River
Y CombinatorBacked by Y Combinator
FREE TEMPLATE

Cash Runway Scenario Analysis Template

Three documents and six sheets, where every trigger is the last month a decision can still be authorized in time to work.

Free download  ·  No account needed

Trigger Dashboard

[Entity] — at 30 September 2026

16.5 months of runway. The last month the round can be authorized is February.

Cash11,850,000Trailing 6-month net outflow718,845
Covenant minimum, at all times1,000,000Quoted runway16.5 mo
Measured intra-month trough660,000Usable runway14.2 mo
Cash floor1,660,000Usable cash10,190,000

Three cases, each ending in a date

CaseSystems revenueCash reaches the floor
Upside, best observed quarter1.23 of trendnot inside 24 months
Base, trend through 24 actualstrendJanuary 2028
Downside, worst observed quarter0.57 of trendOctober 2027

Triggers, armed, in date order

 DecisionAuthorize byDaysBuys
T-1Raise the Series B28 Feb 2027151all of it
T-2Reduce nine roles31 Jul 20273043 months
T-4Stop the tooling programme31 Aug 20273353 months
T-5Open a receivables facility31 Jul 20273040 months
T-3Draw the remaining facility30 Sep 20273651 month

Each date is solved backwards from the October 2027 floor through that decision's own lead time and its own up-front cash cost. The convention everyone uses, act when runway falls under six months, does not fire until May 2027. That is three months after T-1 stopped being executable, because a runway number does not know how long a fix takes.

Page one for this query is 24-month models. Huggle's template builds assumptions, a monthly model, scenarios and a dashboard with a runway warning cell. Founderpath's seven tabs project burn, cash, runway and a break-even month across conservative, base and upside. Both are competent, and both answer the same question: how many months are left. Not one page on the query says how long a fix takes, so not one of them can name the month a decision has to be taken.

Here the answer is a date. Marchetti Robotics, an illustrative warehouse robotics company, holds 11,850,000 and burns 718,845 a month, which quotes as 16.5 months of runway. Its floor is 1,660,000, being a covenant minimum tested at all times plus the measured gap between a month-end balance and that month's intra-month low. Its downside is the worst rolling three months in 24 actuals, 42.8 percent under its own trend. The floor breaks in October 2027, and the Series B takes eight months.

So the trigger is 28 February 2027, solved backwards through that lead time and the cash the raise consumes before it closes. The six-month runway convention does not fire until May, three months after the round stopped being executable. Auditing standards test a plan by whether it is likely to be effectively implemented, and Regulation S-K asks anyone naming a liquidity deficiency to state the course of action beside it. The weekly version of this job is a thirteen week cash forecast.

A month count, and then the four things it hides

Five decisions with solved dates, the raise against the convention, the same reduction in four months, and the downside measured rather than assumed.

Why 16.5 months is not an answer

Marchetti Robotics, Inc. Forty-six people, warehouse robotics, 24 months of ledger actuals to 30 September 2026. An illustrative company.

Cash at 30 September 202611,850,000
Trailing six-month average net operating outflow718,845
Cash divided by that outflow16.5 months

What the quotient contains

 AmountWhere it comes from
Covenant minimum1,000,000Unrestricted cash at all times, equipment facility
Intra-month trough allowance660,000Largest measured month-end to intra-month-low gap, 12 months
Cash floor1,660,000The balance the company actually stops at
Usable cash10,190,0002.3 months of the 16.5 cannot be spent

A month-end balance sitting exactly on a covenant minimum is a breach on the fifteenth, because payroll cleared and the month's largest receipt had not landed. That is what the second row is for, and it is measured from statements rather than estimated.

Three cases, and the dates they produce

CaseBasisFactorFloor date
UpsideBest rolling three months in 24 actuals, Oct to Dec 20251.2333outside 24 months
BaseLog-linear trend through 24 actuals, 2.15% a month1.0000January 2028
DownsideWorst rolling three months in 24 actuals, Jan to Mar 20260.5716October 2027

The quoted 16.5 months lands in mid February 2028, one month past the base case. The floor and the plan's own cost ramp roughly cancel there, which is exactly why the quotient looks harmless.

The triggers are solved against October 2027, not against the base case and not against the quotient. A case nobody would be embarrassed to have used is the one that has already happened once.

Decision Register

Every decision that could move the floor date, with the four properties that set its authorize-by month. Floor date with no action: October 2027.

