Chapter 33: Cash Flow and Working Capital: The Thing That Actually Closes Restaurants

"Profit is an opinion. Cash is a fact." — constructed; a line every operator learns once, usually the expensive way

Overview

Chapter 1 made you a promise, and this chapter is where it comes due: most restaurants that close were, on paper, doing fine two months earlier.

That sentence sounds like a paradox and it is not. It is arithmetic. Profit is a result computed over a period by a set of accounting conventions — conventions that spread a twelve-month insurance premium across twelve months whether you wrote one check or twelve, that book a chicken as an expense when you cook it rather than when you pay for it, and that do not record the repayment of loan principal as a cost at all. Cash is what is in the account on a Thursday afternoon when the payroll file has to be funded by three o'clock. Those two numbers are computed from the same business and they can point in opposite directions in the same month. In this chapter you will watch Bellwether's February do exactly that: a profit-and-loss statement showing $10,694 of profit** and a bank balance that fell by **$10,954. A $21,648 gap, in a business that did nothing wrong that month.

Here is the part that ought to frighten you. A restaurant that is unprofitable has a problem you can see. Prime cost is high, the P&L is red, the flash report screams, and everyone in the building knows something is wrong. A restaurant that is profitable and out of cash has a problem nobody can see, because every report anyone looks at says the business is fine. Chapter 32 gave you the break-even that keeps the lights on. This chapter gives you the number that keeps the doors open, and they are not the same number — Chapter 32 already told you so, in the one line of that chapter you should have written down: accrual break-even is 66 dinner covers a night, and cash break-even is 77.

We are going to do something uncomfortable with Bellwether in this chapter. Up to now the plan has been building — a concept, a lease, a costed menu, a labor model, a break-even. Now we open the bank account. What we find is that the working-capital reserve, the $45,000 line item that has been sitting in the use-of-funds since Chapter 5 looking like a responsible piece of planning, **is effectively gone before the doors open**, because Chapter 9's honest bottom-up pre-opening budget came to $71,300 against a $35,000 plan. The $36,300 gap has exactly one place to come from. What is left is $8,700**, and $8,700 is 5.3 days** of Bellwether's fixed monthly obligations.

Then we build the thirteen-week forecast that finds the week it runs out.

In this chapter, you will learn to:

• Name the four structural reasons a profitable restaurant's bank balance falls, and reconcile profit to cash line by line for a specific month.
• Compute the cash conversion cycle for a business that is paid instantly, and explain why restaurants can have a negative cycle on the sales side and still fail.
• Build a working-capital requirement from the bottom up, and defend it against the rule of thumb.
• Read accounts-payable terms as a financing decision, quantify vendor float in dollars, and price the discount you are giving up.
• Construct a thirteen-week cash forecast week by week, find the trough, and translate the trough into the size of a credit facility.
• Lay out a twelve-month timing calendar and identify the collision months before they arrive.
• Diagnose a cash problem in its first week rather than its ninth.

Learning Paths

🏗️ Opening — this is the most important chapter in Part VII for you, and §33.3 and §33.7 are the reason. Build the thirteen-week forecast for your own project before you sign anything, and treat the working-capital number it produces as a hard requirement, not an aspiration. 📋 Managing — weight §33.5, §33.7, and §33.9. You may not control the capital structure, but you control the timing calendar, the payables run, and whether anybody notices the first warning sign. In many independents the manager is the only person who could have seen it. 🍸 Beverage — §33.2 is yours: a forty-bottle wine list is the slowest-turning inventory in the building and the single largest block of cash sitting still. §33.4's terms discussion matters more to you than to the kitchen, because distributor terms in a three-tier state are often not negotiable at all. 🚚 Small Format — you have less fixed cost and far less cushion. A truck's reserve is measured in days, not weeks, and §33.6's seasonality problem is more violent for you, not less. Read §33.9 twice.

33.1 Why profit and cash diverge, with a worked example

Start with the plan. Here is Bellwether's year-one projection as it has stood since Chapter 4, now carried all the way down past the line most business plans stop at.

Check the arithmetic yourself, because you should always check it: $430,280 + $500,000 = $930,280, which on $1,550,000 is 60.0% exactly. Add occupancy, other operating, and G&A — $95,200 + $217,000 + $46,500 = $358,700 — and total costs are $1,288,980, leaving $261,020. Subtract $69,500 of debt service and you get $191,520.

That is a good-looking plan. It is also not a cash-flow statement, and the difference between those two things is what this chapter is about.

The four wedges

Four specific mechanisms drive a wedge between what the P&L reports and what the bank shows. They are not exotic, they are not accounting tricks, and every one of them is present in every restaurant.

Wedge one: debt principal is not an expense. This is the big one and the one that surprises people most. When you write a check for $5,792 of debt service, the P&L records only the *interest* portion as an expense. The principal portion is a balance-sheet transaction — you are reducing a liability, not incurring a cost. In Bellwether's first year, the $69,500 of debt service splits roughly $39,000 interest / $30,500 principal. That means $30,500 of real cash leaves the building every year and never appears anywhere on the income statement.

Wedge two: owner draws are not an expense either. If the chef-owner takes a distribution — as opposed to a W-2 salary, which is in the labor line — that money leaves the account and the P&L does not blink. Distributions come out of equity. Most owner-operators of independent restaurants take some of both, and the ones who get into trouble are the ones who take draws in a good quarter and think of them as "profit I already earned" rather than "cash I removed from the working-capital pool."

Wedge three: inventory build consumes cash at no cost to profit. COGS on the P&L is usage, not purchases: beginning inventory plus purchases minus ending inventory (Chapter 13). If you buy $9,600 of product in a week and use $8,900 of it, your P&L records $8,900 and your bank account records $9,600. The $700 difference is sitting in the walk-in. Every restaurant that grows builds inventory, and every dollar of that build is cash that the income statement never mentions.

Wedge four: timing. Everything else. Insurance premiums arrive in installments that do not match the monthly expense. The landlord's annual reconciliation of estimated versus actual NNN charges lands in one month for twelve months of expense. Payroll is disbursed every fourteen days into months that are twenty-eight to thirty-one days long, so twice a year a month contains three payrolls. Sales tax is collected daily and remitted monthly. None of these change what a period cost. All of them change when the money moves.

🧮 Run the Numbers

The profit-to-cash bridge, annual.

Take Bellwether's plan year and walk it from operating profit to the change in the bank account, assuming the plan is hit exactly and the owners take a modest distribution.

```text Operating profit (before debt service) $261,020 − Interest portion of debt service (expense) (39,000) ──────────────────────────────────────────────────────────────────── = Pre-tax profit, as the P&L reports it $222,020

− Principal portion of debt service (NOT an expense) (30,500) − Inventory build over the first year (NOT an expense) (7,275) − Owner distributions (NOT an expense) (48,000) ──────────────────────────────────────────────────────────────────── = Change in cash before taxes $136,245 ```

Two things to notice.

First, $85,775 of real cash left the business without ever touching the income statement ($30,500 + $7,275 + $48,000). An owner reading only the P&L would believe they had $222,020 of profit. The account grew by $136,245. Neither number is wrong. They answer different questions.

Second — and this is the one that closes restaurants — the plan year never happens. The plan assumes $1,550,000 of revenue and 60.0% prime cost across twelve even months. Chapter 9 already told us the first quarter runs 66.6% prime on $363,100 of sales, which means the remaining thirty-nine weeks have to average 58.0% just to land the annual number. Q1 prime of 66.6% on $363,100 is $241,825. Subtract that from the annual $930,280 and the remaining $688,455 of prime cost has to sit on $1,186,900 of remaining revenue — 58.0%, to the tenth. Every dollar of that Q1 overrun is a dollar of cash the first quarter consumed and the rest of the year has to replace.

Why this is not an accounting curiosity

I want to be blunt about the stakes, because "profit and cash differ" reads like a technicality and it is the opposite of a technicality.

A restaurant does not close when it stops being profitable. It closes when it cannot fund something — a payroll, a rent payment, a distributor who has put it on credit hold, a sales-tax remittance. Every one of those is a cash event with a date. The P&L has no dates on it. It has a period, and a period is a summary of dates that have already passed.

This is why the pattern from Chapter 1 looks the way it does. A business bleeding three points of prime cost is not, in month fourteen, visibly failing. It is thinner. The reserve that would have absorbed a bad February has been quietly spent covering ordinary weeks, and nobody logged that as an event because there was no event — just fifty-two weeks each of which was a little worse than plan. Then the compressor goes, or the NNN reconciliation is $4,180 instead of the $1,900 you expected, or a hard January runs eight percent light, and the business discovers that it has been operating without a cushion for eleven months.

⚠️ Where the Money Leaks

The reserve that was spent before opening.

This is the single most consequential number in Bellwether's plan, and it has been hiding in plain sight since Chapter 5.

The use of funds budgets $620,000**: construction $310,000, equipment $185,000, smallwares and FF&E $45,000, pre-opening $35,000, and a working-capital reserve of $45,000**.

Chapter 9 built the pre-opening budget honestly, from the bottom up — the manager on payroll fourteen weeks out, the chef ten, the salaried sous eight; four weeks of hourly training and menu rehearsal; licensing and permit fees; the opening inventory; the deposits; the utility connections; the pre-opening marketing; the two weeks of payroll that run before a single guest pays for anything. The honest number was $71,300.

```text Pre-opening, as budgeted in the plan $35,000 Pre-opening, built bottom-up in Chapter 9 $71,300 ──────────────────────────────────────────────────────────────────── Gap $36,300

There are exactly two places a $36,300 gap can come from: more capital, or the reserve. Nobody raises more capital eleven days before opening.

Working-capital reserve, as planned $45,000 Less the pre-opening overrun ($36,300) ──────────────────────────────────────────────────────────────────── Working-capital reserve, in the account on opening day $8,700 ```

$8,700. Bellwether's fixed monthly obligations — the money that leaves whether or not a single guest walks in — are $48,933. Eight thousand seven hundred dollars is 5.3 days of that.

