Posted by: The Tim

Tim Ellis is the founder of Seattle Bubble. His background in engineering and computer / internet technology, a fondness of data-based analysis of problems, and an addiction to spreadsheets all influence his perspective on the Seattle-area real estate market.

54 responses to “Case-Shiller Tiers: Low Tier Dropping Like a Rock”

  1. Kary L. Krismer

    I wonder if what’s going on with the low end is the attraction of first time buyers to some of the bank owned properties. I’ve been saying a long time that buying such properties made a lot of sense for buyers entitled to the $8,000 credit. That’s especially true where the bank has gone in and done all the painting and flooring. That just leaves some of the other things to fix up, and things that can easily be done after you move in when the credit arrives (unlike painting and flooring).

    Which brings up one of the problems with C-S. They don’t know the condition of the property or the terms of the sale. If the price is seemingly too high or too low they just throw it out of the mix, but for the ones that remain they simply don’t know what’s going on with the property. They’re like Zillow in that regard. The sale from 3-4 years ago could have been one where the seller had to fix it up following the inspection, but the one now not. Or visa versa.

    Rate this comment: Thumb up 0

  2. Ira Sacharoff

    RE: Kary L. Krismer @ 1
    It does seem to me that the banks are treating their homes a bit differently. Maybe not making expensive repairs, but cleaning and mowing.
    Maybe a year ago I saw some really disgusting bank owned homes, homes as if the owners disappeared quickly: pots of moldy spaghetti on the stove, and dirty clothes just strewn all over the floor, with an odor of rotten meat. They really should have included that in the listing photos:)

    Rate this comment: Thumb up 0

  3. Kary L. Krismer

    RE: Ira Sacharoff @ 2 – Clearly, especially Fannie and Freddie. But the changes can still be cosmetic. You can have rather nice looking houses that still need a lot of work.

    Rate this comment: Thumb up 0

  4. One Eyed Man

    RE: Kary L. Krismer @ 1

    This is an exagerated example because of the low price Kary, but I think it illustrates your point. My niece closed on that 2000 sq ft REO in Muncie for $40K in the middle of April. As we all know, the banks like to sell “as is” rather than do the repairs. It needs about 18K of work so even though the price was 40K she’ll be into it about 55K after you deduct the 4K credit. In her case neither the CS nor the MLS median will show the repair costs or the tax credit even though they are a huge percentage of the net price she will pay to get a livable house.

    I don’t know how representative her situation is of the low price issue, but its an interesting theory.

    Rate this comment: Thumb up 0

  5. David Losh

    The low tier is still a quarter of a million dollars. I suspect a lot of those properties are spot lots that were being held by builder developers that they are now dumping. Some may be rentals, and the rest bank owned, but let me repeat they are a quarter of a million dollars.

    In 2008 a blocker in Mount Lake Terrace sold for $250K. Is that realistic? Should some one be paying those kinds of prices?

    Rate this comment: Thumb up 0

  6. Cara

    Kary,

    So your guess is that some of the current low-tier downfall is actually just condition related. Trash-outs and deferred maintainence issues on the REOs. Given that on the way up the properties tended to be all gussied up for highest mark-up, and on the way down will include a lot of HELOC borrowers who may not have had the nicest house but were using it as an ATM anyway and are foreclosing or short selling or otherwise leaving now because they could never afford their own ponzi scheme financing…

    Sounds plausible. Impossible to quantify but definitely a systematic difference between then and now.

    Rate this comment: Thumb up 0

  7. softwarengineer

    RE: Ira Sacharoff @ 2

    I Agree Ira

    But even banks will totally remodel a unit at a loss if it doesn’t sell in a couple years at 40% off the peak price….like the house next door to me.

    It did sell at about 20% off the peak price in immaculant condition…..but I would have lost money even doing the labor on my own and just buying materials….LOL

    Rate this comment: Thumb up 0

  8. Kary L. Krismer

    RE: Cara @ 6 – I would agree it’s impossible to quantify, but it is an explanation for why values would be dropping in the area most attractive to those getting a credit.

