The car industry changes hands one small deal at a time

Ask who is buying the car industry and the answer comes back in dollars. It is the least useful answer available.

Deal value is the figure every report leads with, and it is the figure that tells an owner the least about the transaction in front of them. The three levels at which this market can be measured (global, European, national) do not measure the same thing, do not use the same unit, and do not move together. Reading the wrong one is how a supplier concludes the market is frozen in a year when the network around it changed hands twice.

Global automotive deal value is one signature a year, not a market

BGL and REACH tracked the automotive supplier market across fifteen years, 2009 to 2024. Value peaked in 2019 at $61.8 billion, against a low of $8.5 billion in 2010. BGL names the reason for the peak in the same breath: a significant contributor was the transaction that produced Stellantis out of Fiat Chrysler, which it puts at $45.3 billion. One signature, roughly three quarters of the record year.

The deal count behaves nothing like the value. Across the same fifteen years the supplier market averaged about 320 transactions a year, with a low of 212 in 2020, the weakest of the period. So the count moves within a narrow band while the value line swings by a factor of seven.

Note what the low says. In the year the industry physically stopped, it still completed two thirds of a normal year's transactions. Companies change hands during a crisis. What a crisis changes is the price and the identity of the buyer, not the frequency.

That gap is the single most useful thing to know about global automotive M&A. The value series does not measure how much of the industry is changing hands. It measures whether two carmakers happened to merge that year. When they do, the chart spikes and the commentary calls it a boom. When they do not, the chart collapses and the same commentary calls it a freeze, while roughly the same number of businesses change owners either way.

The figure to be most careful with is the one quoted most often. PwC's current automotive deals outlook annualises 2026 value at about $19 billion, the lowest reading in a series that starts in 2019. That number is United States only, and it travels as though it were global. The same report notes that private equity went from around 5% of deal value to 13% while its deal count stayed flat, with an average ticket below $100 million: bigger cheques from the same number of hands, not new money arriving.

More than half of 2025's automotive deals bought chips, electronics and software

Bain's January 2026 read on the year gives the composition rather than the total. More than half of all automotive and mobility transactions in 2025 involved semiconductors, electronics, sensors or software engineering capability. Global deal value rebounded past $35 billion in the first nine months, with average deal value at $1.2 billion, more than double the 2023 level, and scope deals held at about 80% of volume.

Infineon completed its $2.5 billion all-cash purchase of Marvell's automotive Ethernet business in August 2025. A semiconductor company bought a semiconductor business, and it is filed, correctly, under automotive M&A.

This is why totals from different providers never reconcile, and it is not a data quality problem. The industry is buying capability, not capacity. A carmaker that needs a software stack buys a software company, and the transaction enters the automotive series because of who paid, not because of what was sold. Any figure described as the size of the automotive M&A market is therefore a statement about where one provider drew the line.

The deals that made the headlines did not move a franchise point in Europe

The transactions that dominated the trade press over the last eighteen months were large, real, and almost entirely irrelevant to the question an owner is actually asking.

American Axle and Dowlais, the owner of GKN Automotive, combined in a deal that closed in February 2026 and left Dowlais shareholders holding roughly 49% of the resulting group, which is closer to a merger than to a purchase. Asbury Automotive completed its acquisition of the Herb Chambers Companies in July 2025: 33 dealerships, 52 franchises and three collision centres, for a net purchase price of about $1.45 billion, of which $750 million was goodwill, per Asbury's own annual filing.

Not one of those moved a European retail owner. The pressure that will move European owners is building somewhere else entirely, in the German supplier base being resized around a carmaker that no longer needs it, and in the fact that the Chinese groups arriving in Europe have been building and partnering rather than buying. Both produce sellers. Neither has yet produced a headline deal.

European automotive M&A is counted in franchise points, and 49 deals moved 1,039 of them

Europe does not count this market in dollars. It counts it in franchise points, and the unit matters: a franchise point is a mandate to represent a brand, not the building it is represented in, and the difference between addresses, mandates and owners is the whole argument.

Two numbers, from the same source, in the right order.

First, the base is shrinking, and recently. ICDP reports that franchise points for continuing brands across the EU and the UK fell by more than 2% in 2023, by more than 2% again in 2024, and by 4.5% in January 2025 against January 2024. Those three years account for about three quarters of the entire decade's decline. Whatever has been happening to European car retail, most of it has happened since 2023.

