Sizing a restaurant working-capital line in 2026 — how to model the line you'll actually draw

There are two ways restaurant operators get the working-capital line wrong, and both cost real money.

The first is undersizing. The operator qualifies for $75K, takes $50K because it “feels comfortable,” and then a walk-in compressor dies in week two of a slow month. The line is already drawn against payroll. The compressor repair goes onto a merchant cash advance at an effective APR that would make a payday lender blush. The original $25K of unused capacity that would have covered the emergency was left on the table to look conservative on the application.

The second is oversizing. The operator qualifies for $300K, takes the full commitment because “more is better,” and then pays roughly 0.25% to 0.50% per year in commitment fees on $200K of capacity that never gets drawn. Over a three-year line, that’s $1,500 to $3,000 in dead-weight cost for capital that sat unused.

Right-sizing a working-capital line in 2026 is closer to a math exercise than a negotiation. With bank credit tighter than it was in 2024 and alternative lenders pricing more aggressively into the gap, the operators winning at this are the ones modeling against their actual cash-flow shape — not against the lender’s maximum.

What a working-capital line is (and isn’t) for restaurants

A revolving working-capital line is short-duration capital that fills timing gaps. Money comes in late, money goes out early, and the line bridges the difference. You draw, you repay, you draw again. The pricing — typically 6% to 12% APR at bank tier, 10% to 25% at specialist non-bank lenders in May 2026 — only makes sense if the capital is moving.

Good uses for the line:

• Payroll smoothing across a slow week when receivables from a catering contract are 30 days out
• Produce orders on a menu launch where you’re absorbing higher food cost for three weeks before pricing settles
• Equipment downtime cash when a hood fan or POS system needs an immediate fix and you don’t want to touch reserves
• Seasonal inventory buys ahead of a Q4 holiday push or a Restaurant Week menu

Bad uses for the line — and these are the ones that quietly convert a credit facility into a term-loan problem:

• Long-term equipment purchases. A $40K oven on a working-capital line is a financing mismatch. Use an equipment loan with 5–7 year amortization.
• Real estate or buildout. Same problem, longer duration. The line will be maxed for years.
• Covering a recurring monthly deficit. If the restaurant is losing money every month, the line is hiding the symptom. It is not capital — it is debt buying time.

The first test before sizing anything: is the use case actually a timing gap, or is it a permanent capital need wearing a timing-gap costume?

The inputs that actually determine line size

Five inputs drive the right number. Most operators only think about two of them.

1. Fixed monthly operating expenses. Payroll (including the owner draw if it’s a real number), rent, insurance, debt service on existing loans, utilities baseline, software and POS subscriptions, recurring marketing. This is the floor. If revenue went to zero tomorrow, this is what still goes out the door.

2. Variable-with-volume expenses. Food cost as a percentage of revenue (typically 28%–35% for full-service, 30%–38% for casual), variable labor on top of fixed payroll, paper goods, credit card processing. These scale with covers. In a slow month they shrink, but not to zero — and they shrink slower than revenue does.

3. Seasonal peaks and dips. Fine dining peaks in Q4 holidays and dips hard in January and August. Casual concepts peak in summer. Restaurant Week — depending on your market — can mean a 20% revenue lift on thin margins or a 15% dip if your concept doesn’t fit the format. Map the worst-case two-month dip from the last 24 months of P&L.

4. Receivable timing variance. This is the input independent restaurants under-weight the most. Catering contracts often pay net-30. Corporate billing for private dining can run net-45. Delivery platform reconciliation usually clears in 7–14 days but can extend during platform disputes. The line needs to cover the longest realistic gap between booking the revenue and seeing the cash.

5. Planned non-recurring investments in the next 12 months. A new menu rollout with one-time photography costs. A patio buildout under $25K that doesn’t justify a separate loan. A bar program refresh. List them, total them, add them to the model.

The output of this exercise is not “what could I borrow.” It’s “what would I actually draw, and for how long, across a realistic worst-case quarter.”

The quick-math rule for steady-state restaurants

For an operator without an unusually seasonal concept or major planned investments, the working rule that lines up with how bank-tier and non-bank specialist lenders actually underwrite is:

Target line size = (2 to 3 months of fixed operating expenses) + (planned non-recurring 12-month investments)

Walk it through a concrete example. A full-service restaurant with $150K in fixed monthly operating expenses — payroll, rent, insurance, debt service, baseline utilities, software — and one planned investment of $20K for a patio refresh in Q3.

