chapter 2. The formulas and ratios in this chapter are collected in Appendix A; the terms are defined in Appendix I.

34.1 The timing problem: why construction accounting is its own discipline

Start with something that sounds trivial and is not.

A restaurant sells a meal. Cash comes in, food and labor go out, and the transaction is finished in ninety minutes. A software company sells a subscription and delivers it monthly. An auto plant builds a car in eighteen hours and sells it in ninety days.

Now consider Kestrel. We signed the Northgate contract, mobilized on March 3, Year 1, and reach substantial completion on September 18, Year 2. That is 565 calendar days spanning three of Kestrel's fiscal years, on one contract, for $47,500,000.

Here is the question every construction accountant must answer, and it has no obvious answer:

When has Kestrel earned the money?

You cannot say "at the end," because then Kestrel's Year 1 income statement would show enormous cost and no revenue, and Year 2 would show a windfall. You cannot say "when we get paid," because payment timing is a function of the billing cycle, not of the work. And you cannot say "when we spend money," because spending money is not the same as earning it — ask anyone who has ever poured a slab in the wrong place.

Ordinary business accounting does not face this in the same form. Construction faces it on every contract, all the time. That is why construction accounting is a specialty, why there are CPA firms that do nothing else, and why a contractor's financial statements look strange to anyone who has only read a manufacturer's.

The two answers

The completed-contract method. Recognize nothing — no revenue, no cost, no profit — until the contract is substantially complete, then recognize all of it at once.

Simple. Conservative. And it makes the income statement useless. A contractor using it reports a loss in every year it starts long jobs and a fortune in every year it finishes them. Its legitimate uses are narrow: very short contracts, certain small-contractor tax reporting situations, and jobs where progress genuinely cannot be estimated. Almost nobody who matters uses it for financial reporting, because a surety cannot underwrite a company whose reported results are an accident of which jobs happened to close in December.

The percentage-of-completion method. Recognize revenue and profit as the work is performed, in proportion to progress. If a job is 47 percent complete, you have earned 47 percent of the contract's revenue and 47 percent of its expected profit — this period, whether or not you have billed for it and whether or not anyone has paid you.

This is what virtually every meaningful contractor uses, and it is what sureties, banks, and sophisticated owners expect to see. The modern accounting framework describes the same idea in updated language: revenue is recognized as performance obligations are satisfied, and for a typical construction contract that satisfaction occurs over time rather than at a point in time, measured by progress toward completion. The vocabulary has evolved; the arithmetic on the following pages has not. I am going to describe how it works and what it means, and I am deliberately not citing standard paragraph numbers — the standards get amended, the details vary by jurisdiction and by the size and type of the reporting entity, and your CPA is the person who knows which version applies to your company this year.

How progress gets measured: cost-to-cost

If you are going to recognize revenue in proportion to progress, you need a defensible measure of progress. Physical measures exist (square feet of deck placed, tons of steel erected). Labor-hour measures exist. But the dominant method — the one on nearly every WIP schedule you will ever read — is cost-to-cost:

Percent complete = cost incurred to date ÷ total estimated cost at completion

Look hard at the denominator. It is not the original budget. It is not the contract value. It is the current forecast of what the job will cost when it is finished — exactly the cost-to-complete forecast you learned to build in Chapter 28.

Which brings me to the sentence that should reorganize how you think about your monthly cost report.

💡 Aha moment. The entire income statement of a construction company rests on the accuracy of its project managers' cost-to-complete forecasts. Not on the accountants. Not on the CFO. On you. The denominator of the percent-complete calculation is a number a project manager produced, and every dollar of revenue and profit the company reports this period is computed from it.

That is not rhetoric; it is arithmetic. If I tell Owen that Northgate will cost $46,295,200 and I have spent $21,758,744, the company reports 47.00 percent complete. If the honest number were $47,500,000 — because a forecast I did not want to write down was sitting in my head — the true percent complete would be `$21,758,744 ÷ $47,500,000 = 45.81%`, and the company would have just reported more than half a million dollars of revenue it had not earned. Nobody committed fraud. Somebody was optimistic in a spreadsheet, and the optimism became a number on a document a surety relies on.

🔄 Check your understanding. A project manager reports cost incurred to date of $8,400,000 and total estimated cost at completion of $21,000,000 on a $24,000,000 contract. Two weeks later she learns a mechanical scope gap will add $1,400,000 to the job. She does not change her forecast, because she "wants to see if she can absorb it." What has she done to the percent complete the company reports?

34.2 The work-in-progress schedule, column by column

The work-in-progress schedule — the WIP — is a single table with one row per contract in progress and roughly a dozen columns. Competent contractors prepare it monthly; everyone else prepares it at least quarterly. It is the bridge between the project world you live in and the financial statements the company reports. It is the first document a surety underwriter asks for and often the only one they read closely.

Learn its columns and you can read a contractor's health in four minutes.

The columns, defined

Two things to notice before we work an example.

First, only four of those eleven columns are inputs. Contract amount, total estimated cost, cost incurred to date, and amount billed to date. Everything else is a mechanical consequence. That is why the WIP is so revealing — there is almost nowhere to hide. Three of those four inputs are verifiable against the general ledger and the contract file. The only genuinely soft number on the page is total estimated cost, and that is precisely where every construction accounting problem in history has lived.

Second, columns 10 and 11 are the same subtraction with opposite signs. I show them separately because on the balance sheet they land in different places, which §34.3 explains.

Worked: Northgate at month 11

Here is my job as of the January 31 close of Year 2 — month 11 of a 565-calendar-day contract.

The four inputs:

A word on where the cost forecast comes from, because the arithmetic matters. The GMP build-up gives a cost of work of $40,000,000 + $2,900,000 + $900,000 + $1,320,000 + $575,200 = $45,695,200, with the 4.0 percent CM fee of $1,804,800 on top to reach $47,500,000. The approved change orders added $600,000 of cost and $24,000 of fee. So the current forecast is $45,695,200 + $600,000 = $46,295,200.

Now the four calculations, in order.

