A 715p bid for a 577p business.

The seat I took

Castlelake structured a bidding vehicle at 49% Castlelake and co-investors against 51% EU nationals, including Peter Bellew and Mark Breen, to satisfy the ownership and control requirements of Regulation (EC) No 1008/2008. Apollo came in above them at 715p.

Most coverage framed this as a question about easyJet. I wrote it for Athene, a limited partner inside the Apollo fund, facing a live decision about whether to commit more capital. That changes what matters. An LP does not need to know whether easyJet is a good airline. It needs to know whether its general partner is exercising judgement worth continuing to pay for.

Four tests

The analysis runs as a sequence, each test mattering only if the one before it fails.

• Is 715p supported by standalone value?
• If not, what would have to be true to justify it? The Atlas Air precedent, and Castlelake as competitive pressure.
• Are those assumptions realistic, given what operational improvement at a low-cost carrier actually delivers?
• What does that tell Athene about committing further capital?

How the number was built

A discounted cash flow weighted 60/40 against trading comparables, on data pulled across more than fifteen S&P Capital IQ exports.

• Terminal value on exit EV/EBITDA as the primary method, at a peer median of 5.12x, with Gordon Growth as the cross-check rather than the reverse.
• Beta simplified to easyJet's own Capital IQ figure of 1.70, rather than a constructed peer beta adding false precision.
• Cost of debt at a 3.34% weighted average, taken from yields to maturity on the actual bond stack.
• Working capital derived from four years of cash flow history rather than a percentage-of-revenue rule of thumb.

The assumption that changed the answer

The first version carried capital expenditure at roughly 9% of revenue. That is wrong for an airline in a fleet replacement cycle. Corrected to a 14 to 22% range, the free cash flow profile changes materially and the valuation conclusion reverses.

This is the most useful thing in the project. The model did not fail on the discount rate or the exit multiple, the two inputs everyone argues about. It failed on a capex line that had been carried forward without being examined.

What the model cannot tell you

The present value of the terminal value is 197% of enterprise value. The explicit forecast period contributes minus 97%. Nearly all the value sits beyond the horizon anyone can forecast, which is normal for this kind of asset and a reason to hold the output loosely.

The 60/40 weighting is a judgement, not a result. Tested at 50/50 and 70/30, the conclusion does not move. That is worth more than defending the original split.

Written as an MSc Applied Project at Imperial College Business School. The valuation is my own work on public information, and is not investment advice.

← Back to work