River
Y CombinatorBacked by Y Combinator
FREE TEMPLATE

Seasonal Cash Flow Plan Template

Four documents and four sheets that turn two or three years of bank history into the peak funding need and the draw plan behind it.

Free download  ·  No account needed

A seasonal business does not have a cash flow problem. It has a calendar, and the calendar is already recorded, because the bank captured every receipt and every disbursement the last two or three times the season came round. The work is turning that record into a number a lender will act on. Each month becomes a share of its own year, so that growth stops reading as seasonality. The shares average into a twelve-month index, and the spread between the years is reported beside it.

The number most businesses quote is the worst month, and it is the wrong number. On the plan worked through this pack, February is the worst month at 504,042 negative. But February arrives at a business that already drained 381,129 in January, and March takes another 403,608 before the first receipts land. The trough of the accumulated deficit is March, at 993,779 below zero, so against a 150,000 minimum the peak funding need is 1,143,779. A line sized on February is short by 639,737 and runs dry mid-season.

The cushion is not guessed either. Rather than adding a round percentage for safety, the pack re-runs the months before the trough on the worst shape the history actually produced, which sizes it at 244,156 or 21.3 percent. The thirteen week cash forecast pack handles the inside of a quarter and the daily cash position pack handles this morning. This one handles the year, and it feeds the funding conversation.

One seasonal year, from the bank feed to the funding ask

Seasonal Pattern by Month, Peak Funding Need, the cushion behind the ask, the draw plan, and the backtest that justifies it.

Seasonal Pattern by Month

Verdant Yard Supply, an illustrative landscape and nursery wholesaler. Three completed years totalled from the bank feed, each month expressed as a share of its own year.

Month2023 share2024 share2025 shareReceipt indexSpreadRead
Jan2.26%2.28%2.25%2.26%0.03ppstable
Feb2.00%2.11%1.99%2.03%0.12ppstable
Mar5.00%6.72%3.58%5.10%3.14ppvolatile
Apr15.63%17.28%13.80%15.57%3.48ppvolatile
May19.92%17.54%21.42%19.63%3.88ppvolatile
Jun17.08%16.52%18.68%17.42%2.16ppvolatile
Jul12.99%12.71%13.21%12.97%0.50ppstable
Aug8.30%8.21%8.29%8.27%0.09ppstable
Sep6.09%6.04%6.08%6.07%0.05ppstable
Oct4.83%4.76%4.80%4.80%0.07ppstable
Nov3.29%3.26%3.29%3.28%0.03ppstable
Dec2.60%2.57%2.61%2.59%0.03ppstable
Total100.00%100.00%100.00%100.00%13.59pp

The index column is an average of three years, none of which was average. 2024 was an early spring and 2025 late and wet, which is why March, April and May carry 10.50 of the 13.59 percentage points of total spread while August through February all sit under 0.15. Disbursements carry 3.89 points in total: the buying calendar is chosen and the selling calendar is not.

Peak Funding Need

FY2026 plan of 10,000,000 in receipts and 8,800,000 in disbursements, spread on the three-year index. Opening cash 295,000, minimum operating balance 150,000.

MonthPlanned receiptsPlanned disbursementsNetCumulativeBalanceBelow minimum
Jan226,230607,359(381,129)(381,129)(86,129)yes
Feb203,014707,056(504,042)(885,171)(590,171)yes
Mar510,047913,655(403,608)(1,288,779)(993,779)yes
Apr1,556,9571,089,409467,548(821,231)(526,231)yes
May1,962,921976,052986,869165,638460,638no
Jun1,742,434824,113918,3211,083,9591,378,959no
Jul1,297,078683,537613,5411,697,5001,992,500no
Aug826,684592,553234,1311,931,6312,226,631no
Sep607,215536,13671,0792,002,7102,297,710no
Oct479,923569,853(89,930)1,912,7802,207,780no
Nov328,140623,981(295,841)1,616,9391,911,939no
Dec259,357676,296(416,939)1,200,0001,495,000no
Total10,000,0008,800,0001,200,000
MeasureMonthAmount
Worst single monthFeb(504,042)
Cumulative troughMar(993,779)
Peak funding need against the 150,000 minimumMar1,143,779
Shortfall if the line is sized on the worst month639,737

Four months of the plan close below the minimum. February looks like the problem and is not: it lands on a business that already drained 381,129 in January, with March taking another 403,608 before the season opens. The line has to carry the accumulation.