The full $45,000 would have been 27.6 days. Neither is enough. But one of them is a restaurant with three and a half weeks to react to a bad month, and the other is a restaurant that cannot survive a slow week without borrowing.

The mechanism is the point. Nobody decided to spend the reserve. Chapter 1 warned that the construction contingency and the working-capital reserve get confused constantly; this is worse and more common — the reserve was never confused with anything, it was simply the only unspent line left when the pre-opening bills came in higher than the plan said. Every restaurant that opens with an underbuilt pre-opening budget does this, and almost none of them write it down.

What "fixed monthly obligations" actually means

That $48,933 deserves to be built, because it is the denominator of every runway calculation in this chapter and the most useful single number an operator can carry in their head.

Fixed obligations are not a P&L category. They cut across occupancy, labor, other operating, G&A, and debt service. The question they answer is not "what does this cost?" but "what leaves the account this month even if the dining room is empty?"

Add them: $7,933 + $16,800 + $12,400 + $5,792 + $1,450 + $2,300 + $1,150 + $1,108 = $48,933. That is $1,631 a day, every day, before a cook clocks in or a case of produce is ordered.

Two honest caveats about that table, because a method that hides its limits is not a method.

It is not truly fixed. The "minimum hourly crew" line is fixed only in the sense that below it you cannot open. You can close a Tuesday. You can cut brunch. Those are real levers and Chapter 32's daypart analysis tells you which ones to pull. What you cannot do is pull them this Thursday to fund this Thursday's payroll.

It understates the true floor. It excludes food and beverage purchases, which are variable but not optional — you cannot serve a $46 dinner without buying something — and it excludes variable labor above the minimum crew. At Bellwether's plan volume those add roughly $8,274 and $12,467 a month respectively. The $48,933 is the floor beneath the floor.

👨‍🍳 On the Line

What a cash problem actually feels like on a Wednesday.

It does not feel like a crisis. That is the entire difficulty of teaching this.

It feels like this: it is Wednesday morning, the payables run is sitting in front of you, and before you release it you open the banking app. You have never done that before. For eight months you released the payables run without looking, because there was obviously enough. Now you look first.

The next week you look first again, and this time you move the linen invoice to next week. Not because you can't pay it — because you'd rather have the $340 through Saturday. The linen company does not call. Nothing happens. You have just learned something that will cost you a great deal of money, which is that stretching a payable is free.

Three weeks later you are deciding which two of five vendors to pay, and you are choosing by who has called. Two weeks after that you notice that the produce delivery came short and you didn't want to call about the credit memo because you didn't want to have a conversation with that vendor at all.

Nothing on any report changed during those two months. Sales were fine. Prime cost was fine. Nobody in the building except you knew anything was different, and what you knew was not "we have a cash problem" — it was a sequence of six small, individually reasonable decisions.

The countermeasure is embarrassingly cheap and it is in §33.7: a one-page rolling forecast that you update every Monday in twenty minutes. Not because the forecast is accurate — it isn't — but because the act of updating it converts "I'd rather have the $340 through Saturday" from an instinct into an entry on a page you will read again next Monday.

33.2 The cash conversion cycle in a business that gets paid instantly

Every business has a cash conversion cycle: the number of days between paying for inventory and collecting the cash from selling it. It has three components.

FIGURE 33.1 — The cash conversion cycle, and why a restaurant's is strange   [constructed teaching example]

  A MANUFACTURER                                                  cycle ≈ +75 days
  ┌─────────────┐        ┌──────────────────────┐        ┌────────────────────┐
  │ pay supplier│───────▶│ hold raw material    │───────▶│ invoice customer   │
  │  day 30     │  DIO   │ and finished goods   │  DSO   │ collect day 105    │
  └─────────────┘  60 d  └──────────────────────┘  45 d  └────────────────────┘
        ▲                                                            │
        └──────────── 75 days of cash tied up in the cycle ──────────┘

  A MATURE RESTAURANT ON VENDOR TERMS                             cycle ≈ −22 days
  ┌─────────────┐        ┌──────────────────────┐        ┌────────────────────┐
  │ receive food│───────▶│ hold it in the       │───────▶│ guest pays at the  │
  │ pay day 30  │  DIO   │ walk-in ~7 days      │  DSO   │ table, settles     │
  └─────────────┘   7 d  └──────────────────────┘  ~1 d  │ day 1              │
                                                          └────────────────────┘
        The guest's money arrives 22 days BEFORE the vendor's invoice is due.
        The vendor is financing the restaurant. This is the famous structural
        advantage of foodservice, and it is real.

  BELLWETHER, YEAR ONE                                            cycle = +20.9 days
  ┌─────────────┐        ┌──────────────────────┐        ┌────────────────────┐
  │ receive food│───────▶│ walk-in 9 d, bar and │───────▶│ guest pays; cards  │
  │ PAY ON      │  DIO   │ wine list 74 d       │  DSO   │ settle in 1–2 days │
  │ DELIVERY    │ 23.5 d └──────────────────────┘  1.4 d └────────────────────┘
  │ (no credit  │
  │  history)   │        The structural advantage does not exist yet, and the
  └─────────────┘        wine list is the reason the inventory half is so slow.

  Legend:  DIO = days inventory outstanding   DSO = days sales outstanding
           DPO = days payable outstanding     CCC = DIO + DSO − DPO

The formula is the same for everyone:

$$\text{CCC} = \text{DIO} + \text{DSO} - \text{DPO}$$

Days sales outstanding (DSO) is how long it takes to collect. In a restaurant this is close to zero, and Chapter 26 explained why: card settlement runs one to two days, cash is cash, and there are no invoices to chase. Bellwether's blended DSO is about 1.4 days once you account for the small share of event business that collects on terms. A manufacturer would trade a great deal for that number.

Days inventory outstanding (DIO) is how long product sits before you sell it. Here restaurants split violently in two. Food is fast — a scratch kitchen ordering produce and dairy weekly-to-twice-weekly and running a standing protein order, as Chapter 13 specified for Bellwether, turns its walk-in every week or so. Beverage is slow. A forty-bottle wine list, a full back bar, and the backup stock behind them turn a handful of times a year, not fifty.

Days payable outstanding (DPO) is how long you take to pay. This is the number a restaurant's structural advantage runs through — and it is the number a new restaurant does not have.

🧮 Run the Numbers

Bellwether's cycle, in days and then in dollars.

Annual COGS is $430,280 — $334,800 of food (30% of $1,116,000 of food sales) and $95,480 of beverage (22% of $434,000 of beverage sales). Confirm: $334,800 + $95,480 = $430,280. Daily COGS is $430,280 ÷ 365 = **$1,179**.

Inventory. At steady state Bellwether carries about $8,400 of food and $19,300 of beverage — spirits, beer, and the forty-bottle list with its backups. Total: $27,700.

	Inventory	Daily COGS	DIO
Food	$8,400 | $917	9.2 days
Beverage	$19,300 | $262	73.7 days
Blended	$27,700** | **$1,179	23.5 days

Read that table twice. Thirty percent of Bellwether's inventory dollars are food and seventy percent are liquid, and the liquid half turns five times a year. The wine list is not inventory in any operationally meaningful sense. It is capital, parked.

The cycle, year one. New restaurants get COD or prepay terms from almost everybody, because they have no payment history and a distributor's credit department has watched this movie before. Bellwether's effective DPO in year one is about 4 days.

$$\text{CCC} = 23.5 + 1.4 - 4.0 = \textbf{20.9 days}$$

At $1,179 a day of COGS, a 20.9-day cycle ties up **$24,641** of cash permanently. Not once — permanently. That money is in the walk-in and on the wine rack and it stays there as long as the restaurant is open.

The cycle at maturity. Twelve months in, with the broadline distributor on net 14 and the produce and dairy houses on net 7, blended DPO rises to about 18 days:

$$\text{CCC} = 23.5 + 1.4 - 18.0 = \textbf{6.9 days} \rightarrow \$8,135 \text{ tied up}$$

Getting terms releases $16,506 of cash, one time, permanently. That is nearly double the entire $8,700 reserve Bellwether opens with, and it costs a phone call and twelve months of paying on time. It is also, precisely, the thing you cannot have in the quarter you most need it.

The interesting part

Here is what makes restaurants genuinely peculiar as businesses, and the observation is worth sitting with because it explains something about the industry that statistics alone do not.

Restaurants have a negative cash conversion cycle on the sales side and they still fail. The mature-restaurant diagram above is not a fantasy — a settled operation on net-30 terms with a fast-turning walk-in really is financed by its vendors, really does hold the guest's money for three weeks before the invoice comes due, and really does enjoy a structural advantage that a machine-tool manufacturer would kill for. Grocery, hospitality, and foodservice are the textbook examples of negative-cycle businesses.

And roughly a quarter of them still do not reach their first anniversary, with something close to six in ten gone within three years.

Why? Three reasons, and each is a section of this chapter.

One: the advantage arrives late. The negative cycle is a mature-operation property. In year one you have no credit history, you have paid cash for opening inventory, and you have built a wine list. The cycle is positive precisely during the twelve months in which a restaurant is most likely to die. Bellwether's +20.9 days is the correct number for the period that matters.

Two: a negative cycle finances your inventory, not your fixed costs. This is the misunderstanding that does the damage. Vendor float covers the cost of the goods. It does nothing for rent, nothing for salaried labor, nothing for debt service, nothing for insurance — the $48,933 a month that leaves whether or not you sold anything. A negative cycle means you never have to fund a case of tomatoes out of savings. It does not mean you can fund a slow February.

Three: getting paid instantly creates a dangerous illusion of liquidity. Chapter 26 flagged this and it is worth naming as a hazard rather than a feature. Money arriving daily feels like money you have. It is deposited under your name, it shows in the balance, and a great deal of it is not yours: the 7% sales tax you collected today and will remit on the twentieth of next month, the deposits on events you have not catered, the gift cards you have not redeemed, the two weeks of payroll you have accrued and not yet disbursed. A restaurant's bank balance systematically overstates its position, and it overstates it by the most on the days it looks healthiest.

