Monday, August 27, 2007
Home Sales And Freight
These numbers, btw, will continue to be revised for months to come. Generally the revisions are not that great.
Freight dropped 0.7% from May, and 3.4% from June 2006. This is a YoY staggering drop, the largest June drop in the history of the freight index. If you go to Table 10 at the end of the release, which shows freight by quarters, you will see that the pattern shown in 2007 is even weaker than the pattern shown in the first two quarters of 2001. This bodes poorly, and is one of the reasons I am so sure that we are in an inflation-masked recession. Freight is a reflection of primary economic activity.
Passenger TSI is really a diffusion index. In the early stages of a recession some business travel related to sales often rises. By the time it falls, usually businesses have begun to feel the recessionary squeeze and are cutting back on sales. So this is between a second and a third-order effect. In June, passenger TSI fell 0.7%. For 2007 to date, this is the sequence for passenger TSI:
Jan..:-0.1About as classic as you can get, and also a pretty clear indicator of recession. The 2001 recession really began in Q4 2000, and passenger TSI went negative in Q1 2001, so we are a quarter or two later according to this stat. (Note, because of 9/11, Q3 and Q4 2001 TSI is distorted and of no use for comparison).
Now to existing home sales: The most electrifying figure is that adding single-family and condos together, months of supply at the current annualized sales rate reached 9.6 months. Let's just say that no one who still remembers how to blush can maintain that home prices aren't going to continue to drop.
Single family sales dropped 8.3% from the previous month (actual numbers) and dropped 6.5% from the previous July. At the current pace, it would take 9.2 months to sell all of the homes currently on the market, up from 8.6 months in May and 9.0 in June.
Condo sales rose 9.5% from the previous month (actual numbers) and dropped 3.6% from the previous July. At the June sales pace, it would take a staggering 11.9 months to sell the condos currently on the market, up from 9.8 months in May and 10 months in June. Words kind of fail me here; it's difficult to characterize this debacle.
Part of the problem is that foreclosures and forced sales due to impending default and foreclosures are now pushing way more homes on the market. Absolutely every bit of objective evidence seems to show that more and more homes are due to be forced back on the market by resets and recasts for several years. It is mind-boggling to contemplate the situation. We have a sharply reduced pool of buyers due to tightened underwriting and appraisal standards, combined with rapidly growing supply. If no new home were built in the US for a year, market supply would probably continue to rise for the entire year just because of the number of individuals who must sell their homes. We are also in the early stages of recession....
Both condo and single-family sales patterns show that the market in the NE is actually recovering somewhat, and that the market in the South has taken another downturn, while the West is tanking to a historic degree.
Given these conditions, many homes in the West are likely to lose about 35% of their value from the peak. The situation depends on neighborhoods. The worst losses will be seen in the recently developed areas and neighborhoods, whereas the areas with older housing will hold their values better.
The New Home sales report is of almost no use whatsoever. Most numbers on this report are not statistically significant. About the only number that is statistically significant is the YTD 2006/2007 sales comparison, which still shows a drop of over 20%. Also, the fact that the time on market rose again in the last report to 6.1 months indicates that the situation for new homes isn't stabilizing.
Due to very high cancellation rates in the last year and a half, inventory figures in the new home sales report are meaningless. If a sale is ever recorded, it drops off the report entirely, and is not pulled back into the figures if the sale is cancelled and the home comes back on market. After months and months of this, the inventory of new homes for sale is meaningless, because the number of new homes for sale back on the market grows each month. The other side of this aspect of the New Homes sales report is that the sales of these new homes aren't recorded when they finally do go through. The best remaining indicator on that report is how long the new homes have been on market, because of course the homes still being tracked by Census are competing for buyers with the homes not being tracked by Census. When that number stops rising, we can claim that the new home market is stabilizing.
The serious disruptions in the mortgage market have now extended from subprime to Alt-A and conforming jumbos. However, the latest set of "adjustments" is not reflected in either of these reports. As daunting as they are, next spring we may be looking back at this period wistfully as a period of relative stability.
The housing market situation is severe enough that median and average sales figures mean nothing anymore. The distortions in the market compared to a year ago mean that shifts in sales patterns from one region to another and shifts in buying mix (move-ups compared to first-time buyers) are what is being shown in these figures rather than price changes.
For price changes, your best source is Housing Tracker. This provides listing prices for a number of metro areas. The drops shown are real, and always understate the real changes in home values. Many homeowners who must sell persist in listing delusionally high prices. Most of the home buyers since the beginning of 2005 who didn't make more than 10% downpayments are now underwater; many of the refinancers are too. And we have several years to come in which home prices will continue to fall.
A factor few have considered is that the prices are declining and due to decline most in the areas with the highest home values. The proportionate effect on lienholders is therefore more severe than a cursory look at the stats would suggest, because the areas in which prices are stable or rising have very low average mortgage amounts, and the places in which prices are collapsing have high relative loan amounts.
For a display of abject journalistic stupidity, read Kevin Hassett's The Housing Market Crisis May Already Have Passed. The cluelessness demonstrated is why I will always find paying work, although I doubt journalism is in my future.
You beat me to the punch about Hasset. What a tool.
I agree the BTS is awesome, reliable, timely and all that. I use it to time lumber futures contracts, free money. Unless I am mistaken it should be made more clear for your readers that we are talking ton-miles so there is no inflation component like there is with other statistics that discuss the dollar value of items transported.
One more slightly off topic BTS issue. Their transit numbers are becoming increasingly distorted and should not be trusted implicitly like is justified for train, truck and air reportage.
Since you track this stuff, have you noticed the continuing discrepancy between volume measures and currency measures?
Even when I try to back out retail sales for gas, for instance, I always come out with diminishing YoY volume.
But huge swathes of CA are due for extremely large average drops.
Teri, no. I looked at the link and I will put it on my reading list. It looks very interesting.
Let me know when they start jumping on the Street, OK?
You have to be smart with BTS data. They don't run it through a bias filter and chew it for you. Some of the y-o-y decline is explainable. A mild summer and coal shipments to the big plants at the CA/NV border can make a big difference. You don't have to be a rocket scientist to guess what lumber is doing. When fuel prices rise appreciably and persistently there is remodelling. Formerly trucks on roads become containers on flatbeds, etc.
All said you are correct. THere appears to be an appreciable decline in the amount of stuff being shipped and thus presumably consumed through economic activity.
Another note of caution on passenger data. Since the response to 9/11 has been so draconian those trips formerly via air have also remodelled into longer road travel and/or teleconferencing. Then of course the price of gas reduces those road trips as well. Above my pay grade but you should consider the long term trend when making comparisons for our purposes.
Rob, the way I've been doing my adjustments is largely using rail stats, since they give the breakdown in carloads of major commodity groupings. Do you use another method?
Just curious, always learning!
I really prefer the direct trucking and rail stats, and they have seemed to me to be pretty indicative.
Links to this post: