A surplus on the statement of activities and an empty operating account are not contradictory. They are the ordinary result of accrual accounting doing its job: a pledge is recognized when the donor commits, a grant is recognized when the award is made, and the cash lands somewhere between forty-five days and never.
Most finance teams know this. What tends to surprise them is how far the gap can open before anything looks wrong in the reports. Nonprofit Finance Fund’s 2025 State of the Nonprofit Sector Survey, which drew responses from more than 2,200 US nonprofits, found that 52% hold three months or less of cash and 18% hold one month or less. Among the 163 organizations that answered both the 2022 and 2025 surveys, the share with six or more months of cash fell from 36% to 26%.
Cash flow management is what closes that gap: knowing when money actually arrives and leaves, and being able to act on the answer before a board meeting forces the question. The seven non-profit cash flow management strategies below are ordered by dependency rather than importance. The forecast comes first because the rest of them read from it.
Your budget answers what the organization intends to spend and raise over the year. Cash flow management answers whether the money will be in the account on the day it is required. The two diverge whenever revenue is recognized on a different date from when it is received, which for most nonprofits is nearly always.
The statement of cash flows sits somewhere between them, and it is worth being precise about its role. It is a historical document. It reconciles what happened, and auditors and boards rely on it for exactly that. It will not tell you whether the March payroll clears.
|
Budget |
Statement of cash flows |
Cash flow forecast |
|
|---|---|---|---|
|
Direction |
Forward |
Backward |
Forward |
|
Question it answers |
What do we plan to raise and spend? |
Where did cash come from and go? |
Will there be enough cash on a given date? |
|
Typical horizon |
12 months, fixed |
Period just closed |
13 weeks weekly, then monthly |
|
Update cycle |
Annual, with reforecasts |
Monthly or at audit |
Weekly |
|
Fails when |
Timing differs from plan |
Used to predict anything |
Restricted and unrestricted cash are pooled |
All three connect at the annual budgeting process. A budget built without reference to cash timing produces a spending plan the organization cannot actually execute in sequence, which then shows up as a forecasting problem three months later.
Corporate cash flow problems are usually demand problems or collection problems. Nonprofit cash flow problems are more often structural, built into how the money is designated and when funders release it. Four constraints do most of the damage.
|
Constraint |
What it does to cash |
|---|---|
|
Donor restrictions |
A meaningful share of the bank balance cannot legally cover payroll or rent. The balance looks healthy; the available portion may not be. |
|
Reimbursement-based funding |
You front the cost of the program and recover it later. Every month of program delivery is a month of working capital you have loaned to the funder. |
|
Contribution seasonality |
Giving concentrates heavily in the last weeks of the calendar year while expenses run evenly across twelve months. |
|
Grant calendars |
Award periods rarely align to your fiscal year, so a single year contains several overlapping start dates, close-out deadlines, and reporting cycles. |
Reimbursement has been the most volatile of these four constraints. When federal payments pause, organizations working on a reimbursement model continue delivering services and paying staff with no mechanism to recover the outlay until the funder resumes, a dynamic CLA documented in detail during the 2025 shutdown. NFF’s survey found 84% of respondents receiving government funding expected cuts to it.
None of these constraints are new, and none are solvable through better collections. They are conditions to model, which is why the measurement question comes before the strategy question.
Two organizations can report the same months of cash and be in completely different positions. One is waiting on a receivable that will land in three weeks. The other has been running a deficit for two years and is watching the last of its unrestricted balance drain. Reading a few ratios together separates the two cases.
|
Metric |
Calculation |
Common target |
What it hides on its own |
|---|---|---|---|
|
Days cash on hand |
Unrestricted cash and liquid investments ÷ average daily expenses |
90 days minimum |
Says nothing about why cash is low or whether it is recovering |
|
Operating reserve ratio |
Board-designated unrestricted reserves ÷ annual expenses |
25% or more, roughly three months |
Reserves can be designated on paper but already committed in practice |
|
Current ratio |
Current assets ÷ current liabilities |
1.5 or higher |
Restricted receivables inflate the numerator without improving liquidity |
|
Months of unrestricted cash |
Unrestricted cash ÷ average monthly expenses |
3 to 6 months |
A single large unrestricted gift can mask a structural deficit for a quarter |
|
Deficit or surplus trend |
Change in unrestricted net assets, tracked across 3 to 5 years |
Modest recurring surplus |
A single year tells you almost nothing |
Most reporting packages leave out the trend line. A three-to-five-year view of unrestricted net assets shows whether reserves are being built or quietly consumed, and it is the context that makes a single quarter interpretable. Building it into standing nonprofit financial dashboards saves rebuilding the comparison every time a board member asks.
Watch out: a large restricted balance sitting in the operating account is the most common way a liquidity problem stays invisible. Cash-on-hand ratios calculated on total cash rather than unrestricted cash will read as comfortable right up until the point the organization cannot make payroll.
Every ratio above depends on being able to state the unrestricted balance on a given date. Most finance teams can produce that figure for the last close. Far fewer can project it eight weeks out, which is where the useful version of the work starts.
Thirteen weeks is the standard horizon because it is long enough to see a payroll cycle, a grant close-out, and a quarterly payment run, and short enough that weekly estimates stay honest. Beyond week thirteen, monthly granularity is sufficient out to twelve months.
The structural decision is running two tracks rather than one. Restricted and unrestricted cash flow through the same bank account, and a forecast that adds them together produces a single ending balance that no one can act on. Splitting them lets you answer the question the board actually asks, which is how long the organization can operate on money it is free to spend.
|
RUN THE NUMBERS. An organization holds $2.1 million in the operating account and spends $600,000 a month.
Same account, same date, and a materially different conversation with the finance committee. |
Each track carries its own inflow and outflow lines.
A cash flow forecast template is a reasonable starting point for the structure. The limitation appears once you are maintaining the split across several funds and several awards at once, because the reconciliation back to the general ledger becomes the job rather than the forecast.
Building the model, choosing between a direct and an indirect method, and setting the review cycle are covered in more depth in this guide to cash flow forecasting. Classification carries most of the nonprofit-specific work in that model. The rest sits in a single assumption that forecasts routinely get wrong by treating every funder the same.
Forecasts typically apply one collection assumption across all receivables. For an organization drawing from a single revenue type, that is defensible. For one running federal pass-through funding alongside state contracts and private foundation grants, it produces a forecast that is wrong in a predictable direction.
Holding a days-to-cash assumption per funder, and letting the forecast calculate from it, removes most of that error. Build the assumption from your own remittance history rather than from the funder’s stated terms, which describe intent rather than behavior.
|
Funding source |
Typical behavior |
Forecast treatment |
|---|---|---|
|
Direct federal grant |
Drawdown against approved expenditure, subject to appropriation timing |
Per-award drawdown schedule with an explicit approval lag |
|
State pass-through |
Slower, as the state adjusts to federal disbursement |
Longer lag than the direct equivalent, reviewed each quarter |
|
Private foundation |
Scheduled tranches against an award letter |
Date-certain inflow with a modest buffer |
|
Fee-for-service |
Invoiced and collected on commercial terms |
Standard receivable aging |
|
Individual giving |
Immediate, concentrated in December |
Seasonal curve from three years of history |
Once the lag is a named driver rather than a buried assumption, it becomes something you can test. Changing one funder’s assumption from 45 to 90 days and watching the effect ripple through the thirteen weeks is a far more useful exercise than adjusting a single blended figure. The same driver-based approach applies to any input where the organization has a history worth modeling.
|
WATCH OUT. Build each lag from your own paid-date history, not from the terms in the award letter. Stated terms describe what the funder intends. Remittance history describes what the funder does, and the gap between the two is the part your forecast has to absorb. |
Connecticut Green Bank runs its planning against operational drivers including solar production, and consolidated roughly thirty separate reports in the process. Jane Murphy, VP of Finance and Administration, has described the shift as moving the finance team’s attention from assembling numbers to interpreting them. The full account of that implementation is worth reading if your drivers are similarly specific to your programs. Sharper lag assumptions narrow the range of outcomes without closing it. Something still has to absorb the funder who pays ninety days late, and for most organizations that is the reserve.
Reserve targets are well established. The Nonprofit Operating Reserves Initiative Workgroup’s benchmark of 25% of the annual operating expense budget, equivalent to about three months, remains the common floor, with many organizations working toward six. The National Council of Nonprofits notes that a minority of organizations actually hold more than six months.
Targets are the easy part. Policies fail in practice because they specify an amount and nothing else, which leaves the board without a basis for deciding whether a given month qualifies as the rainy day. A policy that holds under pressure covers four things.
The Greater Washington Society of CPAs makes the point that the appropriate balance varies with the organization’s revenue concentration and volatility. An organization with one funder supplying 60% of revenue is carrying a different risk from one with forty funders at similar scale, and the policy should say so explicitly rather than adopting a sector average. Concentration sets the reserve target, and it is also what makes the target worth testing. A reserve sized against average conditions says little about how it holds up when one specific funder disappears.
Scenario planning loses its value when the scenarios are vague. Modeling a "downturn" produces a number nobody trusts. Modeling the loss of your second-largest funder produces a number the board can act on, because it maps to a decision someone will have to make.