⚖️ Code and Compliance

Sales tax and payroll tax are trust-fund money, and they are treated differently from every other debt you owe.

Sales tax is collected from the guest, held by the restaurant, and remitted to the state. In most jurisdictions it is legally characterized as money held in trust — the restaurant is a collection agent, not a taxpayer. The same characterization typically applies to the employee-withheld portion of payroll taxes.

The consequences of that characterization are unlike those of any other obligation on your books:

• Trust-fund taxes are generally not dischargeable in bankruptcy, and unpaid amounts commonly carry penalties and interest that accrue quickly.
• Responsible individuals — owners, officers, and sometimes managers with check-signing authority — can be held personally liable for unremitted trust-fund taxes, regardless of the entity structure you chose in Chapter 8. The corporate veil that protects you from a slip-and-fall claim frequently does not protect you here.
• Many state revenue departments have summary collection powers — levies, license revocation, and in some jurisdictions the ability to close an establishment — that a private creditor does not have.

The practical instruction is simple and I have never met an experienced operator who disagreed with it: sweep the sales tax into a separate account the day it is collected, and do not have a debit card for that account. Bellwether collects $25,417 of sales tax in its first thirteen weeks alone. That money will sit in the operating account looking exactly like revenue during the precise weeks the balance is thinnest.

This varies by state, and in some places by county and city — rates, filing frequency, remittance dates, prepayment thresholds, discount allowances for timely filing, and the definition of a responsible person all differ. The 7% rate used throughout this chapter is illustrative. Verify your own with your state's revenue department and your accountant, and do it before you open.

33.3 Working capital: what you need, and why the reserve is not optional

Working capital is the simplest definition in this chapter and the least useful in isolation:

$$\text{Working capital} = \text{Current assets} - \text{Current liabilities}$$

Current assets are cash, inventory, prepaid expenses, and receivables — things that will become cash within a year. Current liabilities are payables, accrued payroll, sales tax payable, and the portion of long-term debt due within a year.

Here is Bellwether's, on opening day.

🧾 Read the Numbers

```text FIGURE 33.2 — "Working capital, opening day" [the Bellwether plan] THE ARTIFACT Current section of the opening balance sheet, the first Tuesday in April, before service. THE CONTEXT 68 seats, $620,000 project, funded by $150,000 owner injection, a $75,000 tenant-improvement allowance, a $60,000 equipment lease, and an SBA 7(a) note. Construction closed out; certificate of occupancy issued in the second week of March; pre-opening ran $36,300 over the plan.

               CURRENT ASSETS
                 Cash in the operating account                     $8,700
                 Inventory — food                                  $8,400
                 Inventory — beverage (incl. the 40-bottle list)  $19,300
                 Prepaid insurance, licenses, deposits             $6,400
                 ───────────────────────────────────────────────────────
                 Total current assets                             $42,800

               CURRENT LIABILITIES
                 Accounts payable (pre-opening invoices open)      $9,300
                 Accrued payroll (final pre-opening week)          $6,100
                 Current portion of long-term debt                $30,500
                 ───────────────────────────────────────────────────────
                 Total current liabilities                        $45,900

               WORKING CAPITAL                                   ($3,100)

WHAT IT SHOWS Bellwether opens with NEGATIVE working capital. Current liabilities exceed current assets by $3,100 before the first guest is seated. Worse, the composition is bad: 65% of current assets are inventory, and 70% of that inventory is beverage that turns roughly five times a year. The genuinely liquid asset — cash — is $8,700, which is 19% of current assets and 5.3 days of the $48,933 monthly fixed obligation. WHAT IT DOESN'T It does not show the trajectory. A balance sheet is a photograph; the question that matters is whether next week's photograph is better or worse, and only a forecast answers that (§33.7). It also does not show the $23,800 of rent abatement from Chapter 6, which is a real economic asset for the next three months and appears nowhere on this statement. And it does not show the undrawn borrowing capacity that would change the whole picture — because there isn't any yet. THE DECISION Before opening week ends, do three things. (1) Build the thirteen-week forecast in §33.7 and find the trough. (2) Open a separate sales-tax account and set the POS-to-bank sweep. (3) Start the vendor-terms conversation with the broadline distributor now, twelve months before you will be granted terms, so that the clock on your payment history starts today rather than the day you need the float. THE LESSON Working capital is not a number you check. It is a number you fund, at the moment you are raising money, when raising it is possible. Nobody raises working capital in week two. ```

What you actually need: two methods that should agree

There are two honest ways to size a working-capital requirement, and a plan is stronger when it does both and they land near each other.

Method one: the rule of thumb. Practitioner guidance and most lending checklists for independent full-service restaurants converge on sixty to ninety days of fixed obligations, held in cash, after every pre-opening bill is paid. For Bellwether:

• 60 days: $48,933 × 2 = **$97,866**
• 90 days: $48,933 × 3 = **$146,799**

Method two: build it from the bottom. Add the three things the money actually has to do.

Check: $36,300 + $16,142 + $48,933 = $101,375, and $101,375 − $45,000 = $56,375.

The two methods land at $97,866** and **$101,375 — within three and a half percent of each other. That convergence is not luck; it is what happens when a rule of thumb was built by people who had done the bottom-up version enough times to compress it. When your two methods disagree by a factor of two, one of them is wrong and it is usually the bottom-up one, because you left something out.

FIGURE 33.3 — What Bellwether reserved, and what it needed          [the Bellwether plan]

                                  0        25k       50k       75k      100k
                                  ├─────────┼─────────┼─────────┼─────────┤
  Reserve as planned              ██████████████████                        $45,000
  Reserve on opening day          ███▌                                       $8,700
  60-day rule of thumb            ███████████████████████████████████████   $97,866
  Bottom-up requirement           ████████████████████████████████████████▌ $101,375

  Runway on the $48,933/month fixed obligation, with no revenue at all:
     $45,000  →  27.6 days
      $8,700  →   5.3 days      ◀── this is what Bellwether actually opened with
    $101,375  →  62.1 days

  Runway is a coverage metric: it assumes zero revenue, which never happens.
  The forecast in §33.7 is what actually happens, and it is not more forgiving.

Three points about that figure, because it is easy to misread.

Runway assumes zero revenue. It is a stress metric, not a prediction. Bellwether will collect money in week one. The purpose of the coverage number is to tell you how much of an interruption the business can absorb — a two-week closure for a fire-suppression failure, a public-health event, a February that runs twenty percent light. Chapter 8 put business-interruption insurance in the schedule for exactly this reason, and business-interruption coverage typically has a waiting period measured in days that you fund yourself.

Sixty-two days is not generous. It is the low end of the practitioner range. An operator opening a first restaurant with no operating history and a personal guarantee should be looking at ninety.

The trough component is the one people leave out. Everyone can see that the pre-opening budget needs funding and that you want a cushion. Almost nobody funds the ramp — the eight to sixteen weeks during which a new restaurant carries a full staff against partial volume. That is the $16,142, and §33.7 shows you exactly which weeks produce it.

33.4 Payables: terms, vendor float, and the dangerous comfort of stretching them

Accounts-payable terms are the agreement about when you pay. They are the cheapest financing in your business and the most dangerous, in that order.

The common ones:

Vendor float is the cash you hold because of those terms — the money that is in your account only because an invoice has not come due. At Bellwether's plan volume, food purchasing runs $6,438 a week** (Chapter 13) and beverage roughly **$1,836, a total of $8,274** a week. At net 14 that is two weeks of purchases outstanding at all times: about **$16,500 of somebody else's money, sitting in your account, financing your business at an interest rate of zero.

That is the single largest source of free working capital an independent restaurant has, and there are three things to understand about it.

It is not yours. It is a revolving liability that happens to look like cash. Chapter 31 made the same point about sales tax; this is the same category error one step down. If purchases stop — a closure, a pivot, a seasonal shutdown — the float unwinds and you owe it all within two weeks, at exactly the moment you have no revenue.

It has to be earned, and the clock starts before you need it. Credit departments want payment history. Ask for terms in month one, expect to be declined, pay on time anyway, and ask again in month six and month twelve. The single most valuable thing an opening operator can do for their year-two cash position costs nothing and takes fifteen minutes.

It is not available in a three-tier state for alcohol. Chapter 16 covered the three-tier system; the relevant consequence here is that many states impose statutory credit terms on alcohol sales — some require cash on delivery, some cap credit at fifteen or thirty days with a mandatory delinquency list that suspends your ability to buy from any distributor if you go past due. This is one of the few payables where "I'll pay it next week" can stop the beverage program cold. Verify locally.

⚠️ Where the Money Leaks

Two ways payables quietly cost you money.

One: the discount you are giving up. A 2/10 net 30 term is not a courtesy. It is a price. Paying on day 30 instead of day 10 means you paid 2% more for twenty days of money:

$$\frac{0.02}{0.98} \times \frac{365}{20} = 37.2\%$$

An annualized cost of about 37%. If your line of credit charges 11%, taking the discount with borrowed money is a clear win. If you are passing up 2/10 net 30 on $80,000 of annual dry-goods purchases, you are spending $1,600 a year to hold money you could rent for $480. Nobody ever put this on a P&L as a decision, because it isn't a line — it is the absence of a line.

Two: stretching, which is free until it isn't. Moving from a 14-day average payment to a 45-day average payment releases about **$18,000** of cash at Bellwether's volume — 31 days × $581 a day of purchases. It feels like finding money. Here is the full price of that $18,000:

• You lose the discounts. Every 2/10 goes away first.
• You lose the price. Distributor pricing is negotiated, and a slow-pay account does not get the quarterly review it asked for. Two points of food cost on $1,116,000 of food sales is $22,320.
• You lose the service. The credit hold does not arrive as a letter. It arrives as a driver who does not stop, at 6:00 a.m. on the Friday of a 142-cover night with a 40-top on the books.
• You lose the option. Once your DPO is 45 days, stretching is no longer a lever — you have spent it. The next crisis has to be solved some other way.

The diagnostic that matters: a rising DPO against flat sales is one of the clearest early warning signs in §33.9. Terms are a strategy. Stretching is a symptom wearing a strategy's clothes.

One consequence of all this is worth flagging before you read the forecast in §33.7. Every cash projection built so far in this book has quietly assumed vendor terms. This chapter removes that assumption for the first quarter, because a new restaurant does not have them — which is why the purchases column in Figure 33.6 is paid in the week it is incurred, and why the removal costs roughly $16,500 of cash across the ramp.

33.5 The timing calendar: payroll, rent, sales tax, insurance, and the months they collide

A P&L has one date on it — the end of the period. A cash plan has thirty of them.

The timing calendar is the artifact that makes the dates visible: a twelve-month map of every obligation, when it lands, and how big it is. It takes an afternoon to build once and it is the thing that lets you say, in September, "November is going to be tight and February is going to be worse," which is the entire game.

Three structural facts drive it.

Payroll is biweekly and months are not. A biweekly payroll disburses twenty-six times a year. Twelve months × two is twenty-four. Therefore twice in any twelve-month stretch, a month contains three payroll disbursements instead of two. At Bellwether's plan volume a disbursement is $500,000 ÷ 26 = **$19,231. That is an extra $19,231 of cash out in a month where nothing on the P&L changed at all, because the P&L accrues labor to the days it was worked. Given Bellwether's pay calendar — the first disbursement on the Friday of week two, every fourteen days thereafter — the three-payroll month in year one is October, and the next one falls the following April**.

Rent lands on the first and the abatement runs out on a date somebody else chose. Chapter 6 negotiated three months' free rent — $23,800** of abatement, at $7,933.33 a month — with rent commencement tied to the certificate of occupancy. Read that clause again, because it is the kind of term that reads as a win in the LOI and costs real money in practice. Bellwether's certificate of occupancy was issued in the second week of March. The clock started then. Abatement therefore covers March, April, and May, and the first rent payment is due June 1**.

Which means one of Bellwether's three free months — $7,933 of it — was spent in a dark building. The restaurant did not open until the first Tuesday in April. A full third of the abatement produced no revenue at all, because it was consumed by punch list, final inspections, staff training, and menu rehearsal.

Sales tax is collected daily and remitted monthly, one month behind. That lag is a loan from the state to you, and like all loans it has to be repaid on a schedule. Bellwether collects 7% and remits on the twentieth for the prior month. In a growing month the lag flatters you. In a falling month it does the opposite: in January you remit December's $11,088 while collecting only $7,959 of your own. A $3,129 negative swing in the worst revenue month of the year, caused by nothing except the calendar.

🧾 Read the Numbers

