Case Study 2 — Brannock Wall Systems: A Profitable Company That Ran Out of Money
Setup
The company: Brannock Wall Systems — metal-stud framing, gypsum board, taping and finishing, and acoustical ceilings. Founded eighteen years ago by Teresa Brannock, who started as a taper. Forty-one field employees at peak, five in the office. No debt except a $900,000 revolving line of credit at a community bank.
The reputation: good. Brannock's crews were clean, its punch lists were short, and Teresa answered her phone. Three general contractors in the Rivermont market invited Brannock to bid everything.
The trajectory:
| Two years ago | Last year | Trailing twelve months | |
|---|---|---|---|
| Revenue | $11,200,000 | $14,600,000 | $18,400,000 | |
| Gross margin | 12.8% | 12.6% | 12.5% |
| Net income after overhead | $392,000 | $481,800 | $588,800 | |
| Working capital | $1,240,000 | $1,388,000 | $1,240,000 |
Every job Brannock ran in those three years made money. Not one of the seven contracts in the final backlog was a loser. The company failed anyway, in the eleventh month of its best year.
(Every company, person and project in this book is a Tier-3 illustrative composite. The numbers are internally consistent and realistic; they are not a real company.)
What happens
The backlog on the day the phone stopped working
| # | Project | General contractor | Subcontract | Billed to date | Retention rate | Retention held | Status |
|---|---|---|---|---|---|---|---|
| 1 | Rivermont Elementary School #12 | Kestrel (Curtis Boone) | 1,940,000 | 1,844,000 | 10% | 184,400 | 95% complete; punch list open 7 months |
| 2 | Meadowbrook Medical Office Building | Talbot Constructors | 2,480,000 | 2,480,000 | 10% | 248,000 | Scope complete 11 months; retention unreleased |
| 3 | Fairhaven Senior Living | Talbot Constructors | 3,120,000 | 2,184,000 | 10% | 218,400 | 70% |
| 4 | Kettering Corporate Center | Verdel Builders | 4,600,000 | 3,220,000 | 10% | 322,000 | 70% |
| 5 | Orchard Park Schools, Phase 2 | Delacorte Construction (public) | 2,850,000 | 1,995,000 | 10% | 199,500 | 70% |
| 6 | Halyard Point Office Repositioning | Kestrel | 1,650,000 | 660,000 | 10% | 66,000 | 40% |
| 7 | Sable Ridge Logistics Center | Kestrel | 1,260,000 | 352,800 | 10% | 35,280 | 28% |
| Totals | 17,900,000 | 12,735,800 | 1,273,580 |
Foot it: retention is 10 percent of billings on every job, so $12,735,800 × 0.10 = $1,273,580. ✓
Now read that number against the income statement.
Retention held: $1,273,580. Two full years of the company's net income: `$588,800 + $588,800 = $1,177,600`.
Brannock's general contractors were holding more of Teresa's money than her company had earned in two years.
The balance sheet at the moment of failure
| Item | Amount |
|---|---|
| Cash | $71,400 |
| Accounts receivable (excluding retention) | $3,427,945 |
| Retention receivable | $1,273,580 |
| Work in progress and inventory | $180,000 |
| Total current assets | $4,952,925 |
| Accounts payable | $1,384,000 *(of which $611,000 more than 60 days past due)* |
| Accrued payroll and benefits | $328,000 |
| Revolving line of credit ($900,000 limit) | $900,000 — fully drawn | |
| Total current liabilities | $2,612,000 |
| Working capital | $2,340,925 |
Positive working capital. Positive net income. Positive equity. Seven profitable jobs. And $71,400 in the bank against a weekly payroll of about $118,000.
The arithmetic of the failure
Step 1 — Compute the working capital the business actually required.
Required = accounts receivable + retention + WIP − accounts payable
= $3,427,945 + $1,273,580 + $180,000 − $1,384,000 = $3,497,525
Step 2 — Compute the working capital available.
Available = equity-funded working capital + credit line = $1,240,000 + $900,000 = $2,140,000
Step 3 — The gap.
$3,497,525 − $2,140,000 = $1,357,525
That gap is the whole story. It was funded, month by month, by paying suppliers later — which is why $611,000 of accounts payable was more than sixty days past due.
Step 4 — Where the gap came from: growth.
