The Cult of Cars, Racing and Everything That Moves You.
We may earn a commission from links on this page

It Was Japan’s Most Daring Car. Then the Economy Imploded

Illustration: David Grimshaw for Jalopnik
We may earn a commission from links on this page.

“I was driving it with the windows down,” then-product planner and engineer Bob Hall told me over the phone, “and you’d have sworn there was a Ferrari two cars behind you. With the windows up... you couldn’t hear a fucking thing.”

These are charming words to car enthusiasts, because the car Hall—who you might know as the father of the Mazda Miata roadster—was talking about never went into production. We have never heard the sounds it made. We have never even seen one.

But we did come tantalizingly close.

In 1991, some 67 dealers in America were gearing up to put a new top-flight luxury car on the market. It would take on Lexus, Mercedes, Cadillac, BMW—every car company with a full-size flagship luxury sedan at the time—and it’d do it with 12 cylinders under the hood.

It was called Amati. Its parent company was none other than Mazda, a small, independent little car company punching way, way, way above its weight. And then, in a flash, it all disappeared.

Few people have ever laid eyes on an Amati. There are no pictures of it on any Mazda site. There are no prototypes hanging out in museums. (I’ve checked.) There are only glimmers of brochures printed, but never distributed, scanned and hosted on old forum pages. A few spy shots and renderings made their way into a print magazine or two at the time, but without official confirmation from Mazda.

Former employees confirmed to me that there’s still some footage of camouflaged prototypes testing kept in a private collection, but none has been made public. It still has some power in the cultural imagination, probably because imagination is all we have of it.

It evaporated, along with the economic hopes of an entire nation after a decade of unfettered, almost unprecedented growth—and Mazda has never been as ambitious.

The entire Japanese auto industry hasn’t been, either.

The Bubble Era: Behind Japan’s Best Cars Was A Colossal Economic Boom

The scale of the Japanese bubble economy of the 1980s is hard for us to comprehend today. A famous (if apocryphal) real estate assessment at the height of the period valued about one Disneyland’s worth of land in the heart of Tokyo as greater than the combined real estate of the entire state of California. A contemporary report cited by the New York Times actually valued all of the land in Japan at $13.47 trillion, more than the value of all of the land in the United States as a whole. And that figure was in 1988 dollars.

On the back of this real estate bubble, the whole of the Japanese economy surged with growth in the automotive, electronics and financial sectors. And when that bubble burst, it took everything with it, the wildest dreams of Japanese auto industry included.

Ever wondered why Japanese cars seemed to hit a kind of peak in quality in the 1990s? Why the Toyota Supra, Honda’s NSX, the Mazda RX-7 and the Nissan Skyline GT-R were all rapidly developing until a few years into the decade, when they all started cycling into re-heated refreshes and then were altogether canceled? Why it’s not hard to find a 1992 Camry still on the road, why Toyota 2JZ engines are still the go-to choice for big power tuners? These are all products developed in the heyday of the Bubble Era.

Just the same, if you’ve spent the past 20 or 25 years wondering why doesn’t Mazda or Toyota or Nissan or Honda make cars like this anymore, staring at a gleaming NSX Type R or Honda Beat or whatever, it’s the same thing. All of these madcap projects were cut loose after the bubble burst. Only very recently have we started to see a comeback to even approach how things once were.

A turbocharged performance car with all-wheel drive and all-wheel steering is still novel today. Mitsubishi was putting that on sale in 1989. Mitsubishi! It barely makes cars anymore.

Now, I can already hear you complaining that, hey, my new car has a touchscreen, or a heads-up display. Surely no car from the 1980s had that.

Wrong! Touchscreens. GPS. In-dash entertainment. Heads-up displays. Computer-controlled active suspension, engines, and design. Just about every modern convenience made a showing (often a world-first for production) as Mazda, Nissan, Toyota, Honda, Suzuki, Mitsubishi, everyone battled to stand out.

In car terms, the Bubble Era stretches from the mid 1980s into the early 1990s (as it takes a few years to develop a car to get into production), and looking back on things, a lot of these cars come off as total flights of fancy. Mazda and Suzuki partnered up to make a mid-engine economy sports car with gullwing doors. Good luck finding that in a dealership these days.

But since I’m a deeply obsessive person, I’ve spent the past few months (please let me pretend I didn’t start researching this project in 2010 only to drop it because it was taking up too much time) going over press releases, histories and contemporary news reports about these cars and companies. Everything boils down to four factors:

  • Outside tariffs meant to protect western industries backfired and pushed Japanese carmakers upmarket—luxury cars with higher price tags over economy cars would mean better profit margins.
  • A wildly-fluctuating Yen pressed Japanese carmakers to squeeze as much money out of the Japanese home market as possible.
  • The now ultra-competitive Japanese home market forced carmakers to try anything to stand out, exploring bizarre niches and oddball features.
  • Pride.