    Rate this comment: Thumb up 0

  9. Kary L. Krismer

    Tim, I should know this but don’t. How do condos factor into the C-S mix? Are they included at all in the number we typically discuss?

    Rate this comment: Thumb up 0

  10. Daniel

    By Kary L. Krismer @ 1:

    Which brings up one of the problems with C-S. They don’t know the condition of the property or the terms of the sale.

    This is from S&P: “They [the indices] are designed to measure increases or decreases in the market value of residential real estate in defined regions and the overall U.S.”

    Very clearly, all they want to measure is the market value, and for that none of the above is important, once you averaged over a sufficiently large number of sales. If most low value properties are in bad repair now compared to their previous condition, they are clearly less valuable. If you would like to draw more conclusions you would have to supplement the CS data with other information.

    Rate this comment: Thumb up 0

  11. SummitSeeker

    RE: Kary L. Krismer @ 1
    This is consistent with Kary’s comment the other day that outlying areas are getting hammered, but close in desirable places like Ballard are holding steady or only slowly sinking. Since 1/3 of properties fall in each tier, these outlying areas (which are cheap) are heavily weighted in the low price tier.

    The pain train is coming down the tracks, folks. As Renton crumbles, more buyers will go there, stealing demand from other areas, which in turn will cause those areas to decline. Gradually the decay eats its way up the housing food chain. This is crystal clear to anyone watching CA.

    Rate this comment: Thumb up 0

  12. buystocks

    RE: SummitSeeker @ 13
    Very good point. To use myself as an example. I’ve been looking for a house for almost 2 years now; was initially set on purchasing in MI/Somerset. Then included Newport Hills because of the increasing price gradient, and some time later then included the fringe of Renton because of the increasing price gradient.

    Rate this comment: Thumb up 0

  13. Kary L. Krismer

    By Daniel @ 12:

    By Kary L. Krismer @ 1:
    Which brings up one of the problems with C-S. They don’t know the condition of the property or the terms of the sale.

    This is from S&P: “They [the indices] are designed to measure increases or decreases in the market value of residential real estate in defined regions and the overall U.S.”

    Very clearly, all they want to measure is the market value, and for that none of the above is important, once you averaged over a sufficiently large number of sales. If most low value properties are in bad repair now compared to their previous condition, they are clearly less valuable. If you would like to draw more conclusions you would have to supplement the CS data with other information.

    No way is that a valid argument. First, they don’t measure market value. They measure changes in market value. Second, an argument that the positives and negatives of various properties average out, if valid, then changes in mix don’t affect the median and mean either, because they all just average out.

    Rate this comment: Thumb up 0

  14. Kary L. Krismer

    By SummitSeeker @ 13:

    RE: Kary L. Krismer @ 1
    This is consistent with Kary’s comment the other day that outlying areas are getting hammered, but close in desirable places like Ballard are holding steady or only slowly sinking. Since 1/3 of properties fall in each tier, these outlying areas (which are cheap) are heavily weighted in the low price tier. .

    That’s another good possibility. That’s probably even more likely than my bank owned argument because the outlying areas make up a larger portion of the market than bank owned.

    Rate this comment: Thumb up 0

  15. Scotsman

    RE: SummitSeeker @ 13

    This is a great observation. The acceleration of the lower tier downward doesn’t bode well for the coming months as the middle and upper tiers collapse to follow the lowest tier down. It also seems pretty clear than whatever respite we’ve enjoyed over the last several months has come to and end and another downward leg has begun. So much for the spring ’10 bounce?

    Rate this comment: Thumb up 0

  16. patient

    RE: SummitSeeker @ 13 – I agree that it shows that outlying areas are getting hammered if you specify them as low tier which is likely correct in general, BUT it also shows that Kary’s comment that they drag down the C-S overall number is false since in fact the C-S overall number fell almost exactly the same percentage as the High Tier did, 1% vs. 1.1%. So, since most high tier resides in King Co. where Seattle City is a significant part it proves my point that the overall C-S number is strongly inluenced by King Co. with Seattle City and not so much the outlying areas.

