September 27, 2006
Seasonals in new home sales
New home sales could be worse. But don't overlook the seasonals and the fundamentals.
The Census Bureau today released their latest estimates of the number of new homes sold. The seasonally unadjusted figure was 91,000 units sold nationwide in August, exactly the same as the preliminary estimate for July that the Census Department published last month. However, today's release also revised the July figure down to 85,000 units. In other words, since July now turns out to be worse than previously thought, the reported growth from July to August is accordingly positive.
The above graph plots the seasonally unadjusted new home sales data going back to 1989, with vertical lines indicating the August value for each year. It is not unusual for there to be a last gasp of summer that puts the August home sales a little above July. But it has been 20 years since the September figures were any higher than August. The logical direction to expect that things will go from here is down, not up.
Doubtless many analysts will be watching the seasonally adjusted data for an indication of whether the fall slide this year is any worse than usual. But profits and jobs depend on actually selling a home, not seasonally adjusting a statistic, and realistically we shouldn't expect to see sales return to their current levels until next spring at the earliest.
As we watch and wait, it's important also to keep the fundamentals in mind. As Stephen Kirchner notes, yesterday's strong consumer confidence reading is a positive for home sales. And let us not forget that the prime factor responsible for booming home sales a few years back was low mortgage rates, just as the main factor bringing home sales down this year has been higher mortgage rates. On this score, there has been a modest but quite distinct decline in mortgage rates over the past few months.
And on the bearish side is the question of the extent to which buyers and lenders may have over-reached what can reasonably be expected to be repaid if house prices have now entered an era of decline rather than increase. That's a big, and, to me, frightening unknown.
On balance, while Calculated Risk views the new Census data as very weak, to me they appear a bit stronger than might have been expected on the basis of some of the other data that have been coming in recently , . Color me watching, worried, but not panicked.
Posted by James Hamilton at September 27, 2006 10:28 AMdigg this | reddit
During the last big decline in '90-'91 it seemed that the collapse was driven primarily by speculators forced to out of their bad investments. It seems that everyone and their uncle was buying a second home as an investment. This was also a time of relatively high interest rates and the telling factor in the bubble was the wide disparity between mortgage costs and the equivalent rents.
This time around the dynamic seems to be different. It doesn't seem to be driven by speculation so much (except in certain markets like the condo flippers in Florida.) This time it is too many people overextending themselves by spending too much on their primary residence, enticed by the low interest rates. In particular, adjustable rate mortgages are much more prevalent this time around. The other new factor is the use of home equity loans to cash out the equity in homes.
The difference is that last time it was mainly speculators who got burned, but this time it is more likely to be homeowners defaulting on ARM mortgages.
Posted by: Joseph at September 27, 2006 11:08 AM
"It doesn't seem to be driven by speculation so much (except in certain markets like the condo flippers in Florida.)"
I'm not so sure about that. The NAR estimates investment and 2nd homes made up 40% of all purchases. Furthermore, I think you'd be hard pressed to find an area that did NOT lure speculators. Do these sound like your typical investment hotspots: Bakersfield, CA; Boise, ID, Austin, TX?
Posted by: MTHood at September 27, 2006 11:26 AM
What determines mortgage rates? Are they based on Fed rates? If so why are they going down when the Fed is level? And of course they also dropped at the end of last year when the Fed was still raising rates.
Does the Fed determine supply, and the market determine demand; so this represents falling demand? Or are there other factors that affect supply?
Just curious as to whether we might anticipate further falls in rates or whether that will depend on the Fed starting to lower.
Posted by: Hal at September 27, 2006 01:17 PM
Hal, it so happens that I'm currently working on a research paper on precisely that question, which hopefully I'll have finished by the end of October. Bottom line is that the mortgage rate at any given time already incorporates the market's anticipation of Fed policy. The reason mortgage rates came down since June is that the market changed its expectation of fed funds levels for this fall since last June. Because the Fed failed to tighten beyond 5.25 as the market had earlier been assuming, mortgage rates came down from the level that would have made sense with a 5.5-5.75 funds rate.
As for anticipating further drops, my research suggests it's very difficult to do that-- current mortgage rates already reflect anything you could presently factor in.
Posted by: JDH at September 27, 2006 02:03 PM
"The reason mortgage rates came down since June is that the market changed its expectation of fed funds levels for this fall since last June.Because the Fed failed to tighten beyond 5.25 as the market had earlier been assuming, mortgage rates came down from the level that would have made sense with a 5.5-5.75 funds rate"
Why are these 30 year fixed mortgage rates reacting to the Fed Fund rate expectations, when they have a duration comprable to a 7 to 10 year Treasury. Shouldn't the change in mortgage rates be based on the expectations of the longer end of the Swap curve and not the Fed Fund?
Posted by: Street Wise at September 27, 2006 11:07 PM
Mortgage rates reflect supply and demand. From a supplier perspective the mortgage rate anticipates lenders' cost of funds - they obtain money in the market add their costs and profit, and "resell" it to the public.
The credit markets are huge, but there is also one very large gorilla in it - the Fed. All participants in any market with a gorilla will closely watch the actions of that gorilla.