Working-capital intensity: $3,497,525 ÷ $18,400,000 = 19.0 cents of working capital per dollar of annual revenue.
Revenue grew from $11,200,000 to $18,400,000 — an increase of $7,200,000, or 64 percent, in two years. At 19.0 cents per revenue dollar, that growth required:
$7,200,000 × 0.190 = $1,368,000 of additional working capital
Over the same two years the company earned `$481,800 + $588,800 = $1,070,600` and distributed $310,000 for taxes and owner draws, retaining $760,600.
$1,368,000 needed − $760,600 retained = $607,400 short — before a single job went badly.
💡 The mechanism, stated plainly. Every additional dollar of revenue required nineteen cents of cash before it produced twelve and a half cents of gross profit — and the profit arrived sixty to ninety days after the cash went out, with a tenth of it withheld for a year or more. Growth consumed cash faster than profit produced it. That is not a management failure. It is arithmetic, and it is the arithmetic that kills good subcontractors.
Step 5 — Days sales outstanding.
Brannock's average collection period on non-retention receivables had drifted from 47 days three years earlier to 68 days.
$18,400,000 × (68 ÷ 365) = $3,427,945 tied up in receivables
$18,400,000 × (45 ÷ 365) = $2,268,493 if the average had held at 45 days
Difference: $1,159,452 of cash sitting inside other companies' buildings.
Nobody was cheating. Two of Brannock's four general contractors had pay-when-paid subcontracts and slow owners. One paid on the 30th of the month following the month following the work — a completely ordinary cycle that is, when you count it out, about sixty-five days.
The nine days
Day 1. Ridgeline Gypsum Supply, Brannock's board and stud supplier, puts the account on credit hold. The balance is $412,000 and the oldest invoice is 94 days. Ridgeline's credit manager has been calling for six weeks and getting promises.
Day 2. Board deliveries stop at Fairhaven, Kettering and Orchard Park. Framing continues; hanging stops.
Day 4. Teresa calls all four general contractors. She asks each for an early release of retention on the completed scopes — $184,400 on Rivermont #12 and $248,000 on Meadowbrook, both of which are finished work with short punch lists. Talbot's project manager says he needs the owner's approval on Meadowbrook and will "get to it." Curtis Boone says Rivermont #12's owner will not release retention early on a public job, which is true, and does not offer anything else.
Day 6. Verdel Builders, on Kettering, issues a written notice of failure to prosecute the work.
Day 7. Delacorte issues a 48-hour cure notice on Orchard Park and notifies Brannock's surety, because Orchard Park is public work and Brannock's subcontract was bonded.
Day 8. Verdel supplements the Kettering work with another drywall contractor at a premium and back-charges $340,000 — a charge that will be deducted from Brannock's remaining billings and its retention.
Day 9. Brannock cannot fund Friday's payroll. Teresa calls the bank. The line is fully drawn and the bank declines to extend it against retention receivable, which it does not lend against.
The surety takes over the two bonded subcontracts. The unbonded general contractors terminate and back-charge. Brannock Wall Systems files for liquidation eleven weeks later. Final distribution to unsecured creditors, of which Ridgeline Gypsum Supply is the largest, is under nineteen cents on the dollar.
Analysis
What actually killed the company
Three things, and none of them is "unprofitable work."
1. Retention concentration. Retention is invisible on an income statement and enormous on a balance sheet. Brannock's $1,273,580 was 6.9 percent of annual revenue, more than half of total working capital, and more than two years of net income. Because it was spread across seven projects and four general contractors, no single conversation could free it — Teresa needed four separate yeses, and she got zero.
Look hard at rows 1 and 2 of the backlog table. $432,400 of the $1,273,580 was held against work that was finished — one scope complete for eleven months, another 95 percent complete with a seven-month-old punch list. That money was not securing performance. Nothing was left to perform. It was securing paperwork and inattention.
2. Days sales outstanding drift. Twenty-three extra days of collection consumed $1,159,452. Nobody decided to do this. It happened one project at a time, as Brannock took work from general contractors with longer cycles, and it never appeared as a line item anywhere. A subcontractor that does not measure DSO monthly is not measuring the single most important number in its business.