    Rate this comment: Thumb up 0

  17. HappyRenter

    RE: The Tim @ 7

    Sorry, what is usually wrong with blockers? I have seen a couple of those and I always wondered why they are so cheap. For some reason, my wife dislikes them, like this one:

    http://www.redfin.com/WA/Seattle/4034-NE-110th-St-98125/home/112247

    Initially, I liked it but there were too many things I would have liked to change inside.

    Rate this comment: Thumb up 0

  18. Urban Artist

    I might have missed something but what is a blocker? Ballard,Loyal Heights and Crown Hill are still selling pretty well and the prices are not falling very fast. I have seen a few that put a sold sign up and then a week or two later the sold sign is gone and the house is back on the market. Maybe some people have eyes bigger than their bank account.

    Rate this comment: Thumb up 0

  19. David Losh

    RE: HappyRenter @ 19

    A blocker is built out of cinder blocks. In drier climates they are a good good thing. In the pacific northwest they are impossible to adequately ventilate so they have a tendency to mold. and mildew. The outer shell is cinder block, then the studs used to hold the sheet rock on those exterior walls are insulated. It creates moisture that can seep into the studs, along the bottom plate, and saturate the insulation.

    The house in the picture appears to be stick built on site. It is similar in room spaces, but all wood has a better chance to breath.

    Rate this comment: Thumb up 0

  20. Kary L. Krismer

    RE: patient @ 18 – Do you even bother to look at anything? If you were making this argument about the eastside it would actually have some factual support. It’s time to give up your ideas about Seattle proper.

    Rate this comment: Thumb up 0

  21. Jonness

    By Scotsman @ 17:

    RE: SummitSeeker @ 13

    This is a great observation. The acceleration of the lower tier downward doesn’t bode well for the coming months as the middle and upper tiers collapse to follow the lowest tier down. It also seems pretty clear than whatever respite we’ve enjoyed over the last several months has come to and end and another downward leg has begun. So much for the spring ’10 bounce?

    I’m seeing more and more 2004ish prices. It seems we are finally getting to the point of being underwater that has a big impact. The freaking bottom is falling out. Bye bye government handouts.

    Rate this comment: Thumb up 0

  22. Jonness

    By Kary L. Krismer @ 22:

    It’s time to give up your ideas about Seattle proper.

    I refuse to give up my bias. It’s surrounded by water, and they don’t make land anymore. It became a world-class city starting in January, 2003. No way will prices dip below July, 2007. :)

    Rate this comment: Thumb up 0

  23. Daniel

    By Kary L. Krismer @ 15:

    No way is that a valid argument. First, they don’t measure market value. They measure changes in market value.

    Nonsense. You know the normalization point of the index and you know the change so you know the market value.

    Second, an argument that the positives and negatives of various properties average out, if valid, then changes in mix don’t affect the median and mean either, because they all just average out.

    I have never claimed this (learn to read!): All I meant is that any shift in the distribution affects the value so it is naturally included. The only reason I had to include the limit of large statistics is for the current (in the sense of most recent as measured by the index) distribution to be sampled reliably.

    Rate this comment: Thumb up 0

  24. Daniel

    RE: Kary L. Krismer @ 1 – Before I forget it, from the methodology

    “We can control for heteroskedastic errors,
    thereby increasing the precision of the index estimates, by applying weights to each of
    the sale prices before estimating the index points.”

    These errors are the type you talk about. They describe it in some detail and link the original publication.

    Also:

    “The high-tier indices will tend to lie closer to the aggregate indices than do the low-tier
    indices. This is as we would expect, since the aggregate indices are value-weighted and
    hence the high-tier repeat sales figure more prominently in the aggregate indices.”

    This is logical, after all they want to measure the value of a housing portfolio.

    Rate this comment: Thumb up 0

  25. Kary L. Krismer

    By Daniel @ 25:

    By Kary L. Krismer @ 15:
    No way is that a valid argument. First, they don’t measure market value. They measure changes in market value.

    Nonsense. You know the normalization point of the index and you know the change so you know the market value..

    So you’re assuming that if you take the median from Jan 2000 (which I think was 100 w/ C-S) and multiply that by the current C-S number, you’ll get your current median?