 DecisionUp-front cashLeadMonthly effectOne-off in
D-1Raise the Series B185,0008 mo12,000,000
D-2Reduce nine roles412,0001 mo214,000
D-3Draw the remaining facility20,0001 mo(52,178)2,000,000
D-4Stop the tooling programme96,0001 mo165,000
D-5Open a receivables facility35,0002 mo(12,160)640,000

Solved dates

 Authorize byDays leftFloor date if authorized nowBuys
D-128 Feb 2027151outside the horizonall of it
D-231 Jul 2027304January 20283 months
D-431 Aug 2027335January 20283 months
D-330 Sep 2027365November 20271 month
D-531 Jul 2027304October 20270 months

Two million dollars buys one month. By the October 2027 floor date, D-3 has produced 2,000,000 and paid back 626,136 of amortisation plus a 20,000 fee. Drawing does not lower the floor either: the covenant tests unrestricted cash at all times, so the 1,000,000 minimum stands over a larger balance sheet. D-5 adds 459,080 of net cash and no months at all, because the month it would have to bridge needs more than a million.

Both of those get counted as runway in conversation. The register is where a source gets priced instead of asserted.

The two cheap ones, together

D-2 and D-4 authorizedFloor dateAdded
October 2026September 202811 months
February 2027April 20286 months
May 2027, when the convention firesJanuary 20283 months
September 2027September 2027a month earlier

D-1, and the eight months nobody budgets

Raise the Series B · 12,000,000 · authorize by 28 February 2027

The lead time is measured on this company's own Series A, not taken from an article.

StageWeeksCumulative
Model and data room into shape66
First partner meeting to signed term sheet1925
Term sheet to wire934
Authorize to cash348 months

Two ways to decide when to start

 FiresWire arrivesFloor breaksResult
Runway falls under six monthsMay 2027January 2028October 2027Closes three months too late
Authorize-by, solved backwardsFebruary 2027October 2027October 2027Closes in time

On the downside path the runway number first prints under six at 5.99 months in May 2027. It is arithmetically correct and it is three months too late, because a quotient of cash over burn has no term for how long a fix takes.

The condition that can pull the date earlier

Signed systems orders scheduled to ship inside 90 days2,284,000
Base plan for the same quarter2,161,000
Coverage1.06
Threshold that trips T-1 earlybelow 1,800,000

A trigger is a date and a condition, whichever arrives first. Below 1,800,000 the downside case is arriving rather than being modelled, and February stops being the answer.

D-2, and what waiting costs

Nine roles · 412,000 up front · 214,000 a month from the following month

Same nine roles. Same monthly saving. Four different months.

AuthorizedHeadroom in the month of actionFloor dateMonths bought
October 20269,029,616January 20283
February 20275,583,616December 20272
May 2027, when the convention fires3,036,096November 20271
August 2027549,489September 2027a month earlier

The last row is the whole argument. Authorized in August, the 412,000 of severance, accrued time off and employer payroll taxes leaves before any saving arrives, and the floor breaks in September rather than October. The decision does not lose value gradually. It turns negative.

Why its date is 31 July 2027

Last month the reduction still holds the floor through its own landing monthJuly 2027
Week 25 of D-1's schedule, when a term sheet should existJuly 2027

The two land in the same month by construction, which makes the condition on T-2 obvious: no signed term sheet by 31 July. That is the month the answer on the raise is known and the fallback is still worth taking.

Holding a decision in reserve is free only while its authorize-by date is still in front of you. The register turns that into a number per decision instead of a feeling, which is the point of writing it down before anybody needs it.

How the downside was measured

Not a percentage. Every rolling three-month window in 24 actuals, measured against the company's own fitted trend.

WindowActualTrendRatio
Jan to Mar 2026, the worst of 22821,0001,436,4020.5716
Oct to Dec 2025, the best of 221,647,0001,335,4041.2333

History is not symmetric

Measured downside42.8% below trend
Measured upside23.3% above trend
A plus or minus 20 percent case, on the downside2.14x too mild
The same case, on the upside1.17x too generous

Two errors pointing in opposite directions inside one model. The worst window was a supply slip on two subassemblies, and it is inside the trailing twelve months, so it is not an old event the company has grown out of.

Two lines a haircut gets wrong

Recurring revenue keeps its balance. It has not fallen in any of the 24 months, because it is contracted. What a downside removes is the growth. Measured: recurring revenue rose 151,000 over the trailing twelve months against 5,716,000 of systems shipped in the twelve months ending two months earlier, so 26,417 of monthly recurring per 1,000,000 shipped. The downside grows at 57 percent of the base rate rather than losing its base.

Committed purchase orders are already spent. Materials are bought two months ahead on 30-day terms, so November and December shipments are paid for whatever happens.

Month of cashCommittedWhat a flexed model shows
October 2026423,354241,989
November 2026432,456247,192
Cash a flexed downside hands back366,628

A model that flexes cost of revenue with revenue makes a shipment slip look cheaper than it is, which is the one direction a downside case must never be wrong in.

