[An earlier version of this article appeared on Forbes.com.]

Occasionally, you’ll hear an economist say that U.S. home prices are too high in some cities and will likely fall—eventually. These forecasters are meticulous about adding that the next real estate price correction will not be like the 2007-2012 housing bust when U.S. home prices fell by one-quarter and in some cities by one-half or more.

Robert Shiller, the Nobel Laureate economist with a good track record of predicting bubbles, recently said home prices could see a “significant correction or bear market” but, of course, he wasn’t expecting anything like the 2007-2012 bust, according to MarketWatch.

We seem to have a very strong consensus that the next time home prices fall, it won’t be like 2007-2012. No one, however, seems to be talking about what the next real estate bust might actually look like.

Let’s look at the housing bust before the last one: the Savings & Loan Crisis correction. What might that period of falling real home prices in the early 1990s tell us about the next fall in U.S. home prices?

Early 1990s versus 2007-2012

The Case-Shiller Home Price Index shows how incredibly steep the rise and fall in home prices were from The Great Real Estate Bubble. It was the steepest rise and fall in U.S. home prices on record. That’s not going to happen again.

CS USA

Since 2012, however, the increase in U.S. home prices hasn’t been as steep as the early 2000s but nominal home prices are higher now than at the peak of the bubble in 2006.

If we adjust the Case-Shiller Index for inflation (using CPI-U) we see that real U.S. home prices are below the 2006 peak, but they’re nearly 40% above real home prices in 1990, 2000 and 2012. At the 2006 peak, prices were over 50% above the real home prices in 1990, 2000 and 2012.

CS USA IA

More to the point of this article, the nominal home price index makes it look like home prices flattened out after 1990 but real home prices fell.

Real home prices peaked in 1989, the recession hit in 1990, home prices fell 7% from the peak until the end of 1990, the recession ended in the spring of 1991 but real U.S. home prices continued to fade for years until they bottomed out in 1997, down 14% from the 1989 peak eight years earlier.

If history repeats itself (it won’t) and the current cycle peaks exactly 30 years later in 2019 instead of 1989, then real home prices would be down 7% by the end of 2020, they would continue to fall slowly until 2027, and real home prices wouldn’t get up back to 2019 levels until 2030.

Notice that even though the fall in real home prices was two and a half times larger in the 2007-2012 bust, prices fell for years longer following the S&L correction (8 years) compared to the far larger 2007-2012 bust (5 years).

Huge Variability Between Metros in 1990s

Unlike the 2006 bubble which hit most of the country, the 1990 bubble was really only a bubble in a few major metros, for example, in Boston, New York, Los Angeles, San Diego, San Francisco and Washington DC.

CS 5 Cities.png

If we again adjust the Case-Shiller numbers for inflation, the real size of the 1990 bubble for those cities is easier to see.

CS 5 Cities IA

In real prices;

  • Boston didn’t get back to its 1987 peak until 2000 (13 years);
  • New York didn’t get back to its 1987 peak until 2002 (15 years);
  • Los Angeles didn’t get back to its 1989 peak until 2002 (13 years);
  • San Diego didn’t get back to its 1990 peak until 2000 (10 years); and
  • San Francisco didn’t get back to its 1990 peak until 1999 (9 years).

And remember, the 1990s are lionized as an economically prosperous decade.

Synchronization

Back then home prices across the country weren’t as synchronized as they were in the later bust. Many cities didn’t have much of, or any, housing boom or bust at all. Home prices in Portland and Denver actually increased significantly from 1989 to 1997.

Real home prices were up 50% in Portland and 25% in Denver while at about the same time they were down 40% in Los Angeles and 30% in New York, at least according to this inflation-adjusted data. We didn’t see such huge differences between cities in the 2000s real estate bubble.

Observations

What might the early 1990s housing correction tell us about the next real estate price correction?

1. Home prices can boom and bust even without a boom and bust in subprime, interest-only, no-doc and neg-am mortgages.

2. Once started, home price declines (like price increases) seem to become self-reinforcing and can continue on for quite a few years.

3. As metropolitan real estate markets have become more synchronized, booms and bust may have become larger.

4. California seems to be particularly prone to unstable housing prices.

Suggestions

In conclusion, a correction of home prices is likely at some point and the next recession could be a trigger like the 1990 recession was to some degree.

1. Single-family landlords might want to become more defensive.

2. First-time home buyers might want to make sure they’re buying at least a 10-year home, a home that is likely to suit their needs for at least 10 years, in case real home prices fall.

In all likelihood, the next real estate correction or bust will surprise us and won’t be like either the early 1990s or 2007-2012. But I think it’s extremely likely that at some unknown point, the residential real estate market will slowly and surely change from this year’s “fear of missing out” to “fear of losing money.”

Note: All the graphs on this page originally came from this page.

# # #