Ask HN: In plain English what caused the dot com bubble to crash?

Everyone correctly predicted the "New Economy" would change business forever, about 10 years early. Venture capitalists frantically rushed to fund ".com" ideas with billions of collective dollars. To get their money back, the VCs hired investment banks to sell their shares to the public. Those same investment banks had for many years made "buy/sell/hold" recommendations reliably on established companies. Thus, when those investment banks inevitably gave the same .com's they were selling "buy" ratings, the public trusted those ratings. The public was also overcome with daily reports of the hyperbolic growth in the value of these public stocks. These reports fueled the interest in such stocks which caused investors to drive the prices for the stocks higher in a vicious cycle. However, despite such growth in the price of the stocks, and the "buy" recommendations of the investment banks, many of the stocks had business models that could not even theoretically operate at a profit. No matter how many widgets they sold, they lost more money. Another group of these businesses were aimed at providing services TO the "New Economy", ".com" companies.

Eventually, the money from these funding rounds dried up. One-by-one the flawed business models were revealed, and companies began to go bankrupt. As businesses with even viable models relied heavily on the "New Economy" .com's as a customer base, they, too were driven to bankruptcy. Some period after the stocks were sold to the public, investment banks were also allowed to change their ratings, and often did, to "sell". News reports of the ".com" crash began, and sent the public running for the exits. All of these factors combined for another vicious cycle in the value of ".com" stocks in the opposite direction.

ELI5 style; basically there were a number of hype-driven valuations that were way too high, and when the resulting IPO happened, a lot of people lost a lot of money. When this happened enough times, investors got scared of technology and swung the other way.

Pets.com [1] is a great example of this. They spent 11.3 million on marketing despite having a half-million in sales for their first year. Then they IPO'ed way too early, and their initial stock at $11 dollars fell to 50 cents six months later.

We collectively didn't really understand how a lot of internet business models were supposed to work, and that caused a lot of non-viable businesses to float on hype. When the hype went away, all these businesses got culled, and only a few survived (like Amazon). Some of them were just too early for their time, and modern iterations of the same business models were able to take advantage of key differences like the prevalence of mobile devices.

In a sense, its a natural human response to new paradigms and technologies. Internet business valuations from mid-1990's to today follows almost perfectly the hype curve [2] model.

[1]: https://en.wikipedia.org/wiki/Pets.com

[2]: https://en.wikipedia.org/wiki/Hype_cycle

Not to argue with the other interpretations, but fraud was a factor too. Enron, MCI WorldCom and Lucent all perpetrated accounting fraud and put a lot of stress on competitors.

In a lot of ways all bubbles are alike, you could hear echos of the 1920s in the 1960s and of the 1960s in the 1990s. "The money game" and it's sequel "supermoney" tells a great story of the 1960s bubble that is better than anything I've seen about the 1990s.

People paid a very high price to buy companies that weren't really worth that much. When it became clear the companies weren't worth as much as previously thought, everyone lost thier money.

A share just means buying a small part of a company. For example I can buy some small fraction of Apple for around $140 today. Owning a piece of the company entitles me to vote on certain things like who leads the company, and also to get a share of their profits. The share price is basically set by adding up all the future earnings of a company and applying some discount. The fact that you can't really know what future earnings is what makes buying a stock inherently risky.

Many of these comments make it seem like people back then didn't realize it was a bubble. I beg to differ.

I sold my first startup to a publicly traded company in 1999, in an all-stock deal. I remember thinking back then that we were definitely in a bubble, and there was no way it was going to last. The big question was how long it could keep going.

I sold the startup for 10x what I thought it was worth, knowing that the market would probably crash before my stock became free-trading (there was a 1 year hold).

Lo and behold, the market crashed about 3 weeks before my stock became freely trading, I lost 75% of my net worth. It was pretty devastating, but I still made out with more than I expected.

So, the whole point of this story was to tell you that - at least for me and a lot of people I knew, we all knew it was all a big bubble. The question was how long the irrationality could keep going.

The "crash" wasn't a day-long, week-long, or even month-long stock pricing event, but lasted 6+ months from the first tech sector to the last going through the crash.

The first stocks to get hit were the web 1.0 properties in Q1'00 (use Amazon (ticker: AMZN) or Yahoo (YHOO) as a proxy as there aren't many left standing that are easy to find historical prices for) and amongst the last were the comm equipment companies which didn't peak until 6-9 months later (e.g., Cisco (CSCO) or Ciena(CIEN)).