Alright, that last one isn’t totally fair. All car companies are filled with pride. They’re always copying whatever their competitors are doing. That’s why every time BMW comes out with a new four-door crossover coupe, Mercedes debuts another one a year later.

But there was something particular about Japan in the 1980s, and it had the cars to prove it.

The Luxury Flagship That Wasn’t: The V12 Amati

Of all the canceled Bubble Era programs I read about for this article, Amati is the one that got the closest to production. (I wish I had more space to talk about the Nissan MID-4, an all-wheel-drive everyday supercar that predated the NSX, or the great Yamaha OX99-11, an F1 car for the road with tandem seating and a de-tuned racing V12. But let’s keep this tight and leave those stories for another time.)

Amati was announced to the public in 1991, just months before the Japanese economy collapsed. It was a crash was so violent that it ushered in what was prematurely called “The Lost Decade,” a recession that lasted close to 20 years.

Even today it’s hard to get good information on the Amati that wasn’t. Now, there were sister cars to the never-seen Amati flagship. For instance, there was the Amati 500, and Mazda did end up putting a version of it on sale as the car we got here as the Mazda Millenia. (Europe got it as the Xedos 9 and it was called the Eunos 800 elsewhere.)

If you look closely at this brochure from when the Millenia went on sale, you can see how hastily Mazda crammed it into the lineup. You can see Mazda’s normal plans on the right side of the page topped with the 929 luxury sedan, but then there’s the Millenia off to the left, listed as... also a luxury sedan.

The Millenia was about the size of a BMW 5 Series and it had not only sleek styling, but also a very interesting and efficient Miller Cycle V6 engine, an early precursor to today’s super efficient motors.

There was also the Amati 300, aka the Eunos 500 or the Xedos 6, never sold in the States. It was roughly the same as the Mazda 626 we got, but slightly higher upmarket.

And just so everybody is clear, below is a photo of the Mazda 929, sold in Japan as the Sentia. This was not to be sold as an Amati, but the Amati 1000 was like a bigger version of it.

Business reports of the time also suspected that Amati would sell to Americans the car put into production as the Eunos Cosmo, a full-size coupe to rival cars like the Mercedes SL and the BMW 8 Series. Only the Cosmo made a Lexus SC look like a Saturn SC.

It got a three-rotor engine called the 20B-REW, the biggest rotary ever put on sale, the world’s first production car with GPS, and an interior that looked more like a Star Trek: The Next Generation landing craft than something out of a car dealership.

The one we never saw was called the Amati 1000.

It can be confusing to research since the project was initially called Pegasus, and in general everything we think we know about its most alluring detail—its engine—is wrong.

In the 1989 Tokyo Motor Show, Mazda caught the attention of the automotive world with one compact item at one of its stands: a gleaming, four-liter W12.

Ever since, everyone has assumed that this was the engine destined for the Amati 1000. It was powerful enough, with Mazda claiming (in a Tokyo Motor Show-only brochure scanned above and below) it’d make 245 lb-ft of torque. The only horsepower figure I can find is in a news piece on the engine, listing it as 280 HP, the maximum allowed by the Japanese auto industry’s “gentlemen’s agreement” capping power at the time. Still, this would have been reasonably more powerful than the 250 HP 4.0-liter V8 in the Lexus LS400, then the standard of the day.

Now, a 12-cylinder luxury car was rare but not unheard of at the time. BMW put a five-liter V12 into production in 1987. Mercedes responded with a V12 of its own in 1991. But those engines were more like two six-cylinder engines joined together. The BMW even had two ECUs, a holdover from its basic design. This Mazda W12 was different. This was like three engines joined into one.

Look at Mazda’s cutaway drawing and you’ll see that it’s made like three banks of four cylinders, pretty much the rarest cylinder layout of all. It’s complex, sure, but extremely compact. It is, front to back, only four cylinders long rather than six, and would more easily fit under the hood of a normal car. Maybe Mazda wanted people to take it as more production-oriented than they’d otherwise think for such an uncommon configuration. Maybe Mazda just wanted to show off it could do what nobody else even dared attempt.

Only Volkswagen gave the three-bank layout a try a decade later, when it was experimenting with W18s for Bugatti, right at the height of the company’s pre-Recession/pre-Dieselgate megalomania. But even giant Volkswagen couldn’t pull off this configuration, and flubbed down to mashing two narrow-vee eight-cylinder engines together, kind of like what BMW and Mercedes had done years prior. This is all to say that the idea of seeing this Mazda W12 is impossibly romantic.

But it was all a misdirection.

For years the car world figured that, sure, this W12 was going to be the Amati engine, but no. Mazda apparently just was throwing us all for a loop.

“There was never a W12,” as Bob Hall told me himself. “It was always a V.”

Hall was the guy I wanted to talk to most about the Amati program. He was close with Kenichi Yamamoto, the enigmatic head of Mazda through its expansion and the savior of the rotary engine there.

Yamamoto was the one who dreamed the biggest for Mazda, and it was Yamamoto who encouraged and approved Hall’s pet project, a little car that shocked the world called the Miata. An American, Hall went to high school in Japan, worked in Japan for Mazda, and I don’t know of anyone who better understood the company’s actions as they spanned the Pacific.

The Amati, as Hall explained, having worked on the program, was a V12, using pieces sourced from Mazda’s existing (and weirdly overbuilt) V6 program. It used “rotating assemblies of the K series in a new block,” as Hall put it. “This was a completely new block,” he added, though “the valvetrain was the same.”

It made the requisite 280 HP, but, typical of Mazda engines, it was good on power if weak on torque, Hall said. “Faster in 60-80 passing speed,” compared to a Lexus LS400, “not as fast to 60.”

“They had done a 3.6 of the two 1.8s,” Hall said. “The block was designed up to a 5.0-liter. North America would be a 4.6 from two 2.3s.”

Have we ever seen this engine? No. Again, have we ever heard it? No. Have we ever gotten confirmation from Mazda about its specifics? Sadly not.

I’ve never seen any published specs of the so-called Amati 1000 at all, but Hall filled me in as best he could. It would have been as large as, if not slightly larger than the Lexus LS. The 1000 would have been bigger than Mazda’s top-of-the-line car it already had in production, which was the car we got here as the 929. The 1000 was like a stretched 929.

“If you looked at it carefully,” Hall related, “you could see the B-pillar was fatter.” There was more room at the rear, and the hood was effectively drawn out. The big difference was that the 929 was stuck with Mazda’s 193 HP V6 and, while it was a well-made vehicle, was priced and angled downmarket from where the 1000 would have been.

Amati, like the other cars Mazda was building at the time, from the RX-7 to even the little 323, could have been arguably the best car in its field.

What makes Amati so interesting is not only that it never made it, but also where it was coming from.

Daring To Go Big

The first brands to respond to go upmarket were Acura, Lexus and Infiniti from Honda, Toyota and Nissan, respectively. Cars like the Lexus LS and SC were some of the most heavily invested-in cars ever put into production, with unreal spending to make sure they matched the best from America and Europe. Lexus went so far as to buy a house in Southern California to observe wealthy Americans and better understand their desires.

But Honda, Toyota and Nissan were and are big companies, huge in comparison to Mazda. The little automaker had been through a lot over the years, to say the least, and Mazda’s executives remembered not only when the company nearly went out of business after the first Oil Crisis, not only when the company was nearly folded into Toyota in the 1960s by the government, but also when the company was literally nuked. Mazda’s home base is nowhere else but Hiroshima. The light of the apocalypse was within the living memory of much of the company’s top brass.

So Mazda was a scrappy little player and it often acted like it, as Dean Case, who was an engineer at Mazda’s R&D in the U.S. at the time, explained to me in person. He happened to be in town for the New York Auto Show earlier this year, and I hung on his coattails for a day, pestering him with questions about what it was like to work at Mazda back when.

When he worked in Japan for Mazda in the 1980s, he told me while we both rode the R train under Brooklyn, it wouldn’t even pay for new notebooks. People used old engineering drawings flipped over and bound together. The biggest surprise, Dean remembered, was that when Mazda opened up a studio in Irvine in 1988, Mazda didn’t even pay for janitorial staff to clean the office. Back in Japan, every day at 5 p.m., Dean recalled, brooms seemed to whoosh out of nowhere, everyone emptying out their own trash and sweeping their own floors, floor-by-floor. Just the same in Irvine for those first few months, employees were expected to clean the place themselves.

That was the company that was going to take on Lexus, BMW, Porsche. Do you think Mercedes engineers sweep their own floors?

Insiders look back on Amati and laugh, or cry. But at the time it was hotly anticipated, at least in the written word. Remember the automotive market of this era. America was still buying the Oldsmobile Cutlass Calais. The Porsche 911 was still a rattly niche car bought by people who had opinions on elbow patches. Japanese manufacturers were rushing so far ahead of anyone else it was like they had a time machine.

