Where is the Seattle bubble hiding?

[Markor]

I'm getting confused... just where is the big housing bubble in Seattle, in the data? The closer I look at the numbers, I just don't see it.

For example, by Fortune Magazine tells us that the (house price) / (annual rent) ratio in the Seattle metro area is currently 38, whereas the 15-year historical average is 23. To revert to the mean the site predicts that Seattle real house prices will fall by 20%.

But let's check that price/rent ratio of 38. The first rental I see on craiglist is a 4 br 1.5 ba 1970s-style Rose Hill (Kirkland) rambler for $1650 / month. It has an old kitchen and blue & green (i.e. yuck) carpets. ( if the link is still good.) That's a $450K house at the most, so the price/rent ratio is 23, the 15-year historical average. No bubble for this house!

According to the site, the house should be salable at $752K. But never was this house worth that much, not even close. It's a typical rental at a typical rental price. So where in the heck is Fortune getting its figures from?

The median figures so often reported look bogus too. Seattle median house prices have jumped 18% since the end of 2005? Well whatever, because the price of a typical 3 br house has not risen that much since then. Maybe half that much, if that.

What I'd like to see is 30 unremodeled Seattle area houses chosen, and let's find out for ourselves what the actual increase in prices has been since the beginning of 1997, the supposed start of the bubble. Not what is being reported based on seemingly bogus data. The number of houses, 30, should give us a result with a good confidence level. More houses = higher confidence level.

The result should be of interest to most on this site. I'm hoping that we can all pitch in. Here's how:

Using Redfin's past sales search feature, find a typical 3-4 br unremodeled, maintained SFR that recently sold and also sold around Jan. 1997 (plus or minus a year should be okay). Input the previous sale date and previous sale price into a spreadsheet. Input the recent sale date and the recent sale price. Divide the recent sale price by the previous sale price; we'll call the result the gain factor. Subtract the previous sale date from the recent sale date; we'll call the result the number of days. Take the gain factor to the (1 / number of days) power (in Excel, = gain factor ^ (1 / number of days)); we'll call the result the daily factor. Take the daily factor to the power of 365.25, the number of days in a year, and subtract 1 (in Excel, = (daily factor ^ 365.25) - 1); we'll call the result the annualized rate of return. Post it here. Or avoid all the math by just posting the raw data (past & recent dates & prices) and I'll crunch the numbers in one swoop.

I think it's a safe bet that the average annualized rate of return for 30 houses in the Seattle area since about the beginning of 1997 is about 7%, which is a little more than a doubling of prices since then. Zillow thinks the rate is 9.7% for King county for the past 10 years. At that rate a house that sold for $250K in Nov. 1997 would be worth $630K today. I highly doubt that, for an unremodeled house.

[Alan]

Overpriced rental. I'm renting in Bellevue for $1150 for 1200 sqft.

[Markor]

To tell whether or not there's a housing bubble today based on rents, only rents advertised today, and the current realistic value of those houses, should be considered.

[Scotsman]

Still for rent? Maybe it's overpriced? I moved in August '07, found acreage with 2700 square feet and a separate 500 foot studio space, all in decent shape for $1700, right at half what it would cost to buy.

From what I've seen, $1400 would be more appropriate for the home you cite. That would put it's value at $386 by Fortune's standards, still well out of reach for the median income for the area. One point might be that despite "historical" multiples if 23 for the Seattle area, 17 seems to be a more readily acknowledged multiplier. Is Seattle really that much higher than the rest of the country in the long run? Or does their number reflect more recent constrictions in rental markets as everyone bought, without regard for long term affordability?

[The Tim]

Markor, first off, I agree with the other commenters that the rental is likely overpriced. Who cares what they're asking. It only matters what they actually find a renter to pay.

As far as your suggestion of tracking 30 houses, it sounds like what you are describing is essentially exactly what the (which is ) does.

According to the Case-Shiller index, home prices from January 1997 to January 2007 went up an average of 9.4% per year. From January 2002 to January 2007 the average was approximately 10.5% per year. Considering that a "normal," "healthy" market really shouldn't have housing prices increasing more than 3-5% per year, I'd say that's a bubble.

Also consider that from January 1999 to January 2002, Seattle was in the midst of the dot-com / 9-11 economic fallout, having been hit harder than many other parts of the country. In a normal situation like that, prices would decline. However, prices still went up by roughly 6.8% per year over that time period. Thank you Alan Greenspan and ridiculously loose, standards-free lending.

