Case Study 2-1 — Argosy Builders: How a Good Contractor Tripled in Three Years and Died Doing It

Argosy Builders, its people, and its numbers are a Tier-3 composite drawn from a pattern this industry repeats every cycle. The names are invented. The mechanism is not.

Setup

Hollis Prine started Argosy Builders at twenty-nine with a pickup truck, a drywall crew, and a wife who did the books at the kitchen table. Twenty-six years later Argosy was a well-regarded general contractor in a metro about two hours from Rivermont, doing $80,000,000 a year — schools, small medical office buildings, municipal work, the occasional retail center. Argosy had a reputation for finishing on time and for treating subcontractors decently. Hollis knew every superintendent's kids' names.

He also had six superintendents averaging nineteen years of experience, five project managers, a controller named Denise Okada who had been with him for fourteen years, and a surety program of $60,000,000 aggregate and $25,000,000 single project placed through an agent named Marisol Kranz.

Then the market turned hot, and Hollis hired Terrell Vance as VP of Operations out of a national contractor. Terrell was good — genuinely good — and he said the thing that every founder wants to hear: "Hollis, you're leaving work on the table. This market will give us two hundred million if we go get it."

They went and got it.

What Happens

Three years of growth

Every year the company grew, and every year Hollis felt more successful. Revenue tripled. The surety kept raising the program. Argosy moved out of the strip-center office into a building they bought.

Now run the ratios nobody was running.

Sureties and lenders look hard at revenue-to-working-capital, and while every underwriter has their own thresholds, many get visibly uncomfortable somewhere in the range of fifteen to twenty times. Argosy was at twenty-six.

Why didn't the equity keep up? Add the three years of net income: $3,584,000 + $3,534,000 + $2,640,000 = $9,758,000. Equity grew only $14,900,000 − $11,400,000 = $3,500,000. The missing $6,258,000 went two places: roughly $3,900,000 in tax distributions to Hollis on the company's S-corporation income, and $2,400,000 in cash to buy the office and yard property.

Sit with that last one. Hollis bought a building because owning it felt like stability. What he actually did was convert $2,400,000 of current assets into a fixed asset. Sureties do not count real estate toward working capital. On the day the deed recorded, Argosy felt more solid and was measurably weaker.

The two jobs

In Year 3, Argosy took on two projects that were each individually defensible and together fatal.

Meridian Point Hotel — $34,000,000. A select-service hotel. Argosy had never built a hotel. Hotels have a repetitive guestroom stack that rewards a superintendent who has done one before and punishes one who has not: bathroom pods, riser coordination, corridor sequencing, and an FF&E turnover choreography Argosy had never run. The assigned superintendent had five years of experience and had been a lead carpenter three years earlier.

Corbin Yards mixed-use — $41,000,000. A five-story wood-frame podium over two levels of concrete parking and ground-floor retail. It was the largest job in company history and it consumed 59 percent of the newly raised $70,000,000 single-project limit. Terrell called it "the job that puts us on the map."

Both were bid tight. Both started in the same quarter. Both drew on the same thin bench.

The fade nobody saw

Both jobs reported monthly using cost-to-date against budget. Both had schedules of values that were front-loaded — value shifted into early line items so that billings ran ahead of cost.

The result was a cash position that looked fine and a profit position that was fiction.

Overbilling of $6,410,000 was, in cash terms, an interest-free loan from the two owners to Argosy — money collected for work not yet performed. It felt like liquidity. It was a liability that would have to be worked off, and it was concealing an $11,000,000 hole.

Denise Okada had asked twice for a real cost-to-complete on both jobs. Both times Terrell said the jobs would recover in the buyout of the remaining trades. In Month 11 of Year 3, the CPA doing the year-end review would not sign off without one. Denise ran it.

$11,000,000 of fade against $14,900,000 of equity. Argosy had, overnight, lost roughly three-quarters of everything Hollis had built in twenty-nine years.

The surety's move

Marisol Kranz brought the underwriter to a meeting in Month 12. It lasted forty minutes.

The program went from $200,000,000 aggregate / $70,000,000 single project to $85,000,000 / $30,000,000.

Argosy had $61,000,000 of work awarded or in final negotiation that had not yet been bonded. None of it could be executed. Two owners called Argosy's bid bonds; a third simply awarded to the second bidder and Argosy ate the reputational cost.

Then the cash spiral, which is the part people never see coming. Argosy had been quietly running on job borrow — funding the front end of each new project with mobilization and early billings from the newest projects. That works as long as new work keeps arriving. New work stopped arriving in Month 12. Within ninety days Argosy was sixty days slow to every subcontractor on every job.

Thirty-one subcontractors went unpaid on the two bad jobs. Liens were filed on the private work; payment-bond claims were filed on the public work. The surety paid the claims — and then exercised its rights under the general indemnity agreement that Hollis and his wife had signed in Year 0, back when the program was $60,000,000 and signing it felt like a formality. It was not a formality. It pledged personal assets.

Argosy Builders wound down in Month 14 of Year 4. Hollis Prine lost the company and the house.

Analysis

Nobody at Argosy was lazy, dishonest, or stupid. Terrell Vance's read on the market was correct — the work was there. Hollis's instinct to grow was normal. The failure was mechanical, and four mechanisms ran simultaneously.

1. Growth consumed cash faster than it produced it. Every new job requires funding of the first 45 to 90 days of cost before the first payment lands, and permanently parks retention on the balance sheet. A company growing 40 percent a year is perpetually financing the front end of new work. The bank statement looks healthy because incoming money from new jobs disguises the outflow. That is not operating cash flow; that is a treadmill that only works while it accelerates.

2. Supervision was diluted below the threshold where problems get caught early. Average superintendent experience fell from nineteen years to seven while the average job size grew. A nineteen-year superintendent sees the guestroom-stack coordination problem in week four. A seven-year superintendent sees it in week eighteen, when it costs six times as much. This is the least visible cost of growth and the most expensive: you cannot buy a superintendent in a quarter, and you cannot make one in three years.

3. The reporting system stopped telling the truth. Cost-to-date against budget is not a forecast; it is a rear-view mirror. The number that matters is cost-to-complete, and it requires someone to sit down with the superintendent and honestly re-estimate the remaining work. Argosy's system worked at $80,000,000, where Hollis personally knew every job. At $240,000,000 it failed silently, and the front-loaded schedules of values made the failure feel like success. This is exactly why Chapter 28 treats cost-to-complete as the central discipline and why Chapter 32 treats a front-loaded schedule of values as an ethics question and not just a cash-flow tactic.

4. The balance sheet could not carry the program the surety had granted. Sureties raise capacity based on trajectory and relationship as well as ratios, and a rising market makes everyone generous. But the surety's downside protection is indemnity, and when the fade appeared, the surety did the only rational thing: it cut the program and reached for the indemnity. A bonding program is not an asset you own. It is a credit line the surety can withdraw at the worst possible moment, which is precisely when you need it. Chapter 34 works the underwriting arithmetic in full.

What would have saved them

Not slower growth for its own sake — growth is how firms build careers and capacity. What would have saved them is growth the balance sheet could carry:

• Retain earnings. Every dollar left in the company supports roughly ten to twenty dollars of bonding capacity, depending on the surety. The $3,900,000 in distributions and the $2,400,000 building purchase, retained instead, might have carried the growth honestly.
• Rent the building. Real estate is a fine investment and a terrible use of a contractor's working capital.
• Run cost-to-complete monthly, on every job, with the superintendent in the room. Non-negotiable, at any company size.
• Do not take a first-of-its-type building and a largest-ever building in the same year with your newest superintendents. One novelty at a time.
• Hire the CFO before you need one. Denise Okada was a capable controller who was never given the authority a $240,000,000 company's finance function requires. She saw it. She said it twice. Nobody had made it her decision.

Discussion Questions

1. Argosy's reported net income in Year 2 ($3,534,000) was slightly lower than Year 1 ($3,584,000) despite revenue growing by $58,000,000. What does that pattern, on its own, tell you — and what would you have asked to see?
2. The surety raised Argosy's aggregate program from $60,000,000 to $200,000,000 over three years while working capital grew only from $7,200,000 to $9,100,000. Whose responsibility was that decision, and what should Hollis have done when the increase was offered?
3. Front-loading the schedule of values improved Argosy's cash position and concealed the fade. Where is the line between legitimate cash management and misrepresentation? Write the rule you would follow.
4. Terrell Vance was right that the market would give Argosy $240,000,000 of revenue. Was he wrong? Restate his recommendation in a form that would have been correct.
5. If you were Denise Okada in Month 6 of Year 3, holding the suspicion that both big jobs were sliding, what would you actually do — given that you report to Hollis and Terrell outranks you?

Your Turn

Take Argosy's Year 3 numbers — $240,000,000 revenue, $9,100,000 working capital, $14,900,000 net worth — and build a one-page memo to Hollis Prine dated the first day of Year 3, before either bad job started.

Your memo must do three things: state the maximum revenue you believe the balance sheet can safely support and show the arithmetic; identify the two specific ratios you would put on the front page of the monthly report from now on; and recommend a concrete decision about the Meridian Point Hotel opportunity that Hollis could actually act on. One page. No hedging — Hollis will not read a second page, and "be careful" is not a recommendation.