Most fibre investment cases are wrong, but only by a little. The risk-adjusted IRR sits a few points either side of the headline number, and the deal still works. The cases that destroy capital are different. They look fine until you check five specific places — and then they collapse.
Red flag 1: Take-up curves with no shape
If the take-up curve is a smooth exponential approach to a steady-state penetration rate, the modeller has not thought about take-up. Real take-up curves have shape — they pause around saturation in early-adopter segments, accelerate again when pricing tiers shift, and almost always undershoot the modelled rate by 5-15% in year one.
Ask: what is the segment-level take-up assumption, and where did the willingness-to-pay data come from? If the answer is "industry benchmarks," that is a red flag. Benchmarks are starting points, not evidence.
Red flag 2: Civils inflation modelled as flat
Construction inflation has run 4-8% annually in most fibre-active markets for the last three years. Most business cases use 2-3% as the baseline civils inflation. The compounding gap over a 36-month rollout is roughly 12-18% of total civils — which is real money in a capex-heavy programme.
Ask: what civils inflation rate did the model use, and what does the case look like at +200bps stress? A defensible case survives that stress with IRR still above hurdle.
Red flag 3: ARPU that holds steady
Modelled ARPU rarely declines in fibre business cases. In reality, ARPU declines are common — especially after the second wholesale operator enters the market, or after the retail incumbent re-prices to compete. The modeller is not lying; they are just using current ARPU as a proxy for future ARPU.
"On one buy-side diligence, applying a 3% annual ARPU compression — which the market evidence supported — reduced the modelled IRR by 4 percentage points and changed the deal thesis."
Ask: what ARPU trajectory does the model assume, and what is the local market evidence for it? If the answer is silence, you have your red flag.
Red flag 4: The "synergy" line
If the case includes a "synergy" line item — capex savings, operational efficiencies, customer cross-sell — treat it as zero until proven otherwise. In our experience, synergies materialise at roughly 30-50% of the modelled value, and almost always 12-24 months later than planned.
The case should close without the synergy line. If the synergies are real, they are upside. If the case requires them to clear hurdle, the case is a wish.
Red flag 5: No exit-side stress test
Fibre is a long-duration asset. The exit multiple in year 7 or 10 is the single largest determinant of IRR, and it is almost always assumed to be the same as the entry multiple. That is fine when buying a stable asset. It is dangerous when the exit market itself is in flux — which it is, in most fibre markets right now.
Ask: what is the case at -2 turns of EBITDA on exit? If that breaks the deal, the deal was always relying on multiple expansion, not operational performance.
What good looks like
An investment-grade fibre case will give you a clear, single-page sensitivity table covering: take-up rate, civils inflation, ARPU trajectory, exit multiple, and time-to-activation. It will quantify the IRR impact of each, and it will identify the two or three assumptions that matter most. If the case cannot do that, it is not yet ready for an IC.
The good news: every one of these red flags is fixable. The work to fix them is not glamorous, but it is not large either. A pre-IC pressure test is usually a 2-3 week piece of work that pays back many times over in IC defensibility — or, more often, in the deal you decide not to do.