    I think they might actually intend such a thing with respect to individual properties. Take the index when you bought and multiply, but it won’t work. A $200,000 house bought in X/2002 would appreciate differently than a $2,000,000 house bought at the same time.

    The point is it doesn’t show value. And despite using values, neither does the mean or meidan. All three are just a measure of market strength.

    BTW, I do find it a bit ironic that on a site that is so into statistics no one seems to like the statistics of mean and median, but instead like a number that is not even a statistic. C-S is a result of statistical analysis, sort of a black box analysis, but not a statistic in and of itself.

    Rate this comment: Thumb up 0

  26. Kary L. Krismer

    RE: Daniel @ 26 – That last part is interesting. I’ll have to think about that.

    Rate this comment: Thumb up 0

  27. drshort

    By Kary L. Krismer @ 27:

    BTW, I do find it a bit ironic that on a site that is so into statistics no one seems to like the statistics of mean and median, but instead like a number that is not even a statistic. C-S is a result of statistical analysis, sort of a black box analysis, but not a statistic in and of itself.

    This site likes the mean and median just fine when they’re dropping like a rock.

    Rate this comment: Thumb up 0

  28. patient

    RE: Kary L. Krismer @ 22 – The only thing I’ll give up on is that there is hope for you to listen and learn. What I understand you failed to see the bubble and even bought close to peak but you continue to hold on to your failed methodologies and scewed logic. Many people who saw the bubble was attracted to this site early and have learnt a lot about what mix of metrics and methodologies to value.

    Rate this comment: Thumb up 0

  29. Daniel

    By Kary L. Krismer @ 27:

    So you’re assuming that if you take the median from Jan 2000 (which I think was 100 w/ C-S) and multiply that by the current C-S number, you’ll get your current median?

    No. The metric is set by the portfolio value at the time the index was 100. It is like measuring in units of 2000 housing portfolio prices, just like you can measure a distance in meters or feet and like you can measure energy in Joule or somewhat less obvious energy units like the temperature unit Kelvin. Does this easily compare to everyday units? No. Is it sensible for the purpose of measuring and comparing portfolio values? Surely.

    I think they might actually intend such a thing with respect to individual properties. Take the index when you bought and multiply, but it won’t work. A $200,000 house bought in X/2002 would appreciate differently than a $2,000,000 house bought at the same time.

    I could just say RTFM but instead again: They explicitly mention that the intent is NOT to value individual propertis but rather certain types of real estate portfolios. The reason for their tiers is exactly that: They are aware that the low end behaves different than the high end.

    The point is it doesn’t show value.

    Repeating the same nonsense does not make it true: It is an arithmetic stochastic estimator of the aggregate value of well defined real estate portfolios. It measures this values in a “unit” defined by the index point.

    C-S is a result of statistical analysis, sort of a black box analysis, but not a statistic in and of itself.

    It is a black box to those with limited mathematical understanding but to people who actually read and understand the methodology it may just be obvious why it is well suited for estimating portfolio values.

    Rate this comment: Thumb up 0

  30. Masaba

    RE: Kary L. Krismer @ 27

    Kary, you seem to think that a SAMPLE mean and median are the same as the true mean and median. Both of these are ESTIMATES of the true values, and in this way they are very similar to the CS index value, which also gives a normalized ESTIMATE of home prices in an area.

    The only way you could say which of these is better is if you actually computed the mean squared error (MSE) for each estimator. MSE is a sum of two values, the squared bias of the estimator + the variance of the estimator.

    Assuming that your samples are uncorrelated (I’m pretty sure this wouldn’t be true in housing) the SAMPLE mean is unbiased. However, we often talk on this site of the ‘volatility’ of the mean, which equates to the variance. The sample mean can have quite a bit of variance. This makes sense, because in a month with a few number of sales the SAMPLE mean is really susceptible to outliers.

    This is why a lot of people prefer the CS index. And in fact, from a statistical standpoint, the CS index may be a better estimator (if it has lower MSE than the SAMPLE mean).

    If you prefer the SAMPLE mean to the CS index, that is fine. However, your implication that it is better than the CS index from a statistics perspective is just not true.

