OER, CPI, New York Rentals And The Fed: A Strange Love Story

What do apartment rents in NYC have to do with Federal Reserve policy, interest rates, and rising bond yields? According to a recent analysis out of Barclay's Capital by Dean Maki and Julia Coronado, a whole lot more than you might imagine.

When April CPI was released on May 15th, the surprise was to the soft side. U.S. markets rallied, as traders believed rate cuts were coming sooner rather than later. We disagreed with this assessment, noting various Inflation Errors, including the very telling Core/Headline CPI Spread.

Barclay's looked at the key reason for this CPI surprise -- it found it was buried in the way the core CPI gets constructed. The BLS measure of home price inflation is the Owner's Equivalent Rent [OER]; it's what a homeowner could theoretically rent its house out for. That is the key to the housing portion of the BLS CPI calculation -- OER is about 43% of the core CPI measure.

My pal David Kotok, who is Chairman and Chief Investment Officer of Cumberland Advisors, points to a terrific piece of research from Barclays Capital that

disaggregated the OER and found that the bulk of the OER surprise came from the New York City metropolitan area. The CPI is broken down into four regions. Only the Northeast showed a pronounced deceleration in OER. Within the region, NY jumps out so dramatically that Barclays argues it accounts for roughly 75% of the total national deceleration in OER.

Why is this? Given the strength in the Investment Banking, Hedge Fund and Private Equity industries in and around NYC, the local housing market here is doing much better than the national averages.

Barclays surmises that this is tied to the rise in existing home sales in NY. The region has seen a stronger housing sales recovery than elsewhere in the Northeast or in the national statistics. Rising home sales suggest a substitution of ownership for renting. That may be more important than vacancy rates in determining OER.

Thus, Barclays suggests that we not get too excited about a potential Fed easing because of this surprise in OER.

Kotok makes the following astute observation as to what this may mean:

The Fed sees this OER data, too. They incorporate it into their policy decision making. OER is a very large piece (24%) of the total CPI. It is the key to the housing component which is 43% of the total CPI. When you remove the food and energy parts and derive the core CPI, the OER component looms even larger at 30% of core CPI. Note that it is a large 14% of the Feds' preferred core PCE according to Jim Bianco.

At Cumberland we have been proceeding under the assumption that the national housing slump has not bottomed. We saw the upturn in NY housing sales as an exception to the national trend and due to the bull market in the financial sector. We believe that the weakness in housing keeps the Fed from raising rates even though the inflation numbers are still above the Feds comfort zone.

This recent analysis by Barclays Capital gives us some pause. It's not enough for us to change strategy now. But we might alter our strategy if we conclude that the nation's housing sector deterioration is ending sooner rather than later.
The next CPI release is June 15th and will cover the month of May.

Update:
Can CPI go lower, regardless of what inflation actually does?

Yes, according to Barclay's Capital Research. It found Core CPI (also known as Inflation ex-inflation) is being understated for a surprising reason:

1. Core CPI is dominated by Owner's Equivalent Rent [OER].
2. Existing Home Sales in the NorthEast are outpacing the rest of the country.
3. Existing Home Sales in New York are far outpacing the NorthEast.
4. Manhattan Condos/Coops are far outpacing NY.

The deceleration in OER is directly impacted by the strength in the NY City high end real estate markets -- as opposed to homeowner vacancies or rental demand in the rest of the U.S

Barclays writes:

The deceleration in OER has been concentrated in the Northeast, yet the Northeast has the lowest and slowest growing vacancy rate. Meanwhile, the South has the highest and fastest growing vacancy rate, but has the fastest Y/Y pace of growth in OER. These data reinforce our view that vacancy rates have little to do with near-term fluctuations in OER.

The chart is quite telling:

courtesy of Barclays Capital Research

Here's the excerpt from Barclays:

The rapid rise in 2006 and recent deceleration in OER in the Northeast have been driven, to a large extent, by the New York City metropolitan area, which accounts for roughly 75% of the deceleration in the nationwide measure of OER. New York has also enjoyed a recovery in home sales that is far greater than elsewhere in the Northeast region or in the overall U.S. market. Given the high weight the region has on the aggregate OER measure, the bottom line is that stronger housing demand in the Northeast, and in the New York City area in particular, has been enough to offset the upward pressure on OER from other regions, where housing markets remain soft and demand for rental properties stays strong.

There is a great deal of uncertainty surrounding future movements in OER, especially when one metro area appears to be having such a large effect on the national trend. Because overall OER has decelerated, we think it is reasonable to lower our forecast for core inflation; we now expect the core CPI to rise 2.5% this year (Q4/Q4), down from 2.7%. This is a modest adjustment, and reflects our view that the primary driver of near-term fluctuations in OER is demand for rental properties, rather than vacancy rates. The primary downside risk to our forecast is that the housing markets elsewhere in the country will pick up as firmly as the New York City metro area, leading to reduced demand for rental properties and slower OER growth.