Dispersion Cushion

The months up to the trough, re-run on the worst shape the three years actually produced: the lowest observed receipt share and the highest observed disbursement share, at planned volume. Not a worst case, and nothing prevents a fourth year being worse.

MonthPlan receiptsWorst-shape receiptsPlan disbursementsWorst-shape disbursementsWorst-shape balance
Jan226,230224,616607,359620,735(101,119)
Feb203,014198,554707,056741,656(644,221)
Mar510,047357,752913,655951,466(1,237,935)
StepAmountNote
Central need, mean index1,143,779Trough of the planned curve
Worst-shape need1,387,935Trough on the observed extremes
Dispersion cushion244,15621.3% of the central need
Line requested1,400,000Rounded up to the next 50,000, leaving 12,065 of headroom

Adding twenty percent for safety is a guess dressed as prudence. This cushion has a derivation and every input traces to a bank statement, which is what makes it survive the conversation where the facility fee gets quoted.

Line of Credit Draw Plan

Draws in 25,000 increments against the 1,400,000 line, each sized to clear the minimum including that month's interest. Interest at an illustrative 9.25% all-in, which is a sizing input and not a quote.

MonthNetDrawRepayOutstandingInterestClosing cashUsed
Jan(381,129)250,0000250,0001,927161,94417.9%
Feb(504,042)500,0000750,0005,781152,12153.6%
Mar(403,608)425,00001,175,0009,057164,45683.9%
Apr467,5480450,000725,0005,589176,41551.8%
May986,8690725,00000438,2840.0%
Jun918,32100001,356,6050.0%
Jul613,54100001,970,1460.0%
Aug234,13100002,204,2770.0%
Sep71,07900002,275,3560.0%
Oct(89,930)00002,185,4260.0%
Nov(295,841)00001,889,5850.0%
Dec(416,939)00001,472,6460.0%
Total1,200,0001,175,0001,175,000022,3541,472,646

Drawn in four months, peaking at 1,175,000 or 83.9% of the line, with 225,000 of headroom at the peak. Both repayments are funded out of season receipts rather than a further draw, and the line closes the year at zero, which is the self-liquidation a credit officer is looking for. Total interest of 22,354 is 0.22% of planned receipts.

Prior Year Actual vs Plan

The index rebuilt on 2023 and 2024 only, spread across the FY2025 plan of 9,200,000 and 8,100,000 on opening cash of 255,000, then compared with what FY2025 actually did.

MonthPlanned receiptsActual receiptsVarianceVariance %Planned netActual net
Jan208,874214,6005,7262.7%(356,193)(351,800)
Feb188,825189,7008750.5%(472,943)(461,500)
Mar539,299341,800(197,499)-36.6%(315,149)(501,100)
Apr1,513,8321,318,400(195,432)-12.9%523,342254,700
May1,723,3132,046,900323,58718.8%843,3151,078,500
Jun1,545,3301,784,600239,27015.5%792,009988,300
Jul1,182,2101,262,30080,0906.8%553,204610,500
Aug759,308792,40033,0924.4%212,993229,500
Sep558,080581,30023,2204.2%63,29473,000
Oct441,395458,80017,4053.9%(83,864)(82,800)
Nov301,459314,40012,9414.3%(274,003)(278,000)
Dec238,075248,90010,8254.5%(386,005)(392,400)
Total9,200,0009,554,100354,1003.8%1,100,0001,166,900
The call that would have been madeAmountAgainst the 1,209,400 the year reached
Central call, mean index alone1,039,285short by 170,115, or 16.4%
Same index plus its dispersion cushion1,200,000short by 9,400, or 0.8%

Annual volume was close: receipts came in 3.8% over plan. The shape was not. March landed 36.6% under and May 18.8% over, so the mean absolute monthly receipt error was 94,997, or 12.4% of an average month, concentrated in the months already flagged volatile. Both troughs still landed in March. One backtested year is one observation, not a validated error rate.

What's in the pack

01

Seasonal Pattern by Month sheet

Three observed years, each month's share of its own year, the averaged index, and the spread in percentage points. Both index columns sum to 100 percent, which is the check that catches a missing month or a mis-totalled year.

02

Peak Funding Need sheet

The whole cumulative curve rather than a summary figure, so the trough is checkable, plus the worst-shape receipts and disbursements for each pre-trough month and the cushion they produce. Every month below the minimum is flagged.

03

Line of Credit Draw Plan

Draws in the increment the lender requires, sized to clear the minimum including that month's interest. Repayments come from season receipts rather than a redraw, and the schedule closes the year at zero outstanding.

