Monday, March 23, 2009

A Housing Bottom?


In a post about the good housing news today, Kevin Drum says:

It's not clear what caused this, since home prices are almost certainly going to keep falling another 20% or so. In fact, this might even be bad news in a way, since the faster we hit bottom and get back to trend growth, the faster we're likely to see the end of the recession.

I'd like to know what he is basing a further 20% decline on, I don't see it. If you look at a number of indicators, we are at, or very near a housing bottom.

Housing affordability as of December 2008 was at a historic high:

Keep in mind, this affordability metric was based on a median existing home sales price of $180k in December 2008. As of the data released today (for the month of February), the median existing home sales price was down another 8.3% to $165k.

Adding in the stimulus tax credit for buying a home in 2009, we can reduce that median price another $8k and then factor in interest rates at 5%, the lowest in decades. Housing affordability has never been higher.

Now looking at historical home prices, the median existing home sales price is already very close to the historical, inflation adjusted average of $150k:

If prices were to fall another 20% to $132k as Drum suggests, or $124k when adjusted for the tax credit, we would be at the same inflation adjusted median price as the early 80s (when mortgage rates were a whopping 18.5%).

Update: Kevin responds here and I follow up here.


stan said...
March 24, 2009 at 5:20 PM Delete Comment

Who cares about affordability when the supply is still way mismatched to demand? Home ownership was artificially inflated and now must decline as the credit unworthy are driven back out. So of course prices and affordability are going to be great but there are no takers with everyone who can qualify already in a home and that pool is rapidly shrinking. 20% downpayments will be tough on up and coming buyers. You are treating this like a normal cyclical fluctuation. Things have changed. Rent is the new black.

Anonymous said...
March 24, 2009 at 6:07 PM Delete Comment

That 'affordability index' chart is a load of crap -- it means nothing. It's a prop used by realtor associations - and it's crap.

And...I'm not getting your point here. Why should real-estate prices be above levels seen in the 1980s, after adjusting for inflation?

And this your argument for believeing Mr. Drum is wrong?

Mike said...
March 24, 2009 at 6:28 PM Delete Comment

Thanks for the comments. Unfortunately, the National Association of Realtors is the premier source for housing related data. I would disagree that the affordability index "means nothing", it is a rough measure of the number of families who can qualify for mortgages at the median sale price. In other words, the more affordable houses are, the more people can get mortgages, the more buyers (demand).

To the your second point, the charts make my case, look at housing affordability in the early 80s compared to today. Not to mention the $8k incentive in the stimulus package for first time buyers and the current 5% mortgage rates (which is also factored in to the affordability index).

Anonymous said...
March 24, 2009 at 10:12 PM Delete Comment

One of the cheaper homes in my community (East Bay):
3 bed/2 bath
1992 sold at $182,500 (inflation adjusted at 4% = $355,000)
2009 offered at $435,000 after previously being on the market at $525,000.

That looks like 20% overvalued to me. And that's assuming that the market doesn't fall beneath trends after the largest property bubble in history.

The more expensive home loans are only now being affected by alt-a resets and will be in large numbers for the next two years. There's a big price correction still to come in those communities.

If you're looking to buy a $400,000 house then an $8k incentive is worthless if you expect the price to fall by a mere 3%.

Also, are you taking into account the massive wealth destruction that's been going on? My downpayment savings have been eviscerated by the market that's actually collapsed faster than housing prices(!) Meanwhile banks are reluctant to authorize mortgages unless your downpayment is an ever larger share of the house price.

Mike said...
March 24, 2009 at 11:23 PM Delete Comment

A couple points to the last poster:
1) I am looking at national data, there will be a lot of variation from area to area, some areas still have far to fall, and some areas may be starting to recover.

2) Mortgage rates were 8-9% in 1992, they are 5% or lower today. That greatly affects buying power. The monthly payments on a 30-yr fixed mortgage of $355k with 8.5% interest is $2729. The monthly payments on $435k with the current 5% interest is $2335. So the same buyer is getting a much better deal, even though the mortgage principle is 20% more.

James B. Shearer said...
March 25, 2009 at 1:32 AM Delete Comment

Median home sale prices can be misleading as they are sensitive to the mix of sales between expensive houses and cheap houses. When the sales mix shifts to cheaper houses as I believe is the current case this artificially depresses the median price. I believe the Case-Shiller price index which is based on data about sales and resales of the same house is better. And Case-Shiller is what Drum is using.

James said...
March 25, 2009 at 5:13 AM Delete Comment

Mike, here's a chart of Yale University economics professor Robert Shiller's housing data since 1890? (Same recent housing data; different measure of inflation.)

Banking analyst Meredith Whitney recently said housing has fallen 27% since the peak, and will fall at least 40% peak to trough. That suggests we're two-thirds of the way through.

Perhaps Kevin Drum expects overshooting?

Mike said...
March 25, 2009 at 8:29 AM Delete Comment

James Shearer,
I did an analysis using the national Case-Shiller numbers, that index also point to a bottom if I try and predict the January numbers using NAR data. See the post here:

Mike said...
March 25, 2009 at 9:10 AM Delete Comment

Hi James,
Thanks for the comments. Do you know what metric Meredith is using? Both the national median house prices and the national case-shiller numbers are close to the historical averages already. We'll see if she is right, my guess is that the buyer incentives are so great right now that they will cushion the fall and prevent moving too far past the historical averages.

David Pearson said...
March 25, 2009 at 10:15 AM Delete Comment


Affordability is an incomplete measure. The limiting factor in the home buying decision -- for most applicants -- is the down payment.

Look across the economy. The savings rate has been hovering below 5% for years. Besides retirement, the main purpose of savings is to accumulate a down payment for a house. Clearly, there hasn't been enough aggregate saving to generate historical levels of downpayment "buying power". The reason is simple: people were counting on either no-money-down mortgages or equity from their home sale in the case of a "trade up".

So the solution for this "limiting factor" is to save more, but saving more contributes to low demand and unemployment, which tends to depress the demand for housing. How do get out of that vicious circle?

James said...
March 27, 2009 at 12:53 AM Delete Comment

Mike said...
"Do you know what metric Meredith is using? Both the national median house prices and the national case-shiller numbers are close to the historical averages already."

She didn't say, but by doing the math I'm pretty confident she's using Case-Shiller. Nominally, the Case-Shiller national home price index is down 26.74% since the peak, which is 27% rounded to the nearest whole number. According to my math, the Case-Shiller national index has to fall a total of 42% to reach its pre-bubble inflation-adjusted average. Both of those numbers square perfectly with Meredith Whitney's numbers.

The Case-Shiller national HPI currently stands at 139.14 (that's not a dollar value; it's an index value). The nominal peak was 189.93 during Q2, 2006. Dividing 139.14 by 189.93 gives us 0.73, which means the index has fallen 27% (because 1.00 - 0.73 = 0.27) since the peak.

The Case-Shiller index only extends back to 1987, so in my graph I use the OFHEO index to calculate housing values during the 1975-1986 period. The pre-bubble (1975-1999) inflation-adjusted average index value was 110.36. Dividing 110.36 by 189.93 gives us 0.58, which means the index would have to fall 42% (because 1.00 - 0.58 = 0.42) from peak to reach the inflation-adjusted pre-bubble average. Dividing 110.36 by 139.14 gives us 0.79, which means we would fall 21% (because 1.00 - 0.79 = 0.21) from current levels.

Of course, there's no reason to assume we will fall to exactly the 1975-1999 inflation-adjusted average. We might not fall that far, or we might overshoot it. (Personally, I don't believe real estate tends to overshoot.) My math also depends on the current inflationary environment continuing until housing reaches bottom. (We currently have almost no inflation, which is great because it makes the math easier.)

The data I'm using can all be found in this spreadsheet. In addition, Robert Shiller has the spreadsheets for his graphs at

With all that said, I think you might be mislead by the Realtors' recent numbers indicating a recent price decline from $180K to $165K since the fourth quarter of last year. The Realtors' numbers are quite volatile due to the fact that theirs is not a constant-quality index. Their median home price numbers change, not just because the prices of individual houses have changed, but also because the mix of houses sold changes. Look at the Realtors' home price numbers for the past five quarters and you will see lots of bouncing around, unlike Case-Shiller which moves fairly smoothly from one quarter to the next.

Mike said...
March 31, 2009 at 12:13 AM Delete Comment


Thanks for the info, I'm interested to see what the Case-Shiller numbers for January look like. I think they are out tomorrow (Mar 31st). Especially because the NAR numbers were around 165k for both January and February.

As an aside, it would be interesting to know what metric "matters" the most. Is it NAR? Is it CS? is it sales and not prices? etc. For example, if the median prices have stabilized, but the CS index will fall another 20%, what kind of affect does that have on the overall economy?

You have a great blog by the way, don't be afraid to spam:

Mike said...
March 31, 2009 at 12:16 AM Delete Comment

David Pearson,

That is a fair point, I'm not sure how to accurately measure the amount of money the average prospective home buyer has to put towards a down payment.

The FHA still has loans widely available for first time home buyers at 3.5% down, but it is difficult to know what percentage of buyers would qualify for those.

Post a Comment

Our Deal

The markets are crashing, thousands of people are losing their jobs and the government keeps funneling money into broken banks. This little blog will help explain why this is happening and how we fix it.

I'd like to preface this experiment with a contract between myself and my readers. My promise to you is that I will be civil, non-ideological, well researched and honest. In return, I only ask that you think.

Send questions to:

Quick Search

Recent Comments