Buy, renovate, sell: the whole model on one page

30 August 2026 · 11 min read
£235,491capital per deal
7months per cycle
1.7deals per year
53.4%return on capital

A flip is four numbers and one date. Everything else — the contractor who disappears, the survey that finds damp, the buyer who pulls out in week nine — is variance around those four numbers. This piece sets out the arithmetic first, then the places the arithmetic breaks.

The business is not renovation. It is buying below a number you can defend, and being liquid enough to be wrong twice.

The four numbers

Purchase price, renovation cost, holding cost, resale price. Profit is what is left after the first three come out of the fourth.

P = S − ( B + R + H )
Profit = Sale − (Buy + Renovation + Holding)

Return matters more than profit, because capital tied up for a year is not the same as capital tied up for four months.

ROI = PB + R + H · 12m
Annualised return on cash in, where m is months held
The 70% rule. A common ceiling on what to pay: seventy per cent of the after-repair value, minus the repairs. It is crude and it is conservative, which is the point — it is a filter, not a valuation.
Max offer = ARV × 0.70 − R
On a £320k resale with £45k of work, that caps the purchase at £179k

Run it yourself

The panel below is live, and every number on this page moves with it. Drag it by its title bar to keep it on screen while you read.

Deal calculator

£84,509profit
61.5%annualised
26.4%margin
Inside the 70% rule — max offer £179,000

At these assumptions the deal returns £73,309 after fees, on £235,491 of capital tied up for 7 months.

Unit economics of one deal

A flip business is one deal repeated. The question is never what a single property made — it is what one pot of capital earns per year once you divide by the time it was locked up.

Unit economics — per deal

Capital deployed£235,491
Profit per deal£73,309
Cycle time7 months
Deals per year1.7
Profit per year£125,673
Cost of capital 8%£10,990
Return on capital employed53.4%
Two deals a year at a thin margin beat one deal at a fat one. The trap is judging a flip by its profit rather than by how long it held the money.

Where every pound of the sale price goes

Profit is the last slice, and on a healthy deal it is far thinner than it feels while you are spending the other three.

Purchase 54.7%Renovation 14.1%Holding 4.8%Fees & tax 3.5%Profit 22.9%

Where the money actually goes

The rear extension before work started. Damp on two walls, and the reason the asking price had already dropped twice.

Renovation budgets fail in a predictable order. Structural surprises come first and are the largest single risk; they are also the only category where stopping is cheaper than continuing. Kitchens and bathrooms are the most over-specified line items on almost every flip, because they are the ones the buyer photographs and the ones the developer enjoys choosing.

The list below is draggable. Reorder it into the sequence you would actually commission the work, and watch how much of the budget is committed before you have learned anything about the building.

Most people put the kitchen first. On a house with unresolved damp, that is thirty per cent of the budget spent before the wall is opened.

Commissioning order

  1. 1Structural & damp£12,000
  2. 2Roof & windows£9,500
  3. 3Rewire & plumbing£8,000
  4. 4Kitchen£9,000
  5. 5Bathrooms£4,500
  6. 6Decoration & floors£2,000
Committed before the walls are open£0
Same room, eleven weeks later. The floor was the only element kept.

Holding cost is the quiet one

Holding cost is small per month and ruinous per quarter. Finance, council tax, insurance, utilities and the site itself run whether or not anybody is working.

LinePer month7 months
Bridging finance at 0.89%1,55810,906
Council tax (empty)2101,470
Insurance & utilities1451,015
Contingency3002,100

The average residential sale in England and Wales takes 132 days from offer accepted to completion — up from 96 days in 2019.

Landmark Information Group · Residential Property Trends, Q2 2026 · source

Which is why the honest version of the model has time in it twice: once as cost, and once as the divisor that turns a profit into a return.

Every month of delay costs roughly £2,200 here — and pushes the annualised return down faster than it pushes the profit down.

What breaks it

One deal is not a business. The model only works across a portfolio, because the variance on a single property is wider than the margin on it.
A. DevoFive flips, two of them bad
Completion day. The buyer's survey took nine days and found nothing.

Next: financing the second deal before the first one sells, and why bridging is priced the way it is.