Step 1 — Estimated gross profit.

Estimated gross profit = contract amount − total estimated cost = $48,124,000 − $46,295,200 = $1,828,800 Gross profit % = $1,828,800 ÷ $48,124,000 = 3.80%

What it means: on a GMP with a 4.0 percent fee, a 3.80 percent gross margin is exactly what you would expect — the fee, diluted very slightly by the escalation allowance that sits inside cost. No surprise, which is itself information.

Step 2 — Percent complete.

Percent complete = cost incurred to date ÷ total estimated cost = $21,758,744 ÷ $46,295,200 = 0.4700 = 47.00%

What it means: Kestrel has spent 47 cents of every dollar it expects to spend on Northgate. Note this says nothing directly about the calendar. Month 11 of a roughly 19-month contract is about 58 percent of the time and 47 percent of the cost — normal, because the enclosure and interior fit-out are cost-dense and come later. Percent complete by cost and percent complete by time are different animals, and comparing them is a diagnostic we will use in §34.5.

Step 3 — Revenue earned and gross profit earned.

Revenue earned to date = contract amount × percent complete = $48,124,000 × 0.4700 = $22,618,280

Gross profit earned to date = revenue earned − cost incurred = $22,618,280 − $21,758,744 = $859,536

Check it the other way, which should always agree:

Gross profit earned = estimated gross profit × percent complete = $1,828,800 × 0.4700 = $859,536 ✓

What it means: for the eleven months of Northgate, Kestrel's income statement carries $22,618,280 of revenue and $859,536 of gross profit — regardless of what has been billed and regardless of what has been collected.

Step 4 — The billing position.

Over/(under) billing = amount billed to date − revenue earned to date = $23,340,000 − $22,618,280 = $721,720 over-billed

What it means: Kestrel has billed Meridian Health $721,720 more than it has earned — about 1.5 percent of contract. On the balance sheet that is a liability: money collected for work not yet performed. In cash terms it is an interest-free loan from the owner that funds the job, which is good, and it is a loan repayable in work, which is why it is a liability.

Northgate's WIP row, assembled

That is one row. Now look at a company.

The Kestrel WIP schedule — January 31, Year 2

Below are Kestrel's six largest contracts. The full schedule at that date carries fourteen rows; I have shown the six largest so you can foot the table by hand, and the eight smaller contracts are picked up as a single reconciling line in §34.6. The arithmetic is identical at any number of rows.

Foot it yourself. The total row is not decoration — a WIP schedule that does not foot is a WIP schedule somebody edited by hand, and that is a finding all by itself.

• Contract amounts sum to $140,434,000. ✓
• Estimated costs sum to $133,970,200, and `$140,434,000 − $133,970,200 = $6,463,800`, which equals the sum of the estimated gross profit column. ✓
• Company percent complete = $78,556,914 ÷ $133,970,200 = 58.64%. ✓
• Revenue earned sums to $82,214,780; cost to date sums to $78,556,914; the difference is $3,657,866, which equals the sum of the gross profit earned column. ✓
• Billings sum to $83,918,700, and `$83,918,700 − $82,214,780 = $1,703,920` net over-billed, which equals the sum of the over/(under) column. ✓

Now read it, which is a different skill from building it. In four minutes, before any explanation, three things should jump at you:

1. Rivermont Elementary #12 carries a 2.39 percent estimated gross margin on a hard-bid public school. That is thin to the point of being a rounding error away from a loss — and the job is 79 percent complete and 84 percent billed.
2. Sable Ridge is under-billed by $500,000 at 28 percent complete. That is Kestrel's money sitting inside somebody else's building.
3. Brenner Street Parking Structure is 98 percent complete. The right question is how long has it been 98 percent complete? (Three quarters. §34.5 explains why that matters.)

Nobody had to tell me any of that. It is on the page. That is what a WIP schedule is for.

🧩 Productive struggle. Before you read §34.3, work this yourself. Two jobs are each exactly 50 percent complete by cost. Job A is billed at 50 percent of contract. Job B is billed at 58 percent of contract. Both are running exactly on budget. Which one has more cash right now, and which one has more profit? Three minutes. Write down your reasoning before you continue.

34.3 Over-billing and under-billing: what each one is telling you

Those two clumsy phrases at the bottom of a contractor's balance sheet are the most informative numbers on it.

Over-billing — "billings in excess of costs and estimated earnings"

You have billed more than you have earned. The excess appears in current liabilities, under a name that is a full sentence: billings in excess of costs and estimated earnings on uncompleted contracts.

Why it is normal, and often healthy. A contractor with a well-structured schedule of values bills slightly ahead of cost through most of a job. Mobilization, bonds, and insurance get billed early. Material stored on site gets billed on delivery, before installation. General conditions bill on a monthly curve that runs a little ahead of the production curve. The result is that the owner's money funds the work instead of the contractor's. Kestrel's $721,720 of over-billing on Northgate means Meridian Health is carrying three-quarters of a million dollars of Kestrel's working capital at no charge. That is competent cash management, and it is exactly what Chapter 32 taught you to build.

Why excessive over-billing is a warning. Over-billing is borrowed money and the currency of repayment is work. Every dollar of it is a dollar of cost you will incur later with no matching billing. A job that is 90 percent complete and 96 percent billed funds its last 10 percent of cost out of 4 percent of billings, and the difference comes out of the company.

Here is that mechanic, worked on Rivermont Elementary #12:

Curtis Boone's job will consume roughly a million dollars more cash than it generates between now and the day it finishes — even if his forecast is perfectly accurate and nothing else goes wrong. That million comes from Kestrel's line of credit or from over-billings on other jobs. Note where it does not come from: Rivermont #12's own profit, which is $535,000 in total, of which $422,650 has already been recognized.

