Leading Indicators

We've had some really nice attention in the press and blogosphere the past week or so. Here's a quick summary so you can see what people are saying about Altos.

O'Reilly Radar: Tim O'Reilly highlights our real-time data products for Wall Street and the housing derivatives markets. I'll be presenting on this topic at the O'Reilly-sponsored Money:Tech conference in New York February 6-8.

In SFGate.com / San Francisco Chronicle Carol Llyod finds her friends in the market for a home in San Francisco - everyone is surprised how hot the market is. I love this kind of analysis. The data reveal so much more than the headlines can capture. Carol - have your friends email me and I'll set them up with the neighborhood level and insights that we do in our detail reporting.

Dustin thinks we're going to get acquired. And Robbie gives us a nice plug on Rain City Guide.

The bloggers at Redfin have been especially prolific lately. I like the approach Redfin uses for their blogging work. They've got a team of bloggers, each with an intelligent voice, tackling the local real estate markets they're in. None of it is too controversial, but it's solid content, targeted well for their customers.

Altos Links:

National Data for the Housing Derivatives and Case Shiller markets

San Francisco housing market

Some posts just write themselves.

Can't. Help. It. Must. Write. More.

Ugly? You betcha. Do these tell the whole story? Not a chance. In Silicon Valley, San Jose is the dominant market, of course. San Jose is a diverse community, with lots of sub-prime and other crazy loans in the past few years. But also some really great neighborhoods with prosperous, fully employed folks. Lots of big, but not-risky loans too. Here's how the market in a desirable part of town, Willow Glen, is holding up. I've done the price chart in Quartiles so you can see the trends at each price point in the market.

Link: San Jose Housing Market.

Here's a fascinating development for anyone interested in the housing futures markets. Radar Logic has acquired Miller Samuel-- the eponymous home of Jonathan Miller, blogger of Matrix fame and all around good guy.

Radar Logic, in case you haven't been paying attention, has developed the second tradable index for futures and options in residential real estate. Their product, the RPX, differs from the incumbent Case Shiller Index in a few key ways. The CSI updates monthly, where the RPX is a daily update. Case Shiller tracks median price, RPX tracks price-per-square-foot. This difference is ostensibly to better align with the relative value of a property and to avoid entanglement with the changing nature of the properties themselves.

More important in this story however, is not the difference in the trading indexes (each has its own advantages) but the differences in the companies that support these indexes.

With this move, Radar Logic buys itself all kinds of benefits: revenue, diversification, and talent. Miller is arguably the most influential voice in residential property valuation markets today. He's certainly got more readers than Shiller, plus the regular CNBC gig. Radar Logic is building itself into a significant Wall Street player.

Shiller's firm MacroMarkets, on the other hand, remains a couple of introvert finance professors in Cambridge with a reputation as enormously difficult to do business with.

So place your bets. Expect the RPX to over take the CSI in trading volume very quickly. (Both are still thinly traded to date.) Radar Logic may just be making itself into the firm that can fulfill the promise of a real, liquid market in housing futures.

[Disclosures: At Altos Research, we sell data that leads the Case Shiller Index by 3 months. We have trader clients on Wall Street. We have no relationship with MacroMarkets, but wish they'd turn up the heat a bit. RPX stuff coming soon.]

I have huge respect for Jonathan. Great move by Radar Logic. Congrats to both. This is what the market needs.

Jessica at Inman this morning reports that, in a fit of fear of a bursting bubble, the SoCal MLS has stopped publishing it's Days on Market stats.

Even if you give them the benefit of the doubt that a Days On Market stat can be misleading as a standalone indicator of housing market conditions, the move is just plain silly. Bite the bullet guys, sweeping bad news under the rug doesn't make the bad news go away. It just makes it harder to manage intelligently for home buyers and sellers.

So since you can no longer get a view from the SoCal MLS, you'll have to get it from us. And we, of course, don't present DoM as a standalone indicator.

