Posted by Jonathan J. Miller -Thursday, May 10, 2012, 10:50 AM
A few days ago, a post by Felix Salmon at Reuters caught my eye: Chart of the day: Let’s go buy a house! Yesterday he asked me to send similar data for NYC and would run the same chart. I sent over 20 years worth of median sales price and median rental price (face) data for Manhattan and he punched one out: Rent vs buy, Manhattan edition laying the results on top of the US data.
He’s running a payment equivalent adjusting for inflation and he says:
Obviously the Manhattan data series, with fewer transactions, are much noisier than the national series. But broadly speaking, it costs you the same amount to buy a house today, in terms of your monthly mortgage payment, as it did at the end of 2004, when the median sales price was just over $600,000.
Here’s what I told him when admiring his chart handiwork:
We are def moving into a gray area where we are now seeing more and more Manhattan individual apts as cheaper to buy than rent in our appraisal practice – especially coops since they are cheaper than condos. Obviously the problem remains whether the buyer is credit worthy.
Since this analysis is in aggregate, there is not a “tipping” point where the line is crossed and everyone runs out and just starts buying apts (i.e. Justin Bieber tickets).
Posted by Jonathan J. Miller -Thursday, April 5, 2012, 2:02 PM
[click to open press release]
One of my issues with existing national price indices (I have many) has been that they reflect what happened after the fact. That in and of it self is not a bad thing at all. The problem concerns their use by the consumer and media. They rely on them and often have no idea of the severity of the trend lag (as much as 6 months). This lag is interpreted as the current market and then they proceed to mischaracterize or misunderstand what’s actually happening in housing right now.
Jed Kolko, the Trulia’s chief economist has come up with what looks to be a much better way to look at the direction of housing prices by following list price trends which lead home price trends by several months. He’s also created The Trulia Rent Monitor which addresses the same issues on the rental market. Both reports are based on what Trulia does well, aggregating and managing listing information by the boatload.
Trulia Price and Rent Monitors – March 2012 Download
The Trulia Price and Rent Monitors rely on the latest asking price or rent rather than the original to better track the direction of the market. Prices on MOM, QOQ and YOY on based on a 3 month moving average. Here’s the nitty gritty. Love the “technical” and “non-technical” FAQ notes detailing how it works. Jed is very clear that this is not a way to “game” the existing indices like Case Shiller and predict them in advance of their release (aka accurately predict what a 4-6 month old index result will be tomorrow) which serves an entirely different purpose I suppose.
I thought it was particularly interesting that some speculative and depressed markets are showing the most upside swing – i.e. Detroit, Miami, Phoenix. CA still weak throughout. The NYC metro results are consistent with what we are seeing throughout the region, prices down 3.3% YOY and rents are up 6.2% YOY.
From the press release, the Trulia Price Monitor for March 2012 shows:
- Asking prices up 1.4% quarter-over-quarter, seasonally adjusted. This is the first clear indication of a national home-price turnaround. Unadjusted for seasonality, prices were up 2.4%.
- Asking prices up 0.9% in March and 0.6% in February, month-over-month, after bottoming in January 2012.
- Strong year-on-year increases in asking prices throughout Florida, and year-on-year price declines throughout California.
The Trulia Rent Monitor for March 2012 shows:
- Rents up 5.0% year-over-year.
- Rent increases in nearly all large metros, especially metros with faster job growth.
Note: I have been on the Trulia Industry Advisory Board since its inception in 2006.
- Why US Housing Indices Make Terrible Investment Benchmarks [Matrix]
- Asking Prices on the Rise as Housing Recovery Expands [Trulia]
- Trulia Price and Rent Monitors – March 2012 [Trulia]
- Trulia Price and Rent Monitors – FAQ [Trulia]
Posted by Jonathan J. Miller -Wednesday, March 21, 2012, 10:39 PM
Had a great conversation with Aaron Task at Yahoo! Finance’s The Daily Ticker on NAR’s Existing Home Sale numbers and the state of the housing. First time I’ve been on the show.
Posted by Jonathan J. Miller -Tuesday, March 20, 2012, 2:27 PM
One of the by-products of my Three Cents Worth column (3CW to those in the know) is the generation of other data in the process of coming up with a chart.
As a side bar to my Curbed post this week, I also created the tallest chart ever with every residential sale (co-ops, condos and townhouses) from 2003 to 2011 (plus January 2012) in a single chart. I thought it was important to show better context in the market data since we all throw around $88M like it was a generic sales event.