[Markor]

[Markor]

[The Tim]

You're missing the point here Markor. Case-Shiller looks only at actual repeat sales of the same homes. They filter out remodels, teardowns, family sales, and so forth. Read (pdf). It is exactly what you're suggesting we spend a bunch of time doing. Why duplicate research that has already been done, and in much greater detail than I have time to do it myself?

[mike2]

the 15 year average takes into account the current run up, so 1) it's overstated and 2) since it's an average, it doesn't necessarily apply to any one house in particular.

Price to rent ratios are, over the course of a cycle very different between extremely desirable neighborhoods, ok neighborhoods and undesirable neighborhoods (and every variation in between of which there are many). The bubble messed this up a bit since anyone with a pulse could pay top dollar for homes in the ghetto.

It's more meaningful to compare the price to rent ratio of a particular home over the years. For this Kirkland home (which I'd add 4br 1 bath is functionally obsolete) it probably rented for $1300/month in 2002 and cost $250K for a price to rent ratio of 192 (or 16 using the monthly scale). Now presumably it rents for $1650 (doubtful) and sells for $450K for ratios of 272 and 23.

Does that mean the ratio was way below average 5 years ago when it was 16? No. Does it mean that the home is not overpriced because it is "average" now? No.

[The Tim]

I just realized another thing that doesn't really make sense about your original post, Markor. You question the current price/rent ratio calculated by Fortune / Moody's, but you just accept their 15-year historical average. The two were obviously calculated using the same methodology, so if you're going to throw out the current ratio and calculate it yourself, you would have to throw out the historical ratio and calculate that manually as well. It doesn't make any sense to pick and choose which data you're going to accept.

[deejayoh]

Markor -
See below. Any way you measure the market, the measure shows the same trend. Case Shiller, OFHEO, Median price all track.



I was going to do a post on the blog on comparing indices, so I happened to have this Y2Y comparison handy

If you want to tilt against this one, have at it - but I doubt it will be worth the time.

[Markor]

[deejayoh]

[The Tim]

Just for giggles I wasted a few minutes on Redfin looking at past sales and looking them up in the county records. Here are some relevant sales I found in the price range you mentioned above.

Parcel# | Sale Date | Sale Price | Sale Date | Sale Price | Years Between | Average Appreciation

9525100070 | 06/13/1994 | $202,500 | 10/23/2007 | $606,000 | 13.4 | 8.54%
2028700315 | 11/15/1996 | $179,000 | 10/09/2007 | $625,000 | 10.9 | 12.15%
2028700360 | 10/21/1999 | $230,000 | 10/23/2007 | $630,500 | 8.0 | 13.41%
2001700010 | 03/03/2000 | $322,050 | 09/17/2007 | $604,000 | 7.5 | 8.69%
7519000145 | 06/12/2002 | $306,000 | 10/01/2007 | $620,000 | 5.3 | 14.23%
7979900885 | 05/20/2002 | $389,000 | 08/29/2007 | $636,400 | 5.3 | 9.77%

Not really interested in continuing this exercise. If the random sampling of data I pulled up even remotely supported suspicions in the Case-Shiller data, maybe I would, but it doesn't.

[biliruben]

"I submit that a typical Seattle-area unremodeled maintained house that sold for $250K in 1997 is not selling for $614K today (9.4% annual gain), and never sold that high for that matter." - Markor

I don't think you have any clue what you are talking about. I can find example after example, without working too hard at all, of exactly this sort of gain, or more. Spend some time on Redfin. I just sampled 5 properties in QA, and all of them had more than tripled since the 1990s.

Perhaps you are looking at properties that were new in the 1990s? New homes don't appreciate nearly as much in this market for some reason. Perhaps that's your problem. If you focus on ugly 1980s and 1990s Eastside crapshacks where you are greeted by giant, ugly 3-car garages and didn't get maintained, then perhaps you are right. Any house I am looking at, i.e. I would actually like to purchase, has tripled in the last 10 years.

You also need to realize that some remodels are basically just maintenance.

You need to update some to just keep the house at the quality at which you bought it. Otherwise it depreciates. Contrary to popular belief, if you don't continuously throw about 1% of the house's value in it every year, it's worth falls rather rapidly.