    Rate this comment: Thumb up 0

  31. D. in Ballard

    RE: Urban Artist @ 20

    I’ve been paying attention to Ballard, and I’d say that houses that are near perfect are selling and at prices not far below 2007 prices. However, we’ve spotted a lot of well-priced homes that have something ever so slightly wrong with them. No entry way, small lot, bathroom on main floor not well-positioned. These houses don’t seem to be going anywhere.

    Rate this comment: Thumb up 0

  32. Kary L. Krismer

    RE: Masaba @ 32 – The main ways the NWMLS mean and median are “better” than the C-S index is that they are a lot more current and verifiable. But the difference in result is so insignificant that it doesn’t really matter which is “better” in other ways.

    But again, all three are just measures of strength of the market. If you’re trying to determine value of any property from any of them, you’re going about it completely wrong.

    Finally, I’m not sure why you’re saying the mean and median are a sample. They’re the entire population of all listed properties for the period covered. Listed properties are a sub-group of all sold properties, but it’s a very large sub-group. It does “dominate” the whole.

    Rate this comment: Thumb up 0

  33. Masaba

    Kary, I am not going about it ‘all wrong’ to say that the sample mean is an estimate of the average value of all houses within an area. In fact, if every house in an area sold in a single month, then it would be a PERFECT estimate. Furthermore, if only one home in Seattle sold in a single month, then it would most likely be an awful estimate.

    From the CS methodology handbook: ‘The S&P/Case-Shiller U.S. National Home Price Index (“the U.S. national index”) tracks the value of single-family housing within the United States.’ I think many people view the CS index as a normalized estimate of the average value of housing stock. Therefore, if you really want to compare the CS index to the sample mean or median, you should compare their ability to do the same thing; which in this case would be to estimate the average value of homes in an area.

    Your reasons for preferring the mean or median are fine. As you have said, you think that they are more interpretable than the CS value. The ability to interpret an estimate is commonly sighted as a redeeming quality. You have also said that they are localized more to the area you want to look at, which is also good. They are also more timely.

    However, they also have negative qualities. They have more variability than the CS index value. I would also assume that housing sales in a single month are somewhat correlated; therefore, the sample mean or median are probably biased.

    There is nothing wrong with preferring one over the other, but there is a problem when you dismiss one arbitrarily as you seem to do.

    Rate this comment: Thumb up 0

  34. Daniel

    By Kary L. Krismer @ 34:

    But again, all three are just measures of strength of the market. If you’re trying to determine value of any property from any of them, you’re going about it completely wrong.

    Again you repeat the same irrelevancies: Of course using the index for something it is not meant to estimate leads to nonsense. That does not change that the CS index measures what it intends to measure quite well. I guess if I was you I would be very embarrassed: You work in this industry and did not even know what it measured.

    Rate this comment: Thumb up 0

  35. Kary L. Krismer

    RE: Daniel @ 36 – I guess I’m going to have to ask you what you think the C-S index does? How is it relevant to anyone if it isn’t to measure market strength?

    Or perhaps more to the point. If you own, want to sell, or want to buy a particular house, how do you thing the C-S index is at all useful?

    Rate this comment: Thumb up 0

  36. Kary L. Krismer

    RE: Masaba @ 35 – Part of the reason that the mean and median have more variability is that C-S is a three month moving average. Tim a couple of months back did a chart comparing the King County Median 3 month average to C-S. If he sees this he could probably find that a lot easier than I could. As I recall, the median still varied more, but again we’re not dealing with as large of area with the median. Smaller areas will vary more.

    Rate this comment: Thumb up 0

  37. Daniel

    RE: Kary L. Krismer @ 37 – The following is from Case, Karl E. and Shiller, Robert J., Prices of Single Family Homes Since 1970: New Indexes for Four Cities (September 1987). NBER Working Paper Series, Vol. w2393, pp. -, 1987

    You find this part on page 25. Notice that the NAR practice has changed over time but the basic statements remain correct.