In 1989, Nissan brought back the Skyline GT-R, not seen in production since before the days of the 1973 Oil Crisis. And it gave the world the twin-turbo 300ZX, a Corvette that didn’t look like it was built in your uncle’s garage. Infiniti’s Q45 sedan was one of the most advanced vehicles in the world when it came out, and the Lexus LS had nearly unprecedented quality. In 1990, Honda (in America, Acura) took on Ferrari with the NSX. In 1991, Mazda won the 24 Hours of Le Mans, the first Japanese manufacturer to do so.

Japan was taking over the world, car by car.

In America, primarily in California, nearly 70 dealers signed up to sell Amati. It made sense. For decades, West Coasters had seen more and more Japanese cars and trucks dot the roads. They watched them go from rare and wheezy Toyopets and Datsun 1200s to futuristic Supras and King Cab pickups. What was so hard to believe about a Mazda luxury brand?

I’ll say again that it’d be easy just to call this pride. But there was real economic momentum behind it, and Mazda felt there was a solid business case for it in America. After all, Lexus was a success. Acura was a success. Infiniti was... well, it was trying stuff.

These were ventures that were bringing in a great deal of money and prestige. Mazda had every right, or every reason, to go after the luxury market at the time, no matter how much of a moonshot it looks like today.

Not that pride didn’t play into it, but I’m getting ahead of myself.

The Right Car At The Wrong Time: How Amati Died

“Bowing to economic pressure,” John O’Dell wrote in the Los Angeles Times in October, 1992, “Mazda Motors of America said Monday that it is canceling plans to launch its Amati luxury car division in the United States.”

O’Dell had the scoop, having spoken with his contacts within Amati the night before. Precisely 67 dealers had signed on to sell Amati here in the States, O’Dell reported, all of whom were going to be left in the lurch. Mazda had only just made the official announcement confirming Amati would exist in August of 1991. It had expected sales to start in 1994.

Amati had spent more time being anticipated than it had spent actually existing.

All the way back in 1989, Japanese Motor Business wrote that Mazda was going to take on Lexus and Infiniti, though it skeptically expected a V8 and not a V12:

Following the example set by Toyota and Nissan, Mazda acknowledged earlier this year that it is working on a V8 engine with the eventual possibility of producing a luxury car for the US market. Reports in the Japanese press suggested that Mazda would introduce a new V8 engine in the Eunos sales channel that it is to inaugurate this autumn in Japan. The Eunos channel will sell Citroën models, the new Miata MX-5 sports car and a model that has not yet been named, giving rise to speculation about a luxury car powered by a 4.0 liter V8 engine. Alternatively, the new model could be a high-tech three rotor Cosmo (based on the 929) which is expected to appear later this year.

The rationale behind the move into luxury was clear, the publication said. America and other Western countries with auto industries were putting in import restrictions that made it impossible to make a profit on cheap cars like the Corolla. With limited quotas of how many imported cars they could sell, Japanese companies had to not only manufacture stateside but also make more profit per car.

At scale it’s not considerably more expensive to make a Lexus LS than it is to make a Corolla, but you can charge a lot more for the Lexus:

Starting with high volume, low cost models in the 1970s, Japanese manufacturers have been moving progressively upmarket, a process which has been accelerated in recent years by the need to gain more revenue in countries which impose import restrictions and to circumvent the adverse effects of the strong yen.

What set Mazda apart was that it had been pressing harder on reaching into upmarket cars, as well as niche vehicles at home, more than any other Japanese carmaker. In Japan, Mazda ballooned from three to five different sales channels: Mazda, Autorama (selling Fords), Eunos (upscale models and also Citroëns), Autozam (low-cost kei cars in partnership with Suzuki), and the wonderfully-spelled ɛ̃fini (more luxury than the more sports-oriented Eunos).

Differently-branded dealership chains are pretty common in Japan, but little Mazda had built itself as many networks as Toyota, the number one in the market. This is all even leaving out Xedos, Mazda’s branding for its upmarket cars in Europe.

The company barely had the resources to engineer these cars, as Hall recalled, remembering quality control problems springing up as lineups ballooned. “The Flat Rock plant (where Mazda built Fords and Mazdas in the U.S. to skirt import restrictions) had huge problems. The build quality was awful. They blamed it on lazy Americans, then they started building the car in Japan. Same problems. The didn’t have sufficient staff to do all the variants.”

“It was simply out of control,” Mazda research executive Seita Kanai, a senior chassis engineer at the time told Wards Auto in 2006. “We were going in so many directions we damaged our brand. There was no consistency in what we were doing.”

