To call Bitcoin the biggest and most obvious bubble in modern history may be a disservice to its surreality.

The price of bitcoin has doubled four times this year. In early January, one bitcoin was worth about $1,000. By May, it hit $2,000. In June, it breached $4,000. By Thanksgiving, it was $8,000. Two weeks later, it was $16,000.

This astronomical trajectory might make sense for a new public company with accelerating profits. Bitcoin, however, has no profits. It’s not even a company. It is a digital encrypted currency running on a decentralized network of computers around the world. Ordinary currencies, like the U.S. dollar, don’t double in value by the month, unless there’s a historic deflationary crisis, like the Panic of 1837. Instead, bitcoin’s behavior more resembles that of a collectible frenzy, like Beanie Babies in the late 1990s.

But defining and identifying bubbles is harder than it seems (kind of like defining bitcoin). The term technically refers to an asset whose price dramatically exceeds its intrinsic value. But who determines price and value, anyway? Those aren’t scientific concepts with formulas, like gravity or the length of a hypotenuse. They are the co-creation of buyers and sellers whose needs and attitudes are constantly changing.

Sometimes, spotting a bubble is very easy. Imagine three public companies that make shoe leather—Derek Leather, Inc., Joe Leather, Inc., and Becca Leather, Inc.—with the exact same revenue, expenses, talent pool, and customer demographic. Let’s say the market caps for all three companies start the year at $1 billion and Derek Leather and Joe Leather don’t appreciate; meanwhile, the public valuation of Becca Leather climbs to $2 billion, then doubles to $4 billion in a month, and then doubles again to $8 billion in the following week. It would be pretty clear that Becca Lather’s valuation makes no sense with an apples-to-apples comparison to Derek and Joe.

But what happens when an entire industry is a bubble? It becomes harder to make an apples-to-apples comparison, since the entire sector is an incomparable fruit. A good example would be early Internet companies whose valuations soared in the late 1990s and crashed in the dot-com bubble. For years, Internet bulls defended the stock prices of companies like Pets.com by arguing that, due to the rising digitization of the economy and the global nature of the Internet, user growth was a more significant proof of value than old-fangled metrics like profit or revenue. Eventually, a combination of factors—the failure of some large Internet companies, changes to the tax code, rising interest rates, and venture capital exhaustion—contributed to the big pop.

In a way, the emergence of cryptocurrencies is akin the dot-com era, because there is no perfect comparison to illuminate the “real” value of something like bitcoin. It’s a currency (like the dollar), whose owners consider it a long-term store of value (like silver), which is appreciating as if it were a faddish collectible (like a Beanie Baby), and is running on a blockchain platform, which some insist could change the future of everything from legal titles to daily payments (like Internet). How can one be so sure bitcoin is a bubble if we don’t even know what the proper comparison is—dollars, silver, Beanie Babies, or the Internet?

In their great 1982 paper "Bubbles, Rational Expectations, and Financial Markets," the economists Olivier Blanchard and Mark Watson explain why gold is susceptible to bubbles. It’s an explanation that sheds light on the bitcoin frenzy, too. Gold, like bitcoin, is not a company. There are no financial reports, and its investors will never receive dividends. Instead, there are at least two big reasons to invest in gold. First, goldbugs want a hedge against an economic catastrophe or inflation. Second, some people invest in gold simply because they see the price of gold going up. Such an investor “bases his choice of whether or not to hold the asset on the basis of past actual returns rather than on the basis of market fundamentals,” Blanchard and Watson write. In other words, the investor’s story is: The price will go up, because … well, it just went up!

These investors buy gold, not because of any fundamental economic insight or any analysis of value, but rather because they want to catch the train. They see the price rising and they assume they can buy gold, hold onto it as it appreciates, and then offload it to some greater fool before its value declines. In fact, the economic term for this sort of irrational belief is called “greater fool theory.”

Bitcoin is turning into a gaggle of greater fools. Retail investors are jumping into the market to buy bitcoin, in the expectation that they will be able to sell their investments for cash to some other sucker later on. In November, Bloomberg reported that “buy bitcoin” had overtaken “buy gold” as an online search phrase. In December, bitcoin platforms soared up the app charts. Coinbase, an online broker where people can buy cryptocurrencies, is now the top trending app in the Apple App Store. Two similar platforms to oversee cryptocurrency accounts, Gdax and Bitcoin Wallet, are now fifth and eighth on the trending charts.

For the people downloading these apps, bitcoin probably isn’t a philosophical bet on the future of money and society’s relationship to the government, says Christian Catalini, a professor of technology at MIT Sloan School of Management, whom I have spoken to often about bitcoin. “There is a speculative frenzy among retail investors who just want to make a quick buck and the App Store is pretty clear evidence of that,” he said.

There is another important feature of the bitcoin market that could both explain its high valuation and suggest an imminent correction. The crypto market is insanely concentrated. Approximately 1,000 people own 40 percent of all bitcoin in circulation, according to Bloomberg. Just 100 accounts control 17 percent of the market. Many of these accounts have held bitcoin for years because they believe fervently in its value. But if a handful of them sell even a small portion of their shares, it could dramatically move bitcoin’s price, potentially triggering a massive correction, as retail investors (who only bought in because the price was going up) try to sell en masse to avoid losing all of their money. There is an upside to this concentration, however, which is minimal contagion effects. If the bitcoin bubble crashes, it likely won’t spill out into the general economy, like the subprime mortgage crisis did one decade ago.

Smaller bitcoin bubbles have inflated and deflated before, without any macroeconomic effect. In 2011, the price rose from $1 to $30 and then crashed back to $2 all within the same year. “I wouldn’t be surprised with another crash, followed by another growth in line with transactions,” Catalini said. Indeed, the dot-com bubble was an unambiguous frenzy of speculation and financial malpractice. But 15 years later, many of the business propositions that flamed out spectacularly were reincarnated as successful companies. Chewy.com, essentially a modern incarnation of Pets.com, sold for $3 billion earlier this year.

Fifteen years from now, the blockchain, too, might be an integral infrastructure for the digital world. In this hypothetical world of 2033, bitcoin at $16,000 might be an absolute steal. But we don’t live in “hypothetical-world 2033.” This is still real-world 2017. And bitcoin’s last few weeks are the real-world definition of a speculative bubble.