```text FIGURE 33.4 — "The timing calendar, year one" [the Bellwether plan] THE ARTIFACT Twelve-month map of scheduled cash obligations, month 1 (April, opening) through month 12 (March). All figures illustrative and constructed. THE CONTEXT 68 seats, opening the first Tuesday in April. Certificate of occupancy issued the second week of March, which started the three-month rent abatement clock. Biweekly payroll, first disbursement the Friday of week two. Sales tax 7%, remitted the 20th for the prior month.

Mo Month Net sales Pyrl Rent Tax remit Ins Debt Periodic & notes ── ───── ───────── ──── ────── ───────── ───── ───── ─────────────────────────── 1 Apr $95,100 2 ABATED $0 1,450 5,792 opens 1st Tue; partial month 2 May 133,100 2 ABATED 6,657 1,450 5,792 brunch launches 3 Jun 134,900 2 $7,933 9,317 1,450 5,792 ◀ RENT COMMENCES 4 Jul 134,600 2 7,933 9,443 1,450 5,792 patio open; hood clean $680 5 Aug 137,200 2 7,933 9,422 1,450 5,792 6 Sep 133,900 2 7,933 9,604 1,450 5,792 hood clean $680 7 Oct 143,500 3 7,933 9,373 1,450 5,792 ◀ THIRD PAYROLL +$19,231 8 Nov 136,800 2 7,933 10,045 1,450 5,792 holiday event deposits in 9 Dec 158,400 2 7,933 9,576 1,450 5,792 best month; hood clean $680 10 Jan 113,700 2 7,933 11,088 1,450 5,792 ◀ tax remit > tax collected by $3,129; year-end accounting $3,400 11 Feb 106,200 2 7,933 7,959 1,450 5,792 ◀ THE FEBRUARY PROBLEM NNN reconciliation $4,180 Ins. renewal + WC $6,900 License renewals $2,850 Compressor service $1,640 hood clean $680 12 Mar 122,600 2 7,933 7,434 1,450 5,792 first anniversary ── ───── ───────── ──── ────── ───────── ───── ───── ─────────────────────────── TOTAL $1,550,000 25 $79,330 $99,918 17,400 69,504

Sales tax collected across the year: $108,500. Remitted within the year: $99,918. The $8,582 difference is March's collection, remitted April 20 — money in the account on March 31 that is not the restaurant's.

THE COLLISION MONTHS, ranked Feb (mo 11) lowest revenue of the year AND $15,570 of annual lumps Jan (mo 10) revenue drops $44,700 from December; tax remittance exceeds collection Oct (mo 7) third payroll, $19,231 of extra cash out, invisible on the P&L Jun (mo 3) rent commences; the abatement that felt like cushion is spent

WHAT IT SHOWS The obligations are not evenly distributed and the revenue is not either, and the two patterns are almost perfectly out of phase. The four heaviest obligation months are October, January, February, and June. Two of those are among the three weakest revenue months. February carries the year's lowest sales and $15,570 of periodic items that appear in no monthly budget because they are annual. WHAT IT DOESN'T It does not show variable cost — purchases and hourly labor, which are the largest cash outflows and move with sales. It does not show the weekly granularity where payroll and rent actually collide (§33.7 does). It shows scheduled obligations only: no compressor beyond the one budgeted, no health-department capital item, no roof. And it assumes the plan revenue happens, which is exactly the assumption a cash plan exists to stress. THE DECISION In September — not in January — arrange the credit facility sized to February. Move the license renewals to a payment plan if the jurisdiction allows it. Ask the landlord in November for the NNN reconciliation estimate rather than being told in February. Book the hood cleaning for early February so it is paid before the reconciliation lands, not after. THE LESSON Cash problems are scheduled events. You do not need a forecast to know that February is going to be hard — you need a calendar, and you need to have looked at it in September. ```

The calendar's real power is not the arithmetic. It is that it converts a vague dread — winter is going to be rough — into four dated, sized, addressable problems, each of which has a specific action attached to it and a specific month by which the action has to be taken. Vague dread produces nothing. A $15,570 item with a February date produces a phone call in September.

The Monday morning cash meeting

The operating practice that carries the calendar is not complicated, and it is fifteen minutes long.

Every Monday, the manager who owns the numbers opens one page. On it: last week's collections, this week's committed outflows, the next twelve weeks in summary, and the current balance with the not-yours money subtracted out. Fifteen minutes, standing up, same time every week.

What gets said out loud: "We collected $28,400 last week against a $29,800 forecast. Payroll is Friday, $19,200. The insurance renewal is in three weeks. The forecast trough is week seven at $11,200 and we are two hundred dollars ahead of it."

That is the whole meeting. It works for three reasons that have nothing to do with accuracy.

It creates a record of forecast versus actual. After six weeks you know whether you forecast optimistically. Almost everybody does, by about the same margin every week, and once you know your margin you can correct for it.

It puts the trough on the table before it arrives. The difference between a manageable week and a crisis is almost entirely whether you saw it eight weeks out or eight days out. Eight weeks out you can move a payment, delay a hire, take a discount, or call the bank. Eight days out you can only choose which vendor to disappoint.

It stops the private worrying. In an independent, the person who knows the cash position is usually alone with it, at eleven at night, doing arithmetic they will not share because sharing feels like admitting something. A Monday meeting makes it an operating number instead of a secret. Chapter 21 has more to say about what secrets cost a management team.

33.6 Seasonality and the February problem

Every restaurant has a season it dreads. In most of the United States it is the stretch from the second week of January to the first genuinely warm evening in spring, and for a restaurant with a patio it is worse, because the patio is a fixed cost in January and a revenue center in July.

Bellwether opens the first Tuesday in April. Follow the consequence: its first February is month eleven.

That timing is very nearly the worst possible arrangement, and it is worth naming why.

The honeymoon is over. Chapter 9 was explicit that the opening spike is not the baseline. By month eleven the press has moved on, the neighborhood has tried you, the people who came because it was new are back in their rotation, and what remains is the actual demand curve — which is the number the whole business plan was betting on.

The patio is closed. Sixteen seats — nearly a quarter of the capacity — produce nothing from November to April. They still carry occupancy, insurance, and maintenance.

The reserve is gone and has not been rebuilt. Any Q1 cash the business generated went into replacing the $36,300 that pre-opening consumed, or into the ramp, or into inventory. There has been no quarter with enough slack to build a cushion.

The first big periodic bills have now cycled once. The landlord's annual NNN reconciliation, the insurance renewal, the workers' compensation audit, the license renewals — none of these existed in months one through ten. They all arrive together, in the year's weakest month, for the first time, at a business that has no history with them and therefore did not budget them.

FIGURE 33.5 — Bellwether's first twelve months of revenue         [the Bellwether plan]

  Month        Net sales      0        50k       100k      150k
  ──────────  ──────────      ├─────────┼─────────┼─────────┤
   1  Apr        $95,100      ███████████████████             partial month, opens 1st Tue
   2  May        133,100      ██████████████████████████▌     brunch launches
   3  Jun        134,900      ██████████████████████████▉
   4  Jul        134,600      ██████████████████████████▉     patio
   5  Aug        137,200      ███████████████████████████▍    patio
   6  Sep        133,900      ██████████████████████████▊     patio
   7  Oct        143,500      ████████████████████████████▋   ◀ best non-holiday month
   8  Nov        136,800      ███████████████████████████▎
   9  Dec        158,400      ███████████████████████████████ ◀ holiday events
  10  Jan        113,700      ██████████████████████▋
  11  Feb        106,200      █████████████████████▏          ◀ THE FLOOR — month eleven
  12  Mar        122,600      ████████████████████████▌
  ──────────  ──────────
      TOTAL   $1,550,000

  February is 33.0% below December and 26.0% below October.
  Fixed obligations in February: $48,933 — identical to December's.

Check those percentages: $158,400 − $106,200 = $52,200, and $52,200 ÷ $158,400 = 33.0%. $143,500 − $106,200 = $37,300, and $37,300 ÷ $143,500 = 26.0%.

Revenue swings 33%. Fixed obligations swing zero. That is operating leverage from Chapter 32, seen from the cash side, and it is why seasonality is a cash problem rather than a profit problem. Over twelve months the seasonality nets out — the plan still says $1,550,000. In February it does not net out at all.

What February looks like from the floor

Here is the part that connects to Chapter 32, and it is the most useful thing in this section.

February's $106,200 over four operating weeks is $26,550 a week. That sounds survivable. Break it into nights and it stops sounding survivable.

Add the dinner line: $2,116 + $2,668 + $3,404 + $5,428 + $6,394 = $20,010. Add the week: $20,010 + $1,872 + $2,304 + $2,364 = $26,550. Four weeks: $106,200. It foots.

Three of Bellwether's five February dinner services run below cash break-even. Not below plan — below the number of guests required to cover the cash the building consumes. Friday and Saturday carry the entire month, and they carry it by enough that the average — 87 covers — looks fine, which is exactly the problem with averages.

Chapter 32 warned about this from the other direction. It found that the Q1 ramp and the Chapter 20 wage reclassification cluster within four covers of the cash break-even, so that any two of them coinciding puts the requirement at 81 covers with a 14-cover cushion, not 29. February's 87-cover average is six covers above that. Six covers is one four-top and a deuce that did not show.

🧮 Run the Numbers

February: profit $10,694, cash −$10,954.

This is the chapter's thesis in a single month. Same restaurant, same February, two statements.

The profit-and-loss statement, month eleven:

text Revenue $106,200 100.0% Cost of goods sold (usage, 27.9%) 29,630 Labor, all-in (34.5%) 36,639 ────────────────────────────────────────────────────────────── PRIME COST 66,269 62.4% Occupancy 7,933 7.5% Other operating (14.0%) 14,868 14.0% General & administrative (3.0%) 3,186 3.0% ────────────────────────────────────────────────────────────── OPERATING PROFIT 13,944 13.1% Interest on debt 3,250 ────────────────────────────────────────────────────────────── NET PROFIT $10,694 10.1%

A 62.4% prime cost in the worst month of the year, and the business still nets ten percent. The chef-owner would look at that statement in mid-March and conclude the restaurant handled February well. They would be right.

The bank account, same month:

text Cash collected (net sales $106,200 + 7% sales tax $7,434) $113,634 Food and beverage purchases (30,900) Payroll and payroll taxes (two disbursements) (36,600) Operating and administrative (incl. card fees $2,984) (18,384) Insurance, monthly premium (1,450) Rent and NNN (7,933) Debt service (SBA note + equipment lease) (5,792) Sales tax remitted — JANUARY's collection, not February's (7,959) ── the winter lumps ────────────────────────────────────────── NNN/CAM annual reconciliation from the landlord (4,180) Insurance renewal installment + workers' comp audit (6,900) License and permit renewals (2,850) Walk-in compressor service, the cold snap (1,640) ────────────────────────────────────────────────────────────── NET CHANGE IN CASH ($10,954)

The bridge between them. This is the reconciliation every operator should be able to do, and almost none can:

text Net profit as reported $10,694 − Debt PRINCIPAL repaid (not an expense) (2,742) − Inventory build (purchases $30,900 vs. usage $29,630) (1,270) − Periodic items paid in February but expensed across the year (17,350) [cash paid for operating/G&A/insurance categories $35,404 against $18,054 of accrued expense in those categories] − Sales tax timing (remitted $7,959, collected $7,434) (525) + Payroll timing (expensed $36,639, disbursed $36,600) 39 ────────────────────────────────────────────────────────────── Change in cash ($10,954)

A $21,648 divergence in a single month, and every dollar of it is legitimate accounting. No fraud, no error, no aggressive judgment. The P&L is right. The bank statement is right. They are answering different questions and only one of them can fund a payroll.

Now do the thing that actually matters: run this February against an $8,700 reserve. It doesn't matter what the P&L says. A business holding $8,700 cannot absorb a $10,954 month, and February is a month you can see coming from September.

33.7 Building the 13-week cash forecast

The thirteen-week cash forecast is the single most valuable document in restaurant finance and it fits on one page. Thirteen weeks is the standard horizon for a reason: it is one quarter, it is long enough to contain every monthly obligation cycling three times and at least one quarterly item, and it is short enough that you can forecast revenue with something better than a guess.

It is not a budget. A budget says what should happen. A forecast says what the bank balance will be on each of the next thirteen Fridays, and its entire purpose is to find the lowest one.

How to build it

Seven steps. Do them in this order.

1. Start with the actual bank balance, not the balance you expect after the checks clear. Subtract nothing yet.

2. Forecast collections, not sales. For a restaurant these are nearly the same thing — Chapter 26 established that card settlement runs one to two days — but they are not identical, and the difference matters in two places: the week you open (nothing settles before you sell) and the week you close. Bellwether's forecast treats collections as same-week and notes the exception.

3. Add the sales tax you collect, because it hits the account. Then schedule the remittance as an outflow. Do not net it — the whole point is to see how much of your balance isn't yours.

4. Schedule purchases on the week you pay, not the week you order. In year one that is the same week. At maturity it is one to four weeks later, and getting this wrong is the most common error in a first forecast.

5. Schedule payroll on the disbursement date. Not the week worked. Biweekly payroll produces a violent sawtooth and the sawtooth is the point.

6. Schedule every fixed and periodic obligation on its actual date. Rent on the first, debt on the seventh, insurance on the seventh, sales tax on the twentieth, the quarterly hood cleaning in whatever week it falls. Use the timing calendar from §33.5.

7. Compute the running balance and find the minimum. That minimum is the number the whole exercise exists to produce.

Here is Bellwether's, built from the plan.