That’s a harsh memory of how things went, but it’s fair to say that if there was a kind of car, Mazda tried to make it. A mid-engine microcar with gullwing doors? Yes. The Autozam AZ-1. A three-rotor luxury coupe with the world’s first in-car GPS navigation? Yes. The aforementioned Eunos Cosmo. Two different and distinct dedicated sports car platforms? Yes, and at the same time, with both the RX-7 (debuting the first sequential-turbo rotary production car in the final FD generation) and the phenomenon that was the first Roadster, aka the MX-5 Miata.

Mazda also started up “M2,” a boutique section of the company to produce ultra-low volume cars, almost like prototypes but sold to the public by the hundred or the dozen. (Cadillac did something similar with the 1950s Eldorados, but not to this extent.) Apparently internal politics kept this operation from reaching its full potential, but it still managed to produce a number of factory hot rods, all out of a retrofuture building in a hot Tokyo neighborhood.

There was a widebody V6 Shelby Cobra-style Miata as well as other track-prepped RX-7s and coupe-ified MX-5s. (You can read a full story on M2 at Japanese Nostalgic Car in two parts here and here.)

What’s shocking is that somehow, this all worked.

Mazda sales sprinted up to becoming the fourth biggest Japanese automaker, shooting past other smaller makes like Mitsubishi and Suzuki to rest only behind Honda, Nissan, and Toyota. Profits rose to levels that are hard to believe today. In a Japanese Motor Business news briefing from March 1989 detailing that Mazda was using Japan’s first private comms satellite to set up an information network among its dealers, the trade journal also reported a pre-tax profit of 31.2 billion yen, “a gain of 208 percent compared with the previous year.” Net profit was up a ludicrous 241.5 percent, with increased domestic sales first to thank. As Japan’s economy grew, domestic new car demand exploded as well. A new, empowered class of Japanese buyer wanted luxury and sports cars, and automakers were happy to deliver.

Mazda was buoyed by the booming Japanese economy, but this all cost. Mazda ended up rushing out more than two dozen models to fill out these dealer networks at a cost of $3 billion, as Wards Auto reported.

When Amati died, Japan’s Nihon Keizai Shimbun financial paper estimated that Mazda had spent no less than 50 billion Yen—about $400 million in 1992 dollars—on the brand. Australia’s Wheels magazine figured that was a conservative estimate, which is understandable given that it had two cars in development and was setting up a distinct dealership network across the United States as well as a setting up an advanced factory in Hofu to build the cars.

Thanks to conditions beyond its control, all Mazda’s investment was simply washed away.

The Bubble Burst: Why The Economy Collapsed And Killed The 1000

To learn more about how it all collapsed, I called up Roy C. Smith, an Emeritus Professor at NYU’s Stern School of Business. Smith watched the bubble rise, serving as President of Goldman Sachs International Corp through the 1980s. I had to ask a question I realized I never had a clear answer to: How did the bubble form, and why did it burst?

I mean, I started researching this story years ago, in the wake of the 2008 financial collapse. I’m part of a generation that takes it as a given that crashes come and go in the modern world economy, and didn’t think much more than Mazda played the right hand at the wrong time in the boom/bust cycle.

As Prof. Smith explained, the cause of the collapse was more clear than I thought, and the ensuing death of Amati was understandable in its context. Basically, nobody wanted to even think a crash was coming, and when the bubble did burst, it did so with such force that Japan all but closed down any investment, any risk, any car like Amati.

To understand that, you had to go back to the 1970s.

The Japanese economy was, to a greater degree than what we have here in America, centrally planned and supported by the government and its central bank. Hell, Mazda’s CEO going into the tail end of the Bubble Era took his post straight from the Ministry of International Trade and Industry (MITI), the government body that basically guided the country through its economic growth in the second half of the 20th century.

That growth, it shouldn’t be forgotten, was explosive in those years and decades immediately following the second World War, and it only first faltered in the wake of the 1973 Oil Crisis, a global shock. The Japanese economy went from double-digit growth year over year to merely single-digit growth. Half what it had been. Not bad, for sure, but not as good as it had been, and the Japanese government began to try to bring it back up to its old rate with government stimulus. To some degree, it worked, as growth started to climb again before the decade was over. But it wasn’t a real solution. The boom years after WWII were unprecedented. Trying to claw them back was unnatural, impossible.

“Japan, essentially, was experiencing a slowdown,” as Smith put it, simply as he could. The problem was that the rate that the government was trying to push it back to was creating what economists call excess liquidity.

It’s a weird term that means there’s more money in an economy than the economy knows what to do with. Some of the stimulus was getting absorbed with rising prices of consumer goods and services, but in small Japan, there just wasn’t enough. Prices started to balloon elsewhere, particularly in real estate and stocks. From the 1970s and on, prices there were rising unchecked, snowballing faster and faster until they were at an almost unthinkable rate of growth by the mid-1980s.