Among lots of other market data, when we survey a market, we calculate an mean Days on Market vs. a median Days on Market. (The mean, remember, is the average. It'll skew higher if just a few porperties are on the market for super long times. The median is the measure of half the market. So half the homes are on less than X days.) It's fascinating to watch in a changing market, for example, the median drop while the average stays high. That illustrates the freshest properties--and the ones priced right--are turning over quickly while the stale, overpriced, unappealing properties are lingering.

Because I know you're interested, here's a chart illustrating the median Days on Market for some key Southern California real state markets. You can see we're past the seasonal Spring Fling of new properties coming on and the Dog Days are approaching. Though higher than it's been for years, 2+ months is actually not that crazy painful (easy for me to say). This is the median, remember so there are lots of properties hanging around for several (many) months.

[update] Here's average DoM too, note the effect of stale properties staying on the market and skewing the average higher than the median:

Further Research Details available here:

Los Angeles Real Estate Market

Irvine Real Estate Market

Pasadena Real Estate Market

Thousand Oaks Real Estate Market

We're quoted in a fun article by Noam Cohen in today's New York Times.

The topic is the emerging predictive power of internet information. Triggered by the recent murders by a pro-wrestler, and its seemingly prescient reporting in Wikipedia, Noam draws the comparison to the oracles on the Ides of March.

My friend Bill Tancer of Hitwise is also cited. Bill and I gave a talk on this very topic at FOO Camp a couple weeks ago. Bill likes to predict American Idol winners based on search data they capture at Hitwise (and he's good at it, though not infallible.) He's also demonstrated some potentially actionable leading data on housing, unemployment, and the spread of pandemic fears. Bill calls it searchonomics. Here's the point I like to make:

This information serves a Wall Street function as “closer to real time” data for investors, said Michael Simonsen, Altos’s president and chief executive.

“We are really early in the predictive power of the Internet; financial markets are getting a hang of it,” he says. “We have miles to go in all the nuances in capturing information.”

And that's the key, really. Predictive power is a trade-off between advance time and uncertainty. So the more the information is reported in real time, the more it is measured in real time and the more efficient markets get. Funny how in this view, the Bloomberg terminal looks like steam power.

Altos Research happens to be in the business of measuring really, really quickly. In markets where it takes a long time to get the data to people, the internet blows the roof off previously held conventions. Robin Hanson nails it:

Robin Hanson of George Mason University, an expert on using future markets to track public sentiment, said an e-mail exchange that these examples are hardly evidence of predicting the future. Rather, he suggested, it was “a bit newspaper-centric to say that news has not broke ‘publicly’ if it is being discussed online in rumors but has not appeared in a newspaper.” He added that “with more and more kinds of media, there are more and more intermediate levels of info availability.”

This is the crucial dividing line: between reporting on events in as close to real time as possible — which can prove jarring to society, and journalists in particular, but hardly supernatural — and predicting things around the bend.

Cool stuff. Nicely captured by Cohen.

In case you missed it, long term interest rates are rising. In some respects this is actually good news. The yield curve, charting interest rates along their maturity duration, is no longer inverted. Recall that an inverted yield curve means short-term rates are higher than long-term rates, an awkward financial state that commonly signals a coming recession. So a positively sloped curve reflects general economic strength. That's the good news.

The yield curve first inverted nearly a year ago. It was a time of pick-your-poison for the housing market. To get "right", either we were going to see recession, where the resulting joblessness would pummel housing demand. Or we'd be faced with rising long term interest rates, making mortgages more expensive and pummeling housing demand.

Well, the economy has spoken. The bears are capitulating one by one, recognizing that the housing market downturn is not sufficient to drive the economy into a tailspin.

Instead we're faced with something much more mundane in the housing market cycle. Higher mortgage rates. Higher rates, coupled with tighter lending from the subprime cleanup. We've had such low rates for such a long time that returning to normal levels will seem like a foreign country. Every upward move in rates makes homes less affordable.