While the OER fluctuations suggest a somewhat lower run rate on core inflation in the months ahead, and this raises the bar a bit for our call for Fed tightening later this year, we ultimately think the growth and labor market data will be decisive. If growth bounces and the unemployment rate continues to decline, we doubt the Fed will take much comfort from a deceleration in OER, especially if that deceleration is caused by a strengthening housing market.

In other words, CPI is expected to be coming lower, regardless of what inflation will actually be doing.

Source:
Market Strategy Americas: Economic Outlook
US Economics Research and Market Strategy
Dean Maki, Julia Coronado
Barclay's Capital May 17, 2007

Comments

[Johnny Bitc...:]

More B.S. from Ritholtz, here are the flaws:

WRT: When April CPI was released on May 15th, the surprise was to the soft side. U.S. markets rallied, as traders believed rate cuts were coming sooner rather than later. We disagreed with this assessment, noting various Inflation Errors, including the very telling Core/Headline CPI Spread.

Flaws:

1. From March 12th to today the S&P has rallied from 1380 to 1520 - about .20% per trading day

2. The uptrend has been steady and there is no major ramping that occurred on May 15th, as hysterically described by Ritholtz above. For a look at the chart, please refer to:

finance.yahoo.com/char...=^gspc;range=3m;indica...

3. Setting aside for the moment that nothing special occurred on May 15th (as can be ascertained by a simple look at the trend back to March 12th, and contrary to Ritholtz falsely constructed world view), the claim that "traders" believe rate cuts are coming sooner rather than later has no support, except perhaps in the biased minds of statistics slinging chit-chatters. It reminds one of the quote:

Barry Ritholtz uses statistics like a drunk uses a lampost; for support rather than illumination.

Given that this recent uptrend has been steady for a couple of months, with no significant discernible change on May 15th, Mr. Ritholtz appears to be searching for support for his longstanding arguments rather than seeking to understand the data as it is presented.


WRT: When April CPI was released on May 15th, the surprise was to the soft side

4. This claim by Mr. Ritholtz is nothing more than an opinion, and is deceptive on other levels as well:

a. The BLS releases a headline number (which includes food and energy) and a core number which excludes those items. Both numbers are released and available to the markets to interpret. While not perfect, as they include well known flaws, these numbers do have significant history and reasonable comparability for use in trend analysis.

Mr. Ritholtz can make no serious claim as to what the overall market reaction to that single report was, and the trend since March 12 provides some evidence that his attribution is dubious, if not deceptively created to support his anchored views. Or perhaps just to create some kind of false economic melodrama so that his readers return for his breathless views on the latest data - kind of a soap opera for addicted market mavens where Ritholtz trumps reality.

b. In casually referring to the "April CPI", Mr. Ritholtz is referring to the core CPI number rather than the headline number. This is all part of the creation of a phony world that only he can sort out - if he referred properly to the core number and the headline number separately or the CPI report in general, it would be easier for folks to understand the depth of his false and confused logic:

- it is interesting to note that after the April CPI report, Mr. Ritholtz identified correctly (separately from his confused rantings about the market reacting wrongly to the core CPI number) that bond market yields increased. Since this was in direct contradiction to his claim that the markets reacted wrongly to the CPI number, I challenged him on this point. As of yet, he has not responded. I would guess that since he's so busy making phony arguments that support his views about his own false world constructions, he forgets to return to reality for consistency checks - whether it's lies or self-deceptions, the web can get tangled.

WRT: "including the very telling Core/Headline CPI Spread"

5. Previously Mr. Ritholtz attributed the widening spread between the Headline and Core to statistical fraud related to the Boskin commission changes to BLS methods.

As I noticed that the current spread is not out of line with the observed spread in the early 70's and the Boskin commission was a mid 90's phenomenon, I challenged Mr. Ritholtz on his claim, and provided what I believed was a more likely explanation (spiking oil prices then and now).

To this challenge, Mr. Ritholtz also did not respond. I was left to assume this was yet another example of his seeming tendency to reach conclusions in line with his own loosely formed belief system rather than a reasonable analysis of the data.

Now, secondly I would agree with Mr. Ritholtz that the Core/Headline spread is very telling. From Mr. Ritholtz hysterical analysis you might be led to believe that this is some kind of secret statistic discovered by him and a clandestine force of econometricians.

Nope. Just two numbers that are produced monthly by a boring government service. Sorry about that.

And since Mr. Ritholtz tells us how telling something is without telling us the tell, let's get the cat out of the bag:

When the spread between Headline inflation and Core inflation is increasing, it suggests that food and energy costs are rising faster than the other costs.

In Ritholtz falsely constructed world view, these are somehow hidden from the market and he is the superhero exposing the evil. Uh, earth to Barry: the report is available on the internet......lots of people read it.....both Core AND Headline. Take off the suit, Lois needs you to pick up her drycleaning.

Regards,

John.

P.S. In the future, when you are referring to the CPI, please make sure you identify whether you are discussing the Headline number or the Core number. It will help me in identifying the merits and flaws of your arguments.

I will look at Owners Equivalent Rent issue later.

2007 May 23 04:01 PM

[ritholtz:]

Excellent criticism -- its now 1 year later, and pretty much every critique you had proved to be wrong.

2008 Jul 29 04:12 PM