Almost every nonprofit board has been told it should hold three to six months of operating expenses in reserve. Almost none has been told where the number came from, and it does not come from analysis of that organisation’s cash flows. It is a convention — a reasonable default that became a rule through repetition.

The convention is not wrong so much as undifferentiated. A subscription-funded arts organisation with predictable revenue and a fixed cost base faces a completely different liquidity problem from a federally reimbursed human services agency whose costs are incurred sixty to ninety days before the money arrives. Applying one target to both produces a reserve that is too large for one and dangerously small for the other.

What a reserve is actually for

An operating reserve exists to absorb timing mismatches and shocks without forcing the organisation into decisions it would not otherwise make — deferring payroll, drawing an expensive line, or cutting programme in the middle of a service year.

That framing matters because it separates the reserve from other balance-sheet items that get confused with it:

  • A board-designated reserve is unrestricted net assets the board has earmarked. The board can undesignate it. It is a real reserve.
  • Donor-restricted net assets are not a reserve. They are someone else’s money with conditions attached, and using them to cover a payroll gap is a compliance problem.
  • An endowment is not an operating reserve. Its purpose is perpetual support, its spending is governed by policy and often by state law under UPMIFA, and liquidating principal to cover operations is a governance event.
  • A line of credit is not a reserve, though it serves an overlapping purpose. It is contingent on a lender’s willingness to fund, which is least reliable exactly when it is most needed.

An organisation reporting a healthy “reserve” that consists largely of restricted funds or an undrawn line does not have the protection its board believes it has.

Sizing against your own volatility

A more defensible method starts from the organisation’s actual cash cycle rather than a months-of-expenses rule of thumb.

Start with the reimbursement lag. For organisations funded by cost-reimbursement contracts — government human services, many health programmes — the dominant risk is the gap between incurring cost and receiving payment. Measure the actual lag from service delivery to cash receipt, including the tail on slow payers, then size the reserve to fund operations across that period plus a margin for a payer who stops paying.

An agency with a ninety-day reimbursement cycle needs meaningfully more than three months of expenses, because three months of reserve covers the ordinary cycle with nothing left for a disruption within it.

Then add concentration risk. Calculate what share of revenue comes from the single largest funder, and what the organisation would have to do if that funder did not renew. A nonprofit with sixty percent of revenue from one contract has a different reserve requirement from one with twenty funders at five percent each, even at identical budget size.

Then consider fixed-cost rigidity. How much of the cost base can actually be reduced within a quarter? An organisation with a long building lease, specialised staff who cannot be replaced if released, and licensing requirements has less ability to shrink than one whose costs are largely variable. Rigidity increases the reserve requirement because the alternative levers do not work quickly.

Finally, subtract genuine flexibility. A committed, tested line of credit does reduce the reserve requirement — not to zero, but it is not nothing. So does a funder relationship with a demonstrated history of advancing funds. Count these honestly, including how they would behave in a stressed scenario rather than a normal one.

The output is a target expressed as a range with a stated rationale, which a board can revisit as the funding mix changes. That is a materially more useful governance artefact than “we are below the recommended three to six months.”

The policy is the point

A reserve without a written policy tends to be spent. The policy should say four things:

The target and its basis — the range, and why, in terms of the organisation’s own cash cycle and concentration.

Who may authorise a draw, and on what. Ordinarily the board or a designated committee, against defined circumstances. Vagueness here means the reserve erodes through a series of individually reasonable decisions.

The replenishment plan. A draw should come with a schedule for restoration. Reserves that are drawn without a replenishment commitment rarely return to target.

Review cadence. Annually, and on any material change in funding mix.

What funders and lenders actually look at

Lenders underwriting a nonprofit borrower generally look past the reserve headline to months of unrestricted liquidity — unrestricted cash and equivalents divided by average monthly expense — because that measures what is genuinely available. They also read the trend across several years of audited statements rather than a point-in-time figure, and they look at the composition of net assets rather than the total.

An organisation that presents a strong total net asset position built largely of non-liquid restricted funds and fixed assets will find that lenders discount it heavily. Presenting the unrestricted liquid position directly, with the reserve policy attached, is more persuasive than presenting a headline that will be taken apart in diligence anyway.

Institutional funders increasingly accept reserve-building as a legitimate use of general operating support, but the sector remains uneven on this. Organisations should be explicit in proposals that reserve accumulation is a deliberate financial strategy rather than underspending.

Frequently asked questions

Is three to six months of expenses actually wrong?

It is a reasonable starting point and a poor stopping point. For an organisation with diversified revenue, short receivable cycles, and flexible costs it may be more than necessary. For a reimbursement-funded agency with concentrated funding it is likely insufficient. Use it as a prompt to do the analysis, not as the answer.

Can an operating reserve be invested?

Yes, subject to an investment policy consistent with its purpose. Because a reserve exists to be available on short notice, it is ordinarily held in cash, money market instruments, or short-duration fixed income rather than in anything with meaningful price or liquidity risk. Reserves invested for yield have a habit of being least accessible when needed.

How does a reserve interact with an endowment?

They serve different purposes and should be governed separately. An endowment supports the organisation in perpetuity under a spending policy, frequently subject to donor restriction and state law under UPMIFA. An operating reserve is unrestricted and exists to be spent in defined circumstances. Treating endowment as a de facto reserve creates both a governance and, where donor restrictions apply, a legal problem.

What if we simply cannot build one?

Many organisations cannot build a reserve out of programme surpluses, because their funding does not generate any. The realistic paths are a dedicated capital or capacity campaign with reserve accumulation as the stated purpose, general operating support explicitly allocated to it, or a multi-year plan accumulating small annual surpluses. In the interim, a committed line of credit is the practical substitute — arranged while the organisation is healthy, because that is when it can be obtained.

Sources

This is general analysis, not accounting or legal advice. Reserve policy and the treatment of restricted funds should be set with your auditor and counsel.