How Far Will Your House's Price Fall?

It is finally undeniable that house  prices are falling in most of the US.  For those selling their home in this market, the big question now is:  "How far will prices fall?"  Seller instincts are to hold out, insist on a price based on recent sales (within the past 6 months to 1 year), and ride out the storm.  Most sellers seem to expect a resumption of the unprecendented housing euphoria next Spring (2007).

As many readers know, I am a regular author and blog admin at Patrick.net, the top Google-search housing bubble blog.  I approach the real-estate bubble issue from an empirical, data-driven, and fundamentals-oriented approach.

My first introduction to the notion that real estate prices began to outpace supportable fundamentals after reading Yale economist, Robert Shiller.  Mr. Shiller has long been a solitary voice of reason among.  Mr. Shiller is also co-developer of the Case-Shiller Index, upon which the CME housing futures and options products are based.  Let's analyze Shiller's position:

Well over a year ago, Shiller began demonstrating that housing prices had come unhinged from fundamentals.  Traditional arguments supporting sustainable housing prices include income, population, cost of building, income-affordability, and mortgage interest rates.

Before price growth began to slow in 2006, we can see that population has grown at a relatively constant rate.  We can also see that mortgage rates are largely uncorrelated to real house prices.  In fact, house prices rose in the early 1980s while mortgage rates skyrocketed.

A better view of just how far house prices have risen compared to fundamental price support levels is shown below:

Almost unbelievably, when house prices are set to a real-price 1890 benchmark, we see that today's real-prices (12/2006) are 100% appreciated over baseline, while data suggests that instead prices should be closer to 12-15% over benchmark, in real-dollars.  (I realize at this point that some readers will immediately think that of course 2006 prices are twice 1890 prices.  But remember, these are real-dollar price.  Let alone 1890, $1 in 1940 would equal $11.46 in 1/07.  So, nominal-prices of a house would have to have increased 11.46-fold in order for the real-price to remain unchanged.  I cannot overemphasize the importance of thinking in adjusted dollars.)

That means that, overall, houses are overpriced by around 75%!

Many of those predicting a "soft landing" in US house prices are basing that prediction upon the notion that today's prices are permanently high.  In other words, prices may slip from 200% of benchmark to a modest 190%-195% of benchmark. 

The common argument is that, except for the Great Depression, prices have never fallen more than a few percent.  Well, actually prices have fallen by more than 10% as recently as the early 1990s.  But still, the floor underneath falling prices has always been sustainable fundamentals, which was 111% in 1990, and is at best 115% today in 2006.

So how do we get back to fundamentals?

This is the hundred thousand dollar (or million dollar in some areas) question.  Pundits and those with vested interests are insisting upon a soft landing.  In short, a soft landing means that nominal home prices don't fall, instead only real prices fall.  In other words, your house's price tag won't go down, at least not by much.  Meanwhile, your income will rise, rents will go up, overall inflation will creep up, eventually filling the gap.

Let's take a look to see if this is plausible:

Assuming nominal prices continue to grow by 2.5% (most in the industry are calling for a "return to normal", meaning 5%-6% annual increases, so they claim):

Your house price in 2007 is $1,000,000.  Shiller says it should really be $570,000.  Your home is overvalued by 75%, as measured by sustainable, long-term prices.

In 2020, your house would be worth $1,377,511 assuming annual appreciation of 2.5%, and no correction in prices (a soft landing).  Shiller says, in 2020, that house should sell for $806,429.  So your house is still overpriced by 71%!

The question you should be asking yourself is, "do I think inflation will be 71%?"  Actually, inflation would only have to rise by a fraction of that, probably more like 25%, in order to bridge the price gap.  This is because of the way that taxes, rents, and investments work on the math.  But still, do you really think we're going to see sustained inflation of 25% for a decade and a half?

How much will my million dollar home sell for?