💰 Money check. Across the six jobs, Kestrel's aggregate position is $1,703,920 over-billed. That is a real benefit — call it a $1.7 million interest-free loan. At Kestrel's revolver rate of 8.5 percent, carrying that money on the line instead would cost roughly $1,703,920 × 0.085 = $144,833 per year. But do not mistake it for money. Every dollar of it will be worked off, and the schedule of when it works off is the schedule of when Kestrel's cash gets tight. Northgate's $721,720 unwinds through Year 2 as the interiors run. Rivermont #12's $1,120,000 unwinds over the next five months. Fair Oaks's $330,200 unwinds in about six weeks. Owen's cash forecast is, in large part, a forecast of over-billing reversal.

Under-billing — "costs and estimated earnings in excess of billings"

You have earned more than you have billed. The excess appears in current assets.

An asset sounds good. In construction it is usually a symptom. There are four reasons a job is under-billed, and only one of them is benign:

That last row deserves its own paragraph, because it is the most useful diagnostic in this chapter.

🔍 Why this works. Think about what an under-billing is, arithmetically. It is revenue earned > billed. Revenue earned is contract × (cost to date ÷ total estimated cost). Now suppose a job's true cost at completion has risen and the project manager has not updated the forecast. The denominator stays too small, so percent complete comes out too high, so revenue earned comes out too high — while billings, which are driven by physical progress an owner's representative actually verified in the field, do not move at all. The gap opens between them and shows on the WIP as an under-billing. A hidden cost overrun disguises itself as an under-billing. That is why an experienced CFO's first question about an under-billed job is not "when will you bill it?" It is "show me why the cost moved."

Sable Ridge is under-billed by $500,000. Two months ago it was under-billed by roughly $1,900,000 — Nadia Haddad saw it on the Monday operations board in November of Year 1, said so out loud in front of everyone, and gave the team until the twenty-fifth to fix either the schedule of values or the pay application. They fixed the pay application. That is what a benign under-billing looks like: it responds to management.

Brenner Street is under-billed by $64,000 at 98 percent complete, which is a more ordinary thing — closeout costs incurred against a contract balance that is nearly exhausted.

On the balance sheet, they do not net

This trips everybody up the first time.

Over-billings and under-billings are not netted against each other on the balance sheet. Over-billed jobs are summed and reported as a current liability. Under-billed jobs are summed and reported as a current asset. Kestrel's six-job position:

The balance sheet shows $2,267,920 of liabilities and $564,000 of assets. Only the WIP schedule shows the $1,703,920.

🔄 Check your understanding. A contractor with $95,000,000 of annual revenue reports $4,200,000 of billings in excess of costs and estimated earnings, and $3,900,000 of costs and estimated earnings in excess of billings. What should that pattern make you ask?

34.4 Fade and gain

Fade is the decline in a job's estimated gross profit from one reporting period to the next. Gain is the opposite. They are the most closely watched numbers in construction finance and the reason surety underwriters ask for WIP schedules going back several years rather than just the current one.

The definition is arithmetic:

Fade (gain) = estimated gross profit this period − estimated gross profit last period

A negative result is fade. Here is the important part: fade is not the same as a loss. A job can fade $400,000 and still finish profitably. What fade measures is not the outcome — it is the quality of the forecast that produced the outcome.

Rivermont Elementary School #12, across five reporting periods

Curtis Boone's job is a $22,400,000 hard-bid public elementary school: competitively bid, prevailing wage, 100 percent payment and performance bonds. It started in September of Year 0. Here is what its WIP row did.

Read the shape of that, not just the numbers.

The job never reported a loss. It never even reported a bad quarter — the largest single-period fade is $180,000 on a $22,400,000 contract, which is eight tenths of one percent and easy to explain away. Curtis explained every one of them. In April it was "a masonry productivity issue we'll recover in the next phase." In July it was "a timing thing on the sitework closeout." In November, in front of Nadia and everyone else, it was "a timing issue, we'll recover it in the masonry buyout."

Now look at the cumulative column and the GP % column together. The margin has gone 4.91 → 4.11 → 3.59 → 3.08 → 2.39 percent, monotonically, across five consecutive reporting periods, on a job now 79 percent complete. A job that fades in every period for five periods is not experiencing five unrelated events. It is experiencing one thing — an estimate that was wrong at bid — released in increments small enough to survive each individual conversation.

That is the pattern a surety underwriter is trained to see, and it is why they want history.

What a pattern of late fade says about a company

Here is the distinction that matters most, and it is about timing, not magnitude.

Fair Oaks Water Plant Expansion shows a late gain: estimated gross profit rose from $890,000 to $1,040,000 over two quarters, at 95 percent complete. There are two possible explanations, and an underwriter will ask which it is. The innocent one — the true one here — is that a change order settled favorably and the remaining construction contingency was released once the risks it was holding had passed without occurring. The other one is that somebody needed a better quarter. You should be able to name, for every gain on your WIP, the specific event that produced it. "The job is going well" is not an event.

🔄 Check your understanding. Two contractors each finish a $30,000,000 job with a final gross profit of $1,200,000. Contractor A forecast $2,400,000 at 20 percent complete and faded steadily to $1,200,000. Contractor B forecast $1,150,000 at 20 percent complete and gained $50,000 across the life of the job. Which would you rather bond, and why?

34.5 Reading a WIP schedule like a CFO

Building the schedule is arithmetic. Reading it is judgment. Here is the checklist Owen runs, in the order he runs it, and it takes him under five minutes on a fourteen-row schedule.

The diagnostic checklist

1. Jobs at high percent complete with high remaining profit. If a job is 92 percent complete and still forecasting a 7 percent gross margin, the last 8 percent of cost has to come in at or under budget for that margin to survive. It rarely does. Closeout is where cost hides: punch labor, final cleaning, warranty callbacks, extended general conditions, the subcontractor claim you have been putting off. A high forecast margin late in a job is a fade waiting to happen.

2. Jobs whose gross profit percentage improved late. Ask for the event. A settled claim, a released contingency, a favorable buyout on a remaining package — all legitimate, all nameable. If nobody can name it, the forecast moved because somebody wanted it to.

3. Large under-billings. Go back to §34.3. Ask which of the four causes it is, and do not accept "timing" without a date.

