Over the weekend, the Seattle Times published an article titled Bad real-estate loans stack up for smaller local banks, about the exposure local banks have to real estate loans. The gist of the article was that many local banks, squeezed out of the mortgage business by the likes of WaMu and Countrywide, had doubled down on lending to local developers - and thus were exposed to the real estate turn down.
Though much of the attention has focused on financial giants like Citigroup, Wachovia and Washington Mutual, many regional banks are heavily exposed to residential real estate.
For instance, construction and land-development loans make up 61 percent of the loan portfolio at Bremerton-based WSB Financial, parent of Westsound Bank. At First Financial Northwest of Renton, parent of First Savings Bank Northwest, virtually the entire billion-dollar loan portfolio is connected in some way to real estate.
…
For several years, that specialization proved quite profitable. Smaller regional banks often posted growth rates and earnings per share that superbanks would have envied.
But now, with home sales slowing, the regionals are seeing more developers fall behind on their loans, and are setting aside larger sums for expected future loan losses.
The article goes on to give examples of exposure at local banks, including Cascade Bank, Westsound Bank, and First Financial.
The implications of high concentrations of commercial real estate lending at small, regional banks has been a topic of discussion over at Calculated Risk for some time. And as they blogged recently, Fed Vice Chairman Donald L. Kohn has warned of the potential implications:
Concentration risk is another familiar risk that is appearing in a new form. Banks have always had to worry about lending too much to one borrower, one industry, or one geographic region. But as smaller banks hold more of their balance sheet in types of loans that are difficult to securitize, concentration risks can develop. Concentrations of commercial real estate exposures are currently quite high at some smaller banks. This has the potential to make the banking sector much more sensitive to a downturn in the commercial real estate market.
One thing that struck me as I read theses two pieces -just a few days apart - is the nature of our local home-building market. Unlike many of the bubblier cities, the “big” builders such as KB, Lennar, and DR Horton have had little or no presence in our market. The biggest builder around here seems to be Quadrant Homes. Yet as Tim has pointed out in previous posts, we have certainly had a something of a building boom over the past 5 or so years inasmuch as supply has outstripped demand.
Is the inference that the Seattle MSA has had more construction driven by smaller regional developers? As a follow on, what might this mean about the portfolios of our local banks in comparison to other areas? Are they even more heavily focused on real estate than say, regional banks in Southern California? It will be interesting to see what articles follow on after this one. I have a sense we may be seeing the leading edge of a new trend.
Many of you are aware of the “Residential Property Index” published for the top 25 MSAs by Radarlogic. Radarlogic has a similar business model to MacroMarkets with the Case-Shiller Index (e.g., creating housing-based contracts to be traded on a futures market) but there are some key differences in the way that they calculate their index. Here is a quick overview:
Radarlogic’s index is based on ALL sales in a market. They do not ignore any sales that appear to be outliers. As such, Radarlogic does not use a “paired sales” methodology
As a means to make transactions comparable, Radarlogic tracks markets based on the price of a house per square foot, rather than the overall price of the house
Radarlogic reports their index daily - not monthly. The data is published on their site 60 days in arrears (so as of the date of this analysis, prices were available through February 5)
In order to smooth the daily impact of low sales volumes, etc, they offer rolling 7 and 28 day averages. The 28-day average is the one most comparable to other indices.
Like Case-Shiller, the reporting unit for Seattle is the MSA - which includes King, Snohomish and Pierce counties.
The index for Seattle shows a fairly familiar picture. A market that peaked in June of last year, that has fallen steadily since then, and has just recently turned negative on a year over year basis. I was interested in how fast things were off from the peak. When one looks at the rate of decline since the peak in June, things look a little different.
click image to expand
According to Radarlogic - prices in Seattle peaked on June 26, 2007 at $236.16 per square foot. Since then, prices have dropped fairly steadily to their current level of $211.63. That’s a drop of 10.4% in just over 7 months. On an annualized basis, this means prices per square foot in the Seattle MSA are dropping at a rate of 17%. What I also found kind of interesting in looking at this chart is how consistent the rate of appreciation appeared to be between 2000 and 2004. As you can see by the line I’ve added, it was almost linear at ~6.4% per year. If you extend this rate of appreciation through today, you’d have seen prices at about $185 per square foot - or 12% under current levels.