Hey Guiness Book of World Records – check out the chart. ====>
Click graphic to the far right to see it in all its glory. =================>
- Tallest Chart in the History of Manhattan Real Estate [Miller Samuel]
- [Three Cents Worth NY #182] Charting Manhattan’s Sales History [Matrix]
Posted by Jonathan J. Miller -Monday, March 19, 2012, 8:41 AM
Jed Kolko, Trulia’s Chief Economist, let me sneak a peak at his excellent post “Springtime for Housing” as he was placing the finishing touches on it last week. Knowing me all too well, he asked me to set aside one of my pet peeves with econo-housing stats – the use of seasonal adjustments – and focus on the long term view.
To understand the effects of long-term trends or one-time events on the market, housing wonks like to “seasonally adjust” data. That means we strip out the regular seasonal patterns in order to highlight trends or events. This is useful for deciding whether the market is really in recovery or assessing the impact of a housing policy.
He came up with some terrific visuals and identifies some interesting points. One of the most powerful to me was the fact that seasonal sales activity fluctuates a lot more than inventory.
Rising sales and build-up of inventory make for a spring housing market cocktail.
Posted by Jonathan J. Miller -Tuesday, February 28, 2012, 3:00 PM
Analysts have had an easy time forecasting the future results of the Case Shiller Home Price Index with reasonable accuracy. This is because the 3-6 month lag from the actual “meeting of the minds” to reporting period results as well as the lower costs to acquire data to analyze.
If savvy analysts can accurately predict the index in the coming months, then how does this encourage investors to get on both sides of the trade. Call me crazy but if I knew what the index results would be in the coming months, that would mean that most market players would know and then it’s not possible to beat the market (sorry traders, but this is my weak attempt to talk like a trader).
The chart above is a forecast found in a recent housing study. It’s how I currently envision the US housing market given the millions of distressed property sales that need to go through the system.
The S&P/Case Shiller Home Price Price Index’s reason for being, as well as other competing indices were intended to be a benchmark for Wall Street to hedge housing using options (derivatives). This would have served a very logical and useful purpose to be sure but it has been a dismal failure. In one of the most volatile housing markets in history, trading volume has been anemic or non-existent, or at a fraction of the volumes needed to be an efficient market. Still, S&P/CS has become the consumer benchmark as the media grapples with how to characterize the market. The data problems and years of messaging bias over at NAR Research have enabled alternatives like S&P/Case Shiller and Corelogic.
The S&P/Case Shiller December report published today, (2 months after the close of the period, reflecting October-November-December closed sales, reflecting August-Spetember-October contracts) reflects a continuing malaise in US housing even though the 2011 housing market got a reprieve thanks to the robo-signing scandal at the end of 2010 and the anticipated settlement agreement between the major services and the 50 US state attorney generals.
Data through December 2011, released today by S&P Indices for its S&P/Case-Shiller1 Home Price Indices, the leading measure of U.S. home prices, showed that all three headline composites ended 2011 at new index lows. The national composite fell by 3.8% during the fourth quarter of 2011 and was down 4.0% versus the fourth quarter of 2010. Both the 10- and 20-City Composites fell by 1.1`% in December over November, and posted annual returns of -3.9% and -4.0% versus December 2010, respectively. These are worse than the -3.8% respective annual rates both reported for November. With these latest data, all three composites are at their lowest levels since the housing crisis began in mid-2006.
What’s interesting is how many don’t seem to understand what the index represents. Here’s a not uncommon interpretation via AP:
The declines partly reflect the typical slowdown that comes in the fall and winter.
No they don’t – the indices effectively washes out seasonality, both through the seasonal adjustments made to the index and the methodology itself. The non-seasonal adjusted results are virtually the same – and the methodology doesn’t reflect certain truisms like the surge that occurs in nearly every spring market since the dawn of time (or at least since cable tv was invented).
December 2011 S&P/Case-Shiller Home Price Indices [Standard and Poors
Survey: Prices declined in 18 of 20 cities in final months of 2011 [AP/WaPo]
Posted by Jonathan J. Miller -Monday, November 21, 2011, 12:39 PM
[click to expand]
It’s been nearly a year since I got enough energy to talk about a National Association of Realtors Existing Home Sales release. While their editorial side of the association is very good, the research side continues to issue information that simply misleads their agents and the public.
Here’s the release:
Total existing-home sales, which are completed transactions that include single-family, townhomes, condominiums and co-ops, rose 1.4 percent to a seasonally adjusted annual rate of 4.97 million in October from a downwardly revised 4.90 million in September, and are 13.5 percent above the 4.38 million unit level in October 2010.
To those in the real estate trenches, this is confusing. Perhaps it is some sort of skeptical breath of fresh air in a cauldron of bad economic news? The media is confused using the word “unexpected” a lot and agents are confused because this is NOT what is happening on the ground.