The head of Mazda’s American design team at the time, Tom Matano, regularly went back and forth across the Pacific back in those days, and laughed about it over the phone. I called him up at his office at the Academy of Art University in San Francisco’s School of Industrial Design. He could still remember the madcap momentum of the time, particularly in Japan.

“Every couple of years you’d go to Tokyo and it’d be totally different,” Matano said. “It was like they’d rebuilt the whole town. Almost anything would be possible, that mindset. We gotta jump on that wave to get good surfing,” as he put it. “Of course the bank would lend you any money for it, because the outlook was good.”

Anyone could borrow. Factories were expanding. Businesses were investing. Families were buying houses and stocks, financing with margin loans. Institutions were doing it, too. Hell, Mitsubishi bought Rockefeller Center in the heart of Manhattan. Companies were buying so many stocks that they were becoming mutual funds. A toilet seat company was trading at 80-90 times what it should have been, Smith remembered, one of the more wild cases of the Bubble Era financial landscape. Few homes in Japan would even have toilet seats, Smith almost laughed. The company was able to borrow money like mad, and it had to put that money somewhere.

What the company was doing was not reinvesting back into toilet seats; there was no need. It invested in stocks instead. Little surprise, then, that the Nikkei rose six-fold through the 1980s, trading at an average of 70 times earnings, as Smith pointed out in a 2014 blog.

The first warning sign came in late 1989. The world economy stuttered, the interest rate went up. It was because of “something in Germany,” Smith joked. He was talking about the start of the collapse of the Soviet Union.

“The [Japanese] government was trying to support the economy,” Smith said, stimulating things as much as possible, “but it couldn’t do it.” I don’t know if the government could have done more, or if it just didn’t think it was necessary. After all, stimulus had worked before. The economy had been growing and growing. Japan was taking over the world, remember?

“All vulnerabilities were,” Smith paused for a word. “Pooh-pooh’d.”

The economy was only expected to “slow gently in 1989 and 1990,” Japanese Motor Business not-so-presciently described in the fiscal roundup of its June 1989 issue. “There are some signs of overheating which should cool in 1990 as fiscal policy remains neutral.”

The scale of the crash, again, is hard to even describe. “The bottom fell out,” Smith told me. “They dropped 70 percent in a year.”

The country watched bankruptcies flare up where they’d be unthinkable just years, months before. Banks became incredibly cautious. “No risk,” Smith recalled. “Hunkered down.”

“It took two to three years,” he told me, “to even resolve the blast damage.”

So it wasn’t just that Mazda feared Amati would find no customers, it was that it would get no support financially back in Japan. The crash killed that. “It took all of the energy out of expansion,” as Smith said. “Sucked all the air out of the room.”

There was no credit going to a company like Mazda, Smith pointed out, but it might have made no difference. “Who were they going to sell cars to?” Smith said. Everyone who would’ve been able to afford one in Japan was either going bankrupt or watching another company fold.

In America, too, the economy was faltering. The Gulf War started with an oil crisis and ended with a post war recession, and the U.S. Federal Reserve didn’t do much to counter either, as the New York Times recalled towards the end of the mess in 1991. Economists were unsure if they were supposed to be fighting oil-related inflation or war-related recession. Either way, the economy stalled as businesses suffered and buyers became depressed.

If there was ever a wrong time to launch a risky luxury car, let alone an entire risky luxury car brand, this was it.

We Researched Ourselves Out Of A Job: What Working At Amati Was Like

“He had to fire 50 or 60 people,” Joni Gray told me over the phone. “I always thought it was the Japanese who decided. He said ‘No, it was me who decided.’”

Joni Gray, now editor-in-chief at Autobytel, worked as a marketing manager for Amati. She was talking about Dick Colliver, group vice president at Mazda North America. He was the one charged with launching Amati and the one who ran the whole operation.

Once you get beyond its secrecy, catching drawings and rumored specs in bits and pieces, once you see the Amati for what it was, it is still such a shock that Mazda, little Mazda, was trying to pull it off. As I found out talking to people who worked there, it was a shock to the company, itself, too.

“Dick Colliver said,” as Gray recalled, “‘Basically we researched ourselves out of a job.’”

Colliver was a true company man, having been at Mazda since the early 1970s. He actually went on to go to Honda not long after, working there as a top executive and running Acura for a time.

“I took everything to Honda,” Colliver told me over the phone. “All the systems we developed, Mazda didn’t do anything with them. They basically threw them in the dumpster. But I kept them in mind. The image programs, the database programs, the dealer development programs. The people, the processes, I took them with me.”

