How far can U.S. house prices fall?

Published in Housing Finance International Journal.

U.S house prices have become bloated in America’s second house price bubble of thefirst quarter of the 21st century. These prices are now far over the peak they reached in the infamous housing bubble of the early 2000s. According to the S&P CoreLogic Case-Shiller National House Price Index, house prices in September 2025 were at a level of 327.6, compared to the first bubble’s then-monumental peak of 184.6 in 2007. In other words, they are now 77% over the previous top.

This is widely considered a housing affordability crisis, since large numbers of people, especially young families and other first-time buyers, cannot afford houses at this historically normal level of U.S. mortgage interest rates. Said Sean Dobson, CEO of the real estate company, The Amherst Group, “We’ve probably made housing unaffordable for a whole generation of Americans.”[1]

The median price of a single-family residence in 2024 was 5 times the median household income, the Harvard Joint Center for Housing Studies reported, while a historically normal level is more like 3-4 times. Amherst Group’s analysis concludes that if you keep mortgage lending rates and household income at their current levels, to reach the affordability of 2019, U.S. house prices would have to fall 35%.[2]

The “‘Affordability Crisis’ Can’t Be Solved,” a Wall Street Journal headline pessimistically suggested.[3] On the contrary, of course it can: house prices can fall. The whole point of prices is to go down as well as up. But how did they get so high to begin with?

The second extreme inflation of U.S. house prices of this still-youthful century was stoked by the massive, unprecedented and in my view, unjustifiable, buying of mortgages by the Federal Reserve. The Fed purchased mortgage assets for the first time in its history in 2008, as an emergency intervention while the first bubble was collapsing. The Fed promised Congress, and genuinely intended, that such buying would be temporary, would be reversed, and would not affect the longer-run size of its balance sheet.

But instead, the Fed’s mortgage expansion continued until 2022 – for fourteen years. Along the way, it forced 30-year fixed mortgage interest rates to abnormally low and unsustainable levels of below 3%. The Fed’s mortgage portfolio, an asset it had never historically owned, reached the staggering level of $2.7 trillion, three times as big as the whole Fed had been in 2007, as the Federal Reserve turned its balance sheet into the equivalent of the world’s biggest savings and loan. As of November 2025, this portfolio is still huge at $2.1 trillion, with a mark-to-market loss of $323 billion as of the last report on September 30. Its former promises to the Congress of reversing the investment notwithstanding, the Fed cannot sell its huge mortgage portfolio without realizing massive market value losses and pushing mortgage interest rates higher.

When the Fed stopped increasing the size of its mortgage portfolio in 2022, U.S. mortgage interest rates quickly rose to historically common levels of 6%-7%. Many observers, including me, were surprised that this doubling of the interest cost of a house purchase did not cause a sustained, significant fall in average house prices, but national indices of house prices continued an upward trend. However, the volume of house purchases shrank dramatically, to the lowest levels in three decades. The increasing prices were thus on a much reduced volume of transactions.

The rate at which prices were rising slowed steadily in 2024 and 2025. The AEI Housing Center projects November 2025 at a 0.7% year-over-year national house price increase, expecting it to fall in December to zero for the full year 2025. The Center comments, “This will mark the first time since 2011 that December year-over-year house price appreciation has been at about zero.”

Since U.S. inflation is running at 3%, if nominal house prices are flat, in real terms they are falling at a rate of 3%.

The Housing Policy Council, using data from the Federal Housing Finance Agency, has calculated the deviation of U.S. house prices from their long-term, inflation-adjusted trend line. As of the third quarter of 2025, it concludes that house prices were 30.8% over their long-term trend. This is similar to the first quarter of 2007, at the top of the first 21st century bubble, when they were 29.5% over the trend. In that bubble, what happened next was that U.S house prices fell until 2012 by a total of 27% in nominal terms.

In this context, let us consider some cities where mid-tier house prices have already fallen by 15% to 25% from their most recent peak (which was usually in 2022), while the national index was still rising. Wolf Richter has helpfully listed these for us, using data on mid-tier houses from the Zillow Home Value Index:

Oakland, California -25%

Austin, Texas -24%

New Orleans, Louisiana -19%

Lee County, Florida -16%

Manhattan, New York -16%

Sarasota County, Florida -16%

San Francisco, California -15%[4]

Although these are large percentage declines, they are from extreme peaks and leave prices still at high levels. Taking the top two on this list, for example, house prices in Oakland after a 25% fall are still more than 40% over their first bubble peak. In Austin, after falling 24%, they are still more than double their level in first bubble times. This seems to leave room for further declines.

On the cause of the second bubble, consistent with my discussion above, Richter writes, “The below-5% average 30-year fixed mortgage rates were an aberration caused by an explicit policy by the Federal Reserve to repress mortgage rates to trigger the biggest bout of home price inflation this country has ever seen.”

As to what can now be done about the unaffordability of houses, he reaches this sensible conclusion: “The solution of the affordability crisis is many years of rising wages and falling house prices that over time would unwind the crazy home price explosion.”[5]

Where do U.S. house prices go from here? In my view, it seems unavoidably down. But by how much? I don’t know and neither does anybody else. The AEI Housing Center has a preliminary projection of down 1% for 2026, but that would not correct much of a 30% overshoot from trend. The Center adds that risks are to the downside, as they surely are.

Without venturing a numerical guess, I do observe that financial history demonstrates that asset prices after a bubble can go down more than most people, who consider only recent experience, consider possible. The lessons of longer history are that the answer to the question, “How much can asset prices fall?” is liable to be “More than you think.”

1 “The Price of Economic Firefighting,” Fortune, November 15, 2025.

2 Ibid.

3 Wall Street Journal print edition, November 24, 2025, p A2.

4 “The 14 Bigger Cities and Counties with the Biggest Price Declines,” Wolf Street, November 23, 2025.

5 “Mortgage Rates Are Not Too High. What’s Too High Are Home Prices,” Wolf Street, November 29, 2025.