A New York Stock Exchange screen shows the results of the Twitter Inc. IPO in New York, Nov. 7, 2013
Lucas Jackson—Reuters
April 9, 2014 2:07 PM EDT

The envelope, postmarked Franklin, Tenn., contained excellent news: an “emerging opportunity” in a cheap Internet stock. “With a $10 million warchest,” read the glossy flyer inside, “LiveDeal is crushing Groupon in pilot markets!” Further down, in fine print, was more salient information: “The use of research… is done at your own risk.”

Nevermind that LiveDeal is an Internet stock so revolutionary its virtues deserve to be heralded by snail mail. Or that Las Vegas-based LiveDeal has not $10 million but $500,000 in cash, posted a loss twice as large as its revenue last year, and yet trades at 37 times its sales. What matters about this flyer is that I’ve received junk mail just like it before – but not for about 15 years.

Fifteen years ago, we were in a massive tech-stock bubble. Some people are pointing to evidence we’re already in another tech bubble now. Others are certain we’re nowhere near the insane investing we saw in 1999. It’s possible they can both be right.

But I also think neither side of the debate is addressing the real concern, which is that we have been engaging in the very financial behavior that creates and nurtures bubbles. Behavior like complacency toward irrational valuations and explaining away the threat of a bubble. And that makes the bubble deniers more dangerous right now. Because they are making sensible arguments that overlook the risk of irrational investing becoming the norm.

Those who pooh-pooh a bubble often rely on the same arguments. But each one involves a fallacy. Here are some of the more common.

It’s not at all like the dot-com bubble. This is true as far as it goes. Last month, the Nasdaq Composite hit its highest level in 14 years but is still 26% below its high. The IPO market hasn’t been this busy since 2000, yet still shy of record levels. And companies going public have stronger financials, if many are still losing money.

But bubbles don’t repeat in a similar pattern, so we won’t see the same signs of craziness. Nobody’s dumb enough to use a sock puppet as a mascot, or throw a lavish rooftop party. But that doesn’t mean we’re not seeing signs of excess, like parties that aren’t called parties. Other behaviors are consistent, however, like dusting off dot-com metrics like revenue-per-user.

VCs are showing discipline this time. Again, true enough. Historical data shows venture investments are, at most, creeping higher. But they also creeped higher through 1998 before exploding in 1999 and 2000. The explosion came as institutional investors were desperate to invest in VC funds, and if first-tier firms wouldn’t create them, then second- and third-tier firms did.

Venture firms that survived the dot-com bust did so by investing relatively prudently. But once the more promising dot-com IPOs began hitting the market with huge first-day pops in price, less prudent VCs began recklessly funding questionable startups. That’s when things really got cooking.

Tech companies going public are a different breed. So far, yes. And some of them will see the kind of success that Amazon has enjoyed. But not all of them. Once an IPO market heats up, investment banks line up the strongest candidates first, ensuring early pops to prime the market for more. That stokes demand for more IPOs but after a few quarters the candidates aren’t so hot. Box, Airbnb, Alibaba look great. But not every tech IPO can look that good.

The Internet is kicking into high gear, creating opportunities for business models that never existed before. But it’s much easier to spot big trends than to pick individual winners. And besides, there is something dangerous about pushing this idea too far. It’s just a little shy of saying, it’s different this time. That, of course, is the motto of nearly every financial bubble.

Growth today is more important than profits today. This is a slippery argument, and over time a risky one. Again, Amazon thrived despite early losses, and a few like Box are likely to do so as well. But two things: First, it’s is very difficult to pull this trick off for long – Amazon is the exception proving the rule.

Second, there is ample evidence that money-losing startups that went public in recent years have fared poorly. Skype had trouble going public with its losses in 2011, and others like Jive Software, Brightcove, Groupon all went public with losses and are trading below their offering prices. That’s the rule of money-losing IPOs – and the reason why, in sensible times, investors avoid them.

Only a few stocks are overvalued. Many others aren’t. That’s because many of the big-tech giants like IBM, HP and even Microsoft are struggling with incumbent businesses being displaced by younger technologies. What we’ve called the tech industry has been divided into the thriving and the slowly dying, and investors can easily tell them apart. Besides, a bubble doesn’t have to be broad-based to burn investors when it crashes.

But the market has been going down! Even the most sturdy rallies never go up in a straight line. In fact, they are marked by volatile pullbacks. Remember the mini-crash of October 1997? Far from derailing a nascent bubble in technology stocks, it acted as a coiled spring to send it higher. Also, note that even as the broader market has been slumping, recent IPOs like Grubhub and Oopower have been rising.

In short, the market for tech stocks is not in a bubble like the dot-com or real estate bubbles of the past two decades. But it may well be in the early stages of a bubble marked by irrational investments, a bubble that could easily expand out of control if smart people keep rationalizing away the early warning signs.

And there are warning signs, whether it’s a growing tolerance of insanely priced IPOs and M&A deals, or the return of spurious metrics like revenue per user, or even a piece of junk mail touting an unwise stock investment. Such early signs are kindling that, if left to gather, can make for a bonfire later on. No, it’s not 1999 at all. But it may well be 1998 – or something a lot like it.

Contact us at letters@time.com.

Read More From TIME

Related Stories