Arnold Kling  

When Did House Prices Get Out of Hand?

Economists Signing Petitions... The College Fun Puzzle...

One of the talking points among those who deny that Freddie and Fannie played a large role in the housing bubble is that the agencies lost market share in 2003 and 2004, and that is when house prices took off. Some actual data on house prices might offer perspective.

The agency that regulates Freddie and Fannie maintains an index of house prices that is available for download. Using the seasonally-adjusted quarterly data, here are some values:

2002, Q4: 165.47
2004, Q4: 194.47
2006, Q4: 219.96
2007, Q2 (the peak): 224.02
2010, Q1 (the low point): 191.35
2010, Q2 (the latest data point): 196.38

From the end of 2002 to the end of 2004 (the period when Freddie and Fannie were losing market share), prices went up 17.5 percent. Over the next two years, as Freddie and Fannie got back into the market, prices only went up 13.1 percent.

So, was 2003 and 2004 a period when the bubble really took off? At what point did the level of house prices reach unsustainable levels?

If house prices hit bottom earlier this year (admittedly, that is a big if), then the house price index at the end of 2004 was at reasonable levels. The real bubble took off from 2005 through the middle of 2007.

To put this another way, if you stopped lending in 2004, there is a good chance that your borrowers have some equity in their homes--on average, they have retained their down payment plus whatever principal has been paid down on their mortgages. Defaults from loans made before 2005 ought to be much lower than default rates on loans made afterward (and in fact the analysis of loan defaults strongly supports this).

In hindsight, Freddie and Fannie chose the worst possible time to jump back into the market. In hindsight, the rise in house prices in 2003 and 2004 was not a bubble--buying a house in January of 2003 would have given you a capital gain, not a capital loss.

Even setting aside the contentious issue of Freddie and Fannie, I think it helps to have this perspective on house prices. People who write as if the entire run-up in house prices from the late 1990's through 2007 was a bubble are being misleading, unless they want to argue that the worst is yet to come in the housing market. If the worst is over, then house prices did not get out of hand until some time after 2004.

Comments and Sharing

COMMENTS (14 to date)
Jeremy, Alabama writes:

When you are an enormous government-backed market-maker, like Fannie and Freddie, you create your very own bubble.

Fannie and Freddie did not "choose the worst possible time to jump back into the market". They created the market distortion that made it the worst possible time.

gnat writes:

Subprime originations were 8% of orginations in2003,18% in 2004, 20% in 2005 and 2006, and 8% in 2007. Alt A showed a similar pattern. Housing prices peaked in May 2006.
The housing price peak according to the Case Shiller data was May 2006,

Unknown writes:

I think the housing market of prices is good right now in todays economy. You can buy a house a lot cheaper now because the we are in a recession. Back in 2004 house prices were listed at a higher price then they are in 2010 because the economy was not this bad. I think the housing prices werent out of hand before 2004 due to the fact that we werent in a recession like we are now.

I dont think the bubble really went sky high after 2005 because the economy has slowly regressed into a recession, so therefore your house prices are going to go down because real estate agencies want to sell houses not just have them on the market for years and years.

Ironman writes:

Arnold ask

So, was 2003 and 2004 a period when the bubble really took off? At what point did the level of house prices reach unsustainable levels?

The beginning of the U.S. housing bubble may be traced to 2000, reaching unsustainable levels by 2002. The housing bubble then stalled out in 2003 before resuming in a new inflation phase from 2004 through 2005 before slowly stalling out and entering into a deflation phase in 2007.

Yes, housing prices still have some room to fall.

John Thacker writes:


Your link is extremely unpersuasive when it attempts to argue that the facts presented absolve Fannie and Freddie. It argues that since Fannie and Freddie only guaranteed and bought up loans, rather than originating them, that they are not primarily to blame. But clearly the presence of someone willing to buy a loan, and willing to abandon loan standards in order to regain market share, facilitated and enabled those loans.

Futhermore, I'm a lot more concerned with the public losses than private losses (and yes, TARP falls in that category, though they claim that TARP is being repaid.)

