Starting in first gear

Good morning. Twenty-five years ago today, a lot of people who did not have to walked toward a building everyone else was leaving. If you are reading this anywhere near New York, you probably know a name that belongs in that sentence. Take a moment with it. Everything below this line is just metal.

Now, the week. Wednesday's Alpina B7 closed at $17,050, which is $4,200 less than the high bid it turned down in June. The market did not get a second opinion. It got a second listing.

That is the week in one number: the price is what the room says, not what the paperwork says. A cabinet secretary wrote Ford a letter about its Chinese partners, Ford wrote one back with the word "wrongheaded" in the first sentence, and by Wednesday night the White House was siding with Ford. Meanwhile in Surrey, Gordon Murray's new investors put $120 million into his company on the condition that it stays small.

In today's newsletter, we'll get into:

  • Auction Desk: The 2015 M5 Competition Package closing this afternoon, and the near-twin I drive every day.

  • One Big Story: Sean Duffy's letter to Jim Farley, Ford's reply, and the part of the government that agreed with Ford.

  • Engineer's Corner: The hot-vee V8, or why BMW put the turbos where the intake used to be.

  • Garage Economics: Gordon Murray Automotive and the $120 million "investment in further scarcity."

Let’s roll.

THE AUCTION DESK

A near-twin of the M5 I drive every day is on Cars & Bids, and it closes this afternoon

A 2015 BMW M5 Competition Package in BMW Individual Moonstone Metallic over black Merino leather closes on Cars & Bids today at 3:45 PM Eastern. About 61,700 miles, clean Carfax, registered in Georgia and California since new, which is a polite way of saying it has never met a Northeast salt truck. Competition, Executive, and Driver Assistance Plus packages, Bang & Olufsen, no modifications.

The seller bought it in October 2024 and has added roughly 12,000 miles: driven, not stored. The disclosure list reads like most F10 listings: chips, curb rash, a cracked front undertray, tires with 2022 and 2023 date codes, and the corners of the leather dash lifting, which every F10 owner recognizes the way a homeowner recognizes a hairline crack above a door frame. A brake job around 59,000 miles is reported, not documented.

Here is where I stop pretending to be neutral. I own an almost identical car: not this one, but an F10 M5 with the same options, bought a little over two years ago, and it has been my daily driver every day since. It has done two track days at Lime Rock Park, the road course in northwestern Connecticut, and came home both times without a warning light, a smell, or a story. Between track days it does the grocery run and the I-95 crawl with the same indifference. Nothing on it has been anything other than excellent, which is not a sentence I expected to write about a 575-horsepower twin-turbo BMW from the years the internet had opinions about twin-turbo BMWs.

What I would check: put a hand on the dash leather (cosmetic, known fix, price it anyway), ask for the brake receipts, and budget for tires, because four-year-old rubber under 500 lb-ft is a suggestion, not a plan. This is a post-facelift car, which matters to anyone who remembers the 2013 engine recall on the earliest F10s, and the Competition Package is the one to have: 575 hp instead of 560, a 10 mm drop, stiffer springs and bars, sharper steering, and an exhaust that finally sounds like the sticker.

📈 Market signal: The F10 M5 Competition market runs from the high $20,000s to the low $40,000s, with the middle of the cluster in the mid $30s against an original sticker north of six figures. A rare Individual color and a loaded build pull this one up; the mileage and the cosmetic list pull it back. My read is a reserve in the low-to-mid $30s and a result within a few thousand of it. Under $30,000 with the reserve met would not be a soft car. It would be a soft Friday.

ONE BIG STORY

A cabinet secretary wrote Ford a letter. The White House answered it.

On Tuesday, Transportation Secretary Sean Duffy sent Ford CEO Jim Farley a letter expressing "profound concern" about the company's ties to Chinese firms. It named four things: the CATL battery license behind the Marshall, Michigan plant; the Geely-platform cars Ford will build in Valencia; the timeline for moving Lincoln Nautilus production out of China; and a framework Farley allegedly floated at the Detroit Auto Show in January to let Chinese automakers build in the U.S. through joint ventures.