4. A job at 98 percent complete for three consecutive quarters. This is Brenner Street. A job parks at 98 percent for one of three reasons: there is a dispute nobody has resolved and the remaining scope cannot be closed; there is a cost the team knows about and will not book; or the percent complete was overstated earlier and the job is quietly catching up to reality by spending the last "2 percent" for nine months. Whichever it is, the retention on that job is sitting on the balance sheet as an asset and is not moving.

5. Percent complete by cost that does not match percent complete in the field. Walk the job. If the WIP says 60 percent and the superintendent says the building is one-third done, the WIP is wrong, and it is wrong in the direction that overstates revenue. This is the single most valuable thing a project executive can do with an afternoon, and it is the one thing an accountant cannot do from an office.

6. Any job whose estimated gross profit percentage is materially below the company's average. Rivermont #12 at 2.39 percent against a company average of 4.60 percent is a $22,400,000 contract producing less profit than a $12,000,000 one. It consumes the same supervision, the same bonding capacity, and more management attention.

7. The aggregate billing position, and its direction. Net over-billed and rising is usually fine. Net over-billed and flat while revenue grows means the over-billing is not keeping pace and cash is about to get harder. Net under-billed at any scale is a problem to solve this month.

A described diagnostic view

📊 Diagram (described). Picture the WIP schedule plotted as a scatter: percent complete on the horizontal axis, estimated gross profit percentage on the vertical. A healthy contractor's jobs form a loose horizontal band — margins roughly flat from left to right, with a slight downward drift near the right edge as closeout costs land. An unhealthy contractor's plot has a visible downward slope: jobs at low percent complete cluster high, jobs at high percent complete cluster low. That slope is fade, drawn. It says the company's estimates are systematically optimistic and reality corrects them as jobs progress. As ASCII:

GP% 8 | ● Halyard(52%) 7 | 6 | ● Sable Ridge(28%) ● Fair Oaks(95%) 5 | 4 | ● Northgate(47%) ● Brenner(98%) 3 | 2 | ● Rivermont #12(79%) 1 | 0 +--------------------------------------------- 0% 20% 40% 60% 80% 100% PERCENT COMPLETE

Kestrel's plot is a loose band with one clear outlier low and to the right. That outlier is the conversation.

📋 Try it: read this WIP schedule

You have taken a project executive role at Larkspur Builders, a composite regional general contractor doing roughly $95,000,000 a year. It is your third day. The controller hands you the WIP schedule below with four columns filled in and asks you to complete it before the Thursday operations meeting.

Given, for each job: contract amount, total estimated cost, cost incurred to date, and amount billed to date.

Your tasks:

1. For each job, compute estimated gross profit, estimated gross profit percentage, percent complete, revenue earned to date, gross profit earned to date, and over/(under) billing.
2. Foot the schedule. Report the company's total contract value, total estimated gross profit, aggregate percent complete, total gross profit earned, and aggregate billing position — showing the gross over-billing and gross under-billing separately, as they would appear on the balance sheet.
3. Identify the one job that should worry you, and state exactly why in numbers.
4. Write the three questions you would ask that job's project manager on Thursday.

Give yourself twenty minutes and a calculator before you open the answer.

34.6 The three statements, for construction people

Everything so far has been the WIP. Now let us see where it lands. I am going to be brief and construction-specific, and I am going to skip everything a contractor does not need.

The balance sheet — a snapshot on one date

A contractor's balance sheet has a distinctive shape: enormous current assets, enormous current liabilities, and comparatively little in between. Here is Kestrel's internally prepared interim balance sheet as of January 31, Year 2.

Three things a construction person should notice.

Retention appears twice, on both sides. $9,180,000 receivable from owners, $5,880,000 payable to subcontractors. The net $3,300,000 is Kestrel's own money, earned and unpaid, parked until jobs close. You met the carrying cost of that in Chapter 32; here it is, sitting on the company's balance sheet as an asset that pays no interest.

The over- and under-billings tie directly to the WIP. The $2,915,920 and $818,000 on this balance sheet are the company-wide sums from the fourteen-row WIP, including the eight smaller contracts we did not print. `$2,915,920 − $818,000 = $2,097,920`, which is the company's net over-billed position. The six largest jobs accounted for $1,703,920 of that; the other eight contracts contributed $394,000. The WIP schedule and the balance sheet are the same document, viewed from two directions.

Working capital is the number that governs everything.

Working capital = total current assets − total current liabilities = $49,288,000 − $28,955,920 = $20,332,080

What it means: Kestrel has roughly $20.3 million of liquidity available to fund the front end of jobs, carry retention, and absorb surprises. Every bonding decision, every bank covenant, and every honest answer to "can we take this job?" comes back to that number. It was $21,400,000 in November of Year 1. It has drifted down $1,068,000 in two months, and Owen can tell you exactly which rows did it.

The income statement — a movie of one year

Kestrel's audited statement of operations for Year 1, ended December 31:

A few honest notes.

Kestrel, like most closely held contractors, is an S corporation, so the income statement shows pre-tax income and the shareholders pay the tax personally. When you compare contractors' "net income," find out first whether you are comparing pre-tax to after-tax numbers. Many people do not, and it is a two-percentage-point mistake in an industry where two percentage points is the entire business.

Now the fact that should make you sit up. A healthy general contractor's net margin is often in the low single digits. Kestrel's is 2.30 percent. That is not a sign of a struggling company; it is the normal shape of general contracting, where the contractor buys most of the work from subcontractors and earns a fee for organizing it. You saw this in Chapter 2, and here is what it means operationally:

On $410,000,000 of revenue, Kestrel keeps $9,430,000. A single $4,000,000 loss job — one bad estimate, one unrecovered delay, one uninsured event — is 42 percent of the company's entire year. That is the leverage. It is why fade of $410,000 on one twenty-two-million-dollar school gets the CFO's attention, and it is why every chapter in this book that seemed to be about paperwork was actually about this.

The cash flow statement — why profit and cash are different

Kestrel earned $9,430,000 in Year 1. Cash went up $392,000. Here is the reconciliation, which is the whole point of the statement.

