Are the Miller Samuel Inventory Figures Correct???

FIgures get revised in all types of reports from all types of people. Get over it.

Attached are relvant links:

http://www.millersamuel.com/reports/pdf-reports/MMO1Q09.pdf

http://www.millersamuel.com/reports/pdf-reports/MMO4Q08.pdf

and more links:

http://www.prudentialelliman.com/NYCPhotos/retail_reports/mmo2008q3locked.pdf

http://www.millersamuel.com/reports/pdf-reports/MMO4Q07.pdf

http://www.prudentialelliman.com/NYCPhotos/retail_reports/mmo2008q1locked.pdf

Jason10006: the variances in the last 2 quarters are not normal. These are not normal revisions, not to mention the fact that all other inventory numbers from reports dating back to 2005 were carried forward "as is".

do I smell books cooking?

Miller had been keeping two inventory sets, one of which he had published regularly for some time the other which apparently captured additional inventory that I believe he said had to be backtested. Beginning with the 4Q08 reports he started using the latter. His charts have been updated to reflect the use of the second line of inventory, however the archive reports have not been updated.

I asked him about this on his The Matrix blog back when he first posted the 4Q08 figures, and I recall that is how he explained it.

The following is a link to the discussion we had. My question is post #10:

http://matrix.millersamuel.com/?p=2911#comments

StatsChecker - What email address did you send your request to? I never got either one of them. For future reference, you can send any questions to jmiller at millersamuel dot com That email address is also in every report we prepare.

angler7 and jason10006 - thanks for your comments. Angler7 provided a well articulated and proper description.

It's funny how GDP and other federal stats are revised 3 times at each release and no one complains. I try to improve my methodology to capture more data and revise my historical to reflect the change and that is criticized.

thank you for coming by to clarify. i love me some good in-depth review of statistics, but this had a fishy smell to it, particularly the e-mails sent but not responded to issue.

Jonathan, when revised data are published they are advertised as such. When someone just revises data and makes no mention of why or how or where, then it invites suspicion.

You have a 10-year chart in one of your pdf's - the figures reflect the old ones, not the new ones, and no mention is made of possible shadow inventory. Ergo, the reports are not reliable.

Jonathan,

With all due respect, I think your comment about GDP figures being revied 3x without complaint is disingenous. Everybody knows those stats are revised, because the revisions are made public AS REVISIONS. Accordingly, everbody expects those stats to be adjusted when they are first published.

If your report is to be considered the standard bearer for property statistics in the most relevant property market in the US, then I would think that it is relevant to publish to users of the data that you have a practice which results in past data being adjusted, particularly where as is evidently the, the adjustments have resulted in massive deviations from the originally published figures. If anything, the differences demonstrate how crude your "real time" inventory figures are. If you don't want to publish adjustments, then perhaps you should delay the release of your data until it is back tested. You cannot have it both ways.

Furthermore, neither your report nor your website makes mention that there is any "back testing" that applies. Your website sets out the "methodology" for some of the elements in your figures - and where Listing Inventory is defined there is no mention of backtesting or potential adjustments.

I sent the first email on the morning of Wednesday, May 27th to comments@millersamuel.com, which was the address given on your website. After not hearing, I sent a further email on Friday May 29th asking if there would be a response to my original inquiry.

You send an email to a comments@ email address and expect on friday and expect to have an answer by tuesday? Unless it's an autoreply, I don't see how anyone can anyone expect an answer so fast.

no, i sent an email on wednesday and expected merely an auto reply - not even an answer. i was 100 pct not acknowledged. i heard nothing, zip. then i sent a further email on friday, and heard nothing, not even an auto reply, zip. it seems that jonathan miller can answer on a blog. you would think that he can do an auto reply. if he can't, then he should spend less time on the blog and more time setting up his auto reply.

"There are three kinds of lies: lies, damned lies, and statistics."
--Benjamin Disraeli

I agree with StatsChecker re: the comment about GDP.

The "revised" inventory numbers in the Miller Samuel reports, if correct, amount to differences of 25% in terms of additional inventory. If the GDP was revised by the government by 25% up or down, there would be an intense response (i.e., ridicule) by economists, even when these revisions are expected.

A revision in the order of 25% suggests systemic errors in calculation, not merely "adjustments" in data collection...

StatsChecker and polydoa, are you paying MillerSamuel for the data? If not, you might want to consider cutting them some slack and avoiding the conspiracy-theorist tone. I'm pretty sure MillerSamuel doesn't have the data-gathering and -analysis resources of the U.S. government.

just use SE's numbers then.

"just use SE's numbers then"

What's the point. Soon enough inconsistencies will be discovered and SE will be villified as part of the vast real estate conspiracy.

ha ha. Nice one angler

This sort of debate underscores my long held contention that the NYC inventory numbers are an unreliable data point. It is also highly likely the numbers are even more unreliable now than they have ever been in the past.

When the RE market was in its expanding bubble mode listings appeared and often waited for the market to catch up, and it always did. Today we have different forces at work. Many sellers have a bottom line number in their head that they need to achieve, they set their price and then watch the market deteriorate even further. They will never go into contract at their price. The listing sits, gets stale, and eventually expires. They will try their luck in a year or two or three.