The Open Market Committee of the Federal Reserve Board sets their Target rate and then (through the open market desk of the Federal Reserve Bank) engages in their open market activities (buying and selling) to push the market equilibrium price to their target rate - a rate known to most as the Federal Funds Rate. The actual rate fluctuates throughout the day, and has a daily high and low like any other traded item would.
Through this market intervention (manipulation) the Fed dominates the overnight lending rates, thus controlling the shortest end of the yield curve. This influences the supply and demand at other durations to a diminishing degree as the duration lengthens.
Why care about the overnight rate when you are lending long (such as mortgages)? Because the Fed pushes the overnight market in order to achieve other goals that ultimately affect the economy, inflation etc. It is THESE things drive our perception of what the needed long term rates (return on our money) should be.
Currently the bond market seems to be betting on a recession, which should naturally bring declining rates. Of course, the Fed will continue to intervene to artificially steer this natural move. So we must watch not only the fundamentals, but the gorilla as well.
Posted by: Zephyr at September 28, 2006 04:08 AM
The gorilla is almost always late - but he will catch up (and then overdo it - the Fed always oversteers).
"There go my people, I must follow them for I am their leader." - Ghandi
Posted by: Zephyr at September 28, 2006 04:21 AM
"But profits and jobs depend on actually selling a home, not seasonally adjusting a statistic, and realistically we shouldn't expect to see sales return to their current levels until next spring at the earliest."
Thank you for providing the Raw sales data. Your point, in re: Seasonal Adjustments(SA), are exactly on point.
What perturbs me most about SA is lack of definition of methodology along with the purported data.
I do find that you offer some of best available work, when it comes to these types of studies.
Posted by: Mark E Hoffer at September 28, 2006 06:03 AM
Good comments on interest rates.
Have you considered that the FED methodology actually makes their market interventions meaningless concerning inflation. They establish a rate to try to slow inflation but then they adjust the money supply to defend their rate actually undoing any affect of the interest rate change. I am hard pressed to see any positive in the interventions of the FED, but there are a whole lot of negatives.
Posted by: Zephyr at September 28, 2006 06:21 AM
The above comment was from me. I am still having problems posting properly.
Posted by: Dick at September 28, 2006 07:50 AM
I agree, but if inflation expectation theory holds true then why do we have an inverted curve? Under this assumption, you will have an arbitrage breakdown. I am very interested in seeing JDH's study on the relationship of FF to commitment rates.
Posted by: Street Wise at September 28, 2006 12:37 PM
The models for estimating housing starts are subject to wide error margins because housing starts are only a small fraction of the housing stock.
Underlying demand for housing starts is typically measure as:
Houshold formation + empty (holiday/second) homes+ replacement of knockdown homes - mobile homes.
Household formation is a function of population growth, age of population and probably and economic variable. Basically as people get older there are few people per houshold.
Where is this getting us?
Household formation is around 1.5 m and mobile home demand is around 0.3 m so need a lot of second home demand to get to even 1.8 m starts per year.
Posted by: Anonymous at September 28, 2006 06:14 PM
I'm pretty ignorant; but...
It seems to me that the Treasury yield curve represents a closer to risk free yields than the swap rate curve.
(The USD poor prospects is a big question mark here: Greenspan's conundrum?)
If we accept these facts, then, of course investor flight to safety would add fuel to the fire of a perceived future deflation. Or, in other words, treasuries would gain in value while swaps based on riskier underlyings would tend to have the opposite behavior?
Wouldn't you agree?
Posted by: Joe Rotger at September 29, 2006 05:52 AM
...And mortgage lending, is considered safer than corporate lending...
Posted by: Joe Rotger at September 29, 2006 06:05 AM
Do these sound like your typical investment hotspots: Bakersfield, CA; Boise, ID, Austin, TX?
Well, Boise and Austin are both huge for technology investment right now, but I don't think that's what you meant. Incidentally, Texas in general hasn't had home prices increase at all during this bubble-- Houston, Dallas, and San Antonio all have very cheap housing.
Posted by: John Thacker at September 29, 2006 10:46 AM
I agree that flight to quality would cause swap spreads to widen, but I mentioned the swap curve, because swap rates are the funding cost for most mortgage origniators.
"..And mortgage lending, is considered safer than corporate lending..." I don't see how this relates to Mortgage Commitment rates and their relationship to FF futures rate. I think one may be able to find interesting statistical relationship with different rates and the Mortgage Commitment rates, but that does make the expectation of FFs the cause for the change in Mortgage Commitment rates.
Posted by: Street Wise at September 29, 2006 06:07 PM
With this: "Have you considered that the FED methodology actually makes their market interventions meaningless concerning inflation. They establish a rate to try to slow inflation but then they adjust the money supply to defend their rate actually undoing any affect of the interest rate change. I am hard pressed to see any positive in the interventions of the FED, but there are a whole lot of negatives."
No one, with a straight face, can defend the FedRes' actions or its record.
Just to 2x check: Can anyone explain why the American economy needs the FedRes?
Posted by: Mark E Hoffer at October 2, 2006 11:26 AM