🧾 Read the Numbers

```text FIGURE 33.6 — "The first thirteen weeks" [the Bellwether plan] THE ARTIFACT Thirteen-week cash forecast, weeks 1–13 of operation. Operating weeks run Tuesday through Monday. All figures illustrative and constructed. THE CONTEXT Bellwether opens the first Tuesday in April with $8,700 in the operating account — the $45,000 reserve less the $36,300 pre-opening overrun. Rent abated through May (the CO started the clock in March). No vendor terms: a new account pays on delivery. Sales tax 7%, remitted the 20th for the prior month. Payroll biweekly, disbursed the Friday of even weeks, covering the two weeks just completed. Insurance and debt service drafted the 7th. Chapter 9's plan: $363,100 of Q1 sales at 66.6% prime.

Wk Week of Net sales Cash in F&B buy Payroll Operg Scheduled Net chg BALANCE +7% tax lumps ── ──────── ───────── ───────── ──────── ─────── ───── ───────── ──────── ──────── opening $8,700 1 Apr 6 $14,200 $15,194 $6,200 $0 4,249 $7,242 ($2,497) $6,203 2 Apr 13 21,600 23,112 8,100 24,200 4,457 0 (13,645) ($7,442) ◀ 3 Apr 20 27,400 29,318 9,600 0 4,620 0 15,098 7,656 4 Apr 27 31,900 34,133 10,600 22,700 4,746 0 (3,913) 3,743 ◀ 5 May 4 33,800 36,166 10,300 0 4,800 7,242 13,824 17,567 6 May 11 34,600 37,022 10,400 21,100 4,822 0 700 18,267 7 May 18 33,200 35,524 10,000 0 4,783 6,657 14,084 32,351 8 May 25 31,500 33,705 9,500 19,700 4,735 0 (230) 32,121 9 Jun 1 29,700 31,779 9,300 0 4,685 15,175 2,619 34,740 10 Jun 8 27,900 29,853 8,900 18,400 4,634 0 (2,081) 32,659 11 Jun 15 26,400 28,248 8,400 0 4,592 9,317 5,939 38,598 12 Jun 22 25,600 27,392 8,000 17,200 4,569 0 (2,377) 36,221 13 Jun 29 25,300 27,071 8,200 0 4,561 7,933 6,377 42,598 ── ──────── ───────── ───────── ──────── ─────── ───── ───────── ──────── ──────── TOTALS $363,100 $388,517 $117,500 $123,300 60,253 $53,566 $33,898 $42,598

SCHEDULED LUMPS, detailed Wk 1 insurance $1,450 + debt service $5,792 = $7,242 Wk 5 insurance $1,450 + debt service $5,792 = $7,242 Wk 7 April sales tax remitted (7% of $95,100) = $6,657 Wk 9 insurance $1,450 + debt service $5,792 + RENT COMMENCES $7,933 = $15,175 Wk 11 May sales tax remitted (7% of $133,100) = $9,317 Wk 13 July rent $7,933 (July's insurance and debt land in week 14) = $7,933

MEMO — what the $42,598 ending balance is not June sales tax collected, remitted July 20 ($9,443) not the restaurant's Week 13 payroll accrued, disbursed week 14 ($8,300) already earned ───────────────────────────────────────────────────────────────── Genuinely free cash at the end of week 13 $24,855

WHAT IT SHOWS Bellwether runs out of money in WEEK TWO. The balance closes week 2 at negative $7,442 — a cumulative $16,142 below the $8,700 it opened with. The cause is not a bad week; week 2 sells $21,600, which is on plan for a ramp. The cause is the first payroll: $24,200 covering weeks 1 and 2, when week 1 sold $14,200 with a full crew on the floor. Labor is 85.9% of sales in week 1 and 55.6% in week 2, by design, because you cannot open a 68-seat restaurant with half a staff. The forecast breaches again in spirit in week 4 — $3,743, or six days of fixed obligations — and only then does the honeymoon lift it clear. Note also that EVERY week the balance falls is a payroll week or a debt-service week. There are no exceptions in thirteen weeks. WHAT IT DOESN'T It assumes the plan's revenue happens. It treats collections as same-week, ignoring roughly two days of card volume permanently in transit — about $8,500 at plan volume, a constant rather than a variable except in the first and last weeks of the business. It contains no equipment failure, no health-department capital item, and no owner draw. It assumes the walk-in build of $7,275 across the quarter and no more. It ends at week 13 and therefore does not show that week 14 opens with the $8,300 payroll and July's $7,242 of insurance and debt. Most importantly: it is a forecast, which means every number to the right of week one is wrong. Its value is the SHAPE, not the digits. THE DECISION Three things, all of them before opening day, none of them possible after week two. (1) Arrange a revolving credit facility of at least $60,000 — the trough plus a working floor — while the business still has a plan rather than a problem. (2) Re-sequence the pre-opening hire dates so the full hourly crew starts in the week of the soft open rather than two weeks before it; three weeks of crew at 56.6% labor is $15,386 more than the same weeks at the 32.3% plan, and the trough is $16,142. (3) Sell four events into weeks 3–8 on Chapter 29's terms, which pay you before you cook. THE LESSON A thirteen-week forecast does not predict the future. It finds the week the plan breaks, while there is still time to change the plan. Bellwether survives the first quarter and ends it with $42,598 in the bank — and it still cannot get from opening day to its second payroll without borrowing. Both of those sentences are true, and only one of them is on the P&L. ```

FIGURE 33.7 — The same thirteen weeks, as a line                     [the Bellwether plan]

  $45k ┤                                                              ██
       │                                                        ██  ██ ██
  $35k ┤                                    ██  ██  ██  ██  ██  ██  ██ ██
       │                                    ██  ██  ██  ██  ██  ██  ██ ██
  $25k ┤                                    ██  ██  ██  ██  ██  ██  ██ ██
       │                                    ██  ██  ██  ██  ██  ██  ██ ██
  $15k ┤                            ██  ██  ██  ██  ██  ██  ██  ██  ██ ██
       │  ██                        ██  ██  ██  ██  ██  ██  ██  ██  ██ ██
  $5k  ┤  ██  ██          ██  ██    ██  ██  ██  ██  ██  ██  ██  ██  ██ ██
    0 ─┼──██──██──────────██──██────██──██──██──██──██──██──██──██──██─██──
       │      ██
 −$10k ┤      ██  ◀ WEEK 2: −$7,442
       └───┬───┬───┬───┬───┬───┬───┬───┬───┬───┬───┬───┬───┬───┬
          op   1   2   3   4   5   6   7   8   9  10  11  12  13

  The shape is what matters, and it has three parts.
    Weeks 1–2   the ramp trough: full crew, partial volume, first payroll
    Weeks 3–7   the honeymoon: revenue peaks in week 6 at $34,600
    Weeks 8–13  the normalization: revenue fades 27% from the week-6 peak,
                rent commences in week 9, and the balance climbs anyway —
                which is the quarter telling you the business works.

Reading the trough

The trough is the output. Everything else in the forecast is inputs.

**$16,142** is the cumulative amount by which Bellwether's operations dip below its opening balance. That number, and not the negative $7,442, is the working-capital requirement the ramp imposes — which is why it appears as a component in §33.3's bottom-up build. The negative $7,442 is merely what happens when you meet a $16,142 requirement with $8,700.

The trough is almost exactly the first three weeks' labor overrun. Weeks 1 through 3 spend $35,800 of labor on $63,200 of sales — 56.6%. At the plan's 32.3% those weeks would have cost $20,414. The difference, **$15,386**, is 95% of the entire trough. Everything else in the first quarter roughly pays for itself.

That is a genuinely useful finding, because it tells you which lever to pull. You cannot fix a cash trough by selling harder in week one; the room only holds 68 people and you have deliberately restricted the covers. You can fix it by moving eleven start dates.

The trough tells you the size of the facility. A revolving line has to cover the deepest point plus a floor you refuse to spend into. Bellwether breaches by $7,442 against a $25,000 working floor, which argues for a minimum $35,000 facility — and §33.3's $56,375 gap between the $45,000 budgeted and the $101,375 required argues for $60,000. Take the larger number. An undrawn line costs a commitment fee; an undersized one costs the business.

Four ways a first forecast lies to you

Every one of these is a mistake I have made or watched somebody make, and each has a one-line fix.

One: it forecasts sales instead of collections. In a restaurant these are close, but "close" hides the week you open. Bellwether sells $14,200 in week one; if two days of that is still settling on the Monday, the number that funds Tuesday's produce order is smaller than the number on the page. Fix: forecast collections, and in week one assume five days of settled sales, not seven.

Two: it puts purchases on the order date. New operators build the forecast from the order guide because that is the document they have. The order guide is a purchasing document, not a payment document. Fix: build the purchases column from the payment terms of each vendor, and in year one assume COD until a credit department says otherwise in writing.

Three: it forgets that payroll is disbursed, not accrued. A monthly budget divides $500,000 of labor by twelve. A cash forecast disburses it twenty-six times on specific Fridays, two of which fall in the same month twice a year. Fix: put the twenty-six dates on a calendar before you build anything.

Four: it nets the sales tax. This is the most expensive one, because it produces a forecast that looks fine and a business that spends the state's money for eleven months. Bellwether collects $25,417 of sales tax in thirteen weeks and remits $15,974 of it inside the window. Fix: show collections gross, show remittances as outflows, and sweep the difference to a separate account the same day. Then the forecast and the bank agree, which is the only way you will ever trust either.

🤝 Hospitality

The 40-top, read as cash.

Chapter 29 established that Bellwether's events collect 25% at signing, 50% at thirty days, and the balance on the night — and that this is the reason event margin is different from every other revenue in the building. An event cover finances you. An à la carte cover is financed by you: you bought the protein eleven days ago, you paid the cook on Friday, and the guest pays you at 9:40 p.m.

Take the private party on the books for the second Friday in October — 40 guests, a $92 per-person food-and-beverage minimum, 15% service charge. The contract is $4,232. Under Chapter 29's terms:

		Timing
25% at signing	$1,058	eight weeks before the event
50% at thirty days	$2,116	four weeks before
Balance on the night	$1,058	the night

That $2,116 installment is 24.3% of Bellwether's entire $8,700 opening cushion. One event's mid-term deposit is a quarter of everything the restaurant has. Which cuts two ways, and both of them matter.

The good way: four such events sold into weeks 3–8 collect roughly $8,400 of deposits before a single ingredient is purchased. That is the cheapest working capital in the business and it is available to any operator who will make the calls. Chapter 29's outbound calendar is, read from this chapter, a cash-management tool.

The dangerous way: the money arrives before the obligation, which makes it feel like income. It is not. It is a liability you will discharge in food, labor, and rentals eight weeks from now, and if you have spent it on Tuesday's produce order you will cater that party out of that week's cash flow with nothing set aside. An operator who is funding this week with next month's deposits has invented a very expensive loan from their future self.

And now the hospitality part, which is where the two threads meet. That October Friday is the night the grill cook no-shows at 3:40 with 142 covers on the books and the 40-top at 6:30. Every choice costs money in a different column — call someone in at overtime, run the station short, comp the party. Read it as cash and one option changes character completely. Comping the party does not cost you the margin; it costs you the $1,058 balance you were counting on for Friday's deposit, on top of a food cost you paid for on Tuesday. The service recovery is still often the right call — Chapter 23 is right that the second visit is where the business lives, and a 40-top is forty potential regulars and a corporate account. But make it as a decision with a number attached, in a week you can afford it, and not as a reflex at 8:15 p.m. That is the difference between generosity and panic, and the guest cannot tell them apart but your bank account can.

🔍 Check Your Understanding

1. Bellwether's week 6 has the highest sales of the entire quarter — $34,600 — and its net change in cash is +$700. Explain why, in one sentence.
2. The forecast shows an ending balance of $42,598 and "genuinely free cash" of $24,855. What are the two deductions, and which of them is a legal obligation to a third party rather than to an employee?
3. If Bellwether had opened with the full $45,000 reserve instead of $8,700, what would the week-2 balance have been, and would the business still have needed a credit facility?