Read the operating section as a sentence: we earned $9,430,000, and then we put $8,814,000 of it into receivables, retention, and under-billings, and got $6,286,000 back from payables, retention payable, and over-billings. Growth in receivables and retention is where a contractor's profit goes. This is Chapter 32's threshold concept — cash flow is not profit — shown at the company level.

The ratios, worked on Kestrel

Every one of these uses numbers from the two statements above. Every threshold I quote is an approximate industry heuristic, not a rule; sureties and banks each have their own, and they change with the market and with your sector.

The arithmetic, so you can reproduce every one:

• Current ratio: $49,288,000 ÷ $28,955,920 = 1.70
• Quick ratio: ($6,410,000 + $31,480,000 + $9,180,000) ÷ $28,955,920 = $47,070,000 ÷ $28,955,920 = 1.63
• Debt to net worth: $33,835,920 ÷ $27,232,080 = 1.24
• Working-capital turnover: $410,000,000 ÷ $20,332,080 = 20.2×
• Months of backlog: remaining value on the fourteen contracts in progress is $179,034,000 − $105,760,780 = $73,273,220`, plus $61,400,000 of contracts signed but not started =$134,673,220` of backlog. Monthly revenue run rate is `$410,000,000 ÷ 12 = $34,166,667`. So `$134,673,220 ÷ $34,166,667 = 3.9 months`.
• Revenue per employee: $410,000,000 ÷ 340 = $1,205,882
• Return on equity: $9,430,000 ÷ $27,232,080 = 34.6%

Two of these deserve a comment.

Working-capital turnover at 20.2× is the number Owen loses sleep over. It says every dollar of Kestrel's liquidity is supporting twenty dollars of annual volume. That is at or slightly past the edge of where many underwriters get uncomfortable, and it is the arithmetic behind the sentence "we cannot grow into this market without more equity." Compare it to Chapter 2's cautionary example, where the same ratio reached 26.4× shortly before the company failed.

A 34.6 percent return on equity looks spectacular and is not what it appears. Decompose it:

ROE = net margin × asset turnover × equity multiplier = 2.30% × ($410,000,000 ÷ $61,068,000) × ($61,068,000 ÷ $27,232,080) = 0.0230 × 6.71 × 2.24 = 34.6%

The return does not come from margin — the margin is 2.3 percent. It comes from turning the asset base almost seven times a year while running it on borrowed working capital. That is the construction business model stated in one line, and it is also the reason the model breaks so violently when the turns slow down. High ROE built on high turnover and thin margin is spectacular in a good market and lethal in a bad one.

🔄 Check your understanding. Kestrel's months of backlog is 3.9. Its working-capital turnover is 20.2×. Those two numbers point in opposite directions as management advice. What are they each telling you to do, and how do you reconcile them?

34.7 Bonding capacity: how a surety actually decides

A surety bond is not insurance. Insurance is a pooled transfer of risk that expects losses; a surety bond is a three-party credit instrument in which the surety guarantees your performance to the owner and, if it pays a claim, pursues indemnity against you. The surety expects zero losses. That is why they underwrite you like a bank rather than like an insurer, and why your bonding program is a hard ceiling on how much work you can carry.

The three C's

Underwriters have described their analysis in the same three words for a century. Practice varies by surety and by market, but the framework is universal.

Character is not sentimentality. It is the observable behavior of a company under stress, and underwriters have long memories.

The financial statement, and why the audit matters

A CPA can report on a contractor's financial statements at three levels of assurance. The differences are enormous and the words are quiet.

Sureties want the audit for one reason: an audit tests the WIP schedule. The auditor confirms contract amounts with owners, tests cost to date against the ledger, confirms billings, and — critically — challenges the total estimated cost on every significant contract. That last procedure is the one that catches an optimistic forecast, and it is why the year-end audit is when hidden fade surfaces at so many contractors. The audit does not merely cost more. It removes the place the problem was hiding.

The surety also wants personal indemnity. The owners of a closely held contractor sign a general indemnity agreement pledging personal assets. It is signed once, usually early, when the program is small and it feels like a formality. It is not a formality, and it is enforceable for the life of the relationship.

How much capacity — the rough arithmetic

Every underwriter will tell you there is no formula, and they are right. Then they will apply something that behaves very much like one. Present these to yourself as approximate industry heuristics, hedged, not as rules — they vary by surety, by market cycle, by sector, and by the company's history.

Now apply them to Kestrel's canonical $150,000,000 aggregate / $60,000,000 single-project program.

What it means: Kestrel's program is conservatively supported — 7.4 times working capital is well inside the range, and the surety could arguably grant more. But Kestrel's volume is at the top of the range at 20.2 times. That is the honest picture of the company: the surety is comfortable with the obligations it has guaranteed; the balance sheet is stretched by the revenue Kestrel is actually running. Those are different questions and people confuse them constantly.

Where the program stands today

In November of Year 1, uncompleted bonded work was $99,355,000. It is now $115,511,220 while total backlog fell from 4.2 months to 3.9. Kestrel replaced private unbonded backlog with public bonded backlog, which consumes aggregate faster than it adds revenue. That is a real strategic consequence of a sales decision, and it appears nowhere except here.

🧩 What happens to the program if a job fades?

Suppose Rivermont Elementary #12 fades another $900,000 at the next quarter — the masonry recovery does not happen, and there is an unresolved subcontractor backcharge. Before you read on: what happens to Kestrel's bonding capacity?

⚖️ What the contract says. Two places where your company's financial information stops being private.

Owner audit rights under an open-book GMP. Northgate is CM at Risk with a guaranteed maximum price, and like most GMP contracts it gives Meridian Health the right to audit Kestrel's books and records for the cost of the work. That right typically covers job cost detail, payroll records, subcontracts, purchase orders, invoices, and change order backup, and it typically survives final payment by some agreed period. It usually does not extend to Kestrel's fee, its corporate overhead, or its company financial statements — but the boundary is drafted, not assumed, and I have seen it drafted broadly. Read the audit clause before you sign. Then behave, every month, as though somebody will read your cost ledger in two years, because on an open-book job somebody may.

