Manufacturing scale always has been critically important in the automotive industry. Building cars requires so much upfront capital investment that it practically forces automakers to build as many cars as they can. It’s a whole lot easier to spread a $1 billion investment over 250,000 cars a year than 50,000 a year. And yet, we’re fast approaching a future where being the biggest will be a disadvantage.

A century ago automakers such as Ford and General Motors set out to become as vertically integrated as possible. Because they were the largest automakers in the world, making most of their parts and components in-house gave them an advantage in scale that no one else could match.

Half a century ago, Volkswagen decided to concentrate all of its efforts on building one model, the Beetle, to provide it with the greatest scale it could muster. That strategy worked well for decades.

Today automakers are achieving scale by sharing common architectures, powertrains and components across all the different models and brands they own. And they are acquiring competitors; much like Renault-Nissan just did by getting Mitsubishi. R-N-M will probably finish out the year as the largest automaker in the world, giving it greater scale than anyone else.

But we’re fast approaching “Peak Auto” where global auto sales are going to plateau and start declining. In fact, we’ve already hit “Peak Auto” in the U.S. market. A smaller percentage of the American public is buying new cars, and a declining percentage is getting a driver’s license. Sales probably will never again exceed the record they set in 2016.

Several studies predict a big drop off in the seasonally adjusted annual rate in the U.S. market by the mid-2020s, down to only 13 million units. Merrill Lynch predicts it will fall this far due to so many cars coming off lease every year. It predicts we will see 5 million off-lease cars flooding the market every year, driving down prices of used cars, which in turn will cripple new car sales.

Others, like Barclays or RethinkX, forecast sales will plummet as more and more consumers decide to stop owning cars and instead start buying their mobility on an as-needed basis. The emergence of urban aviation, using vertical takeoff and landing craft to transport people and cargo, only will accelerate this trend.

We can debate whether or not this impact will hit in the next decade, the decade after that or the one after that. But it’s coming. And most automakers are unprepared for this eventuality.

But at least one of them is getting ready. GM is bragging to analysts it still can deliver a net profit of $4 billion to $6 billion even if U.S. sales drop by 25%. It says it can break even with a SAAR of only 10 million to 11 million vehicles. More automakers need to take a serious look at their manufacturing footprint and breakeven point.

While scale played such a critical role for over a century, the advantage may now gravitate towards automakers that are profitable with much smaller manufacturing footprints. For example, Mazda is solidly profitable selling only 1.6 million cars a year globally. Subaru is profitable selling only 1 million.

In fact, Tesla probably will have a far easier time ramping up its production to 500,000 vehicles a year than a GM would have trying to downsize to only 6 million a year.

In the U.S. today there are 15 OEMs with 36 different brands selling 310 different models. And that does not include the thousands of trim line permutations they offer. Does the market need so much choice? No it does not. And if you believe that markets are rational, then something’s got to give.

Like the old saying goes, “the bigger they are, the harder they’ll fall.”