The fact that we’re only 8 months past the peak is very important. The papers are still reporting the year over year changes, and interpreting the lack of a significant drop as evidence that Seattle has dodged the bullet. However, as this index (and Case-Shiller, and the median) shows, even if the market is able to trade sideways for the next 4 months - we are still looking at double digit price drops in July.
Since Radarlogic also offers indices for 25 other MSAs, here is a comparison to some other West Coast markets:
As you can see, Seattle again looks most similar to San Diego in the scale of our downturn - and in this index we lag that market by 13 months. Interestingly, San Francisco and Los Angeles seemed to have peaked much closer to Seattle according to this measure. Like Seattle, all of these markets have already booked double digit declines on a square foot basis inside of a year. The Bay Area (San Francisco) seems particularly hard hit on this metric - tracking to nearly a third off in the space of a year.
At the end of the day, this is just one data point - but when I get enough of them all pointing in the same direction, I start to think I’m on to something!
As a follow-on to Tim’s post yesterday about the practice of relisting properties, here is a look at what has been happening to inventory over the past few years, and how the trends have changed over time.
The questions “what is happening with all of those listings” and “how many sellers just give up?” have come up a number of time from readers. The latter question is particularly interesting - because it might tell you something about seller psychology. Logically, one would expect sellers to rarely pull listings in an up market. The rising tide is probably lifting all boats. But in a down market, what happens? Will sellers start pulling listings because they expect prices to go up soon and want to wait out the dip? Will they keep their homes on the market because they need to sell, or because they think things will be worse if they wait?
The NWMLS reporting on inventory is decidedly opaque. They report homes sold, pending, added, and active. What they don’t tell you is how many homes remained on the market, or how many sellers just gave up and took their homes off the market. However, if one looks across months, some pretty basic algebra allows you to back into how many homes are de-listed versus left on the market.
You can get to delisted inventory as follows: [(Beginning inventory + New listings) - (Pending listings (sold) + Ending Inventory) = Delisted]. Take the available stock less what is either sold or left over, and it tells you how many homes disappeared from the listings that month without being sold…
Armed with that, you can calculate how many homes stayed on the market as follows: [Beginning inventory - Pending listings (sold) - Delisted = Stayed]. Take the total starting inventory, and subtract those that sold or gave up, and all you have left is the stuff that stayed on the market!
Using those formulas applied to the last 9 years of inventory data, we can peer inside what has been happening with listings before and during the boom:
On the left axis, we have the average number of homes delisted, sold, added, and staying on the market for each year from 1999-2007. These are displayed as stacked columns, with the reductions from inventory shown as negative, and the additions/remaining inventory shown as positive. On the right axis, we have average inventory throughout the year displayed as the black line.
What you can see on this chart is that new listing activity and sales activity were remarkably even across the past 9 years. New listings averaged about 4500 homes per month, plus or minus no more than 500 units. Sales averaged about 3,000 homes per month and again were typically plus or minus less than 500 units (2005-06 were +~700). Based on the consistency of this data, it appears that the bulk of the of the variability in average inventory levels is driven by whether sellers choose to stay on the market or give up.
During the period for which I have data - the average ratio of delisted/stayed was 1:3. In other words, sellers were 3 times as likely to stay on the market as they were to pull their listing. However, the average is deceiving. Over the course of the boom, this ratio has fallen steadily: from just over 1:2 in 1999 to the current level of 1:5, where it has been from 2003-07. Sellers psychology during the boom years appears to have shifted to where they have been over twice as likely to “ride it out” than they were previously.
I don’t know what the long term averages are, they could be higher or lower. Either way, it will be interesting to see if this behavior changes significantly as the market slows down.