• Two months of fixed expenses: $300K
• Three months of fixed expenses: $450K
• Plus planned investment: +$20K

The right-sized line lands between $320K and $470K. An operator with stable revenue and minimal seasonality should size toward the lower end — call it $325K. An operator with meaningful seasonal compression or catering receivables stretching 30+ days should size toward the upper end — call it $475K.

What this rule does not do: justify sizing to the lender’s maximum. If the operator above qualifies for $600K because the trailing 12-month revenue and DSCR support it, taking the full $600K means paying commitment fees on $125K to $275K of dead capacity. On a 0.375% annual commitment fee, that’s roughly $470 to $1,030 per year — for nothing.

Working capital lines for restaurants in 2026 typically run $10K to $500K in face value, with repayment terms of 6 to 18 months on individual draws. The line itself usually renews annually, sometimes on a 24-month cycle for stronger borrowers.

How specialist non-bank lenders size lines vs. bank tier

Bank-tier lenders — community banks, regional banks, SBA Express lines — will generally underwrite a working-capital line for restaurants that have 24+ months in business, are profitable on the trailing 12 months, and have a DSCR of 1.25x or better. Pricing in May 2026 sits at roughly prime + 0% to prime + 5%, which against a prime of 6.75% puts the all-in rate at 6.75% to 11.75% APR. The catch: decision timelines run 4 to 8 weeks, documentation is heavy, and the line is sized conservatively against historical financials.

Specialist non-bank lenders — the segment that includes non-bank working capital lines built for restaurant cash flow — underwrite faster and look at the business differently. They weight cash-flow patterns from bank statements and processor data more heavily than tax returns, decide in 24–72 hours, and will often work with restaurants that are 12–18 months in business or have one weak year in the trailing 24 months. APRs land in the 10% to 25% range, reflecting both the speed and the risk profile.

The sizing logic also differs. Bank-tier lenders tend to anchor on a multiple of trailing revenue (often 8%–15% of TTM revenue). Specialist non-bank lenders are more likely to size against a percentage of monthly deposits — typically 1x to 2x average monthly deposits, capped against the operator’s stated use case.

The practical implication for operators: if your number from the quick-math rule is materially different from what either lender type wants to offer, the gap is the conversation. Show them the model. A lender who sees a thoughtful sizing exercise will price differently than one who sees an operator taking whatever number is on offer.

Mistakes operators make in 2026

Three patterns keep showing up, and each one is expensive.

Drawing the line for payroll, month after month, with no repayment cycle. The line was supposed to bridge a timing gap. After six months of consecutive payroll draws with no meaningful paydown, it is functionally a term loan being priced like a revolver. The lender will notice this before the operator does — most working-capital line agreements include a covenant requiring 30 consecutive days of zero balance at least once every 12 months. Miss the cleanup window and the line gets called or restructured under worse terms.

Sizing to “maximum qualified” instead of actual need. The lender will happily quote you $500K because their underwriting model says you can carry it. That doesn’t mean you should take it. The commitment fee math punishes oversizing every year the line renews. Before signing, review what restaurant lenders require for working-capital line applications — the documentation requirements tell you what the lender is actually anchoring on, which informs how aggressively they’ll size up if you ask for less than the maximum.

Not refreshing the sizing model after a 10%+ revenue change. Revenue went up 15% year-over-year? Your fixed expenses probably moved too, your receivable timing changed if you added catering, and your seasonal shape might have shifted. The line sized to 2024 financials may be 20% off the right number for 2026 operations. Refresh the model at every annual renewal.

The bottom line

The math is not glamorous, but it’s repeatable. Total your fixed monthly expenses, multiply by two to three, add planned investments, and compare the output to what the lender is offering. If the offer is materially higher, that’s your negotiation room to push for tighter pricing on a smaller commitment. If it’s lower, you have a documented case for why the line should be larger.

The operators who treat working-capital line sizing as a math exercise — not as a vanity metric or a “take what you can get” negotiation — pay less in commitment fees, draw less in emergencies, and never end up funding a compressor repair on a merchant cash advance.

The line should fit the cash-flow gap. Nothing more, nothing less.