04

Prior Year Actual vs Plan

The withheld-index backtest, splitting volume error from shape error so the index is only held responsible for the part it controls. Mean absolute monthly receipt error of 94,997, concentrated in the months already flagged volatile.

05

How the Index Is Derived

The four steps, with the partial-year exclusion and the transfer and loan-draw netting that quietly ruin an index built by hand. Folding last year's borrowing in makes this year's need look smaller than it is. How fast receipts land is partly a collections outcome, worked in the AR collections pack, and when disbursements leave is set by the AP process and payment run pack.

06

Seasonality Notes

What physically drives the shape, which months the dispersion says can be trusted, what changed since the history was written, and four observable triggers that force a rebuild. One trigger is a single customer passing fifteen percent of season receipts, which belongs in the customer credit policy pack instead.

07

Peak Funding Need

Why the trough and not the worst month, how the cushion is derived from observed dispersion, and the backtest that justifies it. Written so a challenged figure still has its arithmetic attached in month seven.

08

The Lender Conversation

The five artifacts to bring and the order to bring them in, plus the four questions that arrive every time with the figure that answers each. Ends by naming what it does not do, which is recommend a facility, a rate or a lender.

How it works

  1. 1

    Send two or three years of bank history

    Statements, CSV exports or a connected feed. River totals each completed year and shows the totals so you can check them against your own records. It names any partial year it excluded, and reports what it netted out as internal transfers or prior-year loan activity.

  2. 2

    The index and its spread get derived

    Each month becomes a share of its own year, the shares are averaged, and both index columns are confirmed to sum to 100 percent. The spread per month is computed alongside, and the gap between receipt and disbursement dispersion is reported as the finding rather than a footnote.

  3. 3

    The trough and the cushion get sized

    Your planned annual totals are spread on the index, the cumulative curve is built from opening cash, and the need is measured from the trough against your minimum balance. The pre-trough months are then re-run on the worst observed shape to size the cushion.

  4. 4

    The draw plan and the backtest close it out

    A month-by-month schedule that clears the minimum including interest, repays from season receipts, and closes at zero. Given three years, the index is rebuilt on the earlier two and tested against the year that has closed, so the cushion arrives with evidence attached.

Frequently asked questions

Why is the peak funding need so much larger than my worst month?

Because deficits accumulate. On the worked plan February is 504,042 negative, but it lands on a business that already spent 381,129 in January, and March takes 403,608 more before receipts arrive. The trough is 993,779 below zero, so the need is 1,143,779 rather than 504,042. The difference, 639,737, is the accumulation.

How many years of bank history do I actually need?

Two completed years produce an index. Three let you measure dispersion properly and run the backtest, which is what makes the cushion arguable. Partial years are excluded entirely: a year with eleven months has an inflated share in every month it does have, and averaging it in contaminates the whole index.

Is a three-year index enough to trust?

For the stable months, yes. On the worked history August through February all sit under 0.15 percentage points of spread. March, April and May carry 10.5 of the 13.6 points of total spread, and the pack marks them volatile rather than pretending the average is the truth. That is exactly what the cushion exists for.

Why compute the cushion from dispersion instead of adding a percentage?

Because a round percentage cannot be defended and the dispersion can. Re-running the pre-trough months on the worst shape the history produced gives 1,387,935 against a central 1,143,779, so the cushion is 21.3 percent for a stated reason. It is not a worst case, and nothing prevents a future year being worse.

Does the pack tell me which facility to get or what rate to expect?

No. It sizes a need, a cushion and a schedule, and the 9.25 percent used to compute interest is an illustrative sizing input rather than a quote. Which instrument fits, how it should be priced and whether to borrow at all are decisions for you and your advisers. The pack produces the analysis they work from.

What if my business changed since the history was written?

That goes in the Seasonality Notes as a named, dated override applied to the plan, with its own row, never by editing the index. A new location, a terms change or a lost anchor account invalidates the shape while leaving the arithmetic intact, and the index stays as the clean record.

What does a lender want to see, and when should I bring it?

Before the season, which is the whole point. Two things get assessed: whether the need is genuinely seasonal, and whether the line self-liquidates. Federal Reserve Regulation A frames seasonal credit around expected patterns of movement in funding needs, and the SBA CAPLines rule covers cyclical, recurring short-term operating capital needs. Both are pattern tests.

Arrange the borrowing before the season

Send the bank history and get the index, the trough, the cushion and the draw plan back with the arithmetic shown.

Edit with AI