Financial disclosure in prequalification. Public owners and many private ones require audited or reviewed financial statements, a bonding letter, and sometimes a WIP schedule as part of prequalification. On public work, submitted materials may be subject to public-records law, though many jurisdictions exempt confidential financial data — the rules vary by state and by agency and you should ask before you submit. Two practical consequences: your WIP schedule may be read by people you did not choose, and a materially false statement in a prequalification submission is a serious matter, potentially exposing the firm to debarment and, on federally funded work, to false-claims liability. These are jurisdictional questions. Bring them to counsel, not to a textbook.

34.8 Why profitable contractors go broke

Everything up to here was vocabulary. This is the argument.

The proposition is not that some contractors report losses and then fail. That is unremarkable. The proposition is that contractors routinely fail while reporting a profit in every single quarter up to the end, and that the mechanism is visible in the WIP schedule months before the failure. Here is the cascade, step by step, with the arithmetic.

Step 1 — Growth outruns working capital

Every new job consumes cash before it produces cash. You learned the shape of that in Chapter 32: you spend from day one, you bill at month end, and you get paid thirty days after that. Call it sixty days of cost carried on average, plus retention that never comes back until closeout.

Work it. A contractor at $100,000,000 of revenue with a 95 percent cost of revenue:

Cash tied up in work ≈ (60 ÷ 365) × $95,000,000 = $15,616,438

Now grow to $140,000,000 — a perfectly ordinary 40 percent year in a hot market:

Cash tied up in work ≈ (60 ÷ 365) × $133,000,000 = $21,863,014

Incremental working capital required by the growth = $6,246,576

And the incremental profit produced by that growth, at a 2.5 percent net margin on the extra $40,000,000, is `$1,000,000`.

💡 Aha moment. Growth is a use of cash, not a source of it. Forty million dollars of new revenue demanded roughly $6,250,000 of additional liquidity and produced $1,000,000. The other $5,250,000 has to come from somewhere: the line of credit, retained earnings, slower payment to subcontractors, or over-billing. A company growing 40 percent a year is financing 40 percent more cash troughs simultaneously, forever, and the faster it grows the further it falls behind. This is why fast-growing contractors fail and slow-growing ones do not — and it has nothing to do with how good they are at building.

Step 2 — Staff dilution

The company that was running eight jobs is now running twelve. The senior superintendents who used to run one job each now oversee two. The project managers who used to have one job have two, and one of the two is a building type the company has never built.

Nothing has broken yet. What has changed is the speed at which problems are detected. An experienced superintendent sees a coordination problem in week four; a newer one sees it in week eighteen, when it costs six times as much to fix. And the person now producing the cost-to-complete forecast has less time, less experience, and more jobs.

Forecasting quality degrades at exactly the moment accurate forecasting matters most. That is not a coincidence — it is the same cause producing both effects.

Step 3 — Fade appears, and is absorbed rather than reported

The first bad news arrives. It is small. It is genuinely arguable. The project manager believes — often sincerely — that it can be recovered in the buyout of the remaining trades, or in a change order, or through a productivity improvement in the next phase.

So it is not reported. Or it is reported at a fraction of its size, to be released later in increments.

You know this pattern. It is Chapter 28's ethics section, and it is what Rivermont Elementary #12's fade table looks like from the inside.

Step 4 — Over-billing covers the gap

Cash is tight, because of Step 1. So the schedule of values gets structured to bill early, stored materials get billed aggressively, and the pay application leans forward. This is not necessarily wrong — until it is being used to fund a hole rather than to fund a job.

What over-billing actually does is move cash from the end of a job to the beginning. Company-wide, it moves cash from jobs that are finishing to jobs that are starting. That works beautifully while new jobs keep arriving.

Step 5 — The jobs end, the over-billing reverses, and the cash does not arrive

Every over-billed job eventually spends its remaining cost against a shrunken billing balance. We worked this twice already: Rivermont #12 will consume $1,007,650 net; the Try it's Vantage Point will consume $700,800 net.

Now do it across a portfolio. If six jobs are finishing in the same two quarters with an aggregate $6,000,000 over-billing, the company must fund $6,000,000 of cost with no billings to match. The only sources are new jobs' over-billings or the line of credit. So the company must book new work faster than it finishes old work, permanently, purely to stay liquid. That is not a strategy. It is a treadmill.

Step 6 — The surety sees the fade and reduces the program

The year-end audit forces an honest cost-to-complete. The fade surfaces. The WIP schedule, which the underwriter has now seen for four consecutive quarters, shows the pattern.

The surety reduces the aggregate and the single-project limit. It does this quickly, and it does it at the worst possible moment, because the moment the fade appears is exactly the moment the contractor most needs to replace backlog.

🚪 The thing to understand about a bonding program: it is not an asset you own. It is a credit line the surety can withdraw, and its withdrawal is correlated with your need for it. Every heuristic in §34.7 is a description of a relationship, not a contract.

Step 7 — Working capital goes negative and the covenant trips

The company cannot replace backlog. Revenue falls while overhead does not. The over-billings reverse. Receivables age, because owners on troubled jobs slow down. Subcontractors, hearing rumors, ask for faster payment or refuse to bid.

Then the bank. Almost every contractor's credit agreement carries covenants — commonly a minimum working capital or tangible net worth, a maximum leverage ratio, and sometimes a limit on distributions. Working capital goes negative, the covenant trips, the line is frozen or called, and the company that was profitable last quarter cannot make payroll this one.

The cascade in one table

Steps 1 through 5 are all visible on the work-in-progress schedule, and Steps 6 and 7 are consequences of what Steps 1 through 5 showed. That is the argument of this chapter. Everything that eventually kills a contractor is legible, in a table, months in advance, to anyone who knows how to read it. Case Study 2 works the entire cascade quarter by quarter on a composite contractor that grew from $80,000,000 to $240,000,000 in three years and reported a profit every quarter until the last one.

