Another intriguing home price report

This is a sucker's rally caused by unprecedented and unsustainable government (taxpayer) subsidies and propaganda. Frankly, the con game has reached a point where I'm not sure I trust the measures anymore. Virtually all of the fundamental market drivers continue to deteriorate (i.e. foreclosure supply in the pipeline, employment, incomes, interest rate pressure, ARM resets/recasts, continuing tight credit by private lenders, the upcoming commercial real estate implosion that will further restrict banks ability to lend, rents, etc.) Nevertheless, the game of "pretend and extend" continues...

If there is one consistent lesson to be learned from history, it is that massive/hasty government attempts to manipulate markets lead to very bad results. This time is no different. There will be another big leg down.

Once again, I will tell everyone Zillow is worthless. Did they call for a national price decline in housing 4-5 years ago? Did the NAR? No. These
institutions have a horrendous track record of accurately calling the direction/trends in housing. The self-serving NAR deserves no credibility whatsoever, and should basically not be allowed to make any comments on the market at all. If so, they should be regulated by the SEC and all realtors should be registered as investment advisors. How can one respect any organization that completely missed the correction we have seen to this point?

Zillow.com's dramatic volatility seemingly with the wind direction has sapped any credibility from its numbers. Zillow.com shows that the values of houses in my neighborhood have increased 10% since June, that's a ridiculous claim: a steady 10% increase in 3 months? that's a 40% increase annually. I also have noticed that as soon as a realtor/seller claims ownership of a property listing on Zillow.com, the ZEstimate suspiciously spikes and then it conveniently gets listed on MLS. That generates a suspicious inflection point in the Market Value Change graph that had previously been in decline or flatlined and suddenly starts spiking steeply upwards. Zillow.com needs to clean up its act to regain any reputation in may have once claimed.

Zillow is the source of this good news? Is this a joke? This is the best you can do, do you get paid to write this? I'm shocked and amazed that Zillow would be used as a source. I think Zillow was big news, but has just become irrelevant. No one cares. It's like writing a paper and using wiki as your source. Isn’t the real issue here a question of whether the casual user of Zillow can really understand the reliability of this information and could a lack of understanding cause public harm?

zillow is junk data, for the most part, unless you live in a subdivision with a lot of activity. Its algorithms really fall apart in New England.

Funny to read what the NAR has to say about extending the housing credit. In classic speaking out of both sides of their mouth, they claim that the credit is all that has kept the housing market afloat so far. Ironically, I actually agree with the NAR more or less (Fed buying down rates, FHA junk mortgages, and dead cat bounce all play a role as well)

Now with the bubble in the stock market, I'm getting scared. Govt is running out of ammo.

So how worthless IS Zillow? I find that many, many sellers out there site the appraisals as gospel. Is it anywhere close for metro Boston areas?

Interesting to see in the Zillow report is which areas of Boston are suffering and which areas are doing fine. With first time buyer tax credits, you would expect lower income area to rebound. But areas that have Y-to-Y gain are Weston, Sudbury, Belmont, Newton, Needham, Arlington, Winchester, Lexington, etc, while Brockton Revere Everett Randolph Lynn etc. still have severe Y-to-Y loss.

Wang - if you are interested in the credibility of people calling housing bottom, there is an interesting recent article in LA times about how these bottom callers see and act in the current market. While I don't have the article link with me, but as far as I remember, Paul Krugman bought an NYC condo, Bill Gross bought a 23mil mansion in Newport Beach, while Shiller said housing maybe at turning point. All in August.

For those on this blog who criticize all the measures of real estate activity based upon their limitations and shortcomings - are there any measures that you find at all worthwhile? Many of you criticize them all including Case Schiller. However, taken in aggregate they show trends and these are significant. Dead cat bounce? Maybe but also maybe not. New Deal spending followed by economic activity fueled subsequent military involvement brought us out of the great depression. You criticize government intervention but you cannot completely discount the role of government, and government intervention is not going away anytime soon - at least not until the next election.

Lance - Regardless on whether or not the USG is doing right or wrong, government spending is never unsustainable. The USG can spend and the Fed can buy debt with new reserves indefinitely. Its happening now. Just this year 1.5 trillion is USG debt has been created not including bailouts, despite virtually no buyers globally. With the world's reserve currency and no need to rely on exports, its silly to have any expectation for long-term Japanese deflation in the USA. If Gold was the world currency then you would be right, but you are a citizen on planet Fiat, and like it or not these are the rules.

If the USG wasnt printing now, 50% or more of all US banks would have gone under, and anybody holding US bond/treasuries would be wiped out. The USG cannot repay its debts, and most savings accounts are actually invested in the US (or municipal) debt. You need to learn that there is nowhere.. savings accounts, gold, real estate.. to hide.

You mention consistent lessons from history. You should observe that the cost of things has been going steadily up for nearly 100 years now. If 4 generations of price increases isnt enough to qualify as a pattern, Im not sure what is.

"Frankly, the con game has reached a point where I'm not sure I trust the measures anymore."

