NCP

NCP's car parks cost almost nothing to run. So how did the company collapse owing £352m?

It filed for administration in March 2026.

The decision that killed it was made in 2002.

____________________

THE SETUP

NCP filed for administration in March 2026 owing up to £352m. Its founders, Ronald Hobson and Sir Donald Gosling, had sold the business for a reported £580m in the late 1990s. The company that collapsed had grown every year since 2022.

• Turnover rose from £173m to £233m between 2022 and 2025. It still lost money every year.

• PwC was called in on 16 March 2026 with 682 jobs and 340 car parks at risk.

• NCP changed owners five times in 24 years: Cinven, 3i, Macquarie, then Park24 and the Development Bank of Japan.

• Cinven raised £600m in 2002 by selling NCP's car parks back to it as a tenant, to help fund its own purchase of the company.

Parking looked like one of the best businesses in Britain for decades. Customers pay before they use the space. A barrier gate and a payment machine do most of the work a member of staff would do elsewhere. Planning rules make it hard to build new car parks in city centres, hospitals and stations, so existing sites face little new competition and prices rise easily because drivers usually have nowhere else to go.

That's why five different owners bought NCP in 24 years and each one could see a growing business with minimal operating costs.

Covid and hybrid working did not kill NCP. Demand fell for a time, but the business stayed sound underneath it. Revenue was still climbing when NCP went under, which is what a Covid explanation can't account for.

One decision made in 2002 was the bullet that eventually killed NCP. For most of the next two decades it didn't look like a problem at all. Then, in the space of a few years, it stopped working completely.

Why it took so long and what finally broke it, is the NCP story.

____________________

THE PLAYBOOK

Play 1: Sell the freeholds

In May 2002, private equity firm Cinven bought NCP from the US conglomerate Cendant for about £820m. Within months, Cinven sold and leased back more than 100 of NCP's freehold car parks for close to £600m. The rent was written to rise with inflation and run for decades.

NCP had owned the concrete outright. After 2002 it paid rent.

While the car parks stayed full, that rent was a small and shrinking share of what each site brought in, so nobody felt the terms were rigid. The £600m didn't stay inside the business. Cinven used it to help fund the original purchase, then sold NCP on to 3i in 2005 for £555m, close to three times its original equity. The freehold carparks were never repurchased.

____________________

Takeaway: A sale and leaseback pays the seller immediately. The company pays for it every year after, for as long as the lease runs and the sites keep earning enough to cover it.

____________________

Play 2: Growth masked the problem

NCP's EBITDA more than doubled to over £42m within three years of the 2002 deal, helped by a push into local government services like traffic enforcement contracts. 3i sold the business on to Macquarie in 2007 for £790m, a gain of roughly £235m in under two years.

Every buyer after that bought into the same theory Cinven and 3i had already proved out: if revenue kept growing faster than the fixed car park rent bill set in in 2002, the lease terms would never actually bite.

Macquarie ran NCP as an infrastructure asset and added roughly £500m of fresh borrowing on top of the rent NCP already owed, betting that scale and disciplined management would cover the old and new deby.

Park24 and the Development Bank of Japan bought in 2017 on a different strategy. If NCP built more government contracts and non-parking revenue, the fixed rent could be spread across a wider business instead of falling on parking alone. For this to work, growth had to keep outrunning the terms fixed in 2002.

For the next fifteen years, it did. Growth was strong and nobody in the ownership chain had a reason to renegotiate those 2002.

____________________

Takeaway: A fragile structure looks exactly like a sound one for as long as growth outpaces it.

____________________

Play 3: The recovery

Between FY2022 and FY2025, NCP's total turnover grew from £173m to £233m and its yearly loss fell from £28.2m to £5.8m.

That recovery wasn't spread evenly. Airports, hospitals and leisure sites bounced back close to pre-pandemic levels. City centre and commuter car parks, the ones tied to the 2002 leases, didn't, because hybrid working had permanently cut how often people drove into town for a nine-to-five.

Inflation made it worse. The rent on those 2002 leases rose automatically with prices, so after 2021 the fixed cost on the weakest sites kept climbing while their income stood still. At the same time, higher interest rates made the debt Macquarie and Park24 had added on top of that rent more expensive to refinance.

Group revenue could look like it was recovering while the sites carrying those costs fell further behind every year. NCP couldn't close them or renegotiate the leases, the contracts didn't allow it. When the next round of rent came due at the end of March 2026, there was no cash left to pay it.

____________________

Takeaway: don't just look at total revenue. Some costs can't be changed once they're signed. Check whether those costs are rising faster than the sites carrying them can keep up.

____________________

ONE MORE THING

By May 2026, 29 of NCP's sites had shut for good. Three of them sat next to the railway station in Grantham and their closure took 590 parking spaces out of the town in a single week.

Q-Park has already picked up 20 of NCP's former sites, negotiating fresh leases directly with the same landlords NCP couldn't get out of. Same buildings, different terms. The locations were never the problem. The contract attached to them was.

Then there's the money. Cinven made close to three times its original investment selling NCP on in 2005. 3i turned that into a further £235m within two years, selling to Macquarie in 2007.

Macquarie and Park24 inherited the lease obligations from that earlier chain of sales, plus the extra borrowing stacked on top. Neither of them chose those terms.

The landlords who bought the freeholds in 2002 never carried any risk. They collected index-linked rent through NCP's growth years and through its final years of losses and they will collect it from whoever takes on those car parks next.

Selling early made money.

Staying meant paying for someone else's decision, that one none of NCP's later owners had actually made.

____________________

THE PAPER TRAIL

BBC News, 1998, on two ex-servicemen selling a £200 bombsite for £580m

3 minute read

The original founding story: how Ronald Hobson and Sir Donald Gosling turned a bombed-out patch of Holborn into the deal that made them "business buccaneers,".

http://news.bbc.co.uk/2/hi/business/68904.stm

____________________

The Guardian, 2002, on English Heritage trying to save a "draughty, weedy and smelly" NCP car park

3 minute read

NCP wanted to knock down a 1929 Soho multistorey and build something worth £80m. English Heritage listed it as a protected monument to early motoring.

https://www.theguardian.com/society/2002/jun/13/urbandesign.arts1

____________________

Express, 2026, ranking NCP's most "extortionate" car parks

3 minute read

£15 an hour in Soho, £49.95 a day in Knightsbridge and drivers on social media calling it "robbery." The prices that made the brand infamous long before the administrators arrived.

https://www.express.co.uk/life-style/cars/2184429/car-parks-most-expensive-ncp

____________________

BBC News, 2026, "I risked a parking fine as it was cheaper than NCP"

2 minute read

A Manchester conference-goer worked out that gambling on a street parking ticket was still cheaper than paying NCP's £35 day rate in Spinningfields.

https://www.bbc.com/news/articles/cr71rng8zz4o

____________________