The ethics of the forecast — where the line is

This is where I have to be plain, because this chapter is the place in the book where the temptation is largest and where the consequences stop being professional and become legal.

Here is the chain. Your cost-to-complete forecast sets total estimated cost. Total estimated cost sets percent complete. Percent complete sets the revenue and gross profit the company reports. Those reported figures become the financial statements. Those statements go to a surety, who extends credit on them; to a bank, who lends on them; and sometimes to an owner, who awards work on them.

🔍 Why this works — and why it is serious. The reason an optimistic forecast is different in kind from ordinary optimism is reliance. When you are optimistic in your head, you are wrong. When you are optimistic in a document that a third party uses to decide whether to guarantee $60,000,000 of your obligations or lend you $15,000,000, you have induced somebody to take a risk they would not otherwise have taken, using information only you had. Every legal regime I am aware of treats that differently from being wrong. The specifics — what constitutes a misrepresentation, what standard of knowledge applies, what remedies exist, what happens on federally funded work — vary by jurisdiction and are a question for counsel. The principle does not vary: knowingly reporting a percent complete or a cost at completion you do not believe is not aggressive forecasting. It is a misstatement of a financial statement that other people rely on.

I want to be careful here, because forecasting is genuinely hard and honest people disagree about numbers all the time. The line is not "your forecast was wrong." Forecasts are wrong constantly; that is what a forecast is. The line is:

• Believing the remaining cost is $4,500,000 and reporting $4,100,000 because $4,100,000 keeps the margin above 3 percent.
• Holding a known scope gap out of the forecast to "see if it can be absorbed."
• Booking a change order as approved revenue when it is not approved.
• Recognizing a claim as revenue on a job when its resolution is genuinely uncertain. (There are legitimate accounting treatments for contract claims, and they are narrow and require specific conditions. Ask your CPA. Do not decide this yourself.)
• Moving cost between jobs — job borrow — so a bad job looks acceptable and a good job absorbs it.

And now the part that most textbooks leave out, which is the management counterpart:

A company that punishes the first honest fade report is manufacturing the outcome in §34.8. If a project manager reports $200,000 of fade at 30 percent complete and is publicly humiliated for it, every project manager in the room has just learned to wait. The next fade shows up at 80 percent complete, at four times the size, when nothing can be done about it. The behavior an organization gets is the behavior it rewards, and the incentive structure around fade reporting is a design decision somebody at the top made — deliberately or by neglect. Case Study 1 is about a company that noticed this and changed it.

⚠️ Safety alert. There is a hard connection between the financial cascade in this section and the material in Chapter 24, and it is not a metaphor. A cash-starved contractor pays its subcontractors slowly. Subcontractors who are paid slowly run thinner crews, defer equipment maintenance, send less experienced supervision, and push their people to make up time. Schedule pressure is a hazard exactly like an unguarded edge — that is what the week-34 scaffold near-miss on Northgate taught us, where the third finding was a crew running behind under an unwritten "make it up" expectation. A company in financial distress is a more dangerous place to work, and the people exposed to that danger had no part in the decisions that caused it. If you are ever in a position to run a company's finances, that is part of what you are responsible for.

34.9 What a project manager actually controls

You do not set the bonding program. You do not negotiate the bank covenant. Here is the list of things you do control, every one of which lands on the WIP schedule.

1. Forecast honestly and early. The cost-to-complete forecast is the only soft number on the WIP schedule and it is yours. Report fade the month you believe it, in the amount you believe, with the reason written down. Early fade is a management problem. Late fade is a company problem.

2. Bill accurately and on time. Missing the 25th on Northgate does not delay a payment by five days; it delays it by a month, because the owner's cycle does not wait for you. And an under-billed job is your money financing somebody else's building.

3. Get changes approved, not carried. Every dollar of unapproved change work you are performing is a dollar you cannot bill and a dollar of cash the company is lending to the owner without a note. This is the entire argument of Chapter 31 restated in balance-sheet terms.

4. Manage retention release. Retention is earned money you are not holding. Know your contract's step-down provision and hit it. On Northgate, retention drops from 10 percent to 5 percent at 50 percent completion. The full effect across the rest of the job — the excess released at the step plus the reduced withholding on every remaining application — is roughly $2,400,000 of cash Kestrel gets to hold rather than Meridian. Missing that step by two months is a real, quantifiable cost.

5. Align subcontract terms with owner terms. If the owner pays you in 30 days and your subcontract pays in 21, you have volunteered to be the bank on every subcontract on the job. Nine days on $30,000,000 of subcontracts is a meaningful, permanent draw on the company's line.

6. Understand that your forecast is a financial statement input. Everything in this chapter reduces to that sentence.

The five questions your CFO wishes you would ask

Walk into Owen's office with these and you will be a different kind of project manager by the end of the conversation.

1. "What is my job's over/under billing position this month, and which direction is it moving?"
2. "Is there any cost in my forecast I have not billed for, and why not?"
3. "Is my job funding the company's cash this quarter or consuming it — and when does that flip?"
4. "How much of my remaining contingency is genuinely uncommitted, and which named risks is it holding?"
5. "If my estimated gross profit moved this month, does the WIP show it, and does the surety know?"

Job costing and the closed loop

One more thing, and it closes a loop that opened twenty-two chapters ago.

The cost data your job produces does not just report the past. It becomes the historical cost database that funds next year's estimates. When Tomás Reyes prices a 132,000-square-foot outpatient building, he is not consulting a national cost book first — he is consulting what Kestrel actually spent, per unit, on the last four buildings like it, adjusted for market and scope.

Which means:

A company's estimating accuracy is a direct function of its job-costing discipline. Coded wrong, the data is worse than useless — it is confidently wrong. Costs charged to the wrong code, general conditions dumped into a trade code to make a line look better, self-perform hours booked against whatever code had budget left: each of those poisons a unit cost that somebody will use to bid a job in two years. That job will be underbid by exactly the amount of your sloppiness, and it will fade, and the fade will show on a WIP schedule, and somewhere a project manager will not get the two people they asked for.

