Results 1 to 8 of 8

Thread: What the recent real estate reports do appear to say

  1. #1
    Join Date
    Mar 2009
    Posts
    13

    Default What the recent real estate reports do appear to say

    I have analyzed/compared the full Warren Group statistics for 2007/2008 and Ben's raw data from MLS using a new method derived from the Warren Group analyst's feedback in one of Alan's posts.

    It appears that ten percentage points of the twenty percentage point drop in the Warren Group's median home price for Wayland can attributed to the fact that in 2007, thirteen more homes were sold in the $630K and above price range.

    In summary, the distribution of sales in 2007 is sufficiently skewed to the high end to pull the median value into a relatively sparsely populated region of the distribution. As a result, the removal of 13 home sales from above the median has the effect of shifting the median home sale by 6.5 units (from a value of $631,000 to a value of $570,000).

    The remaining ten percentage points of the twenty percentage point drop reflected in the Warren Group statistics appear to be attributed to a downward shift in the entire distribution.

    The root cause/possible actions remain to be determined, but I want to try to validate the analysis with the Warren Group, before we put too much energy into that.

    What is curious is that the Warren Group and Ben's MLS statistics are identical for 2008, but Warren Group has 6 more sales and a 5% higher median for 2007. This is not explained by the fact that Warren Group eliminates non "arm's length" sales.

    I have a "magic" spreadsheet with the probability density functions (histograms) for Ben's raw MLS data, but I am having trouble with the attach functionality again - I will try later.

    /Gary

  2. #2
    Join Date
    Nov 2005
    Location
    Wayland MA
    Posts
    1,431

    Default

    As an FYI, Gary's post above continues nicely from this thread.

  3. #3
    Join Date
    Apr 2009
    Location
    We live in Mainstone Farm at Stoneridge Village
    Posts
    12

    Default I will provide some of spreadsheets soon.

    I have just discovered and joined this forum. I was very interested to read the comments since I have twice weighed in on this issue in the local paper. I will share my spreadsheets and what I have learned since my first letter to the editor that was based solely on MLS data as time permits. I am impressed by the comments I have seen and must confess I am not a statistician and cannot comment on the analytical approaches discussed here. I can provide the MLS raw data that may be useful for others to ponder.

    Let me say at the out set that I am professional real estate broker. I am a rather rare bird in the real estate industry as an Exclusive Buyer Agent known as an EBA in the lingo, defined by National Association of Realtor (NAR) as one who does not take property listings. I have just completed a term as President of the National Association of Exclusive Buyer Agents (NAEBA) and was one of its initiators in 1995. I avoid the conflicts of interest inherent in dual agencies that represent both buyers and sellers often in the same transaction. While the state has lowered the standards in 2005 that enable dual agency practice I still function under common law principles of full fiduciary duties to my clients. I am making that clarification so that folks understand that I have no vested interest in any property listings in Wayland or anywhere else for that matter.

    More to come...

    Barry
    Last edited by Barry Nystedt; 04-10-2009 at 04:54 PM. Reason: added words

  4. #4
    Join Date
    Nov 2005
    Location
    Wayland MA
    Posts
    1,431

    Default

    Barry, I've seen your Town Crier columns and found them interesting. The rather simple comparison that I'd like to see is as follows--I'm happy to do it if you have the data.

    1. Map the 2007 and 2008 sales. Google Maps allows this fairly easily--it would probably take a few hours to do on the order of 250 properties.

    2. Find the average and median house square footage for 2007 and 2008.

    3. Find the average and median lot size for 2007 and 2008.

    To be sure, there are many other factors that define a property, condition being an important one.

    Based on analysis by Gary, Kim, and others, which seems to show a 10% value drop and a 10% "apples-to-oranges" drop, I would expect to find that the 2007 houses and lots were larger than 2008, and that the locations were "better." But, those differences wouldn't explain the full 20% Warren Group number.

    It would be especially interesting to do the same thing for several other communities--for instance, Acton, Sudbury, and Weston. Collecting the data is hard (for me), analyzing it is relatively easy (for we of the "magic spreadsheet" crew). Of course, as Gary found, it might be necessary to look at 2006 and 2005 as well.

    For the record, I want to know what is actually going on. If Wayland is truly under-performing relative to peer communities, I'd like to know that so that we can consider how to get back on track. And if Wayland is just one more structurally sound boat on a currently outgoing tide, I'd like to know that as well.

  5. #5
    Join Date
    Apr 2009
    Location
    We live in Mainstone Farm at Stoneridge Village
    Posts
    12

    Default Updated spreadsheet with public record SF, lot size & numerous calculations

    Jeff, per your request I have integrated the initial work I did with some updates including public record living area square footage, and lot size square footage.

    I do not think I can find the time to map the locations in the next few days.