Macro shifts, a declining dollar, protectionism all seem to be lurching us out of the mortgage rate utopia and back to the real world. Time to lock in that 30-year before we get to gasp 7%, methinks.

We've been talking for a while about how the market strength in the Bay Area's housing market has been focused on the economic centers, San Francisco and down the Peninsula, with the market notably cooling the farther you reach into the exurbs.

Thanks to the folks at FortiusOne, who launched GeoCommons at O'Reilly's Where2.0 conference last week, we finally got around to illustrating the heat map of this phenomenon.

In the following snapshot we've created a heatmap of price reductions. Specifically this is our percent-price-reduction stat--for a given zip code, the percent of properties that have had their asking prices reduced. We have some color tweaking to do still, but you can see the picture pretty clearly. The brighter red, the higher the percentage (and the weaker the market).

If we zoom into the San Francisco, San Mateo County Peninsula and the north end of Santa Clara County, you can see the strength in Mountain View, up through Palo Alto, and in the San Mateo/Burlingame areas. Also, demand levels in the City have stayed strong.

You can get details of the markets in Burlingame, San Mateo, Palo Alto, Mountain View etc. on our free research page. From that page, you can go to our map view and look at price trends for each zip code. (Heatmap not yet included.)

The S&P/Case-Shiller index for March 2007 was released today. The San Francisco Home Price index showed it's first increase in a year and came in at 211.09 vs. 210.78 in February. This increase should come as no surprise to our clients (or readers of this blog for that matter.) We whispered to clients on February 26 that the March numbers looked like they were heading higher.

(For the uninitiated, the Case Shiller index tracks existing single family home prices. The index compares prices now to January 2000 where the index = 100. These indexes are then traded as options and futures on the Chicago Mercantile Exchange.)

Broadly speaking, this is not a big surge in San Francisco Bay Area home prices. No

cause for celebration of the end of the bubble burst yet. We see the light trend sustained for a couple more months and then reversing back down by the time the June numbers are announced later this summer. The seasonal spring surge appears to wearing thin already.

In other notable markets reported today with the S&P Case Shiller Index: Los Angeles, Miami, and San Diego were down the biggest of the 10 major metros that are traded on the Merc. Full data table is here.

We'll be attending the Real Estate Derivatives seminar put on by Fritz Siebel and Clare Yang from Tradition Financial Services on May 29 in Manhattan.

TFS is the leading broker of the S&P Case Shiller options and futures products on the Chicago Mercantile Exchange. This seminar is where they discuss the market mechanics, trading strategies, and will probably get into the development of the OTC products forthcoming.

We get lots of questions from readers and clients about the ins and outs of this market, and this seminar is a great place to start for answers. They cover the commercial real estate derivatives as well as residential housing futures markets.

If you're interested in attending the invite is here (PDF file). They have another seminar coming June 19, 2007.

Fritz has been a tireless advocate of this market. His blog is really the go-to source for housing derivatives news and data. Read it here.

Also - We'll have information on our S&P/Case-Shiller leading-data products too. We now have some of the big Wall Street boys building our feed into their trading models. So if you're in that market, give me a buzz. I can share more details at the seminar.

This may come as a surprise to many of our readers, but this spring is shaping up to be a pretty hot market in the Bay Area (especially the Peninsula, with the more desirable spots in the East Bay and South Bay trailing, not far behind). We have plenty of anecdotal evidence coming in from our Realtor clients and friends: multiple offers, three-day sales cycles, etc.

So the question is, where in the stats can one turn to measure this transition? Obviously we see the price charts start to march upward with greater demand. But that's not enough, because prices are influenced by lots of exogenous factors and may lag the actual demand a bit. So we decided to explore a different measurement: Days on Market. In a cool market, many properties sit around for a long time. We watched Days on Market climb upward all last year. Now however, if lot of homes are selling quickly with multiple offers, we should see that fact immediately in the Days on Market indicator.