3. Growth. The $7,200,000 revenue increase required $1,368,000 of working capital and generated $1,070,600 of profit over the same period, $310,000 of which left the company. Teresa was not reckless; she was successful, and success in this industry has a cash price that nobody puts on the invoice.
Add up the counterfactual
$432,400 (retention released on the two completed scopes)
+ $1,159,452 (DSO back to 45 days)
= $1,591,852
against a gap of $1,357,525.
Either fix alone would not have saved the company. Together they would have — with $234,327 to spare. Neither required a single additional dollar of profit, a single additional job, or a single change to how Brannock built walls.
What each party should have done differently
| Party | What they should have done |
|---|---|
| Teresa Brannock | Measure DSO and retention receivable monthly, on one page, alongside the backlog. Set a policy that retention on completed scopes is chased in writing at 30, 60 and 90 days with the contract article cited. Decline the seventh job — or take it and raise equity first. Know the working-capital intensity of the business (19 cents per revenue dollar) and multiply it by every new contract before signing. |
| The general contractors | Release retention on completed, accepted scopes. It costs them the float and buys a subcontractor that survives to finish the other four jobs. Watch the waiver matrix: Ridgeline stopped signing for Brannock three months before the credit hold, and every one of the four general contractors had that signal in a folder. |
| Ridgeline Gypsum Supply | Nothing, really. It carried $412,000 for 94 days and then acted. It recovered nineteen cents. |
| The bank | Also nothing unusual. Retention receivable is a poor borrowing base precisely because its collectability depends on somebody else's punch list. Every subcontractor should know this before it needs the money. |
The transferable lesson
Brannock did not fail because it built badly, bid badly, or managed badly. It failed because its balance sheet was structurally unable to support the volume it had won, and nothing in its profit-and-loss statement said so.
If you are a subcontractor, the number that matters is not your margin. It is working capital per dollar of annual revenue, and if you do not know yours, you are running the business with the most important gauge unlabeled.
If you are a general contractor, the lesson is different and it is about self-interest, not charity. Brannock's failure cost Verdel $340,000 of supplementation, cost Delacorte a bonded takeover and a three-month schedule impact, and cost Kestrel two rebuys at higher prices on Halyard Point and Sable Ridge. The retention that four general contractors would not release totalled $432,400. The failure cost them, collectively, several times that. Chapter 19's point was that you manage the people who contracted to do the work. This is the financial half of it: a subcontractor's solvency is your risk, whether or not it is your problem.
Discussion questions
-
Brannock's income statement showed $588,800 of net income and its balance sheet showed $2,340,925 of positive working capital on the day it could not make payroll. Explain to a first-year field engineer, without using the word "liquidity," how both of those numbers can be true at the same time.
-
Compute Brannock's working-capital intensity (working capital required ÷ annual revenue) and use it to answer this: if Teresa wanted to grow to $26,000,000 of revenue next year at the same margins and terms, how much additional working capital would she need, and how much profit would the growth generate? What does the comparison tell her?
-
The waiver matrix on all four general contractors' jobs showed Ridgeline Gypsum Supply not signing for Brannock beginning three months before the credit hold. Design the one-page monthly report a general contractor's project accountant should produce so that signal reaches a project manager who can act on it.
-
Two of Brannock's seven subcontracts were bonded and five were not. Discuss what subcontractor bonding would and would not have changed here — for Brannock, for the general contractors, and for Ridgeline Gypsum Supply.
-
Curtis Boone told Teresa that Rivermont #12's public owner would not release retention early. That was true. Was it a complete answer? What else could a general contractor do for a subcontractor in this position without breaching its own contract or its duties to the owner?
Your turn
You are the project manager on Fairhaven Senior Living for Talbot Constructors. Brannock is 70 percent complete on a $3,120,000 subcontract, it is your critical-path trade for the next four months, and you have just learned about the Ridgeline credit hold from your project accountant's waiver matrix.
Write a one-page memo to your director of operations that does four things: (1) states the exposure in dollars — what it costs Talbot if Brannock fails at 70 percent, including supplementation premium, schedule impact and rebuy; (2) lists the options, including early retention release, joint checks to Ridgeline, direct payment, accelerated payment terms, and supplementation; (3) prices at least three of them; and (4) makes a recommendation with a decision date.
Then, separately, write the two sentences you would say to Teresa Brannock on the phone. They are harder than the memo, and they matter more.