Ford's reply landed the same evening and skipped the customary Washington register: a wrongheaded attempt to capture headlines, factual errors, the most American automaker. The specifics: Marshall is Ford-owned, Ford-run, and Ford-staffed, with 1,700 jobs, and the CATL deal is a license, not a joint venture. The Valencia cars are for Europe. The Nautilus move is on schedule and was praised by Commerce Secretary Howard Lutnick a month ago. On the joint-venture framework, Farley's answer on Wednesday was "N-O, period."

Then the part that makes this a story. Wednesday night, the White House's own rapid-response account posted that Ford is a great American company doing a tremendous job bringing production home. Within about 36 hours, Transportation criticized Ford, Commerce had already praised it, and the West Wing picked Commerce.

Two things the coverage mostly skipped:

  • DOT does not regulate trade. Tariffs are Treasury and USTR. Investment review is CFIUS. The rule that actually bans Chinese connected-vehicle software from model year 2027 belongs to Commerce. A letter about supply-chain exposure has no rule-making behind it. It is a lever without a fulcrum.

  • The framework fight is about who said what in a hallway. January reporting had Farley raising a controlling-stake JV idea in conversation with Duffy and two other cabinet-level officials at the show, days after the President said he'd be open to Chinese plants that hire Americans. Farley's "N-O" is that Ford never proposed one. Both can be true: an idea in a hallway is not a proposal until someone writes it down, and this week someone did, on the other side.

The takeaway: Ford has spent a year assembling the political version of a service history: Marshall, the Nautilus move, the Lutnick interview, the January plant tour. When the letter came, Ford pointed at the folder. Watch-number: 1,700, the Marshall headcount, the one figure both sides agree on.

ENGINEER’S CORNER

The hot vee: why BMW put the turbos where the intake used to be

Source: BMW

Look down on a conventional turbocharged V8. Intake in the valley, exhaust manifolds on the outside of each head, turbos bolted on out there with them. Cold in the middle, hot on the edges. Easy to service, with one problem: the exhaust travels a long way from port to turbine, and every centimeter of pipe is thermal mass and volume the gas has to fill before the turbo does anything.

BMW's N63, launched in 2008, was the first production V8 to flip it. Exhaust ports face inward, the turbos live in the valley, the intake moves outboard. The S63 in the M5 above keeps the layout and adds twin-scroll turbos. Shorter runners mean less volume and less heat lost between cylinder and turbine, so the turbo spools sooner and the engine answers the pedal faster. The catalysts sit right behind the turbines and light off quicker, an emissions win the brochure never mentions, and the engine gets narrower because the widest parts moved to the middle.

The S63 adds a detail that is easy to miss. A cross-plane V8 fires in an order that gives each bank uneven exhaust pulse spacing, some pairs 90 degrees apart, some 270, and a turbine fed from one bank sees the pulses interfere. BMW's cross-bank manifold feeds each turbo from cylinders on both banks, chosen so the pulses arrive at each scroll 180 degrees apart. That is why the S63 sounds flatter than an American V8 and pulls like a much bigger one from 1,500 rpm.

The cost is the part the forums were right about: heat. Two turbochargers in a pocket under the intake, the one place on the engine with the least airflow, cook everything nearby: hoses, wiring, the valley pan, the oil sitting in the turbo center sections after shutdown. BMW answered with air-to-water intercoolers on top of the engine, a dedicated low-temperature cooling loop, and, revision by revision, more cooling wherever it would fit. The other bill is access. On a conventional V8 a turbo swap is a bolt-on. On a hot vee it starts with removing everything above the engine.

Which is why Lime Rock matters more than the spec sheet. It is a 1.5-mile road course with no stretch long enough for the engine to get a clean breath of cold air. If the cooling were marginal, that is where it would show. Twice through, it did not.

The takeaway: The hot-vee bets that throttle response and packaging are worth more than ease of service, and the industry took the bet: Mercedes-AMG, Porsche, Audi, and Cadillac all followed. The early cars paid for the learning curve in repair bills, and those bills wrote the depreciation curve that puts a 575-hp sedan in the $30,000s. The engine is not the risk it was. The reputation still is, and reputation is what you are buying at this price.