    “The NAR publishes the median sales price of existing single family homes. That figure depends on the characteristics of homes that are sold in a given period as well as on the level of prices. The NAR is careful to point out that their numbers are not meant as an index of appreciation. [...] Despite their warnings the popular press often interprets their figures as appreciation.”

    The Case Shiller index measures appreciation of a housing portfolio (for the stupid among you: not of single houses) while the median is entirely ill suited to do so.

    It is also interesting to note that the appreciation was lower than the median price would suggest in all markets investigated in above paper Sometimes even only a forth of the rise in median price!

    How hard is it to get that in your head?

    Rate this comment: Thumb up 0

  38. Kary L. Krismer

    RE: Daniel @ 39 – How hard is it to get into my head? Are you purposefully trying to be difficult, trying to be an A-hole, or is it impossible for you to understand a simple set of questions? Go read #37 again and try to come back with an appropriate response, if your brain is capable of doing more than quoting stuff.

    Rate this comment: Thumb up 0

  39. EconE

    Being that the banks basically “own” the housing market now. Wouldn’t it be pretty easy for them to “game” the median? They can do it city wide or even within a particular zip code where the variance can be high. If they want the median to go down…roll out a few more crack-houses to sell. If they want it to go up, they can sell some of their “better” stock.

    Shiller? Is he going to run a historical check for the selling price of EVERY property?

    I think you either have to do your own homework…or have your agent do it for you.

    Regardless, I think that the best agents will be the ones to tell you why you DON’T want to buy a particular house rather than the other way around.

    Rate this comment: Thumb up 0

  40. Daniel

    By Kary L. Krismer @ 40:

    RE: Daniel @ 39 – How hard is it to get into my head? Are you purposefully trying to be difficult, trying to be an A-hole, or is it impossible for you to understand a simple set of questions? Go read #37 again and try to come back with an appropriate response, if your brain is capable of doing more than quoting stuff.

    Why would I _want_ to answer your questions? You made some clearly nonsensical and moronical statements to which you ignore my replies and you have proven that you have no clue what the CS index actually measures. Now you try to distract from that with some nonsense questions and expect me to reply… Maybe your clients do but I won’t fall for that.

    Rate this comment: Thumb up 0

  41. April Seattle Newsletter: Seattle Prices Dip (Surprise, Surprise) | Redfin Seattle Sweet Digs

    [...] the full conniption on the pricing drop, visit SeattleBubble, a blog dedicated to the idea that Seattle homes have long been [...]

    Rate this comment: Thumb up 0

  42. Kary L. Krismer

    RE: Daniel @ 42 – Fall for what? Your name calling says a lot about you and the limits of your abilities.

    Rate this comment: Thumb up 0

  43. David Losh

    RE: Kary L. Krismer @ 44

    Sorry Kary, but you’re not following the presentation of facts. Case Schiller is an Index. It’s meant as a broad stroke interpretation of values. It is making assumptions that price, sales data, is what determines value.

    Case Schiller has been proving to be a vary bad market indicator. It is, as you are pointing out, as useless as looking at median pricing.

    Banks do manipulate the market place, banks do use this Index to determine market strength. The Index is a financial tool used solely to demonstrate market trends.

    Rate this comment: Thumb up 0

  44. Kary L. Krismer

    By David Losh @ 45:

    Banks do manipulate the market place, banks do use this Index to determine market strength. The Index is a financial tool used solely to demonstrate market trends.

    How is that different than what I’m saying, which is it’s a measure of market strength?

    BTW, one other issue with C-S, and/or a reason why it would vary in different amounts from mean/median, it is ignores new construction, because there would be nothing to pair it to.

    Rate this comment: Thumb up 0

  45. David Losh

    RE: Kary L. Krismer @ 46

    It’s much different, it’s an Index. It’s a list of numbers used to engineer lending formulas. We can lend X against Y because the market says that people will spend this much more, or less, over a period of time. The Index itself is also based on the idea that there is a market psycology. Schiller has said this repeatedly over the years.

    More plainly put, the Index is a meaningless list of numbers used for manipulating the market place. It has nothing to do with the housing units, or sales, or sales data, it has to do with creating a market trend. It’s an Index. It can be manipulated.

