The conglomerate was supposed to be dead, a relic of a bygone of era of corporate America. Investors, we have been repeatedly told, want smaller, nimbler, more focused companies.
And yet there is Amazon.
Just when it seemed that sprawling empire building had gone out of vogue — eulogies were written last week for General Electric after Jeffrey Immelt, its chief executive, retired under pressure from shareholders — Amazon announced that it was buying Whole Foods for $13.4 billion.
The deal will put Amazon in the brick-and-mortar grocery store business, which will exist alongside an ever-increasing bevy of disparate interests: the selling of everything from electronics to toothpaste online; payments and credit; cloud computing; production and distribution of movies and television programming; book publishing; shipping and logistics operations; and on and on.
It is actually a myth that conglomerates disappeared. They are now just dressed up with a bit of Silicon Valley flair, and dress down in the boardroom, with chief executives who wear sneakers.
Amazon is just one of these new-economy conglomerates. Alphabet, the parent company of Google, is another. Facebook is quickly becoming a conglomerate, too.
Michael C. Jensen, a Nobel laureate and professor emeritus at Harvard Business School, famously — and successfully — made the case in the 1970s and 1980s that conglomerates like RJR, which owned tobacco and food brands like Nabisco, wasted "billions in unproductive capital expenditures and organizational inefficiencies."
That is very likely true of today's tech-enabled conglomerates, too, which are spending, and often losing, tens of billions of dollars annually on all sorts of projects and acquisitions that may or may not turn out to be successful. But investors are seemingly willing to give these new behemoths a free pass in the name of growth and innovation — until they aren't.
If there is any lesson from the last breed of industrial conglomerates, it is that there is a natural life cycle to most of them.
The model begins like this: A company that is successful in one area turns itself into a conglomerate by using its free cash flow to finance the development or acquisition of businesses in other areas — at first, ones that are similar to their current business, and later often ones that are farther afield. And then the company does this again and again.
When such an economic machine works, it works extraordinarily well. But when any one of the major levers in the machine breaks or even stalls, the entire enterprise comes under pressure. Shareholders start complaining that the sum of the parts would be worth more separately than together.
"You look at companies that got really big in the world, the record is not very good," Charles T. Munger, vice chairman of Berkshire Hathaway — the world's largest conglomerate — told investors several years ago.
His business partner, Warren Buffett, stunned shareholders this year when he suggested that he expected shares of Berkshire to rise immediately after his death because of speculation that the company would be broken up and thus would be worth more. (He and Mr. Munger both believe that Berkshire is better off intact, but Mr. Buffett thinks investors' knee-jerk reaction will be to believe the opposite.)
When it comes to Amazon (or Alphabet, or any of the new conglomerates), the question is whether there is something fundamentally different about these businesses given their grounding in digital information — especially as they expand into complex brick-and-mortar operations like upscale supermarkets.
In an age of big data and artificial intelligence, are businesses that look disparate really similar? And can one company's leadership really oversee so many different businesses? When does it become too big to manage?
Google's own internal list of top-10 principles seems to include an anti-conglomerate provision: "It's best to do one thing really, really well."
Leaving aside the question of whether that maxim may have been more suitable to the Google of a decade ago, it certainly hasn't stopped the company from jumping into all sorts of businesses outside its flagship search and advertising business. Some of these businesses — which include Android, YouTube, Waze, Nest Labs, self-driving cars and internet distribution — have been more successful than others. Most of these were brought under the Google umbrella through acquisition, evidence of how the company has used the enormous proceeds of its advertising business to subsidize its entrance into all sorts of other enterprises.
Facebook, too, is expanding its offerings through acquisition, buying Instagram, WhatsApp and Oculus VR. And for now, there are no obvious signs that looking after a wider array of businesses is weakening the management at Alphabet, Facebook or, for that matter, Amazon.
In fact, a recent article in the Yale Law Journal made a compelling case that Amazon has built perhaps the ultimate economic mousetrap — one impervious to the natural life cycle of a conglomerate, but one that might ultimately prove to be anticompetitive.
The author, Lina M. Khan, a Yale Law student who has written about antitrust law and competition policy, argued that Amazon had created a "platform market" and can use its size and scale to subsidize its entrance into new businesses through predatory pricing.
"The economics of platform markets create incentives for a company to pursue growth over profits, a strategy that investors have rewarded," Ms. Khan wrote. "Under these conditions, predatory pricing becomes highly rational — even as existing doctrine treats it as irrational and therefore implausible."
Even more provocatively, she contends that Amazon's role as both a distributor and cloud provider for many of its competitors gives it an unfair advantage. "This dual role also enables a platform to exploit information collected on companies using its services to undermine them as competitors," Ms. Khan wrote.
Amazon has thus far been left alone by regulators because it has helped reduce prices of most products. It is not a natural monopoly. "Can prices ever be 'too low'?" the Federal Trade Commission asks on its website. "The short answer is yes, but not very often."
On its webpage about predatory pricing, the agency explains how the government thinks about competition. "A firm's independent decision to reduce prices to a level below its own costs does not necessarily injure competition, and, in fact, may simply reflect particularly vigorous competition."
That may be true. But in Amazon's case, it has the potential to become so dominant in so many areas that its impact could be more than simply lowering prices for consumers; it could put large companies out of business.
Were that to happen, this new breed of Silicon Valley conglomerates may become more powerful — and resilient — than the 20th-century conglomerates of yore.
If Amazon were to use its acquisition of Whole Foods, along with the profits from its other businesses, to lower prices so much that it put out of business a company like Walmart, with its 1.5 million United States employees, would that be a good outcome? Or is that just a natural part of capitalism?
The view of Amazon's chief executive, Jeff Bezos, is clear. The man who is assembling the 21st century's most fearsome new conglomerate once explained his view of competition this way: "Your margin is my opportunity."