How Much Should You Hold in Cash Reserves?

A payroll date does not move because a large invoice is expected to be paid. Neither does a loan debit, tax payment, or processor fee. Cash reserves exist to cover that difference between expected cash and cash that has actually cleared - without forcing the business into a rushed borrowing, payment-delay, or headcount decision.

For an operating company, the question is not whether a reserve is prudent. The harder question is how much cash is genuinely available, how much must remain untouched, and how quickly a downside case can consume the balance. A useful reserve policy answers all three from reconciled transactions and a day-level forecast, not from a revenue multiple or a single bank-account screenshot.

What cash reserves are meant to protect

Cash reserves are liquid funds held beyond the amount needed for planned near-term disbursements. They provide capacity to absorb timing gaps and adverse events while the business continues to meet its obligations.

That definition matters because a bank balance is not automatically a reserve. It may include customer deposits that must be delivered against, sales-tax liabilities, payroll taxes, restricted loan proceeds, or cash already committed to bills and payroll. Treating all cash as available is how a company can look healthy on Monday and face a preventable shortfall on Friday.

A reserve protects against operational realities such as a delayed customer payment, a weaker settlement cycle, an unexpected repair, a seasonal demand decline, or a month where payroll lands before receivables. It is not primarily an investment portfolio or a substitute for fixing unprofitable operations. If the base forecast repeatedly consumes the reserve, the operating model needs attention.

Start with available cash, not the headline balance

The reserve calculation begins after reconciliation. Every connected bank account, card account, processor clearing balance, and cash-equivalent account should be accounted for. The system should identify what has cleared, what remains in transit, and which balances are restricted or committed.

A practical view separates four amounts:

| Cash measure | What it answers | |---|---| | Ledger cash | What cash accounts show after reconciliation | | Restricted cash | What cannot fund normal operations | | Committed cash | What is already allocated to known near-term obligations | | Available cash | What remains to fund operations and the reserve |

Available cash is the starting point, not the target. The target comes from the business's actual cash obligations and risk profile.

For example, a company may have $600,000 across accounts. If $90,000 is a customer deposit, $65,000 is reserved for payroll taxes, and $180,000 is required for payroll, rent, and approved vendor payments before the next expected collection, it does not have a $600,000 reserve. It has $265,000 of uncommitted liquidity before considering uncertainty.

That distinction is especially important when payment processors settle several days after a sale. Revenue may be visible in the P&L, but the cash may still be in a clearing account, net of fees, refunds, and holds. The reserve should be based on settlement timing that has been observed and reconciled, not assumed.

Set the target from cash burn and timing risk

The familiar rule of thumb is three to six months of operating expenses. It is a reasonable starting point for discussion, but it is too blunt to operate from. A business with annual contracts collected upfront has a different risk profile from a project-based firm whose top three customers pay 60 to 90 days after invoicing.

A better reserve target starts with recurring, unavoidable cash outflows: payroll and related taxes, rent, debt service, core software, insurance, essential vendors, and minimum owner draws where applicable. Use cash outflows, not accrual expenses, because the reserve must meet dates on the bank calendar.

Then adjust the coverage period for the realities of the business:

• A stable business with diversified customers, predictable settlements, and low fixed obligations may operate safely near the lower end of its range.
• A company with concentrated receivables, seasonal demand, lumpy project revenue, or long collection cycles needs more coverage.
• A business carrying variable-rate debt, covenant requirements, inventory commitments, or regulatory exposure should add a specific buffer for those risks.
• A fast-growing company may require a larger reserve even when profitable, because hiring and working-capital needs arrive before new revenue converts to cash.

The point is not to create a universal percentage. It is to make the assumptions explicit and testable.

Use a day-level forecast to test the reserve

Monthly forecasts can hide the exact failure a reserve is supposed to prevent. A month can close cash-positive while still containing a 10-day period in which payroll, a debt payment, and vendor debits land before the largest customer collection.

The operating test is therefore a daily liquidity floor. Build the forecast from expected receipt dates, processor settlement schedules, scheduled bills, payroll calendars, recurring debits, loan amortization, and known tax payments. Every projected movement should retain a source, owner, date assumption, and connection to its underlying transaction or rule.

A simple policy might define a minimum operating floor of $150,000 and a strategic reserve of $300,000. The forecast must remain above $450,000 under the base case. If it falls to $410,000 for four days before a customer pays, the business has not met its policy, even if the month-end balance recovers to $700,000.

This is why direct-method forecasting is useful for reserve management. It models the expected movement of money into and out of accounts on the dates it is expected to occur. It does not infer near-term cash from net income and hope timing works itself out.

Run scenarios before the decision becomes urgent

A reserve target becomes credible when it survives scenarios that reflect how the business actually breaks. Start with the events that would change cash timing or cash magnitude materially: a major customer pays 30 days late, a processor increases its hold period, bookings fall 20%, a key supplier requires a deposit, or two planned hires start before a delayed contract closes.

For each scenario, measure three outputs: the lowest daily cash point, the date it occurs, and the number of days below the required floor. Those outputs create a direct operating choice. The company can postpone hiring, accelerate collections, renegotiate payment terms, draw on a facility, or accept the risk knowingly. What it should not do is rely on a forecast that cannot explain why the low point moved.

Scenario planning also prevents an unhelpful habit: treating every cash surplus as excess cash. A balance may exceed the target today but be required to cover a seasonal trough six weeks from now. Conversely, a reserve that appears thin may be adequate if signed collections clear before the next major outflow and the downside cases remain above the floor.

Put the reserve policy into operating controls

The reserve should be governed like any other financial control. Define the target, minimum floor, covered obligations, approved uses, and escalation path. Assign an owner - often the controller, CFO, or finance lead - and establish a review cadence tied to the cash forecast.

The policy should also state what happens when the floor is breached. For example, a breach might automatically require a 13-week forecast refresh, a review of discretionary spend, collection outreach on specific invoices, and approval for new commitments above a defined threshold. The objective is not bureaucracy. It is to turn early warning into a repeatable response.

Data discipline is part of that control. If a forecasted receipt changes, the change should be visible. If a transaction is classified differently, the resulting cash view should be replayable from its rules and source rows. A finance team cannot defend a reserve decision if the underlying numbers shift without an audit trail.

This is the operating logic behind platforms such as Shekl: transactions are reconciled to $0.00, rules are deterministic, and every forecast movement can trace back to the ledger evidence that supports it. AI may help a team investigate or explain, but it should never silently alter the cash calculation.

Avoid two common reserve mistakes

The first mistake is holding a large reserve without distinguishing liquidity from confidence. Excess idle cash can be appropriate when risk is elevated, but it also has an opportunity cost. It can defer debt reduction, necessary equipment, product investment, or a planned growth initiative. The decision should follow tested downside exposure, not an arbitrary desire for a larger bank balance.

The second is treating the reserve as untouchable in every circumstance. A reserve is designed to be used when a defined event occurs. The discipline lies in knowing when a draw is warranted, documenting why it happened, and rebuilding the reserve through a specific plan once conditions normalize.

A cash reserve is not a static number on a board slide. It is a living commitment between the company’s obligations, its collection reality, and its tolerance for risk. Keep that commitment connected to reconciled data and a forecast that shows every date, and the next decision can be made before cash turns into an emergency.