    Rate this comment: Thumb up 0

  46. Lurker

    Nice one Bubble, you got linked in Redfin!

    Rate this comment: Thumb up 0

  47. Daniel

    By Kary L. Krismer @ 46:

    BTW, one other issue with C-S, and/or a reason why it would vary in different amounts from mean/median, it is ignores new construction, because there would be nothing to pair it to.

    For the purpose of the index (measuring the value of real estate portfolios versus some reference point, for example at the time of investment) this is not an issue, it is a necessity. Your comment is exactly the type of statement I referred to previously. What you do is like pointing out that a tennis racket isn’t a good can opener: It is true but neither surprising nor insightful.

    Rate this comment: Thumb up 0

  48. Kary L. Krismer

    By Daniel @ 49:

    By Kary L. Krismer @ 46:
    BTW, one other issue with C-S, and/or a reason why it would vary in different amounts from mean/median, it is ignores new construction, because there would be nothing to pair it to.

    For the purpose of the index (measuring the value of real estate portfolios versus some reference point, for example at the time of investment) this is not an issue, it is a necessity. Your comment is exactly the type of statement I referred to previously. What you do is like pointing out that a tennis racket isn’t a good can opener: It is true but neither surprising nor insightful.

    I guess you are just trying to be difficult. Of course it’s a necessity for C-S to do it that way–as I pointed out there’s no pair to the current sale. But that’s irrelevant to my point that the NWMLS mean and median include all sales, not just resales like C-S. That means that even if they did cover the same geographic area you would still not expect a near perfect correlation. If the NWMLS did report just resales by themselves, you would expect that to be a better correlation to C-S than the all sales median. That it’s a “necessity” for C-S to exclude new construction doesn’t matter. It’s a difference and that was my point.

    Rate this comment: Thumb up 0

  49. Daniel

    By Kary L. Krismer @ 50:

    I guess you are just trying to be difficult. Of course it’s a necessity for C-S to do it that way–as I pointed out there’s no pair to the current sale.

    That is not why it is necessary: New construction needs to be excluded as you want to measure appreciation of construction (or rather real estate including the land). For appreciation, new construction is irrelevant and has to be excluded. The fact that it is included makes the median over all sales useless to measure appreciation.

    It’s a difference and that was my point.

    I know that was your point: figuratively you just repeated _again_ that a tennis racket is indeed different from a can opener.

    Rate this comment: Thumb up 0

  50. Kary L. Krismer

    RE: Daniel @ 51 – I actually would prefer that the NWMLS publish an official median that excludes new construction, because I don’t deal in new construction. Although my guess is new construction doesn’t affect the median as much as it does the mean–there are too few new construction sales to have a very significant effect on median.

    But even if they did publish such data, it wouldn’t be useful for determining appreciation of specific property, just as C-S is not useful for such a purpose. But it would be a better indicator of market strength in the area I work.

    Rate this comment: Thumb up 0

  51. Kary L. Krismer

    RE: Kary L. Krismer @ 52 – I just did a search of April sales and July, 2007 (peak) sales, King County SFR. Even during the peak new construction only affected the median about $5,000, where today it’s about $2,500. Both short sales and REOs have more of an effect than new construction, and REOs affect the median more than short sales (both because REOs tend to be cheaper than short sales and there are more of them).

    Numbers from NWMLS sources, but not compiled are guaranteed by the NWMLS.

    Rate this comment: Thumb up 0

  52. Realogics: “In-City Housing Demand Trends Directionally Positive” | Seattle condos on Urbnlivn

    [...] about that, the non-seasonally adjusted numbers show Seattle still falling across all price bands, Case-Shiller Tiers: Low Tier Dropping Like a Rock. For a predominantly condo firm like RSIR we’d have talked about how the Case-Shiller index [...]

    Rate this comment: Thumb up 0

Leave a Reply

Do you want a nifty avatar picture next to your name, instead of a photograph of Tim's dog? Just sign up with Gravatar, and make sure to use the same email address in the form below. It's that easy!

Please read the rules before posting a comment.

You have 3 comments remaining on this post.

Archives

Find us on Google+