The Index Doesn't Measure Your Car

Two numbers from this year don't agree with each other.

In the first quarter of 2026, Hagerty's Supercar Index rose 19 percent. The Ferrari Enzo gained 109 percent. The 288 GTO gained 106 percent. Over the same stretch, Hagerty's Index of Affordable Classics, which tracks thirteen cars priced around $40,000, moved zero. Flat, after losing one percent the quarter before. The broad Hagerty Market Rating opened the year at 58.28, the lowest reading in about fifteen years.

So the collector car market is booming and the collector car market is dead. Both are true, depending on which market you mean.

‍There is a book that explains exactly why, and it was published fifteen years before this happened. It is not written for you. That turns out to be the useful part.

What the book actually is

Better Than Gold: Investing in Historic Cars is Dietrich Hatlapa's attempt to treat rare classic cars as an asset class instead of a hobby with prices attached. Hatlapa is a former managing director at ING Barings. In 2007 he founded Historic Automobile Group International in London, and since December 2008 HAGI has published monthly indices tracking the top end of the collector market. The HAGI Top index follows fifty benchmark models across nineteen marques. There are marque-specific indices for Ferrari, Porsche, and Mercedes-Benz. The whole thing runs off a proprietary database of more than 100,000 transactions, sourced from private sales, dealer sales, marque experts, and auction results.

The book is where that machinery gets explained. Second edition, 2014, HAGI Publishing. Three hundred fifty pages, thirty-three chapters, seventy-six charts, fourteen appendices of price data. It opens with a legal disclaimer of the kind you'd find in an investment fund prospectus, which tells you something about the seriousness of the exercise. In the preface Hatlapa notes that classic cars differ from financial markets in one respect that no index can capture: they are a passion.

List price was £125. You can find copies used, sometimes for less, sometimes for more.

I purchased the book a few years ago to learn about cars as an investment class and the history of this top end market. I quickly discovered it was aimed at readers or investors with a much much higher net worth than me. Nonetheless, there were some really interesting nuggets that I learned.

Your car is excluded on purpose

HAGI's methodology only counts cars in excellent condition. Hatlapa has said as much directly: if a car isn't good, it doesn't go in the data. A rusty 300 SL with a replacement engine is not the same investment as a correct one, so the index refuses to average them. The methodology also throws out cars with extraordinary competition history, because a car Stirling Moss actually raced can be worth double the same model without that history, and one outlier would distort the whole series.

Read that as an owner rather than as an investor and it says something plain. Every index in this business measures condition #2 or better. Hagerty's indices average condition #2 values. HAGI's average excellent examples. The condition grade this publication is named after, #3, imperfect and used and mechanically sound and driven, appears in none of them.

An index has to compare like with like to be effective. Ironically, this turns these top end cars into a commodity of sorts. But it means the price charts you see quoted in the Financial Times describe a set of objects that your car is not a member of, and never will be, no matter how well you maintain it. Fifty models. Nineteen marques. Excellent condition only. That is the universe.

Which is genuinely freeing, if you let it be.

The price corner

The single most useful idea Hatlapa's group has produced for a normal owner is the one they call the price corner. It's the point in a car's life where it stops behaving like a used car and starts behaving like a collectible. Before the corner, value tracks depreciation, wear, and how much life is left. After it, value tracks rarity, provenance, and originality.

Most cars take twenty to twenty-five years to reach that turn, and the great majority never turn at all. They just keep going down until they're gone. Think about that. In 2000, some cars from 1975-1980 turned the corner. In 2026, cars from 2001-2006 are turning the corner.

The corner is real at every price point, though, and that's what makes the concept portable. A 1999 Miata turned the corner. So did the VR6 Golf. So did the C6 Z06, which Hagerty put on its 2026 Bull Market List at a fraction of what the same list's Ferrari costs. Identical to what happens with a 300 SL, but the ceiling is about three orders of magnitude lower.

The practical use of knowing this: it tells you what money you're spending. Before the corner, every dollar you put into a car is consumption. After the corner, some fraction of it is preservation. Neither one is investment. The corner just changes the recovery rate from roughly zero to something above zero.

What actually drives value, and what you control

Hatlapa's value drivers are not mysterious. Low production numbers. Racing pedigree, if the model has any. Documented ownership and service history. Originality of components. Low mileage, which HAGI treats as one of the most important parameters of all. In a December 2025 interview he was still hammering the same three things: mileage, documented service history, original parts.

Sort that list by whether you can do anything about it.

Two are fixed the day you sign. You cannot change how many were built, and you cannot retroactively enter your car at Sebring. You also cannot un-drive the miles, and you shouldn't want to, because a car you don't drive is the problem this whole blog exists to prevent.

Fixed after you buy is not the same as irrelevant before you buy. If you're shopping, it costs nothing to weigh production value. The threshold is what matters. A trim package that moved 40,000 units is not rare. Neither is a special edition that amounts to a badge, a stripe, and a numbered plaque. Genuine scarcity in this era means four figures in the market you're actually buying in, and sometimes low four figures.

Five that clear that bar. All 2000 to 2007, all selling well under $50,000 today.

2003 Audi RS6, C5. 8,081 built worldwide, roughly 1,200 of them sold in North America. Sedan only here, automatic only, twin-turbo 4.2 V8. Classic.com has the average sale around $20,000.

