A modest independent workshop on a wet British high street

Cases · 21 September 2026

Three decisions, not three averages

The same arithmetic, pointed at different questions. A retirement in engineering, a seasonal guest house, and a clinic partnership where the buyer’s plan is trying to become the seller’s value.

Case one. Porth Engineering and a retirement

The decision is not “what is the company worth in the abstract”. A founder wants to retire, a production director is willing to buy a majority with bank support, and a son who does not work in the business holds a minority that must be treated fairly. Nobody has hired an auction. The number has to survive a kitchen table and a bank.

The essays already did the work. Brought onto one page, equity value on a going-concern basis is:

Method Enterprise £ Equity £
Discounted cash flow4,566,3373,946,337
Precedent, 6.2x, upfront4,464,0003,844,000
Comparables, 5.68x4,089,6003,469,600
Fair net assets, in placen/a745,000

Net debt of 620,000 is the bridge on the first three rows. The asset row is already an equity figure. A retirement sale to an insider, supported by a bank that will not pay for a strategic bidder who has not turned up, should live between the comparables equity and the cash-flow equity, with the precedent as a reminder of what a cleaner auction might do. This desk would describe a negotiating range of roughly 3.47 million to 3.95 million of equity, and would refuse a single-point “valuation” of 3.7 million. The asset floor of 745,000 tells the minority holder something else important: even a gloomy outcome is not zero, and even a cheerful outcome is mostly goodwill, which a bank will not mortgage with enthusiasm.

What the range does not do is price a minority discount. The son’s shares are not a controlling stake. If the transaction is a sale of the whole company, every share can be priced pro rata to the whole, which is often the fair instinct inside a family. If the son is being bought out and will have no control and no easy exit, a discount for that lack of control and marketability is a separate argument. It should be written as its own percentage with its own reasons, not stirred into the enterprise multiple where nobody can see it.

Case two. Kernow Guest Houses and the weather

Two small houses on a Cornish estuary, traded through a fictional partnership. Occupancy is a function of the school holidays, the rain, and one excellent breakfast cook. A five-year discounted cash flow here is mostly a weather forecast wearing a discount rate. The primary method should be maintainable earnings, with the cash flow used only as a sensitivity if someone insists on seeing one.

Reported EBITDA is 186,000. It is not maintainable. The bridge:

Normalisation £
Reported EBITDA186,000
Add room profit if the owner’s suite is let28,000
Deduct underpaid family wages to a market level(20,000)
Deduct a storm insurance receipt inside EBITDA(42,000)
Maintainable EBITDA152,000

186,000 plus 28,000 is 214,000. Minus 20,000 is 194,000. Minus 42,000 is 152,000. The family-wage line goes down, not up, because the two family members are paid less than a market wage. EBITDA has been flattered by their cheap labour. A buyer who will have to hire will not keep that gift. The insurance receipt is not a season. It leaves.

There is also a 60,000 capex overhang: roofs and boilers deferred. That is not an EBITDA adjustment. It is cash a buyer spends after completion, so it comes off the equity bridge, once.

The multiple used in this classroom is 5.0 times. It is not a published lodging multiple. It is a low figure chosen because the trade is seasonal, owner-dependent, and small. 5.0 times 152,000 is an enterprise value of 760,000. Interest-bearing debt is 110,000 and there is no surplus cash worth arguing about, because the bank balance turns over with deposits and suppliers. Equity value before the overhang is 650,000. After the 60,000 of deferred work, 590,000.

A cash flow model that grew 186,000 of reported EBITDA at 4 per cent forever would have valued a storm and an underpaid family. That is the failure mode. Normalise first. Then decide whether a forecast adds anything a multiple has not already said. Here, it does not.

Case three. Two clinics, and synergies changing sides

A fictional two-site dental partnership. Associate-led, transferable contracts, no bank debt, 70,000 of surplus cash. Maintainable EBITDA, already adjusted for a market principal’s time, is 410,000. A consolidator is in the market for practices. The partners have seen a slide claiming “clinics are selling for ten and a half times”.

Stand the methods apart.

A financial buyer, or a valuation that refuses to gift the seller the consolidator’s head-office savings, might pay 8.0 times standalone EBITDA in this classroom. 8.0 times 410,000 is 3,280,000 of enterprise value. Net cash of 70,000 is added, because net debt is negative. Equity is 3,350,000. The 8.0 is an assumption of this illustration, sitting inside a made-up clearing range of 7.5 to 9.0 for similar revenue quality. It is not a statistic.

The 10.5 times headline the partners heard would imply enterprise value of 4,305,000 and equity of 4,375,000 once the cash is added. The gap between 8.0 and 10.5 is 2.5 turns, which is 1,025,000. That gap is a plausible size for buyer synergies: one central billing team, recruiting the partners will not do, and a brand the partners do not own. If the seller’s model quietly lifts margin to the buyer’s post-deal margin and then applies 10.5 times as well, the synergies have been counted twice and the partners will be disappointed in the first meeting.

The walk-away analysis this desk would write is short. Standalone equity around 3,350,000 is the value of the partnership as a partnership. Anything above that is a share of a named synergy, negotiated, and preferably paid partly as an earn-out so the share is only received if the synergy arrives. The asset base of chairs and fit-out is a floor well below either number and matters mainly if a site’s lease is the real constraint. It does not get a vote equal to the earnings.

What the three cases share

If you are holding a real letter from a buyer, a bank, or a family solicitor, the desk can talk about which essay is the closest analogue. It will not turn that conversation into a formal opinion. The limits are set out on the about page and in the terms.