For a mid-market nonprofit, three scenarios cover most of the realistic exposure. Running all three off the same set of outputs keeps them comparable.
|
Scenario |
What it tests |
What it should return |
|---|---|---|
|
Largest funder delays 90 days |
Working capital, not solvency |
Lowest projected unrestricted balance, the week it occurs, and how much of the gap a line of credit would have to cover |
|
Largest funder lost at renewal |
Structure of the expense base |
Which programs cover their own costs, and the run rate twelve months out |
|
Demand rises while funding holds flat |
Gradual erosion rather than a single event |
The quarter the unrestricted balance crosses the reserve floor |
Of the three, the last is the most common in practice and the least often modeled, because nothing happens on a specific date to prompt it. Reporting more than those figures per scenario tends to obscure the comparison rather than sharpen it.
Running these as standing scenarios rather than one-off analyses is what makes them useful, since the underlying assumptions change as funding does. The broader mechanics of scenario planning apply, with the caveat that the variable worth flexing is usually a specific funder relationship rather than a market condition. Run the funder-loss case a few times and it tends to return the same verdict: too much of the revenue base rests on one relationship. Diversification is the standard answer, and its timing is where the plan usually goes wrong.
Diversification advice usually stops at reducing dependence on any single funder, which is correct and incomplete. Revenue sources differ in how long they take to become spendable cash, and a diversification plan that ignores the timing can worsen the near-term position even as it improves the risk profile.
|
Revenue source |
Time to cash |
Effect on near-term liquidity |
|---|---|---|
|
Individual giving |
Days |
Improves immediately, but concentrated in the fourth quarter |
|
Fee-for-service |
Weeks |
Improves steadily once the program reaches scale |
|
Private foundation grant |
Months |
Neutral in the near term; cash arrives on the award schedule |
|
Government reimbursement |
Months, after outlay |
Consumes working capital before it returns any |
|
Endowment or planned giving |
Years |
No near-term effect; supports long-horizon stability |
An organization adding a government contract to reduce reliance on individual giving is taking on more working capital pressure in year one, not less. That can still be the right decision, but it is a decision to fund deliberately rather than discover in month four. Pairing a slow-converting new source with a faster one, or sizing the reserve to absorb the ramp, is the practical response. Where the expense base sits during that ramp matters too, since nonprofit operating expenses rarely flex as quickly as revenue does. Sequencing decisions of that kind are easier to defend when you can state how much cash the organization can actually spend on a given date. That figure already has a formal definition, and most organizations calculate it once a year without ever using it.
Under FASB ASU 2016-14, nonprofits already disclose both quantitative and qualitative information about the financial assets available to meet general expenditures within one year of the balance sheet date, along with how the organization manages its liquid resources. The AICPA guidance on the disclosure spells out what qualifies as available and what does not.
Read closely, that requirement describes the same calculation as a well-built forecast. Available financial assets are unrestricted, liquid, and unencumbered by board designation. A forecast that already separates restricted from unrestricted, and already flags board-designated reserves, contains the number. Most organizations calculate it twice: once during the year for management, and again at year-end for the auditors, from a different source, with a different result.
Reconciling the two saves duplicated work, but the larger gain is in the qualitative half of the disclosure. It requires a stated liquidity management policy, and the most credible version of that policy describes a process the organization actually runs. A footnote explaining that liquidity is reviewed weekly against a rolling forecast, with a board-approved reserve floor and defined drawdown triggers, is materially stronger than one asserting that management monitors cash carefully.
Where the two do diverge is worth investigating rather than reconciling away. A forecast that consistently reports more available cash than the year-end disclosure supports usually means something is being treated as unrestricted in the model that the auditors classify otherwise. The underlying accounting treatment governs, and the forecast should be corrected to match it. Both the forecast and the disclosure describe money the organization has already committed or is still free to commit. Neither catches a commitment before it is made, and those decisions get made outside finance.
Program directors commit cash constantly through hiring decisions, vendor contracts, and purchase timing. When their only reference point is an annual budget line, those commitments get made against an allocation rather than against an actual cash position, and finance discovers the effect after the fact.
A program director with $180,000 remaining in a budget line reasonably concludes there is room to hire. Whether the organization can carry that salary through a ninety-day reimbursement gap is a different question, and one they have no way to answer from the information in front of them.
Three pieces of information close most of that gap.
Communication Service for the Deaf approached the participation problem this way and cut its budget cycle in half, with Ben Daniel, Director of FP&A, pointing to distributed input as the change that made the difference. Whether the same structure suits your organization depends largely on how many budget holders you have and how much of the expense base they control. The CSD implementation is a reasonable reference point for organizations of comparable size.
Board reporting sits above this and answers a different question, which is whether the organization is on track overall rather than whether a specific commitment is affordable this month. Both are needed, and conflating them tends to produce board reports that are too detailed for governance and too slow for operations. Each of the seven strategies assumes someone is looking at the forecast on a fixed schedule. Without that, the model ages quietly and the first sign of trouble arrives with the bank statement.
