Hi! I’m Aziz. My newsletter contextualizes America’s housing market using data visualization—hit reply and let me know what you think.

The current housing cycle began with accelerating prices in early 2020, continued with prices dipping in 2022, and ended early this summer when prices returned to their peak. This period was extreme by three measures.

First, from February 2020 to the peak in 2022, prices soared by 45%—a rate that was matched only during the 2003-09 cycle.

Second, the acceleration, peak, fall, and recovery played out unusually quickly. The 1977 and 1989 cycles took five years. The 2003 cycle took ten years. This one played out over three.

Third, all regions participated in the volatility. In 2003-09, large cities drove most of the price action. This time prices boomed in even the sleepiest of small towns.

In the above chart, each bubble corresponds to a U.S. Census Bureau metropolitan or micropolitan statistical area and is sized proportionally to its population.

The location of each bubble on the x axis represents it’s volatility: how much prices in that region changed, on average, over rolling three month periods in each cycle.

Large cities (over 5 million residents) are colored yellow, all other regions are blue. Bubbles to the right experienced more extreme price swings.

A comparison to the 2003-09 cycle helps put the current one in context.

2020 vs 2003: Sprint vs marathon

In both 2003-09 and 2020-22, prices increase by 45% (in both cases, I date the beginning of the cycle to the moment when prices began accelerating at a faster-than-trend pace).

But these gains stretched out over four years in the 2003 cycle, whereas in this one they were achieved in just over two.

In the last cycle, prices collapsed after peaking in 2006, and continued to fall for more than five years as the Great Financial Crisis (GFC) played out. This time prices fell for only 7 months.

In the last cycle, after prices bottomed in 2012 , it took almost five years for them to recover to their prior peak. This time the recovery took five months.

In one very important sense, this cycle was less extreme than 2003-09: gains were fueled less by speculation, high leverage, weak borrowers, and explosive mortgage products, than by growing incomes and changing tastes. Unsurprisingly, this time around prices fell much more modestly (-5%) than in 2003-09 (-28%).

But because the timeline was so much more compressed during the pandemic, the month-to-month prices changes were larger. If the 2003-09 cycle was a marathon, this one was a sprint.

There were specific catalysts that sped up this cycle.

The pandemic ignited a frenzy of demand and higher prices. Fed hikes crushed affordability and pushed prices back down. Rising incomes and a dearth of supply drove them back up.

No town left behind

There is another reason this cycle felt so extreme: participation was widespread. In the last cycle, many areas—especially small towns—didn’t see much price volatility. The areas the whipped around the most tended to be large cities.

This time around, no region was immune from the volatility. Maybe this is because urban residents were fleeing to less densely populated areas. In any case—as the chart below shows—in 2020-23 (blue bars), small towns were slightly more volatile than large cities. Prices soared in even the sleepiest small towns.

This contrasts starkly with 2003-09 (yellow bars), when large cities were clearly more volatile. The old adage “real estate is local” remains true. But real estate has been less local than in the last cycle.

Is a reversal in the cards?

In the most recent data, prices are rising and there is no relationship between gains and population size. If the pandemic was the rising tide that lifted all boats, there is no sign yet of the tide receding. For now, booming prices is small towns are sticking.