What's in the pack

01

Trigger Dashboard sheet

One row per trigger. The decision it authorizes, the owner, the authorize-by date, the days remaining, the leading-indicator condition with its current reading, and the status. A trigger is the date or the condition, whichever arrives first.

02

Decision Register sheet

Every decision that could move the floor date, with the four measured properties that set its date. Lead to first cash and where that lead was measured, up-front cash and the month it leaves, the monthly effect once landed, and any one-off inflow.

03

Space rule

A trigger is a date, and the date is solved backwards. It fixes how the floor is built, that burn and both scenario factors are measured rather than chosen, and that an uncommitted source of cash is not runway. It governs every prompt here.

04

Base, Downside and Upside models

Three sheets, 24 forward months each, differing in exactly one input. Full cash statement per month with the floor and headroom on every row, ending in the month closing cash falls below the floor. That date is the output, not a month count.

05

Burn History sheet

Twenty-four actual months by line, with the rolling three-month ratio against trend on every row so both scenario factors are auditable rather than asserted. It carries the quoted runway and the usable runway side by side, and the gap between them.

06

Scenario Assumptions

Where every figure came from, one line each. The floor as two numbers that add up, the measured attach relationship, the committed purchase orders, the cost basis. Ends with what is deliberately in no case, so nobody assumes an unsigned pipeline was modelled.

07

Trigger and Decision Note

One record per decision, with the arithmetic behind its authorize-by date and the cost of postponing it. Includes the comparison against the six-month runway convention, and the month a late decision turns net negative. Refreshed after every close.

08

Runway Slide

One slide with speaker notes, generated from the four sheets rather than retyped. The same slide goes to the board and to the investor, because two versions of the runway is how a company loses the room. Pairs with the board package.

09

Four prompts, in dependency order

Set the floor, build the three cases, solve the authorize-by dates, refresh after the close. The last one re-solves every date against the new month and reports which ones moved, because a date that never moves is not being maintained.

How to use it

  1. 1

    Download the files, or open the pack

    Three Word documents and six CSV sheets, with no account. Or open the pack in River and hand it eighteen months of ledger plus every credit agreement. The ledger sets the burn and both scenario factors; the agreements set the floor and usually name at least one decision. Statements help too, since the intra-month trough has to be measured rather than guessed.

  2. 2

    Set the floor before quoting anything

    The covenant minimum and the measured trough, as two numbers that add up. It changes the answer before any scenario does, and it is the fastest part to establish. Where there is no agreement it is one payroll cycle plus undeferrable payables.

  3. 3

    List the decisions, then solve the dates

    Include the weak options. A decision worth one month belongs in the register precisely because somebody will otherwise call it three. Get a raise's lead time from your own last round, and severance from real terms rather than a rule of thumb.

  4. 4

    Re-solve after every close

    A bad month that becomes your new worst window moves every date earlier at once. Report which dates moved and by how many days, and retire anything past its date instead of quietly extending it.

Frequently asked questions

Is this template free?

Yes. Three Word documents and six CSV sheets, no account, no card, no trial. Edit with AI exists for anyone who would rather hand over a ledger and a credit agreement than fit a trend line and solve five dates by hand. The rest of the library is in free templates.

Why not just track months of runway?

Because a month count cannot be acted on. It is cash over burn, and it has no term for how long a fix takes, so a threshold on it fires after the fix has expired. In the worked example the six-month convention fires three months after the last month the round could have been authorized.

How is the authorize-by date actually calculated?

By running the downside cash path with the decision applied in each candidate month in turn. The date is the latest month where closing cash stays at or above the floor from now through the month that decision's own cash effect lands. The downside is not a chosen percentage; it is the worst rolling three months in your own 24 actuals.

Why is the floor not zero?

Because you stop at the highest of your binding constraints. Here that is a covenant minimum tested at all times plus the largest measured gap between a month-end balance and that month's intra-month low. In the example that is 1,660,000, which removes 2.3 months from the quoted figure.

How is this different from a thirteen week cash forecast?

Horizon and question. A thirteen week forecast works weekly on receipts and disbursements and asks whether payroll clears. This works monthly over two years and asks by when a structural decision has to be authorized. Run both; they answer to different people.

Where does the actual cost reduction get built?

Not here. This register holds one line with a measured up-front cost and a monthly effect. Building the candidate list, reading contracts for notice periods and separating in-year cash from run rate is the cost reduction plan tool. Covenant headroom by test date is its own forecast.

Does this replace the reforecast decision?

No, and the two triggers are different animals. Reforecast triggers decide whether this month's miss warrants rebuilding the plan. These decide by when an action has to be authorized to still work. A month can trip one and not the other.

Turn the runway number into a date

Download the blank pack as Word and CSV files, or open this exact pack in River and let it set the floor, measure both scenario factors from your own history and solve every authorize-by date.

Edit with AI