Doesn’t it seem odd that the rate of contract failures nearly doubled, yet sales increased? Not logical.
A higher rate of contract failures has held back a sales recovery. Contract failures2 reported by NAR members jumped to 33 percent in October from 18 percent in September, and were only 8 percent a year ago, so we should be seeing stronger sales,” Yun added.
The reason for all this gobblygook? NAR Research continues to over emphasize “seasonally adjusted” (methodology is not disclosed) results in a market that has been subjected to unprecedented external influences such as tax credits and downwardly managed mortgage rates.
No seasonal adjustments [black line]
In the above chart, the orange glow overlay on the black line reflects non-adjusted sales activity annualized from August to October over the past 3 years. The latest segment shows a sharp decline in sales relative to the prior 2 years. That is what real estate professionals and consumers are generally feeling since that is what is actually happening across the country.
Seasonal adjustments [red line]
The seasonally adjusted red line has been slowly eroding in 2011 and a headline saying “October Existing-Home Sales Rise, Unsold Inventory Continues to Decline” is completely out of context. Inventory is declining because it always declines in the last several months of the year.
Contrary to the institutional culture baked-in over at NAR Research, this simply creates more mistrust from the membership and consumers – and more importantly, it does not help agents sell more properties, the key reason for the existence of this trade group.
Posted by Jonathan J. Miller -Tuesday, October 25, 2011, 4:26 PM
Got a chance to speak with Tom Keene and Ken Prewitt in studio this morning where we covered a lot of ground.
The S&P Case Shiller Index was a released just as I came on the show, as well as what is driving the Brooklyn housing market, why is NYC fairing well, what ails the national housing landscape, will it have further to fall, what are the problems with relying on CSI, foreign buyers and whether Tom is looking at a kick up in his rent next year. Always fun.
Miller Samuel’s Miller Says Home Prices to Fall 5-10% (13 min) [Listen]
Posted by Jonathan J. Miller -Monday, October 24, 2011, 2:00 PM
I’ve written about the problems with the Case Shiller Index at length – here’s my 3 part series I wrote during my DC/Baltimore MRIS gig:
The index wasn’t created for the widespread consumer use it currently enjoys. It was designed for Wall Street to hedge the housing market. The right idea but the methodology wasn’t effective – only one investment bank ended up buying the rights to use the index.
What is the reason for investor reluctance to hedge housing using this or other indices?
Too much transparency. That sounds weird even though the CS index methodologies are partial hidden in a “black box”, the index and competing indexes end up being highly predictable.
It’s tough for options traders to get jazzed about any kind of housing index when the significant lag time between market activity and reporting allows for reasonable predictions of where the index will move.
A lot has changed since 2006 when the index began to be traded. Real estate data moves a lot faster now. The CS index, with its extensive use by media, has merely evolved into a tool for data companies to leverage their own predictive analytics.
Zillow happens to be one of many analytics firms and economists that can get public relations traction by predicting the results of the index results days in advance:
Zillow: Case-Shiller likely fell 3.8% in August [HousingWire]
Posted by Jonathan J. Miller -Tuesday, June 28, 2011, 11:01 AM
Its the end of the month and time for an S&P/Case Shiller Home Price Indices release.
Never any sleight meant to Bob, whom I personally admire, in my CS rants but toward the way the index is used and perceived. My Case Shiller report aversion has been well documented.
Consistent with the Case Shiller release confusion and lack of context, the June release of the February, March, April average closing price of December, January, February contracts show two things:
My friend Les Christie covers the month over month change which is the fundamental message of all CS releases.
They seasonally adjust so the fact that this really reflects January activity and that tends to see an uptick was therefore not a factor. This shows how useless the index is. The declines through December (their May release) reflect the tax-credit aftermath where the market was artificially depressed after being artificially inflated in early 2010.
Washington DC continues to be the strongest of the 20 cities covered, but MRIS/RBI & moi reported the like time frame in early February (just about 6 months ago).
Emphasis was placed on the year over year comparison, which ordinarily I’d agree with but since early 2010 house market was artificially pushed higher by the federal home buyer tax credit, it really isn’t.
The headline really should say closed housing prices in February, March, April in 20 U.S. Cities Fell 4%.
Here’s some sage insight in the piece:
“Home prices are still easing but the declines are not dramatic any more,” said Harm Bandholz, chief U.S. economist at Unicredit Group in New York, who correctly predicted the year- over-year drop. While month to month changes show “prices have basically bottomed and are moving sideways,” he said “we’re a long way away from significant increases in house prices.”
Two approaches to the same release and both are correct for different reasons and because of the 6 month data lag, we have no better understanding of the state of housing this month.
Next Page »