Specifically, when watching a transition from cool to hot, we should see median DoM drop first while the mean DoM stays high. (Remember that median implies half the homes on the market are on for less that this number. And mean is the average of all properties, so a handful of homes that have been for sale for, say, 12 months keeps the average high.)

For example, our client Bart Marchioni recently sold a place in Willow Glen San Jose in 7 days when the average home has been on the market for over 100 days! Good, fresh properties come available and get sold right away, but the stale ones are still sitting around. That's your transition.

So we dove into the Bay Area housing market data to see where the hottest markets are emerging. We looked for zip codes with greater than 15 single family homes for sale, then we looked for those with the greatest percentage difference between the median days on market and the average DoM. What we found confirms what we're hearing from clients are the hottest markets. Here are the 10 hottest Bay Area housing markets ranked by percent difference in median and mean days on market:

City (click any city name for the free research page)	Zip	Median Price	Median Time on Market (approx)*	Mean Time on Market (approx)*
San Jose	95130	$810,000	newly listed	8 weeks
San Carlos	94070	$1,098,000	less than 1 month	12 weeks
San Leandro	94579	$598,888	less than 3 weeks	8 weeks
Los Altos	94024	$1,649,000	less than 1 month	11 weeks
San Francisco	94110	$849,000	less than 3 weeks	7 weeks
Palo Alto	94301	$2,599,000	less than 5 weeks	13 weeks
San Mateo	94402	$1,388,000	less than 6 weeks	15 weeks
Oakland	94618	$1,124,000	less than 1 month	9 weeks
Sunnyvale	94086	$758,900	less than 3 weeks	7 weeks
San Ramon	94582	$1,098,000	less than 5 weeks	11 weeks

* Days on Market rounded to nearest week. See also this post on recent DoM methodology changes we did.

So in West San Jose (Cupertino and Saratoga schools!) half the properties on the market are newly listed. But the average property has been on the market for two months. Sales rates have stayed high so this is not driven by a buildup of new inventory.

Note the theme we've been talking about for a while here: ALL of these markets are close in. These top ten zip codes are from a sample of several hundred in 12 Bay Area counties. Four of these are Peninsula towns, two in the South Bay and three in the East Bay. If the trend holds, and the number of buyers continues to grow, expect the market strength to bleed further out from the Bay Area's geographic heart. If it does, you'll be the first to know.

Last week while discussing the subprime mortgage meltdown and it's effects on the housing market on Bloomberg TV, I made a comment that host Rhonda Schaffler jumped on and caught me by surprise. I said something to the effect of this:

We're very early in the effects of this market meltdown's impact on the housing market. We're likely to see more bankruptcies and financial trouble for the lenders in the short-term. Then we'll see lending standards harden, regulation increase, criminal activity exposed... all these factors will conspire to make it significantly harder for first-time buyers for many years.

My intent was to emphasize the long-term nature of this problem for the housing market. Rhonda focused on the criminal activity comment. "When you say 'criminal activity' are you saying you know something beyond what we know already?" she asked.

Rhonda wanted to make sure I wasn't directly accusing anyone of criminal activity. Of course, I only meant to point out that in the frothy bubbly markets, these things have a way of rearing their ugly heads after the fact.

Exactly one week later, this news alert crossed my desk from the WSJ:

NEWS ALERT
from The Wall Street Journal

March 2, 2007

New Century Financial is the target of a federal criminal inquiry into its accounting and trading in its securities, and said its auditors could warn of "substantial doubt" over the home lender's ability to remain in business. Fremont General, another big lender, said it plans to stop making subprime residential loans and is in talks to sell that business.

The news came the same day federal bank regulators announced a crackdown on loose lending standards on subprime home mortgages.

It's a tough subject to be right about, but there you go.