Strategies that depend on a review cycle fail quietly when the cycle is not assigned. Naming the owner and the decision each review drives is what keeps the forecast from becoming a document nobody opens between board meetings.
The weekly review carries the most operational weight. It decides whether to accelerate a drawdown, delay a payment run, or draw on a line of credit, and those options narrow as the gap gets closer. The quarterly review is where concentration gets re-checked, since a funder can grow into a dependency without any single decision creating one.
Weekly review is the cadence that lapses first and matters most, because it is the only cadence fast enough to catch a timing problem while options remain open. Organizations that have moved to rolling forecasts generally find the weekly update takes minutes once the model reads directly from the accounting system, and takes a full day when it does not, which decides on its own whether the cadence survives.
Failures here are rarely analytical. They come from a small number of habits that persist because nothing forces a correction until the position is already tight.
|
Habit |
How it shows up |
The correction |
|---|---|---|
|
Counting restricted cash toward runway |
A months-of-cash figure that is confidently wrong and that the board has no basis to challenge |
Report unrestricted cash separately in every liquidity metric |
|
Forecasting from the budget |
The forecast inherits planned amounts but none of the timing information that makes it useful |
Drive the model from receivables aging and award drawdown schedules |
|
Updating quarterly |
A ninety-day refresh cycle cannot catch a sixty-day payment delay in time to act on it |
Weekly for the 13-week horizon, monthly beyond it |
|
Holding a reserve nobody will authorize |
Defended past the point of usefulness, then drawn in a crisis with no replenishment plan |
Write drawdown triggers and an authorization path into the policy |
|
Treating liquidity as an audit task |
The available-assets figure gets calculated once a year and never used for management |
Produce it from the forecast and reconcile at year-end |
Each of these is a symptom of the forecast sitting apart from the rest of the planning process rather than inside it. Where cash flow work connects to the wider nonprofit financial planning cycle, the corrections tend to happen on their own, because the assumptions get revisited whenever anything else does.
None of the seven strategies require software. They require a model that stays reconciled to the general ledger, and that is where spreadsheets tend to break down once an organization is running multiple funds across several awards.
Limelight imports fund dimensions directly from Sage Intacct, NetSuite, Microsoft Dynamics, and Blackbaud, and preserves them through planning and reporting. The fund accounting itself stays in the accounting system, where it belongs. What changes is that the restricted and unrestricted split in your forecast comes from the same dimensions your accounting system already maintains, rather than from a mapping table someone updates by hand.
Operating expense planning in Limelight
Three of the seven lean on that connection more than the rest.
Organizations running on Blackbaud alongside a separate ERP tend to hit the reconciliation problem earliest, since the fund structure exists in more than one place. Limelight’s Blackbaud integration addresses that specific case. For organizations with a single accounting system and a handful of awards, a well-maintained spreadsheet may hold for some time, and there is no particular urgency in replacing it.
If you implement only one of these strategies, separate restricted from unrestricted cash in whatever forecast you already maintain. It requires no new tooling and it usually changes the runway figure enough to reset the conversation with your board.
The reserve policy follows naturally, because once the unrestricted position is visible weekly, the question of what triggers a draw stops being hypothetical. Scenarios, diversification sequencing, and the liquidity disclosure all read from those same two foundations.
What changes once the system is running is the timing of the conversation. Cash problems announce themselves eight weeks out instead of arriving as a surprise, which is generally the difference between choosing among options and taking whichever one is left.
|
SEE YOUR CASH POSITION IN REAL TIME. Limelight connects to your accounting system and preserves fund dimensions through forecasting, scenario planning, and reporting. Book a demo to see how it handles restricted and unrestricted planning for organizations like yours. |
It is the practice of tracking when cash actually enters and leaves the organization, then acting on that timing. It differs from budgeting, which plans annual revenue and expense without addressing the dates money moves.
Three months of unrestricted operating expenses is the common floor, with three to six months preferred. Organizations dependent on reimbursement-based or highly concentrated funding generally target the upper end of that range.
Accrual accounting recognizes pledges and grant awards when committed, not when received. A surplus on the statement of activities can therefore coexist with an operating account that cannot cover payroll.
No. Donor-restricted funds cannot cover general operating costs, so including them overstates runway. Reserve and days-cash calculations should use unrestricted, liquid, unencumbered assets only.
Weekly for the rolling 13-week horizon, monthly for the longer view. Quarterly updates cannot catch a 60-day payment delay early enough to leave useful options open.
Any tool that reconciles to your accounting system and preserves fund dimensions. This comparison of cash flow forecasting tools covers the main options and where each fits.