    I decided I would look at public record square footage because listing brokers often exaggerate the number and often include finished basement areas which do not add that much value and never equate to above grade living area. One Wayland agent who has moved out of town with whom I had some contact on the issues with the assessorís data a few years ago told me out right that Wayland agents lie about square footage. She called me upset about my first letter to the editor and the approach I had taken. Well, I decided to check every transaction. I was amazed to find the amount of errors in the MLS data on these Wayland listings. Very frustrating to my use of the data on market trends, not only that, but it is misleading to the public in most cases when the agent cites public record as the source in the MLS listing sheets even when the square footage is wrong. I have added that information to the spread sheet. When I do a market analysis for a client I always check for and use the public record living area unless there is really solid evidence to the contrary like an addition to the property than may not have been reflected in the public record. Based on the public record correction the median $ per square foot change would be down by only 1.6%. What is tricky about this number is that usually a smaller property would have a little higher dollars per foot value.

    Based on the data so far, I would be surprised if the decline is actually as high as 10%.

    Another factor in this apples to oranges data is the house style. In doing a market analysis you would not use ranches to value colonials or vice versa.

    In terms of the land, when consumers are buying they do not generally make any distinction between the house value and the land value that is characteristic of assessments. The qualities of the lot are important as are the other amenities of the home but they look at it very much as a whole package without really accounting for the value of the land. If there is an over sized lot or an extra piece of land it can actually be seen as a disadvantage for many folks.

    Hope this is helpful to the number crunchers on the forum. I will be very interested in their perspectives.

    Let me know if I can provide further data.

    I started to explore mapping software that would take the spreadsheet data and apply it to a map but it is quite expensive and I could not get a timely answer on their consumer line.

    Barry
    Attached Files Attached Files

  6. #6
    Join Date
    Apr 2009
    Location
    We live in Mainstone Farm at Stoneridge Village
    Posts
    12

    Default

    I just reviewed the MLS sales again visually by style of house arranged in sale price ordrer to see if there were any near identical sales that could be used in a paired analysis. I really did not find any, amplifying the apples to oranges issue.

    There was only one resale at 5 Sylvan Way which showed a 3.7% decline, obviously not enough to draw any conclusions from.

    In my spreadsheet if you subtract the decline in the average public record living area from the average price decline you get a decline of 4.3%. If you subtract the decline in the median public record square footage from the median price decline you get a 1.3% decline.

    There are no absolutes in determining real estate values. When relocation companies acquire property they usually order two appraisals and if they are not within 5% of each other they order a third and sometimes more. That procedure is over and above a Brokers Opinion of Value.

  7. #7
    Join Date
    Mar 2009
    Posts
    13

    Default Warren Group conversation

    I talked to a person from Warren Group on Friday. It was the same person Allen talked to.

    In Summary:

    1. The Warren Group is a data collection company, not a data analysis company- that is where both their expertise and reputation are based.

    2. There can be "statistical anomolies" in 1yr data that affect two year comparisons (2007-2008 changes). We may have uncovered one.

    3. The four year data (2005-2008 changes) is more representative of depreciation as a result of the real estate bubble bursting than two year data (2007-2008 changes). The big picture also would have to take into account how much a town's prices inflated during the bubble.

    4. The Warren Group does include sources beyond MLS, including for sale by owner and bank sales, but neither seem a good explanation for the additional high end sales they recorded relative to MLS in 2007. There is the possibilty of an error in their data, but I chose not pursue this path for now.

    5. They gave Allen more detailed data (I only have their medians for every year) if we want to do more analysis.

    6. One should not expect the same level of rigor out of Boston Magazine as one would expect from an economic/real estate publication. (Of course, impact is a function of readership, so this doesn't help us with the perception issue.)

    Logical next steps would be:

    1. Look at 2006 (and possibly 2005) Wayland data to see if it more closely resembles 2007 or 2008- to better determine which year is the "anomoly".

    2. Compare Warren Group and MLS 2007 data more closely looking for evidence of a mistake.

    3. Look at data from other towns to see if they have exhibited similar behavior, but still dropped less.

    I think we should strategize a little before we do much more additional work.

    This overall problem seems to exhibit traits of Heisenberg's uncertainty principle - the more carefully we attempt to specify one dimension (the attributes of a sale), the smaller the sample size and hence the higher the degree of uncertainty in another dimension.

  8. #8
    Join Date
    Apr 2009
    Location
    We live in Mainstone Farm at Stoneridge Village
    Posts
    12

    Default MLS Sales Information for 2004 to 2008

    You may not have had a chance to review the spreadsheet I submitted previously which identifies the individual transactions. I do not believe Banker & Tradesman has made a mistake. They draw their information from public records. The MLS data calculated for median price is pretty close to their data. They have more transactions because they include direct sales that includes for sale by owners as well as family and paper transfers that may not reflect market values.

    I have attached summary reports I downloaded from the MLS starting in 2004 which was at least the peak year of this cycle for number of sales. the S&P/Case-Shiller Home Price Index shows the price peak in this region to have occurred in Sep '05.

    Through the MLS I can also access Banker & Tradesman raw data from the public record though with much less detail. If you would like me to provide that let me know.

    Hope this is helpful.

    Barry
    Attached Files Attached Files
    Last edited by Barry Nystedt; 04-13-2009 at 05:31 PM. Reason: Add attachment

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •