The Housing Market and the Economy:

Liz Ann Sonders, Chief Investment Strategist, at Schwab, has an excellent summary, complete with neat charts, of the state of the housing market, the current outlook, and the potential ramifications for the economy as a whole. In short: not a pretty picture. This chart (click for larger view), in particular, is worrisome if one is invested in equities.

Delurking (mail):

Look at 1994 to 1995 (Which is actually 1995-1996 for the S&P). Why should I expect 2005-2006 (Actually 2006-2007 for S&P) to be different? She even says "As cited by many a homebuilder CEO lately, there is something very unique about this cycle: It's the first time in American history that a severe housing downturn has been led not by rising unemployment or falling incomes, but by too much inventory and speculation."

So, why should I not expect that investors will dump houses and rush to equities. I'm not predicting anything, just saying that plotting this short-term correlation is next-to-useless in predicting equity prices.
9.28.2006 10:44am
frankcross (mail):
This is probably just showing that both are driven by the same outside economic third factor. Take out the one year lag, and the association would look about the same, maybe better. Moreover, the chart shows one quick housing market plummet at the very beginning, and equities apparently increased at that time
9.28.2006 11:00am
DavidBernstein (mail):
The worrisome thing is that there may indeed be an "outside economic factor" that is causing the housing crunch, and will also cause an equity slide: a more-than-quintupling of short-term interest rates. I'm not making a prediction, though. Even Schwab, thus far, is only suggesting underweighting stocks by 5%
9.28.2006 11:07am
JoshL (mail):
Agreed with both of the people above. Housing "investors" and speculators are really no different from people speculating in futures at this point. If anything, cheaper housing will allow better quality of life and possibly pump more of the money that might have been used to buy living space (for people planning to actually live in their houses for a long period of time) into equities, or at least help in paying off debts.
9.28.2006 11:10am
Tom952 (mail):
Looking across to the left side of her graph, the SP500 would have to fall to 600 or so to follow the housing trend. That action would require investors to eschew buying stocks while the S&P PE ratio halved. I would therefore be surprised if the S&P 500 Index did indeed continue to track the NAHB housing market index.

I'll opine that we simply had a housing speculative bubble, and it will take time for fundamentals to reassert. Meantime, as commented here and elsewhere, builders are going to continue to build even as prices come down, past the point where they break even - i.e. they will overbuild as they have in the past, trying to make money as fast as possible in the downturn. When prices reach the point where builders lose money on new homes, they will stop saturating the market.
9.28.2006 11:25am
Professor Bernstein, please explain what you mean by "a more-than-quintupling of short-term interest rates." Seems to me that if that happens we will all have more problems than housing prices declining by 30%.
9.28.2006 11:33am
I don't like graphs with questions marks on them. Just the facts, please.
9.28.2006 11:35am
Arbusto Spectrum (mail):

cheaper housing will allow better quality of life and possibly pump more of the money that might have been used to buy living space (for people planning to actually live in their houses for a long period of time) into equities, or at least help in paying off debts.

Josh - if you look at the statistics on how much cash-out refis have contributed to American's spending over the last five years, you might rethink that premise.
Not to minimize Prof. Bernstein's contributions on this topic, but the blog I read for perspective on the real estate market is
9.28.2006 11:40am
Arbusto Spectrum (mail):
Anonassociate --
Three month LIBOR (London Interbank Offer Rate) was 1% in mid 2003; it is now 5.36%. That is likely to present a significant problem for people who overleveraged themselves to buy houses using 3/1 or 5/1 ARMS, on which the interest rates will step up significantly.
9.28.2006 11:47am
Chris 24601 (mail):
If there really were compelling reason to think that stocks will tank in the next year, prices should be down now already. I'm very hesitant to be so sanguine that I've got better information than the market. I'd also like to see what the chart looks like pre-'94, especially given the way the graph starts.
9.28.2006 12:03pm
I would like to see that graph plotted against interest rates/mortgage rates.
9.28.2006 12:27pm
David W Drake (mail):

My comment exactly. The interest rate is probably the "outside economic factor" alluded to in some of the other comments.
9.28.2006 1:50pm
Carl Carlson (mail):
The "quintupling" is deceptive, if you are implying that mortgage rates must have quintupled as well. The average 1-yr. ARM rate (a very short rate) has gone from 3.88% in September 2003 to 5.54% today. Higher, yes, but not even double the previous rate, let alone 5x. I don't know if Prof. Bernstein was alluding to the LIBOR rate as you suggest, but the notion that interest rates have skyrocketed is just not true.
9.28.2006 2:01pm
Carl Carlson (mail):
Just for clarity, those numbers come from Fannie Mae.
9.28.2006 2:02pm
Thanks for the LIBOR data--I didn't know that and it is indeed interesting. But to say short-term interest rates have quadrupled without discussing the rise in mortgage rates at a minimum implies a similar rise, which is not true.
9.28.2006 2:07pm
Dan G (mail):
I moved recently. And, after selling my house I took my equity and put it in a bank CD rather than buying a new one down here. I'm too young to have been through a previous bust, so I have no idea how to try time time the market. But, while things haven't fallen much yet, I'm feeling pretty glad so far...
9.28.2006 2:35pm
Arbusto Spectrum (mail):
Beware the rates you see quoted for 1 year ARMS, as lenders (i) get paid a fee for originating, averaging at 0.6% of the size of the loan for a 1 year ARM and (ii) put some of that fee back into offering you an artificially low "teaser" rate to make the product look more attractive. In other words, that 5.54% does not reflect what the true cost of the mortgage would be after the reset period.
Many people stretched themselves to buy homes they could not really afford, but since the monthly payments at the teaser rates were so low, it created the illusion that they could. Have mortgage rates doubled? No. Have they increased to the point that there are a lot of people who will have difficulty making payments. Just look at the delinquency data, which will be followed soon by the foreclosure data.
There was a nice article in today's Journal (Wall Street) about a fairly sophisticated mortgage fraud scheme as well....
9.28.2006 2:51pm
Sailorman (www):
As for me... I'd like to see the part of the graph she cut off at the left. Because I'd want to be sure she wasn't "cherry picking" her statistics, which it looks like she is.

It's entirely possible that housing and stock are linked sometimes. But they don't look like they're linked ALL the time by any means. So unless she wants to ALSO address how you can tell an upcoming "linked year" from an upcoming "unliked year" then the statustics are worthless. Will next year look like 1994 in reverse? Who knows?
9.28.2006 3:15pm
Another theory is that stocks will go UP when housing goes down:

a wide range of real estate experts agree that much of the current weakness in sales and home prices is due to investors pulling out of the sector. And some stock market experts say that the money being pulled out of homes is finding its way into stocks.
9.28.2006 3:19pm
Carl Carlson (mail):
You're wrong on this one Arbusto. As Freddy Mac's website makes clear, the fees on a 1-yr ARM have remained steady, at 0.6 points in September 2003 and September 2006.

Comparing apples-to-apples, the rate increase on a 1-yr. ARM has increased less than 2% in the past 3 years, which is the point I'm making. Sure, the APR is slightly higher than the rate because of those points and fees, but it hasn't gone up any more than the rate, since fees have remained flat.

Looking elsewhere on the site, I see that 30 yr. Fixed rates have increased a grand total of 26 basis points (0.26%) from August 2003 to August 2006. The fact is, short term rates mean almost nothing with regard to first mortgage pricing. It's the people who took out giant 2nd mortgages based on prime in the past few years who are seeing their rates and payments go up dramatically.

However, thanks partially to the increased value in their homes from appreciation during that time, many of those borrowers are able to wrap their old 1st and 2nd into a new 1st mortgage and actually reduce their current blended rate.

Sorry to rain on the parade of negativity, but them's the facts...
9.28.2006 3:23pm
DavidBernstein (mail):
You can call it "2 points" or you can call it "50%", which is much more pertinent. Also, lenders can't afford very low teaser rates anymore, which makes it harder to shoehorn unqualified buyers into homes they can't afford, though they are trying to make up for that with teaser payments artificially lower by negative amortization loans. Also, when people used to question how they would afford these loans when they readjust they were told, don't worry, prices are going up, you'll just refinance when it becomes necessary. What are they being told now?
9.28.2006 4:13pm
Arbusto Spectrum (mail):
I made no reference to any change in up front fees (points) charged in connection with ARMS. It does not surprise me that fees on ARMS have remained constant over time, as lenders are typically willing to trade off up-front fees for rate to accommodate the desire of the borrower.
What I said is that originators of ARMS offer artificially low teaser rates for the initial fixed rate periods of the loans as an inducement. In other words, if the rate will reset to the one year treasury + 275bps, they set the initial rate at T+125. This creates the illusion that interest rates have to increase in order for the mortgage rate to increase, which is unfortunately not the truth. If you don't believe me, read a standard ARM mortgage disclosure statement -- it's right there.

As it relates to mortgage pricing, 30 year fixed rate mortgages are based of the ten year treasury rate, so you are correct in saying that they are not affected by short term rates. Unfortunately, to pick a recent month, say March 2006, 41% of residential mortgages originated were ARMs. They are not second mortgages, they are first mortgages. The average fixed period on them is about 2 years. These borrowers will have a rude awakening arriving in their mailbox rather soon.

Don't worry about raining on my parade -- I got tons of grief from relatives when I told them the NASDAQ was overvalued in '99 and '00.
9.28.2006 4:24pm
Carl Carlson (mail):

A "50%" increase is NOT more pertinent. If rates went from .25% to .5%, that would be a 100% increase, right? But the actual DOLLARS that a $100,000 loan would go up would be a whopping $250 a year.

In the current 1-yr. ARM case, a $250,000 loan has seen monthly payments (fully amortized) go from around $1,190/mo. @ 3.98% to $1,425/mo @ 5.54%, for a $235/mo. increase on a $250,000 loan.

Is $235 a "50% increase" on $1190? Not even close. Try less than 20%. Sure, there will be some people who can't make the increased payment, but the way you choose your numbers overstates the case.

Note that none of this has to do with property VALUES, just what it costs to borrow money. If values go down sharply, then many borrowers will have loan-to-value ratios at dangerously high levels, and will have a hard time qualifing for good rates.

But the rise in interest rates is not severe enough itself to be the cause of any precipitous drop in property values, in my opinion. If unemployment creeps up, then we'd have a different story.
9.28.2006 4:46pm
DavidBernstein (mail):
I remember when the Fed Funds rate went up from 3% to 6% before the NASDAQ crash, and I remember reading lots of articles about why this wouldn't affect the price of technology stocks. But there is something about a liquidity squeeze that tends to deflate,or even pop. bubbles.
9.28.2006 6:17pm
Professor Berntein, with all due respect you flatly don't know what you are talking about here. If I recall correctly (and I used to work a quant desk, and so I DO recall correctly) FFR was at 6% in the begining of 1995 and varied from a low of 4.75 to a high of 6.50 or so over the next 6 years, before starting its descent in 2001. To be sure, from 98 to 2000 it rose from 4.75 to 6.50, but nothing like what you claimed and the correlation was nothing like what you assert. Come on.
9.28.2006 7:09pm
Just to back up my facts, I looked it up. Good to see my memory has not faded over time. Link to historical ffrs.
9.28.2006 7:11pm
DavidBernstein (mail):
I misremembered the exact rise, I didn't look it up, but I absolutely remember a spate of article in 99-00 discussing how rising interest rates wouldn't have any effect on the technology stocks. But the Fed Funds rate did indeed start rising in late 99, and peaked just when the NASDAQ bubble was apopping. The theme of the stories in the business press at the time was that rising interest rates usually are bad for stocks, but that for various reasons (that didn't even make much sense at the time), the technology stocks were to be exempt from this trend. It makes perfect sense to me that when you have a liquidity driven bubble, reducing liquidity will end it.
9.28.2006 7:26pm
DavidBernstein (mail):
For example, here's a transcript from CNN, March 1, 2000:

HAFFENREFFER: The Nasdaq certainly had a very impressive -- continuing to run up here. Blue chips seemingly two days in a row out of the woods. Any insights?

SMITH: Well, you know, your -- Christine, just a moment ago, talked about how the seasonals or the -- what we've usually seen is the market rally in the first days of the month. That will probably continue to be here. I think we're in the midst, actually, of a reflex rally in the Dow, which will carry us higher over the next few days. But when you've got rising interest rates, it's very difficult for the old economy, represented by the Dow, to move higher because it's a very interest-rate sensitive index, unlike the Nasdaq which, in current market conditions, is not interest-rate sensitive. So, to mean, "the great divide," as you called it a few moments ago, is going to -- is set to continue.

MARCHINI: The Nasdaq is not interest-rate sensitive. That does seem to be some of the wisdom about the new economy. But do you think the Fed is going to keep raising rates until it becomes interest-rate sensitive?

SMITH: At some level, everything is interest-rate sensitive, and I do think that they're going to raise it. I don't know if it's going to be enough to really hurt the Nasdaq; I think enough to slow the advance or maybe even cause it to dribble lower.

I'm looking more for something like 50 to 100 more basis points or a half to a full percent rise by the Fed from here, and that's going to continue to win in the broad market, and I think it will blunt the advance of the Nasdaq.
9.28.2006 7:30pm
DavidBernstein (mail):
Here's another, Dallas Morning News, 2/26/00:

The specter of rising interest rates has proved to be a formidable headwind for many of the stocks in the Dow and S&P, while the Nasdaq seems oblivious, said Bill Meehan, chief market analyst at Cantor Fitzgerald &Co.

The Federal Reserve board has raised short-term interest rates four times over the past year as a pre-emptive move against inflation. Since July, the federal funds rate -- a key indicator of the direction of interest rates -- has moved from 4.75 percent to 5.75 percent.

"A lot of people think higher interest rates don't affect technology stocks," Mr. Meehan said, "but if the Fed raises rates enough there is no question that the economy will slow, and obviously these tech companies will sell less of their products."

Some analysts are expecting the Fed to raise rates at least two more times this year, including at its next meeting, March 21. But some Wall Street analysts argue that many of the "New Economy" companies of the Nasdaq don't have debt on their books, and because of that they aren't affected by rising interest rates.
9.28.2006 7:32pm
I don't doubt that people were saying that but a rise from 5.00 to 6.50 is not--was not--enough to cause a stock market collapse. The underlying fundamentals (or lack thereof) caused the NASDAQ crash not an incremental hike in fed funds which in any event takes months to work its way into the economy. Just because foolish commentators said that tech stocks were exempt from rate hikes, and rates were hiked, and tech stocks crashed, does not mean that b caused c. Look at a really long-term (30 years or so) ffr chart and then you might see some trends that could cause a crash--but not the 99-00 hikes.
9.28.2006 7:35pm
The events that caused the Stock market to crash in 99-00 and what is causing the housing market to slump in 2006 are exactly the same factor: they are/were perceived to be overvalued.

Just about everybody knew both markets were overvalued at least 18 months before they started to decline. It may well have been interest rate hikes that triggered the actual decline in both markets, but the stock market is not going to follow the housing market down unless there is a perception that it has become significantly over valued. And I don't see any such perception is becoming widespread.
9.28.2006 8:07pm
Just about everybody knew both markets were overvalued at least 18 months before they started to decline.

Come on, this is supposed to be at least a quasi-libertarian site. Have some respect for efficiant markets. If that statement was true there would have been massive shorting and the NASDAQ would have crashed, well, 18 months earlier, or right about when "everybody knew" it was overvalued. That's just sloppy reasoning. I'll grant you it is an inapt analogy to the housing market, mostly because no one has come up with a way for me to short a 1BR in Clarendon 1021 yet...
9.28.2006 8:16pm
Arbusto Spectrum:

If that statement was true there would have been massive shorting and the NASDAQ would have crashed, well, 18 months earlier, or right about when "everybody knew" it was overvalued

I disagree with you on that point. Not everybody knew the market was overvalued, but a hell of a lot did, and a lot of those who did short it got their a55es handed to them as the valuations became increasingly absurd. If you know the market is overvalued, but you don't know what catalyst will crack the bubble, it is exceptionally difficult to short the market. Sure, you can buy puts as opposed to taking a naked short position, but while that strategy mitigates your downside, it increases your need to have the timing correct.
Unfortunately, from what I hear from friends in the technology M&A business, the morons who do valuations based on "page views" are coming back into vogue.... That does not bode well.
Every cycle is different, and I don't think there is necessarily a strict correlation between the housing market and the S&P. However, there are a number of reasons to be cautious:
(i) slowdowns in the housing market ARE pretty well correlated with recessions since the early 1970s;
(ii) the construction industry has accounted for a large portion of the jobs created since 2001; a decrease in the amount of construction is likely to lead to layoffs in that sector (as well as ancillary businesses like mortgage brokers, furniture salespeople, etc);
(iii) there is no question that a significant portion of consumer spending has been driven by cash-out refis, as opposed to earned income.

I am not an economist, and I don't have a crystal ball; but, there are a lot of smart people out there, including many who were ahead of the curve on the stock market bubble, who have raised red flags that are at least worth paying attention to.
9.28.2006 9:54pm
Arbusto Spectrum:
I believe that someone actually has developed a real-estate market hedging product, so you very well may be able to hedge your risk. Another way would be to short the equities of the subprime mortgage originators....
9.28.2006 9:56pm
You know, I looked into that for a while. I was and remain unconvinved, mostly because real estate is such a local phenomonon. And I don't see any correlation between my house in Lyon Village and a subprime mortgagae originator.