All of this is completely separate from whether any of these companies had viable business models. Some did, like the telecoms, their suppliers, semiconductors, enterprise software. Some did not, like pets.com.

A very small number of dot-coms survived, and ultimately thrived after the crash (AMZN). Though you would have had to have enormous foresight to properly select your investments during the boom to not lose a ton of money if you bought in this frothy period.

All of the company fundamentals were well known at the time. It's not like the enormous cash burns of certain types of businesses was hidden from view. It was a badge of honor at the time (also true today in a few places). Some of that cash made it into other businesses that had been around for almost 100 years, like HP or Lucent, but juiced them, and they suffered withdrawal symptoms when their customers stopped spending (or even paying their bills).

To a fundamental investor, the perplexing thing about the whole era was looking at one of many not particularly unique companies that was trading at 25x annual sales, growing 50% a year, and losing money (which is a ludicrous valuation for a public company). Yet you'd find it at 40x three months later.

To get back to your question, having been close to the situation, I can't point to a specific event that started the cascade, but once it started, there was no stopping it. To a fundamental investor in early 2000, the outcome was perceived as being inevitable, but impossible to predict when it would happen. If you'd asked them at the time, they would have told you that the tech market had detached from fundamental behavior back in 1997/98 earlier with no signs of slowing down.

As an aside, if you were willing to look outside tech, it was an amazing buying opportunity in pretty much every other sector of the stock market. Industrials, energy, commodities, real estate, and so on were trading at discount prices even in the midst of this enormous bull market in tech/telecom and related stocks. Unfortunately that is not the case today and it's hard to find any sector that could be considered cheap.

In 1997, congress lowered the tax on capital gains making non-dividend-paying stocks cheaper than dividend-paying stocks. It made it very attractive to invest in companies that didn't pay dividends like tech company IPOs which were popular at the time because of the growth of computers and the internet. There was a boom of personal investing as average people were quitting their jobs to become day traders. Stocks were rising uncorrelated with business fundamentals but because of the speculation of them increasing even more in the future.

The "crash" happened when some large companies became too nervous with the overvaluation and put in large sell orders for their own stock. This caused everyone else to panic worried that the stocks they owned were worthless and that they would lose everything if they continued to hold onto them.

The bubble "burst" in 2003 when congress changed the tax laws so that capital gains was taxed at the same rate again.

As I often do, I'll be a bit of a contrarian and take a different angle.

People believed the dot com bullshit because amazing technology gains were being made by technology. In 1990 you still had typing pools and paper and manual processes.

The mid 90s were when everything got computerized. People were making millions rolling out PCs and email. Lotus Notes was transforming business.

People saw the internet as the next big thing, which it was. But the technology wasn't there yet and the ability to conduct meaningful commerce was undermined by encryption policy.

There are already much more detailed answers than this one is going to be but: all bubbles pop. The very nature of the bubble is people are walking around with assets that are misvalued. The greatest trick in finance though, isn't spotting overvaluation: it's figuring out when reality is going to kick in. Every time a price defies gravity (housing, tech stocks, the UK economy), there's plenty of people who see the crash coming. But timing is everything.

In a nutshell: Unrealistic, contrived business models at way too high company valuations.

Lies, which couldn't be sustained forever, as in most bubbles. The most common kind of lie in this era was advertising sales fraud, in which ad sales (and rates) were wildly inflated by what were really swaps of rights to run ads that had a net null effect on the companies but were trumpeted as massive revenues, and massive ongoing increases in revenue, although in fact no money was changing hands, at all.

In 1996 The FED was calling the Internet an "irrational exuberance" because the global economy wasn't doing well. Then, 3 major events led to the crash. This is why the bubble burst:

- 1997 Asian crisis

- 1998 Russian crisis

- 1999 Launch of the Euro (flop)

April 14, 2000 was the End of a 3 week crash. The global market trend officially switched, the rest is history.

The right person to read to learn about this is Andrew Odlyzko at UMN: http://www.dtc.umn.edu/~odlyzko . He draws some amazing comparisons between the dot-com bubble and the railway mania of the 1850s.

The dot com boom was based heavily upon a myth: that the internet was doubling every 100 days. Various executives at WorldCom / UUnet kept repeating this for years, even though it was only true for a few months in 1996. Since WorldCom / UUnet was the biggest ISP at the time, the business world (and especially investors) believed this.