Many people in the San Francisco Bay Area are sitting in homes they assume they can easily sell for $1,000,000.  In fact, the house my wife and I purchased in 1996 (and have long since sold) for $365K now boasts a Zillow-value of $1.285M!  Some comparable homes sold in early 2006 for over $1.3M.  Let's say those owners are "pricing to sell", and want to get out in 2007 for "only one million".

Shiller predicts that it will take until roughly 2011 for prices to fully correct to fundamentals, or the full 75%.  In order for this to occur, there will need to be 20% or greater drops in house prices in some years, probably the early years.  A hard landing.  So those sellers would be lucky to get $800,000 for that "million dollar home".

But what about 2008, 2009, 2010?  Well, if you wait all the way until after the dust has settled, you should be able to pick up that million dollar single-family home for a mere $585,000.

And does $585K really sound that out of line?  I mean, a home purchased for $365K in 1996, and then sold for $585K in 2010 is very much in line with historical expectations.  But that's really the key:  historical expectations.  Bubbles like this have the effect of distorting expectations.  It was only a few years ago that normal people did not expect to retire wealthy thanks to their house.  They expected to sell or bequeath their home for a nice appreciation, about 1% over inflation plus a little extra in highly valued areas.

Final Points

• We are not going to see sustained ultra-high inflation of over 20%.  So house prices cannot stay this high as a function of fundamentals.
• We are not living in some great new tomorrow -- a new paradigm.  One would have thought after the embarrassment of "The Long Boom" or "Dow 36,000" people would have learned to be skeptical of new paradigm talk.
• Fundamentals always win in the end.
• Maybe Shiller is wrong.  He has been before.  But it is very unlikely that he is completely wrong.  He just may have some dates wrong.  Maybe we'll get lucky and the nastiness to come will stretch on until 2020 or beyond.  That's something wonderful to look forward to.

Comments

[jonathan]

great article, Randy.
It would be nice to have a graph that would have real-dollar housing price graphed on top of another line that showed inflation, so we could see which declines in the past led to nominal dollar value loss, and which ones did not.
Btw you are missing the word 'question' in the first sentence of the "So how do we get back to fundamentals?" paragraph. (feel free to delete this comment later)

[randolfe]

Jonathan,

Thanks for the comments. That is a great idea to create a house price real-dollar plot against inflation. I'm pretty sure I have all that data already in spreadsheet form in the HSBC study I've linked in previous articles. Look for that graph in a future article.

[BitterRenter]

Robert Shiller\\\'s book of famous title which just happened to come out with impeccable timing was based upon his work which was originally published in July of 1996:

http://www.econ.yale.edu/~shiller/data/peratio.html

His conclusion: \\\"It appears that long run investors should stay out of the market for the next decade.\\\"

On July 1, 1996, the S&P 500 Index was 671. Today, a \\\"decade\\\" later, the S&P Index closed at 1,410 (17% dividends NOT included).

This is the guy whose prediction we are all sitting in rentals over?

God help us!

[randolfe]

BitterRenter,

Of course no predictions are entirely accurate. Economists are notoriously bad at timing, usually being too early in trends (or otherwise underestimating the persistence of irrational group behavior).

I would point out that your nominal 671 to 1,410 includes 17.3% inflation, which erodes real returns quite a bit, even including dividends.

[Cheddah Yetti]

If you include dividends and 17% inflation, real CAGR would have been 7.5% (roughly). I know of plenty of Pension Funds that would gnaw off their right arm for that rate of return.

I'd also add, there's "not entirely accurate" and then there is nearly 100% wrong. I'm not saying he's going to be directionally wrong this time, but his past order of magnitude miss is relevant when evalulating his current long-run prediction.

Cheddah

[Joe]

I thought the boom was over in 2003, was glad to unload 1800sf in Miami Shores for $345K... silly me.

I did correctly predict the tech bubble - only again, I was a few years early.

The housing bubble will deflate more slowly than tech stocks, nature of the product. It will be interesting to see how this plays out.