I just completed a segment with Rhonda Schaffler on Bloomberg Television The Bloomberg Report this afternoon. Topic was the subprime mortgage market shakeout and it's implications on the housing market. Bottom line is this: The subprime crash hits the low income and low-end of the housing market. As of today, the crash has not impacted pricing in the broader housing market. But we see the ripple effect extending several years into the future as the first time home buys are squeezed out now don't become trade-up buyers in the future.

They stream the live TV but I haven't found links to the recorded shows. Let me know if you can find it. The segment broadcast from about 3:10 - 3:20 Pacific time today.

We're quoted in BusinessWeek today discussing the potential ripple effects of the sub-prime mortgage meltdown into the broader housing market.

"The majority of the subprime business is with first-time buyers. So it may take several years to shake out," Simonsen says. "But when it comes
time to sell and trade up we may find that the low end has been squeezed out." In other words, a meltdown in the subprime market could affect the supply of future buyers for years to come.

Based on the evidence we're seeing currently, troubles in the subprime market have not percolated into the rest of housing market. And have not, generally speaking, impacted prices even at the low end. It would appear that current economic strength continues to isolate the housing market from problems in the low-end mortgage world. In fact, in lots of the areas we measure, we're seeing surprising strength of housing demand and price resilience. (More on this phenomenon soon.)

That's why it may take several years for the impact of this change to play out. I used the legs of a stool metaphor with Justin at BusinessWeek. Even though we're seeing one leg weakened, the housing market is currently steadied with a strong economy, still-low interest rates, great employment prospects, and a strong stock market. If in 18 months, we see an economic downturn, that's when we could get hit by the ugly stick.

Bonus subprime mortgage links: Dan Green covers the subprime crisis better than anyone out there. And CalculatedRisk diligently calculates the risks on every nuance of the meltdown.

Some readers have asked lately about a recent jump in our Days on Market statistic. I thought it makes sense to explain it here. We recently did a methodology change where we increased the look-back window for watching properties.

By making this change, we more accurately reflect the actual total time on market, even for properties that have come on and off the market, especially including those that have been off the market for 30 days before being relisted.

What we found was that in some markets, we can find a significant chunk of properties that have really been on the market for long periods and these properties were previously under-reported in our stats. We've seen a notable decline in the number of properties relisted (with the sole intention of "refreshing" a listing) in Silicon Valley since last fall when the local MLS changed its rules. In San Jose California we've observed a drop from nearly 25% of homes on the market having been relisted down to 9%. That's real progress.

Those rules however do not impact a property that for example was on the market all last summer, was pulled off over the holidays and put back on the market in January. A soon-to-be-announced Altos partner calls these properties "stale fish." We've adjusted our nets to catch more stale fish. We're still dolphin-safe though.

This change effected our the trend charts in most communities a little bit. In some areas, though, you might notice a big jump in the DoM stat that week in January 2007. This change is the reason for the jump. Inset is a snapshot of the current Days on Market trend for San Jose Homes.

Lately we've seen a lot of web searchers looking for answers to "when will the housing market turn around?" My gut reaction is, "you're looking for a bottom already?"

But to answer that question appropriately, we have to sift out what we mean by "housing market." Which part of the market do you care about? Are you a builder? A Realtor? A home buyer in an established neighborhood? A home seller in the exurb construction boom? Someone trying to hedge against a recession in 2007?

Altos Research deals primarily in the market conditions of existing homes, their prices, and their supply and demand. However, the housing market is much more vast than that. Much of the media coverage of the housing market focuses on new construction, and specifically the numbers of properties being built. So lets lay out the numbers to be tracked, and how they can be tracked:

Market Measure	Source	What does the data say?
Existing Home Prices	Altos, Case-Shiller, OFHEO, NAR	Generally down, locally occasionally up. Likely under-reported. Lagging data, like OFHEO, still says up nationally, but that's several months old.
Existing Home Sales & Inventory	Altos,