I haven't trusted any measures put out by the government or any group with a vested interest in making things appear better than they are for many, many years. But, it definitely has become completely absurd in the last year or so. Fortunately, the real numbers are out there, you just will not find them in the MSM.

"If so, they should be regulated by the SEC and all realtors should be registered as investment advisors."

Leaving aside the fact that the SEC has become as corrupt as Wall Street, with their inaction in investigating the outright fraud occurring today, I absolutely agree. A mutual fund cannot make any claims about future returns, but yet realtors can cherry pick data to make an absurd claim such as "home prices historically appreciate at about 8%" (sure, if you look at 1998 to 2006).

Katie from Zillow here. Just wanted to clear up a few things and talk a little bit about what we may be seeing in the Boston market.

Hung Wang: Zillow launched in 2006, so we weren't around 4-5 years ago. We tend to be very cautious when analyzing housing data, as you'll see farther down in this message.

Paul: What neighborhood do you live in? Are you looking at individual Zestimates or the Zillow Home Value Index for your neighborhood? I'd love to be able to see what you're seeing. You can email me at katiec -at- zillow.com if you like. As for spikes in Zestimates before a home goes on the market, there is probably a good explanation for this. I can assure you that we don't tweak Zestimates for anyone. But when a homeowner claims a home or a real estate agent puts a home up for sale, they can update the home facts. Public record facts often won't capture recent improvements, and we get the bulk of our data from public records. So we allow homeowners and real estate agents to update home facts, and when that means details like the square footage or number of bedrooms and bathrooms changed for the better, the Zestimate will likely rise.

Mike v, charles and First Time Homebuyers: It helps to understand a bit how we derive our data. Zestimates (estimates of individual homes' values) are calculated using a variety of public record information (square footage, bedrooms and bathrooms, the most recent sale price of the home itself). Other information, like recent comparable sales, is used to train the algorithms. We track the accuracy of Zestimates by comparing them to actual sales, and you can see accuracy numbers by county by clicking on the "Zestimates Values and Accuracy" link at the bottom of any page on Zillow.com.

The Zestimates, which are intended to be used as a starting point for a buyer or seller, are then aggregated into the Zillow Home Value Index. Because Zestimates are calculated without bias, meaning they are just as likely to be low as they are high, when we aggregate them and look at the median, it creates a very accurate index. Important to note: Our set of data looks at the value of all homes. This is different from NAR's which looks at the median price of homes that have sold in a given time period, and Case-Shiller's, which relies on a weighted repeat sales methodology and includes foreclosures (Zillow excludes foreclosures from our index).

Onto the Boston housing market: In many areas across the country, we're seeing year-over-year home value declines shrink. In the City of Boston, we saw home value declines remain flat in July, for the first time in 40 months. Importantly, the larger Boston metro region is not there yet (home values fell 2.4%). The Boston area's housing market hit its peak about a year before the rest of the country's, so that could be a factor in why there are hints of the area reaching a bottom before many others.

But as for what this means long-term, we're also predicting a long, bumpy bottom for most areas of the country. And we would hesitate to even call a true bottom based on one month of data. We believe that when the bottom does arrive, it could last years. It will likely be characterized by fits and starts, as inventory in the form of foreclosures, which are expected to crest again, and homes belonging to sidelined sellers hit the market. There are many other factors, from the expiration of the first-time homebuyer tax credit to mortgage rates, that will also affect sales and inventory and, consequently, home values.

We are watching Boston closely, as it's one of the only major cities we've seen so far to achieve a flat year-over-year change in home values.

Bostonrunner,

Any statistical measure is a simplification, and when you look at it, it's important to know and acknowledge its limitations and biases. Saying an analysis has flaws doesn't render it useless, it just means that you should be looking at it through a particular lens.

My biggest problem with Zillow is how often it gets crucial information wrong. Even things that should be relatively easy to determine based on a map, such as school districts, are routinely wrong. Since the quality of schools are a major driving force in real estate prices in the Boston suburbs, and all it takes is a map to get it right, I feel like Zillow just isn't putting in much effort when it comes to Boston.

My biggest problem with Zillow is that their methodology is secret. Most of the other data providers (Case-Shiller, Warren Group, NAR, etc.) disclose how their numbers are computed as well as their underlying assumptions. Furthermore, their calculations are relatively simple and straightforward. This makes the information much easier to understand and interpret.

My theory for why Zillow's numbers are such obvious nonsense... I suspect their model is based mostly on regression analysis. This is not necessarily a bad approach. But it does have a major flaw: it is based on historic data, typically 10-20 years worth. Think about that for a minute. With the exception of a few rough years in the late 1980's, house prices generally followed a very stable upward trend right up until the peak of the bubble. If this was the sample that Zillow analyzed, then it stands to reason the resulting formula is probably pretty good at predicting prices in a stable bull market environment but probably not very good at predicting prices during a game-changing market collapse like we are now seeing.

In the words of Warren Buffet: “Beware of geeks bearing formulas”... Good advice.