[ausman]

You mean 71% inflation over 15 years, not 71% inflation per year.

And I think you mean 25% (excess) inflation over 15 years, which is not really that unlikely.

You don't really mean 25% inflation/yr do you?

[randolfe_]

ausman,

Thanks for the corrections.

[GK]

I don't disagree with the article. But I have a few questions :

1) Why can't the normal rate of appreciation be the same as nominal per capita GDP growth (4-5%), rather than just CPI (2-3%)? Shouldn't per capital GDP growth be a factor?

2) I think GDP growth is a reason that stocks return 8-10% a year. Why can't housing do 4-5% a year in that case (sorry if the question is worded simplistically)?

3) Some places (like Cleveland, OH), have had only 3-5% annual appreciation for the last 20 years. Thus, are these already at the fundamental median and relatively immune to correction?

4) Why is there no inverse corelation between mortgage rates and prices? I would think that since higher mortgage rates corelate exactly to higher monthly payments per a fixed loan amount, higher mortgage rates lead to lower affordability per fixed loan amount, and thus lower prices.

Why is this not the case?

[randolfe_]

GK,

Good questions. I'll do my best to answer them.

1. Oversimplifying things a lot, housing is part of the equation that produces GDP. All GDP is is the summation of the parts of the economy. So housing cannot be 100% of GDP or there'd be nothing else. (In actuality there is all kinds of voodoo that goes into how housing is counted in GDP, but suffices to say it is only a part of the equation).

2. Taking #1 into consideration, if some parts of the GDP grow faster, then other parts have to grow slower, or else GDP would be 8-10%.

Another important point is that stocks are "productive" -- their growth reflects creation of wealth in the economy. Houses do not produce anything directly. They are, in a technical financial sense, big savings accounts individuals own.

The actual equation, along with very rough allocations, is:

Employee Compensation (70%) +
Proprietor Income (9%) +
Rental Income* (2%) +
Corporate Profits (12%) +
Net Interest (7%)

Keep in mind most employee compensation is consumption, because Americans save little, no (or right now negative) amounts of their compensation income. Also (*) Rental income includes all housing. Homeowners are computed to earn what they'd have to pay themselves to rent their houses. This is a complicated thing in and of itself.

3. No. Cleveland and other low-inflation areas are not immune. This is also complicated, and I don't understand the entire model, but it involves the fact that certain areas have different local-inflation conditions than other areas. So, Cleveland by holding the "median" price may well be overpriced too. That is, without a bubble, Cleveland would have declined relative to other areas holding their median values.

4. I'm not really sure. The HSBC data shows this pretty well. There is a connection between mortgage rates and home prices, but it is not a simple correlation. It is probably more of a catalyst factor, maybe psychological in nature more than financial. There are clearly periods when mortgage rates rose at the same time housing prices rose, and vice versa.

-- These are just my opinions. Seek a professional, licensed financial adviser before making financial decisions if you are uncertain.

[randolfe_]

One more thing, I strongly recommend the Mankiw book I link to the left for gaining an understanding into your questions. The first couple of chapters cover most of these topics, and it's pretty accessible reading.

[Shaoky Taraman]

Shiller's analysis is VERY flawed... why does he ASSUME that everything should be pegged to 1890 prices?

Just because those are real dollars?

What if in 1890 home prices were actually UNDERvalued?

Afterall, there was a lot more available land back then, particularly in now predominantly urban centers.

Did he ever think of that?!

[randolfe_]

Shaoky,

I don't understand your point. 1890 is just a baseline. It doesn't matter whether prices were over or under valued in 1890. If they were *under*valued in 1890, then there would just have been a dramatic rise in subsequent years.

Or are you saying that prices were UNDERvalued from 1890 until around 2002, when suddenly we discovered that everything should be 2.5 times more expensive?

I suspect that you don't understand inflation dynamics. Do a bit of reading before criticizing Shiller's model. There are plenty of grounds for criticizing it. You have failed to raise any of those issues.