2004 Volkswagen R32, Mk4. 5,000 cars for the US in a single model year, manual only. Volkswagen's own figure for global Mk4 R32 production is about 14,000. Average sale around $26,600.

2004–2005 Mazdaspeed MX-5. 5,428 across two model years, and the second year got cut short at 1,428 cars when a fire hit the factory. Cheapest entry here by a wide margin.

2003–2004 Mercury Marauder. 11,052 total, split 7,839 and 3,213. Higher count than the rest of this list, and still nothing against the millions of Panther-platform cars it shares a frame with. Hagerty values a good 2004 around $13,500.

2006–2007 BMW Z4 M Coupe. 4,275 worldwide, 1,815 for North America. S54 engine, manual only, hydraulic steering, and the last of that. Average sale around $34,000.

Now for the part that keeps this from being a shopping list.

Production volume sets a ceiling. It does not set a price. The Mk4 R32 is the cleanest illustration available: 5,000 cars, near-identical build sheets, and recorded sales running from $6,300 to $104,000. Z4 M Coupes run $12,250 to $80,500. A driving Marauder is $13,500 and a 257-mile Marauder brought $46,200 at Barrett-Jackson. Same production number in every case. The whole spread is condition, mileage, and paperwork.

Which is this piece's argument arriving from the other direction. Rarity decides whether a ceiling exists. You decide where under it you land.

The second warning is more practical. Low production means low parts production. Model-specific parts for an RS6 or a Mazdaspeed are not sitting on a shelf somewhere, and a few of them are gone for good. Before you buy anything on this list, spend an hour on the model's forum reading about what's discontinued. If the answer is “a lot” or includes something that fails on a schedule, the rarity is a liability and you should price it that way.

That leaves two variables you control. Documentation and originality. Both are free or close to it, and both are things most owners handle badly.

Documentation means the file. Every receipt, every invoice, the date and the mileage, the name of the shop, what the part number was. Photographs of the work while the car is apart, because that's the only time anyone can see it. Records of the previous owners if you can reconstruct them. This costs you a folder and ten minutes per visit. At sale time it is the difference between a buyer taking your word and a buyer taking your paperwork, and that difference is real money on a $30,000 car.

Originality is harder, because it conflicts with driving. Some of it doesn't. Keeping the original wheels in the attic when you fit something wider costs storage space and nothing else. Keeping the pulled airbox, the original head unit, the factory jack and the tool roll, the owner's manual with the dealer stamp. The rule that follows from Hatlapa's data is not "don't modify." It's "don't throw anything away."

The corollary is less comfortable. Cosmetic money does not come back. A repaint on a driver-quality car costs more than it adds. So does a full interior. The work that pays at this level is mechanical and documented, which happens to be the same work that keeps the car running, which is the whole reason to own it.

The bubble chapter

The section of the book most worth an evening is the one on the late 1980s.

Prices for Ferraris and other blue-chip cars went vertical, driven partly by Japanese money and partly by speculators who never intended to drive anything, and then the market broke. Values on some cars fell by more than half. People who had borrowed against them lost the cars. It took the better part of two decades for the top of the market to get back where it had been.

Hatlapa documents it with the indices reconstructed backward, which is more useful than the anecdotal version, because you can see the shape of the thing. Steep run-up concentrated in a narrow band of cars. Broad market barely participating. Collapse from the top down.

Compare that to now. 2025 set an all-time record for collector car auction and online sales at roughly $4.8 billion, up ten percent. In the same window the Hagerty Market Index fell to 171.04 in March 2026, its lowest in over four years and seventeen percent off the December 2022 high. The British Car Index dropped six percent in a quarter to a five-year low, with the Jaguar XK120 down twelve percent by itself. The money is in the room. It's buying eleven cars.

I am not calling a crash. The point is narrower and more useful: the top of this market and the middle of it have decoupled before, and when they decouple, the middle is where the pain shows up first. If you bought a car in 2022 on the theory that it would hold value, the data says you were wrong, and the data also says you were never in the segment the headlines were describing.

Who should buy the book

Not most of you.

At £125 for a fifteen-year-old reference work about cars you will not own anytime soon, the value is poor unless you're doing something specific with it. If you're advising clients, dealing at the top of the market, or writing about valuations, it's a foundational text and you should have it. If you own something that plausibly sits inside HAGI's universe, buy it. If you're curious about how a market gets turned into an asset class, it's the best example I know of the process being done carefully by somebody who actually likes the underlying object.

If you own a 1990s coupe with 140,000 miles on it, borrow it or skip it. Read HAGI's monthly commentary and Hagerty's quarterly index notes instead. Both are free. Both will tell you what the top of the market did last month.

But take the framework, because it translates. The reason the framework matters isn't that it lets you predict what your car will be worth. It's that it tells you, with more rigor than anything else available, that your “Driver Quality” car is not an investment and was never going to be one. Somebody built a proper financial instrument out of collector cars and had to exclude yours to do it.

Which means the case for keeping it has to be that you drive it. Document the work, don't throw parts away, and stop refreshing the price guide. The value you can actually capture is measured in miles.

My 911 was a roach when I bought it and it had no pedigree to speak of. Yes I have turned it into a sort of resto mod, but I take it out and drive it on rallies on gravel. I have a growing file of all of the work I have done and ordered a Porsche certificate of authenticity, but I don’t think about what it’s worth. I think about tinkering on it and the ability to hop in and take it for a ride without worrying about my investments.