(1: Week 6 is a payroll week — $21,100 covering weeks 5 and 6 — so the highest-revenue week of the quarter barely breaks even on cash; the sawtooth is driven by the disbursement calendar, not by sales. 2: June sales tax of $9,443, which is a trust-fund liability owed to the state, and $8,300 of week-13 payroll already earned by employees but not yet disbursed. The sales tax is the third-party legal obligation. 3: $45,000 − $16,142 = $28,858 at the week-2 close — never negative. It would still want a facility, because $28,858 against a $48,933 monthly fixed obligation is 17.7 days of coverage and the forecast contains no equipment failure, but it would be arranging one from strength rather than from a negative balance.)

33.8 Lines of credit, bridge financing, and borrowing before you need to

There is a rule about borrowing that everyone in this industry learns and almost nobody learns early: the time to arrange money is when you do not need it.

This is not a proverb. It is a description of how credit works. A facility is arranged on the strength of a business's history and its plan. A business in a cash crisis has a history it would rather not discuss and a plan that has just been falsified. The same restaurant, six months apart, is two different applicants, and the one with the problem is the one that gets the worse answer — or, more commonly, the slower one, which in a cash crisis is the same thing.

What the instruments actually are

A term loan — Bellwether's SBA 7(a) note — is a fixed amount, borrowed once, repaid on a schedule. It funds assets: a build-out, a hood, a hearth. It is the wrong instrument for working capital, because working capital is not a one-time need, it is a revolving one that rises and falls with the season.

A revolving line of credit is a committed maximum you can draw against, repay, and draw again. You pay interest only on the drawn balance, plus a commitment or unused-line fee on the undrawn portion — commonly a fraction of a percent per year. It is the correct instrument for seasonality, for the ramp trough in §33.7, and for the February problem in §33.6, and it is the instrument most independents do not have because nobody told them to ask.

Bridge financing is short-term money against a specific, identified future inflow — an insurance settlement, a landlord reimbursement of a TI allowance, a tax refund. It has a defined end. If you cannot name the event that repays it, it is not a bridge, it is a loan you are describing optimistically.

Equipment financing — Bellwether's $60,000 lease — funds a specific asset and is secured by it. It does not help with working capital and should never be used for it.

Merchant cash advances and daily-remittance products deserve their own treatment, below.

Sizing a line

Two anchors, and take the larger.

Anchor one: the trough. From §33.7, Bellwether's forecast dips $7,442 below zero and $16,142 below its opening balance. A facility has to cover the trough and leave a working floor you refuse to spend into — call it $25,000, roughly two weeks of fixed obligations. That is a **$35,000 minimum**.

Anchor two: the requirement gap. From §33.3, the bottom-up working-capital requirement is $101,375 against $45,000 budgeted — a **$56,375 shortfall.** A facility sized at $60,000 closes it with a small margin.

Take $60,000**. The undrawn cost, at a commitment fee in the range of a quarter to a half point, is on the order of **$150 to $300 a year. That is roughly what Bellwether spends on linen in a fortnight, and it is the difference between a business that can absorb a February and one that cannot.

The discipline that makes a line safe

A revolving line is a tool that turns into a trap in exactly one way: the balance stops going back to zero.

That is the whole discipline, and it fits in three rules.

Rule one: the line must clean up. A properly used seasonal line is drawn in January and February and repaid by June. If your line has not touched zero in twelve months, it has stopped being a line of credit and become a term loan with a variable rate and no amortization schedule — which is a considerably worse product than the term loan you would have negotiated deliberately.

Rule two: never fund an operating loss with a revolver. A line covers timing. If the business is structurally unprofitable, the line does not solve the problem; it postpones the diagnosis and adds interest to it. The test: does your thirteen-week forecast show the drawn balance being repaid inside the horizon? If not, you have a P&L problem wearing a cash problem's clothes, and Chapter 39 is the chapter you actually need.

Rule three: draw on a schedule, not on a feeling. Decide, in advance and in writing, at what balance you draw and how much. "We draw $15,000 when the forecast shows the balance dropping below $20,000, and we repay in the first week the balance exceeds $45,000." That is a policy. "I'll draw if it gets tight" is not, and the operator who improvises it draws late, small, and repeatedly.

⚠️ Where the Money Leaks

The merchant cash advance, and why it is the last mistake.

When a restaurant is out of options, the offers arrive. They arrive by email, they arrive through the payment processor, and they are extremely easy to accept — often funded in forty-eight hours with no collateral and almost no underwriting.

A merchant cash advance is not structured as a loan. It is a purchase of future card receivables at a discount, repaid by a daily holdback — a fixed percentage of every card batch, taken before the money reaches your account. The price is quoted as a factor rate rather than an interest rate, which is the crucial thing to understand, because a factor rate is not comparable to an APR and is not meant to be.

An illustrative structure: $50,000 advanced at a 1.35 factor** means **$67,500 repaid. At a 12% holdback on Bellwether's card volume — roughly $3,990 a day at plan — that is about $479 a day, and the advance is retired in something like 141 business days, call it four and a half to five months.

So: $17,500 of cost on $50,000, over about five months. That is 35% for five months on the full amount, and because the balance amortizes daily, the effective annualized rate is far higher than the simple annualization suggests. Industry and regulatory commentary on these products routinely describes effective APRs in the high double and triple digits. Compare with the $150–$300 a year that an undrawn $60,000 line costs.

And the structure is worse than the price. The holdback comes off the top of every batch, which means it takes the most on your best days, and it takes it during the exact weeks — January, February — when card volume is lowest and each dollar matters most. An operator who takes an advance in February to survive February has committed 12% of every card dollar through the following summer, which is when the business would otherwise have rebuilt its reserve.

The honest framing: a merchant cash advance is not financing. It is the sale of next season at a discount, and it is very hard to take only once. If you are reading the offers, the decision in front of you is not "which advance" — it is the one in Chapter 39.

Emergency programs are real, and they are not a plan

The COVID-19 shutdowns produced two federal programs restaurant operators will remember: the Paycheck Protection Program (PPP), which provided forgivable loans tied to maintaining payroll, and the Restaurant Revitalization Fund (RRF), created specifically for foodservice. Both were real, both were consequential, and both are matters of public record.

Two structural lessons, which are the transferable part.

The RRF was oversubscribed. Demand exceeded available funding, and many eligible applicants who filed correctly and on time received nothing. That is not a criticism of the program; it is the nature of a fixed appropriation against an industry-wide shock. Design your business so that its survival does not depend on winning a lottery.

The operators who got money fastest were the ones with clean books. Both programs required payroll documentation, tax filings, and bank records on short deadlines. Restaurants with a current chart of accounts, filed returns, and a bookkeeper who could produce a report in an afternoon applied in week one. Restaurants whose records were a shoebox applied in week six, if at all. Chapter 31's weekly discipline turned out, unexpectedly, to be disaster preparedness.

Program rules, eligibility, and availability change constantly and vary by jurisdiction, and nothing here should be read as advice about any current program. If a relief program is open when you need one, work with your accountant and read the actual guidance.

33.9 Early warning signs, and what to do in the first week you see one

Cash problems announce themselves. The signals are behavioral before they are numerical, which is why the operator is usually the last person to name what is happening — you cannot see your own habits changing.

Here is the list. Read it as a diagnostic instrument, not a scare piece.

Three of those deserve elaboration.

Rising DPO against flat sales is the cleanest quantitative signal in this chapter. It is objective, it is computable monthly from data you already have, and it precedes every other symptom. If your average payment period moves from fourteen days to nineteen while sales are unchanged, you have borrowed roughly five days of purchases — about $5,900 at Bellwether's volume — from people who did not agree to lend it and will eventually notice.

"You have stopped counting inventory" is the one operators recognize with a wince. It is the same avoidance Chapter 1 described in the restaurant that ran 34.5% food cost for eleven months while believing it was 30%. Cash pressure makes counting feel like a luxury — there is no time, the manager is on the floor, the count takes two hours. Those two hours are the highest-return two hours in the building precisely when things are tight, because a cash problem and a variance problem are frequently the same problem, and Chapter 34 is about to show you exactly how much of one is hiding inside the other.

"An advance offer is starting to look reasonable" is a state of mind, not a number, and it is the most reliable late-stage indicator I know. Nobody in a healthy business reads those emails.

What you do in the first week

Not the first month. The first week. In this order:

1. Build or update the thirteen-week forecast. Two hours. You cannot act on a problem you have not sized.
2. Separate the money that is not yours. Sales tax, payroll withholding, event deposits, gift-card liability. One account, one sweep, no debit card.
3. Call three vendors before they call you. Operators dread this and it is almost always the easiest conversation of the week. A distributor's credit department deals with restaurants constantly; a proactive call proposing a specific plan on a specific date is treated completely differently from a missed payment followed by silence.
4. Fix the largest controllable line for four weeks. Chapter 19's schedule is the fastest lever in the building — it moves within seven days. Chapter 13's purchasing moves within two weeks. Nothing else moves in a month.
5. Arrange or draw the facility. If you have a line, draw deliberately per the policy in §33.8. If you do not, start the conversation this week rather than next month.
6. Stop the draws. Every distribution during a cash crisis is a decision to fund a household with the business's survival money, and it is the decision owners regret most.
7. Tell your accountant. Not at year-end. Now.

🧮 Run the Numbers

The cheapest cash in the building is the cash you are already losing.

Chapter 34 quantifies Bellwether's exposure to leakage — over-pouring, unrecorded comps and voids, shorted deliveries, register discrepancies, over-portioning, unclaimed credit memos — at $53,122 a year, against a control package costing $4,849.

Read those two numbers through this chapter and they change character entirely.

• $53,122 is 32.6 days of Bellwether's $48,933 monthly fixed obligation. The leak, unaddressed, is a full month of the building's existence, every year.
• It is 94% of the $56,375 working-capital shortfall identified in §33.3. The business does not need to raise the entire gap. It needs to raise part of it and stop losing the rest.
• The controls return roughly eleven dollars of exposure closed per dollar spent — $53,122 ÷ $4,849 = 11.0. There is no financing product in §33.8 that competes with that, and no revenue initiative in Part VI either.

This is the argument for reading Chapter 34 as a cash chapter rather than an ethics chapter. Controls are not an accusation. They are the cheapest source of working capital an independent restaurant has, they require no application, and unlike a line of credit they do not have to be repaid.

🔍 Check Your Understanding

1. Why is a rising DPO against flat sales a better early-warning indicator than a falling bank balance?
2. An operator says: "We're fine — the line of credit covers us." What single follow-up question tells you whether that is true?
3. Bellwether's February shows $10,694 of profit and a $10,954 decline in cash. Which of those two numbers would a lender, a landlord, and a payroll processor each care about, and why does that matter for how you report your own performance?