That is the loop. Chapter 12 and Chapter 13 built the estimate from unit costs; Chapter 28 tracked what those units actually cost; this chapter shows where the consequence lands. Your cost report is not paperwork. It is the raw material of the next bid and a line in this year's financial statements at the same time.

Spaced Review

Answer these before you look back. Retrieval before restatement — that is the point.

1. From Chapter 28: what is the difference between cost-to-date and cost-to-complete, and why does it decide whether you are making money? Write your answer, then check: cost-to-date is history and cannot tell you anything about the future; cost-to-complete is a forecast of what remains, and only it can tell you whether the job's estimated result has moved. Now add this chapter's layer: the cost-to-complete forecast is the denominator of the percent-complete calculation, so it does not just tell you whether you are making money. It tells the company how much revenue and profit to report, and it sets a line on the balance sheet.

2. From Chapter 32: what is retention costing you, and what is the cash trough? Recall first. Retention is earned money withheld — on Northgate, 10 percent until 50 percent complete and 5 percent after, sitting on Kestrel's balance sheet earning nothing while the company borrows at 8.5 percent. The cash trough is the low point of a job's cumulative cash position, early, before billings catch up with cost. This chapter's addition: every job has a trough, so a company running fourteen jobs is financing fourteen troughs at once, and growth means financing more of them simultaneously. That is §34.8, Step 1.

3. The deep callback — from Chapter 2: what are the two limits in a bonding program, what is backlog, and why does growth kill contractors? Try to state all three before reading on. The two limits are the aggregate (total uncompleted bonded work at once) and the single-project cap — and different pursuits fail different ones, so you check both, every time. Backlog is remaining contract value not yet earned, expressed in months of revenue. And growth kills contractors because it consumes working capital faster than it generates profit, dilutes the supervision that catches problems early, and requires ever more new work to fund the reversal of the last batch of over-billings. You have now seen the arithmetic behind every one of those sentences.

Project Checkpoint: The Willow Street Company View

In Chapter 28 you built the month-6 cost report for the Willow Street Community Center with a cost-to-complete forecast. In Chapter 32 you built the schedule of values and pay application #6. This checkpoint takes those two documents you already have and turns them into the one row your CFO would see.

Your deliverable: the Willow Street WIP schedule entry at month 6, with a written explanation of any fade.

Pull four numbers out of the work you have already done:

Then produce the row. Show every calculation:

1. Estimated gross profit = contract amount − total estimated cost, and the percentage.
2. Percent complete = cost incurred ÷ total estimated cost. Carry it to two decimals.
3. Revenue earned = contract amount × percent complete.
4. Gross profit earned = revenue earned − cost incurred. Verify it equals estimated gross profit × percent complete. If those two do not agree, you have an arithmetic error — find it before you go on.
5. Over/(under) billing = billed to date − revenue earned.

Then write three short items:

(a) The cash verdict — two sentences. Is Willow Street helping or hurting the City of Rivermont's contractor this quarter? Show the arithmetic that supports it: remaining to bill = contract − billed versus remaining cost = total estimated cost − cost to date. If remaining cost exceeds remaining billings, the job will consume cash for the rest of its life, and you should say so plainly.

(b) The retention line. Willow Street holds 5 percent retention with no step-down. Compute retention held at month 6 (billed to date × 0.05), and state what it will be at substantial completion ($6,800,000 × 0.05 = $340,000, plus 5 percent of approved changes). Note the 425-calendar-day contract term and estimate how long that money sits.

(c) The two-sentence fade explanation for the CFO. This is the deliverable that matters. If your estimated gross profit moved between your Chapter 28 forecast and today, write the two sentences you would say. Sentence one names the cause specifically — not "productivity," but "the underground utility relocation of the existing 8-inch water main ran 14 days long and added $38,000 of unrecovered cost because the differing-site-condition notice was late." Sentence two states what you have changed so the next number is more reliable. If your gross profit did not move, write the two sentences that say why you believe it, naming the two largest remaining risks and what is holding them.

File all of it behind your month-6 cost report in the notebook, labeled Company View — Month 6.

Next up: Chapter 35 moves the notebook into the digital coordination world with a BIM execution plan — the level-of-development matrix, the clash-detection process, and a 4D sequence for Willow Street.

Chapter Summary

The WIP schedule, in one place. Four inputs, seven derived columns.

Reading a WIP schedule — the seven-point check.

1. High percent complete with high remaining margin → a fade waiting to happen.
2. Late gain → name the event that produced it, or it did not happen.
3. Large under-billing → which of the four causes: timing, unbilled changes, billing discipline, or a hidden overrun?
4. A job parked at 98 percent for multiple quarters → a dispute, a hidden cost, or an earlier overstatement.
5. Percent complete by cost ≠ percent complete in the field → walk the job.
6. A job well below the company's average margin → it consumes the same supervision and bonding for less return.
7. The aggregate billing position and its direction.

The financial reference card.

The failure cascade, in seven steps: growth outruns working capital → staff dilution degrades forecasting → fade appears and is absorbed rather than reported → over-billing covers the gap → the jobs end and the over-billing reverses → the surety reduces the program → working capital goes negative and the covenant trips. Steps 1 through 5 are visible on the WIP schedule months before Step 7.

And the sentence to carry out of this chapter: 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 is going to make a $60,000,000 decision using a number you typed on a Thursday afternoon.

What's Next

That closes Part VI. You now have the full project-controls toolkit: cost control and forecasting, schedule control and recovery, earned value, change orders, progress payments and cash flow, claims and delay analysis, and — as of this chapter — the company-level finance that all of it feeds.

Part VII changes altitude in a different direction. Where Part VI asked how do we measure and control what we are building, Part VII asks how is the work itself changing: coordinating a building in a model before it is coordinated in the field (Chapter 35), building to energy and sustainability standards that are now code rather than aspiration (Chapter 36), and the sectors that run by different rules — residential, heavy civil, and the technologies arriving on job sites right now. Bring the financial lens with you. Every technology decision in Part VII is also a capital decision, and now you know how to read the balance sheet it lands on.