Second, the base is not just shrinking, it is being redistributed. In 2024, 49 transactions moved 1,039 franchise points in Europe (ICDP, reported by Motor Trader). That is an average of twenty-one mandates per deal, which tells you these are not showroom sales. They are groups buying groups. At the top of the resulting structure, by revenue, sit Emil Frey at £15.8 billion, Penske at £8.8 billion and Hedin at £7 billion.

There is a cheaper way into the same network, and it explains part of the shrinkage above: a brand that wants European coverage can take mandates rather than buy the companies that already hold them. Every mandate awarded that way is one that does not need to be bought.

In 2024 the foreign buyer was Europe's exception, and in the United States it is a different trade

Of those 49 European deals, 8% had a foreign buyer. Which is to say that 92% of the buying was done by someone already in the market, on our arithmetic from ICDP's published share.

That single figure disposes of the most common assumption in the room. European dealer groups are not being bought by foreign capital. They are being bought by their neighbours, and usually by the neighbour one size larger.

The American market looks different in volume and identical in shape. Kerrigan Advisors counted 454 dealership buy-sell transactions in the twelve months to June 2025, a record, and 220 in the first half of 2025 against 204 in the first half of 2024. Values sat 75% above pre-pandemic levels and 17% below the 2022 peak. And two-thirds of buyers in that half-year already had a presence in the market they bought into.

The two counts do not compare directly, and the difference is instructive rather than awkward: 454 counts individual dealership transactions in one country, while 49 counts corporate deals across a continent, each carrying twenty-one mandates. What does compare is the identity of the buyer, and on both sides of the Atlantic it is a local operator getting bigger. What that buyer is underwriting is a stream of future earnings whose inputs include, increasingly, what the electric cars on the forecourt will be worth in three years.

What that buyer is actually paying for is visible in the Asbury filing quoted earlier, which is unusually explicit for this market. The Herb Chambers businesses carried $3.2 billion of revenue in 2024. Of the $1.45 billion net price, about $610 million went to real estate and leasehold improvements and about $85 million to vehicles, fixed assets, parts and supplies. The remaining $750 million was goodwill: the portion of the cheque that buys no building, no vehicle and no shelf of parts. A buyer at this level is writing two cheques at the same time, one for a property portfolio and one for a trading position, and the two are priced by entirely different logic.

It also becomes clear who is not at that table. Private equity's share of US automotive deal value roughly doubled to 13%, on a flat deal count and an average ticket below $100 million. Those three facts together describe a buyer writing larger cheques into a narrow band of assets, not a fund sweeping up dealer groups. For a European group of ordinary size, the probable buyer is not a fund. It is a trade buyer who already knows the territory, and who is weighing your business against the two others they could buy instead.

Italy changes owners faster than most, and nobody counts by how much

Here the trail stops.

KPMG counted 1,441 completed transactions worth €79.4 billion in the Italian market in 2024. That is every sector at once, and no automotive breakdown is published. Search for a count of Italian automotive M&A and the result is not a hard number to find. It is a number that does not exist.

The reason is not mysterious, and it matters more than the absence. These transactions are not secret. A change of control in an Italian company is filed, and the filing is public. What does not happen is the aggregation: nobody collects those filings, sorts out which of them are automotive and publishes the total, because no provider has a commercial reason to do it for a market this size. The number is obtainable. It has simply never been assembled.

This is not a gap in a dataset, then, so much as a job nobody has taken. It is also the reason Italian owners make the largest decision of their working lives using European averages and American headlines, neither of which describes their position. The retail consequence is documented: Italy has been adding showrooms while losing the families who owned them. The industrial consequence is documented too, in the supplier base around a plant that has been told to find other customers without anyone publishing who is buying those suppliers.

When the number that would inform a decision has not been produced by anybody, it has to be produced for the decision. That is what commissioned research is for, and it is the honest answer to a question the public sources do not cover.

The level that decides your deal is always the smallest of the three

Global tells you the weather: whether capital is willing, and what it is currently willing to buy. European tells you the shape of the buyer: domestic, already present, and acquiring mandates twenty at a time. Neither tells you what your business is worth to the person most likely to buy it, because on the evidence of every level above, that person operates within a short drive of you and appears in none of these series.

Which is the practical use of the three levels together. They narrow the field. A market where 92% of buyers are domestic and two-thirds already trade in your postcode is not a market you prepare for by watching deal value. It is one you prepare for by knowing precisely which three or four groups are one size larger than you, what they have bought in the last two years, and what they still need.

That work is the deal. The rest is weather.