The CME Housing products are much too thinly traded and short-term to be of value at this point, though eventually they may be more useful. For what it is worth, they appear to be predicting a 6% nominal decline in housing prices in D.C. by next August, which is certainly less than Professor Bernstein and others would have you believe but, as I said, I don't think they are of much value at this point.

I remain convinced that the best hedge againt NoVa real estate declines, particularly in the SFH Arlington/McLean/Fairfax market, would be some kind of constructed put on government contractors and the like. I've more or less become convinced that as long the the DHS and DOD money keeps flowing, nothing's really going to happen to the upper end of the NoVa housing market.

I suppose if I wanted to pay enough for it I could get some derivatives desk to write me a collar on north Arlington single family home prices, but it would probably cost me half my gains to buy it!
9.28.2006 11:06pm
DavidBernstein (mail):
Actually, a 6% nominal decline, plus additional seller concessions (being willing to fix things that come up in inspections, paying closing costs, etc.), plus 4% inflation, equaling maybe a 12-15% real decline, sounds about right.
9.28.2006 11:13pm
Arbusto Spectrum:
Uh oh, I didn't realize there were predictions here. I will consult with my crystal ball. I think your assessment of what keeps the NoVa market afloat is probably accurate, though an outflow of lobbyists might also lead to a correction. And, as Goldman Sachs has accrued $532,000 in compensation per employee through the first nine months of the year, I suspect that the Manhattan and Greenwich markets will also be OK for a while....

My sense is that the fed has two choices - (a) leave rates at a level that stems inflation, which will most likely create an issue for the housing market as well as those people who are overleveraged; or (b) cut rates to allow nominal house prices to hold, so as to avoid a wave of defaults, but thereby allowing a much higher level of inflation.

Given that the federal government is the biggest debtor of all, my bet is on option (b).
9.28.2006 11:24pm
While the article is interesting in detailing in great detail how we are all doomed, doomed, she says, going to hell in a handbasket over the falls, yet, I have to laugh at "the recommendation":

This remains one of the reasons we maintain our recommendation to slightly underweight U.S. stocks in favor of cash (see the Schwab Market Perspective). In Schwab parlance, "slightly" means five percentage points of your total portfolio value. So we are not recommending clients completely replace their stocks with cash, just that they make a minor adjustment.

Move that 50% down to 45%, because we are all doomed, DOOMED!
9.29.2006 1:49am
Professor Bernstein, so now you are only expecting a 6% nominal/12-15% real decline in prices? That's not even significant, statistically speaking. Seems like you are expending a lot of noise on what is really just market noise? And that is in contrast to what you have at a minimum implied here, herehere, here and in many other places.

If all you are really expecting is a 6% nominal decline, then I think a correction from you is in order.
9.29.2006 8:35am
DavidBernstein (mail):
Within the next year, that's entirely plausible. But there's also ALREADY been a nominal decline of well more than 6% from the August 2005 peak, so if we're talking about a total 25-30% real decline by next August, that's quite a contrast to predictions of indefinite 8-12% annual nominal increases the "experts" were predicting last Fall. I don't have a good enough crystal ball to predict the ultimate outcome, but my best guess would be a 50% real decline over a period of 3-8 years, starting at the peak of August 05, which could be a result of significent nominal declines in a short period of time, or of relatively minor nominal declines over a longer period of time.
9.29.2006 9:28am
Seems highly suspect to me. Matching up two lines on a graph does not constitute a rigorous analysis.

These folks really need to scale their data to returns instead of levels, work with excess returns by removing interest rate changes, use a larger sample size that includes more recessions, and finally (assuming the hypothesis still holds) perform a test for cointegration.
9.29.2006 11:31am
qsi (www):
I call shenanigans on this.

The main problem is that the two lines represent totally different kinds of indices: the S&P has no maximum possible value, while the NAHB is a diffusion index that by defition will have a value between 0 and 100. If we take the graph at face value, it would imply that the S&P would have a maximum value of around 2,000, a 50% or so increase from current levels. That's clearly nonsensical.

It would also mean that with the NAHB housing index at 30, the S&P 500 would have to fall to 354 by next year, a drop of 74% from current levels, or 10.5% a month, every month for the next year.

If you look at the longer history of these timeseries, you see that the purported relationship breaks down completely.

The housing market is deflating from its bubbly state, but this kind of graphic superposition does not make too much sense.
9.30.2006 5:13pm