Colliver took Honda and Acura to years of record sales and lasted there until early 2009. Even in the wake of another global financial downturn, he had managed to get Honda’s sales up, the only automaker to do so, as the New York Times reported. If anybody could set up a car company and run it in hard times, it was Colliver. But keeping Amati going was an ask too great.

As Gray put it, there was a significant gulf between Mazda’s Japanese operations and its operations in America. This isn’t uncommon in overseas companies with American arms, but it was like nothing Gray has seen since, neither covering the industry as a journalist, nor at her later work at Hyundai and then Honda through the rest of the 1990s.

Mazda in Japan wanted Amati. Mazda in America had to figure out how it was all going to work, and sell. And even from the start, few in America thought it could. At least that’s how they talked to me about it years later.

“I came from Mazda,” Gray explained. “Just being there in the advertising department, even when they were just researching for Amati at the time. They really didn’t think the Amati, the Pegasus it was called, they didn’t think it would get funded. It was just a move to get Dick Colliver out to pasture. That was how it was rumored,” Gray sighed. “That was three years before we launched it.”

The way Gray talked about it, the more time people spent at the Amati program in the States, the more they became convinced it was too bold a project for Mazda. Everyone suspected that Mazda was seen as too basic a brand to support a luxury offshoot. Mazda and its 323s and 626s just didn’t have the strength to hold up against Lexus, Lincoln, Mercedes, whatever. Amati didn’t make sense, even to those charged with getting it going.

But there was a seed of doubt even in these pessimistic looks, as people within the company told me. Mazda itself defied what made sense. It had the Miata. That was an out-of-nowhere success, a higher-value kind of car than anyone would have expected from Mazda, a hugely profitable turn for the company that hinged on attractive design. Maybe the Amati was going to be the same.

“That was a time when there was ‘The Miata Effect,’” Gray explained. “If Tom Matano [in design] and Bob Hall [in product planning] would just do a bunch of redesigns with the look and feel of the Miata, we’d have the same effect that the Miata would have. And that would circulate across the whole American economy.”

So the Amati project continued, with Mazda halfway doubting its own judgement.

Amati planned out how its interiors would feel, how its branding would look, how its dealerships would be laid out. “It was pretty serious how much money they were willing to spend,” Gray said. “Having gone through now the sparse years, the spending was crazy.”

Hall remembered “unbelievably fantastic” proposals for Amati. Kimono silk interiors instead of leather. Lacquerware instead of wood.

Gray, by contrast, remembered those proposals stalling when they reached American focus groups that Amati put on. Silk became leather again, lacquerware turned back into wood. It all seems a little silly now, Mazda playing with money that wouldn’t last, Amati prepping a car that would never see production.

“We were right at the point when we had to commit $50 million to tool the bigger car,” Dick Colliver told me, driving to an appointment. I had gotten a hold of him through the company he set up in 2009: Amati Consulting. Colliver was optimistic about Amati as a concept.

“I think the project would’ve worked,” he said. “What got them was they tried to launch five cars at the same time.”

A lot of circumstances led to Amati shuttering, Colliver pointed out, the economy being a big part of it. But despite the ambition to go as big as it would as fast as it would, Mazda just wasn’t prepared, or able, to commit what it needed. When it came time to either go forward knowing Amati needed $50 million it wasn’t going to get, Dick pulled the plug. “I told them, it just wasn’t going to work.”

By 1993, Amati was dead. By 1995, Ford had a controlling stake of Mazda, increasing its shares up to 33.4 percent to rescue the company from complete financial ruin. Ford had stepped in before during past crises, taking a 24.5 percent stake in 1979, and having worked with Mazda since 1974. (The two began parting ways in the 2000s recession, then fully in 2015.)

Mazda wasn’t alone here, of course. All the Japanese automakers suffered. But it was only the middle of the 1990s, and the dreams of an independent Mazda that had seemed so Pegasus-like had turned Icarus instead.

The Heart Without A Body: A V12 Without A Car

In making my first round of calls for this story, I hit on a bit of a postscript to the Amati 1000, something I was never looking for.

Other people have written about Amati in the years since it shuttered, but I had never read anything about the kind of zombie afterlife it went through entirely behind closed doors.

To start, the economic boom was never as strong in America as it was in Japan, as Matano recalled, and a fissure had started to grow between the two sides of the company. As Mazda in Japan became more ambitious through the end of the 1980s and into the 1990s, Mazda in America became more cautious. And communication between the two was not as strong as it probably should have been. Many higher-ups sent over to America weren’t even strong English speakers, as Gray pointed out, which was unusual for a Japanese car company at the time.