Our fearless leader...

On Feb. 28th 2007 testifying on Capitol Hill, "I do not believe a housing downturn is a broad financial concern or a major factor in assessing the economy."

Ben Bernanke

Is this really the best we can get?

Hun

Katie,

Great post..very informative and open. Appreciate the information as I've always liked Zillow and was curious to some of the methodologies...clearly its an imperfect game, but real human appraisals weren't much better the past 20 years....

My biggest problem with Zillow is that their methodology is secret. Most of the other data providers (Case-Shiller, Warren Group, NAR, etc.) disclose how their numbers are computed

Exactly. Which means, Zillow numbers are not data. Period.

If we ever get to audit the fed, i think this whole house of cards we call the US economy is going to go right down the toilet. I would like to see this happen, because i would rather deal with the pain now than pass it on to my children like the past 3 generations have done.

Katie,

Concur with AJ.. not sure why the venomous posts are so prolific on the boston housing market, but it has been the case for years. Anger over the cost of housing and any uptick in the market seems to bring out uneducated speculation as to why this isn't the case.

Having lived on both coasts in high priced housing areas it was clear to me that in 2004 the boston housing market was no where near as over-priced as other markets. If you consider incomes in San Diego and LA compared with Boston the difference is marginal yet the cost of housing in Boston has been significantly more reasonable.

For those who have never left beantown or the "greater metropolis" this may seem like a stretch, but if those same people really consider boston to be a major city in the U.S. property is cheap!

Data don't lie. Home prices are up. End of stroy. Zillow is not perfect but better than NAR.

One thing I've always wondered about various estimates is whether they include age, condition and floorplan. I've seen 2,200 Sq ft houses that have 30% more usable floor space than comparable older homes that have various hallways and staircases that seem poorly planned (perhaps due to construction technology constraints).

Wow Katie, really great and informative post. I know Zillow can't disclose its formulas, but it is nice to know what does and doesn't matter. I disagree with a few of the methods you've discussed (such as excluding foreclosures from the data) but that just means that I can compensate for it when I'm looking at zillow's results.

kz (#9): Your post was very thoughtful and insightful. I agree with most of what you say. Long term, I see no way out of this mess other than inflation. And as you point out, there is considerable precedent for this. Every fiat currency in history has eventually gone to zero. Period.

But here's where we disagree:

You mention consistent lessons from history. You should observe that the cost of things has been going steadily up for nearly 100 years now. If 4 generations of price increases isnt enough to qualify as a pattern, Im not sure what is.

Prices have not gone up “steadily” for 100 years. The great depression in the 1930's caused severe deflation. Data is hard to find, but general consensus is that homes-- like other asset classes-- fell dramatically in price. Many plausible estimates (such as Case-Shiller) suggest home prices fell 40-50%. Then there was also the late 1980's crash when home prices in the Boston area fell 20-30% over a 3-year period. You may disagree with my numbers, but the point is this... The overall trend for prices has been up, but there have been significant corrections along the way. The pattern looks like a saw tooth. We are in one of those corrections now... And it is almost surely worse than the late 1980's mainly because fundamentals are much more out of whack, and interest rates are already essentially at 0% so the Fed can't lower them more to stimulate. Prices are headed MUCH lower in the next few years-- particularly in higher end markets where government subsidies are absent.

Bottom line: Cash is king in the short term, and cash is trash in the long term...

My guess was that Zillow comes up with a house value in a similar way that a realtor would come up with an asking price for a home. It looks at homes within a X mile radius (in the same zip code or something) and then finds "comparables" that have recently sold. By comparing the details between the comparables and the house in question, they can adjust the price even more. Llike if the comparable has 2 bedrooms and the house in question only has 1.5 baths, it reduces the price by X dollars. It's not that the algorithm is terribly flawed, it's that the data is incomplete (the data is coming from the county). It's your classic problem with computer systems: garbage in, garbage out. And, the algorithm can't take into account things like curb appeal and craftsmanship. But if the house is in a neighborhood where all the houses were built at the same time or in a similar fashion, it doesn't matter too much.

Even if one were to trust the data by NAR, the Fed, Zillow, and any other "housing mafia" organizations, which I don't. At a minimum one should be abale to realize this little rally (if we dare call it that) is a bear market rally. Of course with "cash for shacks" suckers will be drawn in. Just like any other subsidy or perceived "value" people will step up to the plate. I would use "cash for clunkers" as a leading indicator for "cash for shacks". It will be interesting to see how much auto sales drop (probably dramatically) after this "bait" has been removed. Historically subsidies have proven to be miserable failures in free market economies. Cash for shacks will just be one more...

Hung Wang: How big of an effect do you think the $8k credit is having on prices? If it's only raising prices by $8k, that's almost just noise in the data. Clearly, your real worry is that the limiting factor for people using the credit is the down payment, rather than the monthly payment, and the $8k credit is turning into a $40k (or even $160k if people use a 5% down payment) bump in the price of the house. How much do you think the bump is?