GARAGE ECONOMICS

Gordon Murray's investors gave him $120 million to build fewer cars

Every business plan you have ever read asks for capital to grow. In December, Gordon Murray Automotive announced a $120 million investment from Halo Cars Group, and the investors described it as not an investment in scale but in further scarcity. Managing partner Tarik Ouass, now on the board, was GMA's first customer, took the first T.50 built, has bought one of every car since, and is understood to be behind the five-car S1 LM commission. A month before the deal, the first S1 LM sold in Las Vegas for $20.63 million, the highest non-charity auction price ever paid for a new car. The customer bought a piece of the factory. Here is why that is rational.

Run the order book. T.50: 100 cars at £2.36 million, about $3.2 million, before taxes. T.50s: 25 cars at a little over £3 million, north of $4 million. T.33: 100 cars at about £1.37 million, roughly $1.85 million, plus 100 Spiders at more. Add 24 Le Mans GTRs, five S1 LMs, and the 64-car S1 from The Quail (Issue #13), all sold out, and it is roughly half a billion pounds, around $675 million, of contracted revenue across about 400 cars, with a roadmap sketched to 2039. A lot of money and a tiny number of cars, and both halves of that sentence are the point.

The fixed costs are the same at 100 cars or 1,000: one bespoke Cosworth V12 (3,994 cc, 12,100 rpm, 178 kg) amortized across four models, one £50 million-plus ($68 million) factory. That math only closes at 400 units if each sells for millions and the secondary market confirms, loudly, that the number was too low. A $20.63 million result for a new car is not a problem for GMA. It is the advertising budget, and someone else paid for it.

Now the investor's side. Ouass owns one of everything, and the value of that collection depends on one variable: how many more of each GMA builds. A growth investor would want the T.33 run doubled. Ouass wants the opposite, and he sits on the board to make sure of it. Vertical integration in reverse: the customer acquiring control of supply. McLaren, which chased volume and ended up rescued and then sold, is the counterexample a few miles down the road in Woking.

The risk is the same thesis in a different hat. Revenue arrives on delivery, not on order, and the CEO left in November with the CFO running things on an interim basis. A scarcity business is a delivery business with better margins, and delivery is what a small factory finds hardest.

The takeaway: Most carmakers are valued on how many cars they might sell. GMA is now capitalized by an investor whose returns depend on how many it won't. Watch-number: 2039, the last year on the roadmap. Thirteen years of product from one engine and one man's name, at fewer than 40 cars a year, is either the most disciplined plan in the industry or the most fragile. Probably both.

🏁 THE COOL-DOWN LAP

  • The Alpina result, in full. The 2013 B7 xDrive from Wednesday closed at $17,050. On its first listing in June, the high bid was $21,250 against an unmet reserve. Same car, three months later, $4,200 less. The Garage Economics case study wrote its own epilogue: in a limited-variant market, the first room is usually the best room.

  • Volkswagen put a price on last week's vote. The restructuring the supervisory board approved 20-to-0 will cost about €16 billion, per a person familiar with the plan: roughly €10 billion for up to 60,000 job reductions, about €1 billion each to wind down Emden and Zwickau, about €2 billion each for Neckarsulm and Hanover. VW declined to comment.

  • Cadillac is un-pivoting. U.S. sales fell 22 percent in the first half, and dealers are bracing for a gap until revived gasoline models arrive in 2027 and 2028. The all-EV plan lasted about as long as a 2021 press release.

  • Volvo cancelled the EX40 for the U.S. and will keep the aging XC40 alive with a hybrid refresh in 2027. Two EV retreats in one week, one from GM and one from Geely, is a pattern, not a coincidence.

  • Brampton countdown: nine days. The Stellantis expiry at Brampton Assembly lands September 20. Nothing new to report, which at this point is news.

Share the Downshift. Forward them this email. Every subscriber from your link gets you closer to the Daily Downshift starter-pack.

That's a wrap. See you Monday— same inbox, lower gear.

Was this email forwarded to you? Subscribe free at Daily Downshift. Drop a gear. Read the good stuff.

Consider this the feedback loop: reply with what worked, what didn't, and what you want more of. Reader replies are the single biggest input into what gets written next.

Marsel

What did you think of today's email?

Your feedback helps me create better emails for you!

Login or Subscribe to participate

Reply

Avatar

or to participate