(1: Because a bank balance is a level and DPO is a rate — the balance can be propped up for weeks by stretching payables, delaying a hire, or drawing a line, all of which show up in DPO first. DPO is measuring the behavior; the balance is measuring the result. 2: "Has the balance been at zero in the last twelve months?" If not, it is not a line, it is unamortized term debt funding an operating loss. 3: A lender looks at profit and debt-service coverage; a landlord looks at whether rent clears on the first; a payroll processor looks at whether the draft funds on Thursday. Two of those three are cash tests. Report both numbers, always, and never let the profit number stand alone.)

🍽️ The Business Plan

Checkpoint 33 of 40 — the Cash Flow section.

This is the section of Bellwether's plan that the rest of the document has been avoiding.

What this chapter contributes:

Bellwether — Cash Flow and Working Capital (constructed teaching example; all figures illustrative)

1. The thirteen-week cash forecast. Figure 33.6, in full, as an appendix exhibit: weeks 1–13 of operation, opening balance $8,700, ending balance $42,598, with the memo lines showing $9,443 of unremitted sales tax and $8,300 of accrued payroll deducted to arrive at $24,855 of genuinely free cash. The forecast's central finding is stated in the plan in one sentence: the balance closes week two at negative $7,442, a cumulative $16,142 below the opening balance, and the cause is the first biweekly payroll landing against a deliberately restricted opening volume.

2. The working-capital requirement, bottom-up.

Component	Amount
Pre-opening overrun beyond the $35,000 budget (Chapter 9's build: $71,300)	$36,300
Deepest cumulative operating deficit, weeks 1–13	$16,142
Operating floor: 30 days of fixed obligations	$48,933
Required at opening	$101,375
Budgeted in the use of funds	$45,000
Shortfall	$56,375

Cross-checked against the practitioner rule of sixty days of fixed obligations — $48,933 × 2 = $97,866 — the two methods agree within 3.5%.

3. The reserve, stated honestly. The plan's $45,000 working-capital reserve is **$8,700 on opening day, because the $36,300 pre-opening gap had nowhere else to come from. At $48,933 of monthly fixed obligations that is 5.3 days of coverage**, against 27.6 days for the full reserve and 62.1 days for the bottom-up requirement. This is the plan's largest single weakness and it is named as such in the document rather than buried.

4. The February problem. Bellwether opens the first Tuesday in April, so its first February is month eleven — after the honeymoon, before the patio, with the annual periodic bills arriving together for the first time. February revenue of $106,200 is 33.0% below December and 26.0% below October, against fixed obligations that do not move. Three of five February dinner services run below the Chapter 32 cash break-even of 77 covers. The month shows $10,694 of profit and a $10,954 decline in cash — a $21,648 divergence, fully reconciled in §33.6.

5. The timing calendar. Figure 33.4, as an exhibit: twelve months of scheduled obligations with the four collision months identified — February (lowest revenue plus $15,570 of annual items), January (revenue down $44,700 from December, and December's larger sales-tax remittance landing against January's smaller collection), October (the three-payroll month, $19,231 of extra cash out that the P&L never records), and June (rent commences; the abatement, which began at the certificate of occupancy in March, is spent).

6. The facility. A $60,000 revolving line of credit, arranged before opening, with a written draw-and-repay policy: draw when the rolling forecast projects a balance below $20,000; repay in the first week the balance exceeds $45,000; the balance returns to zero at least once every twelve months. The undrawn carrying cost is on the order of $150–$300 a year.

7. Cash-flow policy. Trust-fund money — sales tax, payroll withholding, event deposits, gift-card liability — held in a separate account with a same-day sweep and no debit card. Thirteen-week forecast updated every Monday in a fifteen-minute standing meeting, forecast-versus-actual tracked from week one. Vendor-terms conversation opened with the broadline distributor in month one, revisited at month six and month twelve.

What this checkpoint does not settle.

Quite a lot, and the plan is stronger for saying so.

It does not settle where the $56,375 comes from. Additional owner injection, a larger facility, a smaller build-out, a delayed hearth, a shorter pre-opening — each is a real option with a real cost, and the plan presents them rather than pretending the gap is closed.

It does not settle whether the 58.0% prime cost required across weeks 14–52 is achievable. Chapter 9's Q1 at 66.6% on $363,100 leaves $688,455 of prime cost to sit on $1,186,900 of remaining revenue. Chapters 11, 13, and 19 built the systems; whether a first-time ownership team executes them is a question no forecast answers.

It does not settle the February revenue assumption. $106,200 is a constructed estimate. If February runs 10% light — $95,580 — the month's cash decline moves from $10,954 to roughly $17,000, and the facility is not optional in any sense.

It does not settle what happens in the second February, which arrives with a rent escalation, a lease that is one year closer to its five-year option, and whatever the reserve looks like by then.

Open questions carried forward:

1. Where does the $56,375 working-capital shortfall come from, and what does each source cost? (Chapters 5, 39, 40)
2. How much of the $53,122 leak exposure can controls actually close in year one? (Chapter 34)
3. Does the events calendar produce enough advance deposits to flatten the ramp trough? (Chapters 27, 29)
4. If the first February goes badly, is that a seasonality problem, an execution problem, or a business that does not work? (Chapter 39)
5. What does the completed plan look like to somebody reading it cold? (Chapter 40)

Conclusion

Chapter 1 promised that a profitable restaurant can run out of money, and named cash timing as one of the four mechanisms that actually close restaurants. This chapter cashed the promise on a specific business.

Bellwether's plan year shows $261,020 of operating profit and $191,520 of cash after debt service. Its February shows $10,694 of profit and a $10,954 decline in the bank balance. Its first thirteen weeks end with $42,598 in the account — of which $24,855 is actually free — and its second week ends at negative $7,442. Every one of those numbers is correct. They describe the same restaurant.

The divergence is not mysterious once you know where to look: debt principal never appears on the P&L, owner draws never appear on it, inventory build consumes cash at no cost to profit, and everything else is timing. Timing is not noise. It is a calendar, and the calendar is knowable in advance — the three-payroll month, the abatement that ran out on the landlord's clock rather than yours, the sales tax that remits one month behind and therefore punishes you in exactly the month revenue falls, the annual bills that all chose February.

The instrument is the thirteen-week cash forecast, and its purpose is not accuracy. It is to find the week the plan breaks while there is still time to change the plan. Bellwether's forecast finds week two, and having found it, offers three responses that all had to happen before opening day: move eleven hire dates, arrange a $60,000 facility, and sell four events on terms that pay you before you cook.

The number the chapter turns on is **$8,700** — a $45,000 working-capital reserve reduced to 5.3 days of coverage by a $36,300 pre-opening overrun that nobody decided to incur. Not a scandal. Not fraud. The most ordinary event in the industry: a bottom-up budget that was honest and a plan number that was not, meeting eleven days before opening, with the reserve as the only line left to take it from.

Which brings us to the next chapter, and to a claim that should now read differently than it would have three chapters ago. Chapter 34 is about financial controls — cash handling, comp and void authorization, inventory variance, the daily sales report. It reads like an ethics chapter and it is not. Bellwether's leak exposure is $53,122 a year against $4,849 of controls: nearly eleven dollars closed per dollar spent, 32.6 days of fixed obligations, and 94% of the working-capital shortfall this chapter just identified. There is no financing product in §33.8 that comes close.

The cheapest working capital in a restaurant is the money it has already stopped losing.

Key Terms

Cash flow vs. profit — profit is an accounting result computed over a period under accrual conventions; cash flow is the actual movement of money in and out of the bank account, on dates. They diverge structurally in a restaurant because debt principal and owner draws never appear on the P&L, inventory build consumes cash at no cost to profit, and periodic obligations are expensed evenly but paid in lumps. (Ch. 33)

Working capital — current assets minus current liabilities; operationally, the cash and near-cash a business needs to fund the gap between paying for things and being paid for them. Distinct from the construction contingency, and funded at the moment capital is raised because it cannot be raised later. (Ch. 33)

Cash conversion cycle (CCC) — days inventory outstanding plus days sales outstanding minus days payable outstanding; the number of days between paying for inventory and collecting cash from selling it. Mature restaurants often run a negative cycle and are financed by their vendors; new restaurants, with no credit history and a slow-turning beverage inventory, do not. (Ch. 33)

Accounts-payable terms — the agreement governing when an invoice is due: COD, net 7, net 14, net 30, EOM, or a discount term such as 2/10 net 30. Terms are a financing decision priced in dollars, not a clerical detail. (Ch. 33)

Vendor float — the cash a business holds because purchases have been received but the invoices are not yet due; free, revolving, and not the restaurant's money. At Bellwether's $8,274 a week of purchases, net-14 terms are worth about $16,500 of permanent working capital. (Ch. 33)

Seasonality — the predictable variation of revenue across the year against fixed obligations that do not vary. A cash problem rather than a profit problem, because it nets out over twelve months and does not net out in February. (Ch. 33)

The 13-week cash forecast — a week-by-week projection of collections, disbursements, and the resulting bank balance over one quarter. Its purpose is to locate the trough — the lowest projected balance — early enough to act on it. (Ch. 33)

Runway — the number of days a business can meet its fixed obligations from cash on hand with no revenue; cash ÷ (monthly fixed obligations ÷ 30). A stress metric, not a prediction. (Ch. 33)

Reserve — cash held deliberately and not spent on operations, sized to a specific requirement rather than to what was left over. Distinct from the construction contingency and from an undrawn credit facility, though all three do related work. (Ch. 33)

Line-of-credit discipline — the written rules that keep a revolving facility from becoming unamortized term debt: draw and repay on stated triggers, never fund an operating loss, and return the balance to zero at least once every twelve months. (Ch. 33)

Spaced Review

1. Without looking back: name the three items that consume cash without ever appearing on a profit-and-loss statement, and state which one is largest for a restaurant carrying an SBA note.
2. A restaurant's thirteen-week forecast shows a trough of negative $12,000 in week five and an ending balance of $60,000 in week thirteen. The owner says the forecast proves the business is fine. What is wrong with that reading, and what should the forecast cause them to do this month?
3. From Chapter 32: Bellwether's accrual break-even is 66 dinner covers a night and its cash break-even is 77. Explain in one sentence why the second number is higher, using at least one item from this chapter.
4. From Chapter 13: Bellwether purchases $6,438 of food a week. If the broadline distributor moves the account from COD to net 14, how much one-time cash does that release, and why is that number not the same as a profit improvement?
5. From Chapter 1: the failure research finds roughly a quarter of restaurants do not reach their first anniversary and something close to six in ten are gone within three years. Using this chapter's material, explain why that pattern is more consistent with a cash-timing story than with a bad-food story.
6. The recurring question: an operator has $18,000 in the account, a $9,400 payroll on Friday, a $7,933 rent payment on the first, and a $6,200 produce and protein order that has to be placed Tuesday to serve the weekend. They also have $11,300 of sales tax collected sitting in that same account. What is the actual position, what are the options in order of preference, and which option is not an option?