Stress-testing growth debt against a ramp that runs half speed

Debt taken against revenue that already exists is serviced by revenue that already exists. Debt taken against revenue you expect is serviced by an expectation, and expectations about new revenue are wrong in one direction far more often than the other. That is the single most common way a business that is doing fine ends up in trouble — not a bad decision, a decision whose timing assumption was optimistic by a normal margin.

The repair is a stress test with three specific outputs: the revenue needed to stand still, the depth of the cumulative hole, and the month it becomes unmanageable.

The base case

The growth carries incremental fixed costs of its own — supervision, space, insurance — of 9,200 a month. Contribution margin on the new revenue is 38 per cent.

That number alone changes most conversations. A plan that reaches 87 per cent of target is normally described as a success. Here it is the break-even.

The two ramps

Plan ramp, reaching 42,000 by month six and holding:

• First month with positive net: month 6
• Deepest cumulative hole: -29,143
• Cumulative at month 12: -14,083

Slow ramp — twelve months instead of six to mature, peaking at 70 per cent of plan:

• First month with positive net: never, inside the first year
• Deepest cumulative hole: -75,301, still deepening at month 12
• Cumulative at month 12: -75,301

The plan case needs about 30,000 of cash absorbed and returns to positive in month six. The slow case needs 75,000 and has not turned. Same decision, same debt, a ramp assumption that was half as fast and a peak 30 per cent lower — neither of which is an unusual miss.

Test it against your actual liquidity

Suppose you hold 55,000 of cash and have 40,000 undrawn on a line. Available liquidity is 95,000.

• Against the plan case hole of 29,143: comfortable.
• Against the slow case hole of 75,301: it fits, with 19,699 left, and only if nothing else goes wrong for twelve months — no tax bill, no equipment failure, no slow-paying customer.

And a third number: with zero new revenue, the payment plus the new fixed costs consume 13,809 a month, so the liquidity carries the expansion for 6.9 months before it is gone.

If your liquidity does not cover the slow case, you do not have a financing problem. You have a sizing problem, and the answers are to do a smaller version, to delay until you hold more cash, or to secure a standby facility before you commit rather than after the hole opens.

Two failure signatures to recognise

A slow ramp and a bad idea look identical in month three and completely different by month nine. The distinguishing evidence is whether the leading indicators are moving. New enquiries rising while conversion holds means the ramp is slow. Enquiries flat after the spend, or conversion falling as volume rises, means the demand estimate was wrong, and additional time does not fix a wrong estimate.

The second signature is margin. If new revenue arrives roughly on plan but contribution margin comes in below the 38 per cent assumption, the stand-still revenue number rises and the break-even moves further away with every sale. At 30 per cent rather than 38, the stand-still figure goes from 36,338 to 46,029 — above the planned peak, which means the expansion cannot break even at full volume. That is worth checking monthly, because it is silent.

Running the test on your own numbers

1. Write down the incremental fixed costs the growth brings, separately from the debt payment. Almost every plan omits some: supervision, additional insurance, software seats, utilities, extra accounting.
2. Compute the stand-still revenue: (incremental fixed costs + debt payment) divided by contribution margin. Express it as a percentage of the planned peak. Above 70 per cent, the plan has no margin for error.
3. Halve the ramp speed and cut the peak by 30 per cent. Not a doomsday case — a normal miss.
4. Run the cumulative cash line for 18 months under both, and record the deepest hole and the month it occurs.
5. Compare the slow-case hole to your available liquidity, net of anything already committed.
6. Compute the months of survival at zero incremental revenue. Under twelve months is a warning; under six is a different decision.

What to change when the test fails

• Reduce the size. One crew instead of two, one shift instead of a site, a shorter lease.
• Lengthen the term. A 60-month term instead of 48 lowers the payment and raises total interest. Against a slow ramp, survival beats efficiency.
• Ask for interest-only through the ramp. Concretely: interest-only for six months on this facility removes roughly 3,000 a month from the deepest part of the hole.
• Phase the commitment. Sign the lease but delay the fit-out; hire one person and prove the ramp before the second.
• Arrange standby liquidity while the file is strong. Availability costs an unused line fee. Applying at month nine of a slow ramp, with a deteriorating cash position, costs far more and often fails.

What to refuse

Refuse a structure whose payments assume the ramp. Refuse a facility with a fixed total repayment where early repayment saves nothing, because a fast ramp should be rewarded. Refuse to sign while the only model you have is the one where it works.

And write the slow-case numbers into the file you keep. When month five looks bad, the question that matters is whether it is bad relative to the plan case or bad relative to the case you funded, and only one of those is an emergency.