[Shaoky Taraman]

My point is that why is he looking at 1890 as a starting point... because that is the earliest data available? If that's the case, then the entire analysis is junk. What if the data was only available from 1977-79? Then we'd be looking in great shape now.

...and, maybe even 2.5 times is not enough!

If you actually think you need to only look at inflation dynamics instead of core forces and trends than you better read this...

http://economistsview.typepad.com/economistsview/2005/09/do_stickyprice_.html

[randolfe_]

If data was only available from 1977 - 1979 then there wouldn't be a long enough of a sample series to establish any meaningful analysis. I suspect that you don't have a solid understanding of how statistical regression techniques work. I cannot help you there other than recommend you take a course or read a text.

If you actually look at the analysis, the regression is only performed started after WWII. If they'd started the actual regression in 1890 then things would look even worse because the Great Depression and WWII period drag everything way way way down. If you don't believe this, read "The Grapes of Wrath".

The choice of 1890 dollars as baseline is arbitrary. You could as easily pick 2007 dollars and adjust everything according and come up with the same mathematical answers. Sorry. Keep hunting and you'll find the legitimate criticisms of his model.

As for the article you link: Do you know what the "Phillips Curve" is? The author seems to not, or at best be living in the past. Despite his opening assertion, this common statement on the Phillips Curve is a truism today:

Most economists no longer use the Phillips curve in its original form because it was shown that it simply did not work.

[Shaoky Taraman]

I think we are starting to get petty. I have read the Grapes of Wrath and I have seen the movie 3+ times... that period was hard for most, but not all. The wealthy that had PLENTY of real estate or money from other sources weathered the storm... and maybe that's the ultimate point. Most folks tend to get in over their heads... but I'm going to hang in as long as possible and continue to buy real estate because at some point in the future there will be another peak in real estate. There always is... and maybe I should actually encourage threads like this instead of disputing them because if enough people scare away from a sure thing, it will only create better opportunities for those of us that are intelligent enough to think for ourselves and continue to buy even when everyone else is running like a coward.

[randolfe_]

Shaoky

Please, by all means, keep encouraging threads like this! Facts and data are inconvenient, I know, but you provide a great opportunity for educating people.

Your common-wisdom is wrong. Categorically wrong. Here is what happened in the Great Depression and WWII period:

Decile analysis of incomes shows the top 1% of wealthy families went from about 24% of aggregate wealth ownership (over 90% of which was real estate), to about 6% ownership. So the rich lost 3/4 of their wealth, which was mostly control and ownership over real estate. You do know, I hope, that there were whole neighborhoods of mansions that stood empty. I leave it as an exercise for you to research what pejorative they called those towns.

The WWII years following the Depression were even harder on the rich. An entire class of poverty-stricken Americans, in the bottom 1/3 of the wealth distribution and owning less than 0.5% of the nations real estate, shot up to the top 1/3 of wealth by the end of the war, and owned over 50% of the nations' real estate. The factors in who gained were having at least 5 members in the family of wartime working age (roughly 15 or older), and no debt prior to or during the deflationary recession.

The rich, on the other hand, lost another 40% of their real estate holdings because the wartime economy did no allow them to service their debt with capital gains, so they had to liquidate real estate to raise cash (and sold it to those families earning 5+ incomes). The only rich who did well during this period were those who owned (as in family-owned) large stakes in companies deemed critical to the war effort or homeland stability (like sugar refiners or munitions makers).

The lesson from the Depression, which you raised not I, is that if you were a renter with no debt you did exponentially better than anyone holding real estate financed with debt.

Sorry, you disprove your own position. Try another tangential argument to support your position.

I don't think there will be a "Depression" anyway so it's a pointless straw man. I do know that I could *already* buy a home today where I live for roughly 2005 prices, so not buying in 2006 even adding in inflation would have made me financially _worse_ off, regardless of how long I held the house (which averages 5 years in the Bay Area).