As a follow on to Tim’s post yesterday about December stats - the thing that jumped out at me from his post was the growth in inventory. Given that there have been a few comments on the blog about home sales picking up, I thought a data-driven look at what is happening with active listings would be interesting. This data is taken from Tim’s inventory logs, which are available at the upper-left corner of the blog. I used the feed in the third column, which has proven to be the most accurate in tracking month-end inventory levels.
As you can see in the chart above, the growth in listings has been incredibly steady throughout the month - nearly linear. The trend lines through the listing trackers show that we’ve been adding 53 SFH listings and 22 condos for a total of 75 new listings a day thus far this month. If you extend this trend through January 31, and add it to the reported end of month listings for December 2007, here is where we will end up:
In my perusals of real-estate blogs this morning, I ran across a link to Housing Predictor, a site which has recently published a forecast for the top and worst performing housing markets in 2008. Housing Predictor describes their business as follows:
We examine more than 20 micro market characteristics in each market place, including income levels, employment trends and changes, school enrollment, business trends, regional political influences, real estate sales history and current housing market velocity. The factors are all considered to come up with each local market forecast.
All of the information is gathered from hundreds of independent sources the staff develops without bias for the real estate industry to issue the forecasts.
Today Housing Predictor is consulted by the nation’s foremost investment houses, mortgage companies, real estate companies and most importantly consumers for our forecasts. Housing Predictor maintains more than an 85% accuracy rating with its forecasts.
Here is how they show Portland and Seattle in their forecast.
Given what I have seen in the performance of the Case-Shiller indices for the two cities – I found the prediction of a 12.6% spread in appreciation rates to be, well – interesting.
Prior to 2001, year over year appreciation rates in Seattle and Portland seemed to oscillate on different frequencies – but since then have tracked each other very closely. The spread in appreciation between Portland and Seattle has not been that great since 1991. I guess anything could happen – but I suspect the returns in at least one of these markets are going to challenge their claim of 85% accuracy. Perhaps it will be worthwhile to delve more deeply into similarities and differences between these two markets in a later post.
Edit:
Note that this piece has now been the subject of a cheap-shot post by our good friend Ardell over at Rain City Guide - who referenced it in her post entitled “Seattle Bubble Says Seattle Markets Going UP!“.
Most of you, I hope - caught the sarcastic references to the source - for those how didn’t (e.g., Ardell), you might want to read this article
Quick post, NWMLS numbers have been released for December - showing that both Seattle and King County SFH have gone negative on a year-over-year basis.
The median selling price in December was $455,975 in Seattle and $435,000 in King County, down 1.5 percent and 1.1 percent, respectively, from December 2006, according to the Northwest Multiple Listing Service. The city price was down 9 percent from a high of $501,000 in August, while the county price was off 9.6 percent from a peak of $481,000 in July 2007
Edit:
Here are a couple other tidbits. First number is Seattle, second is King County
- Pending sales off 23% and 34%
- Closed sales off 14% and 28%
- Inventory up a whopping 70% and 61%, respectively
It’s hard to put a positive spin on a month as dismal as this one, but based on the articles posted so far - the press release appears to try with a few “Yun-esque” Realtor quotes.
In a news release accompanying the numbers, area brokers said flooding and above-normal precipitation contributed to the typical seasonal slowdown, but expressed optimism about the market.
“Traffic at open houses between Christmas and New Year’s was the heaviest we’ve seen in a long time,” said Dick Beeson, a listing-service director and broker/owner of Windermere/Commencement Associates, in Tacoma. “I believe the bottom has arrived in the Puget Sound marketplace and from here on prices will stay level or advance slightly in 2008.”
J. Lennox Scott, chairman and chief executive of John L. Scott Real Estate, said several factors point to a market turnaround.
“Interest rates are down, sellers have adjusted their prices, apartments are full, job growth is strong and there is a pent-up demand of buyers coming into the market,” he said.
The full text of the NWMLS press release is available over at Seattle Housing Buzz.