At the same time Lucent was doing some creative accounting where they would sell equipment super-cheap on long-term repayment plans with a buy-back offer. They showed inflated sales and no liabilities: this too made it look like their networking business was growing at an astounding rate, which supported the WorldCom claim.

Now, if it had been true (that the internet was doubling every 100 days continuously for 4 years), then it would have made perfect sense to invest in bizarre dot com stocks: there was so much potential upside that there was no way you could possibly be paying too much for any stock. A company that has an exponential growth with no end in sight is a very good buy, regardless of how silly it might look now.

So that's precisely what investors did.

Unfortunately, the myth wasn't true, and it was becoming obvious by 2000. At this point there was no reason to continue investing in business solely because they had a strong internet presence. If internet growth was "only" 100% per 18 months (which was probably closer to the truth) then even established bricks-and-mortar companies would be able to move fast enough to capture significant internet sales. This is indeed what has happened since.

Without fresh cash injections, the crazier dot coms died, taking with them large numbers of companies that supported them with hardware, development and (especially) network infrastructure.

People bought Aeron chairs until they ran out of money.

At the time, no one knew what a dotcom was or how much it was worth. Many people guessed way too high because they were told that things were different now and at any moment these companies would begin vomiting money in a way never seen in history. This was untrue.

People were tricked into paying high prices for stocks with low earnings.

The crash was inevitable due to an unsustainable boom in stock prices.

But the TIMING of the crash is the most interesting thing about the 2000 dot com crash, and I very rarely see it attributed as described in the following analysis.

TL;DR: People think the Y2K event was a bust, a non-event, but actually it caused the Dot-com crash! (At least it caused it to occur when it did, in March/April 2000...) (Hear me out.)

And in theory, it should have all been forseeable. But nobody at least that I've heard of saw it coming.

First remember that in any gold rush (e.g. dot-com online riches) where it's hard to pick who the winners will be, an easy bet is the companies selling the shovels/infrastructure (in the dot-com case, servers, etc). You don't have to be smart enough to know where the gold is to get rich off the infrastructure "boom" and that was a common play for many investors of that era.

The late 90s was a huge time for UNIX companies who were not yet cannibalized by Linux and had larger server businesses. HP, IBM, DEC and most notably Sun Microsystems (whose development of Java added some marketing sizzle to their already leading UNIX server marketshare). And there were software players like Oracle also. Webservers and even more large database servers backing those webservers were a multi-billion dollar business. And the sizzle of that sexy internet business was used to sell more boring conventional backend database servers to companies who were increasingly getting on the UNIX/TCP-IP bandwagon in the 90s. The server companies made an attractive stock offering for brokers trying to pitch high-growth but still somewhat safe stocks to an unsuspecting clientele just looking to get in on the action. (The owner of the pizza shop near me was always quizzing me about tech stocks recommended by his broker. So the hype had gotten pretty high even on "Main Street". He suffered significant losses btw.)

However, in the first 2-3 months of 2000 all the hardware sales for all the big server players started to drop off a cliff, relatively speaking. (You'll see the dot-com crash often dated to occurring in late March, early April). With such a noticeable off-trend sales drop, the high PE multiples of Internet poster-children like Sun Microsystems, justified only by continued massive growth were suddenly not sustainable and contagion among these internet high-flyers suddenly came into sharp relief and question, spawning a panic/stampede ("who will be next"? "get out while you can!", etc).

Now why did the hardware sales start crashing? It turns out it wasn't the Dot-com Internet companies slowing their orders (at least not until their stocks started crashing too)... it was all the companies who had built up a huge secondary infrastructure of hardware+software for Y2K testing. To verify that production equipment would properly function when clocks rolled over from 12-31-99 to 1-1-00, many companies just bought small (or even fullsized production-grade) servers to duplicate their production gear and tested it would function properly by rolling the clocks forward in advance of the date and ensuring the software handled the transition.

With Jan 1st 2000 in the rearview mirror, many medium and large companies slowed their hardware server purchases for a year or so while they re-purposed their Y2K test servers for normal business use.

For Internet dot-com companies (even though they were a hardware play) like the flagship internet server company Sun Microsystems, this resulted in a significant drop in sales, and the awareness of this started hitting the market as those companies started to issue guidance about 1Q00 earnings.

As I said, once Sun and others started dropping, the Internet play itself was considered vulnerable and everything came back to earth.

Its nothing unique to dot com. Its just a buble. Same with gold price ups and downs.