Bubbles are hard to spot, even in hindsight

People use the same flawed reasoning to identify tech bubbles and housing bubbles.
Bubbles are hard to spot, even in hindsight
Photo by Alexas_Fotos.

There’s a long history of people underestimating tech startups. Back in the mid-2010s, I wrote several pieces defending the valuations of companies like Uber, Slack, and Snap against critics who argued that they couldn’t possibly be worth so much. All three companies are now worth more than at the time I wrote those pieces.

In my 2017 Vox piece about Snap, I pointed out that people have been under-valuing technology companies for many years:

Almost every one of today's big internet companies — Google, Facebook, and Twitter — faced skepticism in their early years. Back in 2007, Kara Swisher expressed astonishment that Facebook could be worth $15 billion. A year before that, a Slate headline read "$1 Billion for Facebook? LOL!"

Today, Facebook is worth $900 billion—more than twice its value when I wrote those words.

A common mistake people made in all of these cases was to focus on the company’s current revenues or profits, which were often tiny compared to the startup’s market capitalization. When someone invests in a technology startup, they aren’t making a bet on the company’s profits in that particular year. They’re making a bet on the company’s long-term growth potential.

People sometimes make a similar mistake when thinking about real estate markets. The conventional wisdom holds that rising home prices in cities like Phoenix and Las Vegas in 2005 were an unsustainable bubble. Critics point out that home prices rose much faster than rents during this period and they take that as a sign of market irrationality.

But that’s too simplistic. If you anticipate that rents will rise in the future, it makes sense to pay a premium for a home now. And if a bunch of homebuyers make that same calculation, that can cause home prices to rise a lot faster than rents.

But that doesn’t necessarily mean there’s a speculative bubble. When the Great Recession happened, it crushed home prices in Phoenix and Las Vegas. But someone buying a house in Phoenix had no way of knowing we were on the verge of the worst recession in 70 years, or that the Phoenix metro area would be especially hard-hit by that recession.

A housing boom that wasn’t a bubble

In January 2004, I moved to Washington DC for the first time, renting this house with two roommates. The neighborhood, called U Street, was pretty rough. We were burglarized twice in the 18 months I lived there. But it was showing early signs of gentrification. People had started opening trendy bars on U Street, two blocks to the south. And more educated and affluent tenants—like me—were starting to move in.

As a result, home prices were soaring. Our landlords had purchased the house just a few months earlier (in 2003) for $446,000. That was more than triple the $126,000 it had sold for just five years earlier.

If you’d been looking at the ratio of home prices to rents during this era, you could have easily concluded that there was a bubble in U Street housing. But you would have been wrong. In 2009, in the depths of the Great Recession, the home sold for $616,000. Today it’s worth around $1 million.

DC homebuyers in the early 2000s were making a bet that continued migration from the more affluent parts of DC—and young professionals who would move to DC in future years—would drive up demand for urban housing. They were correct, and their bets paid off handsomely.

The bullish case for Phoenix circa 2005

Homebuyers in Phoenix during the early 2000s were making a similar bet. The sprawling Phoenix metro area had been growing robustly for decades. In the late 1990s, demand for Phoenix real estate became especially strong thanks to the rising cost of housing in California. Every year, tens of thousands of working- and middle-class Californians fled Los Angeles and San Francisco for relatively affordable homes in and around Phoenix.

That was driving a massive building boom. The Phoenix area added more than 40,000 new homes per year between 1997 and 2002. That jumped to 53,000 in 2003 and 63,000 in 2004.

In the early years of this boom, price gains were fairly moderate. Inflation-adjusted home prices rose 27 percent between January 1998 and January 2004. Much of this gain could be explained by falling mortgage rates, which enabled consumers to buy more expensive homes with the same monthly payment.

But then home prices rose another 63 percent (adjusted for inflation) between the start of 2004 and the start of 2006. Interest rates were not falling during this period, so that can’t explain the surge. The conventional wisdom held that this was a speculative bubble, with people buying homes in hopes of selling them to others at even higher prices.

But I don’t think it’s hard to imagine why a reasonable person would have bought a home at the 2006 peak. A median home in Phoenix cost around $250,000 in 2006, compared to $600,000 in Orange County, California. That kind of price differential was driving migration from California to Arizona.

Phoenix was responding by building houses at a rapid rate, but the area’s low-density development style and lack of high-density transit options limited how big the Phoenix metro area could get. If migration into Phoenix continued, demand might outstrip supply. If that happened, home prices would rise more quickly. Eventually, continued migration from California would push Phoenix home prices up toward Los Angeles and San Francisco levels.

And obviously, if you think that’s what’s going to happen in the future, you want to buy a home now while it is still relatively cheap. This may explain why home prices suddenly jumped upwards in 2005 and 2006: homebuyers became convinced that Phoenix was destined for Los Angeles-style housing scarcity rather than a return to housing abundance.

As it turned out, this bet was wrong, but for a reason that would have been hard to predict in advance. The US economy started to fall into a recession in 2007. As I argued last week, the Fed botched its response, cutting rates too slowly and turning what could have been a mild downturn into the worst recession in 70 years.

Ironically, one reason for the Fed’s slow response was the perception that the US was in the middle of an unsustainable housing bubble. In a sense, then, the Fed’s belief in a housing bubble became a self-fulfilling prophecy.

The Great Recession tightened mortgage lending. It also dramatically slowed migration into Arizona. Net migration from California to Arizona fell from 53,000 people in 2006 to 13,000 in 2008. Phoenix home builders cut back production, but not quickly enough to avoid a glut of unsold homes.

But the California-to-Arizona flow resumed after the Great Recession, reaching 31,000 in 2019. That has helped push Phoenix home prices up again. They’ve grown much faster than the national average. Today, they are just a few percentage points below the 2005 peak, on an inflation-adjusted basis.

An Amazon bubble?

Another tech startup people underestimated in its early years was Amazon. The company’s share price soared to what seemed like absurd highs in 1999, only to fall more than 90 percent in the early 2000s.

The company almost didn’t survive the dot-com crash. In February 2000, Amazon sold $672 million in bonds to overseas investors. The next month, the NASDAQ peaked and funding for tech startups started to dry up. Amazon was burning cash at the time; without that $672 million war chest, it likely would have run out of money, just like famous dot-com failures like Pets.com and Webvan.

If Amazon had failed, everyone today would agree that the company was ludicrously overvalued in 1999. Instead, thanks to some good luck, the company turned out to be dramatically undervalued. Today, Amazon’s market capitalization is about 30 times higher than the 1999 peak.

I view the difference between Phoenix and U Street homebuyers in the mid-2000s similarly to the difference between Amazon and Webvan investors in the late 1990s. With the benefit of hindsight, it’s easy to say that the buyers of Amazon stock and Washington DC homes were savvy, visionary investors, while buyers of Phoenix homes and Webvan stock were fools caught up in speculative mania. But I think that underestimates the amount of uncertainty and historical contingency involved. To a large degree, the first set of investors got lucky and the second set got unlucky.

And for this reason I think policymakers should be cautious about trying to identify asset price bubbles and making policy based on these identifications. The correct value of an asset depends on what happens in the future, and no one can predict that ahead of time. It’s better to focus on maintaining a steady flow of goods and services, and let asset prices take care of themselves.

Full Stack Economics depends on readers to pay our bills. We have no investors and accept no advertising. Please join us!


Only members can comment.
Please subscribe to a free or paid plan or sign in to join the conversation.