Just how big the gap was became painfully clear after the official closure of Amati. Mazda in America discovered that Mazda in Japan had not only finished designing the V12 planned for the Amati 1000, it had tooled up for production.

While the American Amati team was thinking about designing dealerships and testing with focus groups, engineers in Japan were absolutely ready to go.

Tooling up a new engine is not a small cost, and it was one that needed to be recouped. Mazda tried to sell the engine to Jaguar, Hall told me. It was a conversation made possible through their shared connection of Ford. But Jaguar refused to put a Japanese engine into its British cars.

Unsure of what else to do, Mazda in the States drew up a four-door coupe to stuff the V12 into. “Mazda is in the shitter,” as Hall explained, “we’d better do something with it.”

Called Project A007, Mazda actually completed a full-size mock up of this car in 1994 with an interior buck as well. Its four doors were butterfly doors, and it had four seats, the two in the back mounted so low that the transmission tunnel acted as a center armrest.

Basically, Mazda designed the Aston Martin Rapide a decade and a half early. But it wouldn’t see the light of day.

“When Ford stepped in they basically flattened that,” Hall told me. “The RX-8 was the basic idea, but adapted to use FD [RX-7] components. The big V12 used the FD’s front and rear suspension and that was it.”

This was the final end. The Amati 1000 flagship was dead in ‘92 along with the whole brand, its engine dead a few years on.

Mazda, struggling on its own, got wrapped tighter into Ford and worked more and more as its development wing for economy cars and drivetrains. Mazda went from taking on the world on its own, to engineering global platforms for a parent automaker.

Only in the past few years has Mazda picked up its old sense of independence, with wonderfully strange high compression engines to show off as its own.

A Wound Not Yet Healed: Nobody Wants To Talk Amati

What was interesting in conducting all of these interviews was the sense not only of loss but also of defeat in so many of the people I talked to. We can think of the Amati as this incredible could-have-been car. Those who worked on it in America were more dismissive. Case, Matano, Gray and Hall all say Mazda never invested enough. Mazda didn’t plan enough. Mazda didn’t have the cachet or the funding to pull it off, even if it tried to do things right.

Mazda’s people here in the States weren’t able to even connect me with someone in Japan. Amati is a touchy subject over there, I was told.

Again, it’s why you won’t find info or pics on Mazda’s website or press materials. You won’t even find Amati stuff at Mazda’s corporate museum, John O’Dell told me.

O’Dell was that reporter at the LA Times who broke the story that Amati was shutting down. I asked him if he got the sense, talking to people at Amati then at Mazda later, if it still felt sore.

“Well, yeah!” he said. “Face is big in the Japanese culture. It was a loss of face. To decide to do that, to make announcements, to tell the world you’d do it, and then not be able to... I think it would be an open wound to an American company. Edsel is still an open wound” to Ford, he said.

But even O’Dell had grown skeptical, too. Mazda was such an economy company in the States. He, like others I talked to who were inside Mazda at the time, feared that Amati was too great a leap for Mazda, that its reputation couldn’t support it.

“They look at what Lexus did for Toyota,” O’Dell told me, “But I think it would have killed them if they’d done it. To have to support a second channel with separate dealers. In the ’90s to 2002? They would have ended up like Mitsubishi and damn near disappearing. Like Daihatsu.”

Even The Perfect Car Company Can Go Broke: The Amati Legacy

I went into this Amati story looking to do a kind of business report with pretty pictures. I wanted to explain that a car company can do everything “right” by car enthusiasts and still almost go broke. I wanted to show that Amati seemed like such a smart move at the time, only to be undone by an economic crash outside Mazda’s control.

But I didn’t expect to hear so many of those who worked on Amati to denounce it, to say it was foolish, or doomed.

After all of that, I don’t know if I can believe them all. The power of the Amati fantasy is so strong, I still feel the light of Amati has dimmed to those who touched it, reduced by decades of, I don’t know, disgrace is too strong a word. There’s still hurt there.

Maybe it will eventually heal, and Mazda will pick back up some optimism and pride about their V12 flagship of the Bubble Era. Maybe then we’ll finally see the 1000 in all of its glory, as big as our imagination.

In 2016, those who made up the American staff of Amati threw a 25th anniversary party for the marque they’d worked for but never got to see go on sale. Old employees got together, about 20 of them, bringing along various pieces of official Amati gear.

This was stuff they’d been given in the years when they worked for the car company that never launched, accessories meant to sold in Amati dealerships. A hat with the W-shaped Amati logo. A polo with “Pegasus” embroidered in gold. A full set of Amati golf clubs didn’t make it out, but someone at least kept a picture.

Joni Gray was there. Dick Colliver was there, too. At the reunion, he cried.