Today, The Donald is considered a political force, a charismatic populist who has tapped a deep vein of frustration with the gridlock entrenched by waffling career politicians. So, if you believe the experts, Trump probably won’t win the nomination, but he’ll maintain strong support, probably in the 20% range, all the way to the Republican convention in July 2016. Get used to it, folks, goes the drumbeat, Trump is here to stay.

Yet for students of financial markets, the Trump phenomenon strongly resembles a bubble. A stock market craze can go on for a long time before it fizzles; a lot longer than Trump’s sudden rise. So, by stock market standards, his durability doesn’t mean he’s here to stay at all. In fact, you never know how long a bubble will last; only that it’s bound to pop.

Of course, political and financial markets operate differently. The Trump candidacy may not be a bubble at all. But investors do behave crazily at times, and so do voters. And the same psychology that drives shares to unsustainable heights appears to be elevating Trump: the view that the “world has changed,” and that once a bandwagon gains this much momentum, it just has to keep rolling. Anyone who has followed past financial frenzies wouldn’t mistake a sudden spike for a durable trend. Using stocks as a good guide, the Trump candidacy is far more likely to be a bubble than the prognosticators are telling us.

Just for fun, let’s compare Trump’s rise with the turn-of-the-century Nasdaq bubble. On June 30, Trump was languishing in 8th place, at 4.2%, in RealClearPolitics’ poll average. By August 6, he had jumped to a wide lead at 24%; he remains near that level now. Hence, Trump’s support multiplied a meteoric five times in five weeks.

Almost two decades earlier, investors’ infatuation with tech stocks sent the Nasdaq on a record ride that, by stock market standards, amounted to a Trump-like sprint. From April 1997 to March 2000, the index leapt from 1194 to 5132, or 4.3 times. The Nasdaq then famously retreated to 1100 by October 2002, a drop of 79%. It took more than 12 long years, until March 2015, for the Nasdaq to breach 5000 once again.

The Nasdaq frenzy, and most financial bubbles, share three characteristics with Trump’s rise. The first is the “extrapolation,” or “momentum,” effect. It happens with stocks all the time. What goes up often keeps going up, what’s sexy stays sexy—for a while. A herd mentality drives investors. In our continuous news cycle, political predictions follow the same dynamic. We assume that what is happening today will keep happening tomorrow.

The second factor is the “world has changed” effect. During the tech bubble, we were assured that the Internet would create never-before-witnessed earnings growth and that tech stocks thus deserved price-to-earnings multiples of more than 100. Today, we’re hearing that although stocks look overpriced by most conventional measures, they’re really a bargain because of permanently low “real” interest rates, a highly questionable conclusion. In politics, the new theory is that the Republican Party is suffering a durable, structural split between voters supporting the maverick Trump and those backing the mainstream candidates. The “this time it’s different” view goes contends that an amateur politician can sell radical views on immigration and trade on the strength of a flamboyant, in-your-face persona.

The third issue is the “not paying close attention” phenomenon. In a stock market bubble, investors are driven by enthusiasm, not numbers. They follow the crowd instead of soberly assessing how fast earnings really need to grow in order to justify the explosion in prices. Similarly, at this early stage in the campaign, it’s more about showmanship than fundamentals. It isn’t clear voters are yet weighing the true worth of the candidates, measured by their policies for restoring economic growth, ensuring national security, and positions on thorny issues such as immigration. When they need to start paying close attention, will voters really be able to picture a leader as different, some would say outrageous, as Donald Trump in the White House?

To unite politics and stocks, let’s create a kind of price-to-earnings (P/E) ratio for the candidates. In finance, the P represents, say, the S&P’s current price, and the E its recent earnings. If the P/E is extremely high, the P, or price, can still keep rising if the E increases at an extremely rapid rate. If not, it could be a bubble, and the bubble will pop with a steep fall in prices.

Here’s our political P/E, applied to Trump. His poll numbers—the P—far exceed those of his rivals. So the P is exceedingly high. What about the E? Let’s say it stands for “excellence,” or the intrinsic worth of the candidate based on their positions, credibility, and magnetism. So far, Trump’s E is hard to assess. By conventional measures, it would stand below candidates with substantial records of accomplishment, like governors and senators. So Trump’s P/E is most likely really, really lofty.

One of two things needs to happen. Either Trump’s E, or “excellence” factor, his enduring appeal as a potential president, must expand rapidly. Or this is a bubble, and his poll numbers need to fall sharply.

Of course, many great companies not only fulfilled but exceeded high, seemingly impossible expectations, including Google and Apple. Keep in mind that those are new economy titans that virtually invented their industries. By contrast, politics is an old economy game, and the Grand Old Party epitomizes the oldest of the old style. In politics, the polished professionals almost always win. That’s not The Donald. Trump’s candidacy is not a surefire bubble just yet. But if Donald Trump were a stock, a really expensive stock, you’d have worry that these shares are way overvalued.