Exonerating Fannie and Freddie is no different from exonerating politicians from sins of either commission or omissions. After all, if Fannie and Freddie aren't guilty because they didn't make the loans, but only jumped in and were willing to guarantee them, then there can be no blame assigned to Bush or other politicians for failing to regulate more (even if you grant After all, they didn't make those bad loans either.

Thomas DeMeo writes:

It is important for us to distinguish between a housing price bubble and a housing spending bubble. We do not have houses with prices that are too high. We have houses which are too expensive.

The improvements to the nation's housing stock are real. We have been ratcheting up building codes for some time, and have been relentlessly putting elements into our homes that our parents never considered necessary.

There was some wiggle room with respect to prices, and that has been squeezed out. From here, we can have cheaper homes, but we will need to build them more cheaply.

gnat writes:

@ John Thacker
I meant to point to the just origination data that I cited from Inside Mortgage Finance, Figure 1.1, not the Karl Smith argument. I find this more balanced view from Jim Hamilton most persuasive:

ThomasL writes:

I agree with Ironman. I think prices were already out of whack by 2004, and have room left to fall, maybe a good deal of room.

Billy writes:

Based on this...

Shouldn't we be talking about the mid-90's for the start of a housing bubble?

Eric Rall writes:

Is the housing index you're using an index of real prices or nominal prices? If it's based on nominal prices, then an index of 194.47 in 2004 represents the same real price level as an index of 224.75 in 2010.

HoldingTight writes:

House prices are "out of wack" when buyers assume mortgages more than 2.5 times income. This has been true for most home-buyers for over 40 years. That has not bothered the economy until now because increases in home values has ameliorated the stretch needed to buy a home. Now that this phenomenon no longer applies, we are seeing the consequences. This problem is worse where home prices were highest (e.g., California, Florida).

Back to simple economics now ...

Chris W writes:

Don't forget that there's a huge inventory and "shadow" inventory of houses yet to be sold. If we forced them to all sell, say, over a period of 6 months, what would the housing price index say? I don't even want to guess.
Add to that the fact that the federal government is desperately trying to prop up the housing market to ease the pain of the downturn. If all that support went away today, what would the housing price index tell us? Again, I don't even want to guess how low it would get.

Brian Clendinen writes:

It has been know for a while housing prices historical increased at the same rate as wages. However, it has also been know that the average house size has become larger, and nicer over the years. Studies Case and Shiller have done showed before the late 70's, housing prices increased at inflation for existing housing. Then some major metropolitan areas in the late 70's early 80's started to buck the trend, Boston, New York, areas of California. Although the data is not conclusive it appears the increase was primarily due to government regulation which started limit housing growth and increasing the cost to build new housing. So with the exception of regions which have significant regulation/cost hurdles to increasing the supply, housing price should be at real wage. The Case-Shiller index should be at inflation.

If the bubble started in 1997, using CPI-U, then the index should be at 108, not the 160 (Composite-10) it is currently at. If the bubble started in 2004 then the index should be at 191. However, I due think it is interesting to note that inflation rose faster ever year than existing home prices from 1989 to 1997. Also we need to realize the index is only based on the ten largest markets. So in 1997 maybe housing prices were to low?

My thoughts are a 130 index is about were housing prices should be (this is an addtional 19% fall), however with the bad job market and typical behavior of markets to over correct, I think adding an additional 10% reduction would be reasonable. I don’t think it is unreasonable to say prices could drop 40% (real price, this is a mid 90's index) from where they are today, especially if there is a stagnate job market for the next few years. However, the real curve ball is how much of the limitations on increases in supply will limit the drop. I don’t have any idea how to estimate that impact.

What I think will be very interesting is state budgets if property values due fall that far. Considering the huge anti-tax populist movement, the chances of states being able to raise taxes are slim. So there is a very good chance of state spending shrinking unless the federal government continues to fund ever increasing portions of a states budget.

Steve Sailer writes:

Don't look at prices, look at defaults by vintage.

Mortages originated in 2006 had worse default rates than mortgages originated in 2005, which were much worse than 2004's. And, in California, 97.5% of all the defaults by late 2009 were of 2004 or later vintage.

So, if Bush had put the clamps on right after he was re-elected in 2004, we would have gotten away okay. It was really 2005 through early 2007 that killed us in terms of dollars defaulted.

Comments for this entry have been closed
Return to top