Those who need to sell are more likely to be realistic and price accordingly -if even at a loss. This is the real inventory. In the absence of some elegant algorithm to filter out the stale, overpriced and unmotivated listings I just stick to very limited searches and manually weed out the fat. It gives me a much clearer picture of the state of the market.

You guys are absurd. The first set of numbers are preliminary, the second revised when they CAN be. Its not some grand conspiracy. This very topic has been discussed on other real estate blogs and by Miller himself - even the NYC real estate board has discussed this. If you want more reliable figures, you read the second set.

If you want a real comparison to similar government stats being WAY off, look at the various measurements ofunemployment. They can indeed be over 25% off.

...or for that matter home price changes - the FHA numbers are often WAY off from other measures, because they do not include non-FHA loans.

And if you're really interested in basing your entry into the market on inventory trends, I highly recommend developing and maintaining your own data set. You can debate the big number all day long but in the end what is it really telling you about the market for the kind of apartment you want in the neighborhood you want?

Amen jason10006!

the numbers are valuable to monitor general trends, but spinnaker makes a good point. the only way to know what's happening in YOUR sub-market is to follow it.

it also helps if you can read a few languages. there are plenty of listing that are not being picked up that are advertised in the spanish, chinese, russian, etc... papers exclusively.

All good points. I completely agree that inventory as a primary data point is not reliable as a basis of understanding the health of a real estate market - not because of the way it is collected, but because of the growing level of shadow inventory.

Currently there are more new development units in shadow inventory (new units not yet offered) than the total number of re-sale and new dev listings combined and its growing. That's why inventory is not growing as fast as sales have fallen. As far as local neighborhoods go, I agree its all about local. When a property is listed for sale, inventory matters in the context of what is competing with that specific apartment at that time.

JM... Would you then offer that the primary competition for the shadow inventory will be new and resale condo developments? Clearly if this is the case then one ought to steer clear of this submarket and not lose sight of the fact that the more established (read: prewar) prime markets would and should fare better by comparison.

Why "pre-war"? Any building that is sold out is in the same boat. People in pre-war buildings loose there jobs just like those in Chelsea condos built in 2006 and sold out in 2008.

no, spin is talking about the potential nondos. spin, the thing is that sales of those units in the past few quarters have significantly elevated the numbers. you and i might have realized that the median number is skewed, but it had major psychological impact on the way up. and it will become skewed to the lower side as the condo inventory remains in the shadows for lack of ability to sell, and that will have major psychological impact on the way down.

ok.

although jason he did write new and resale. and your point was a different one. sorry.

"That's why inventory is not growing as fast as sales have fallen."

I don't understand this.

Nonetheless, if there is more shadow inventory (are we talking just Manhattan, or NYC in general?) than listed inventory, and listed inventory is about 11,000 units, and only about 4,000 units are selling a year, then there is a 6-year supply of apartments available for purchase.

Quite the figure, no?

jason... I'm using the pre-war term as a generic catch all for anything that isn't new or a recently new development condo. I could have chosen my words a little better, but the point I'm trying to make is that when the shadow inventory begins to be released for sale the predominant impact will be to those newer developments, not the established co-op's. For the most part they are different buyer pools. Economic impact is another matter and as you say it doesn't matter who who you are or where you live, the effect is generally the same across the board.

I like nondos. I agree with your assessment AR re: we are about to be skewed to the negative side. Median price was another one of those meaningless stats sellers exploited on the run up.. Now it'll be the buyers turn to exploit. The numbers waiting in the wings will send shock waves. Mostly meaningless shock waves, but shock waves nonetheless.

I have already seen that, because while I am looking to rent, many of the places i am seeing are owner or sponsor offered condo/nondos, so I pay attention to all this. I definitely have seen resales in some of the bldgs I have looked at for 20% less than the sponsor offering price of new units. Of course i am perusing mostly harlem , lic, and fidi, but I am sure the point is the same in Tribeca.

Steve,

I note from the Urbandigs site that the most recent 30-day inventory is 11,092 and the most recent 30-day sales are 973.

I've been pretty startled by how robust the past 30-day sales figure has been. Although there is typically a modest seasonality to sales numbers, Miller Samuel analysis has suggested it has only be modest. Annualizing 973 brings you to a very robust figure.

NWT: the fact that i don't "pay" for miller samuel reports has nothing to do with the fact that they are doing a disservice to the actual point of it all - which is to assess the market. btw, for all of you who think i have alleged a conspiracy - you're dead wrong. i have identified a material difference in 3 inventory stats. I have simply asked for an explanation. miller samuel and douglas elliman splatter the media with their quarterly reports - and their website has links to all the times they are quoted in the media. to me this means that they are seaking coverage - seaking to be the main source of info on nyc real estate. if they are going to do this, they should disclose when and why they adjust figures in arrears - especially when those adjustments are so large in terms of the numbers and percentages.

And speaking of an explanation - Jonathan, if you are still reading this - you have explained nothing. you have failed to explain what "back testing" is and you have failed to explain why you don't disclose that a fairly crucial figure is adjusted in later reports.

"Currently there are more new development units in shadow inventory (new units not yet offered) than the total number of re-sale and new dev listings combined and its growing."

That's a shockingly huge statement. Do you have any sources/numbers to back it up with?