I have a lot of respect for the research of Christopher Cagan at First American CoreLogic. His latest paper, a 118-page study of the impact of ARM adjustment, was picked up by the Wall Street Journal, in a story which itself was picked up by Dean Baker:
The WSJ decided to highlight a 3-month old study to tell readers that everything is fine with the mortage market… the problem with the WSJ assessment is that it is based on the assumption that house prices do not fall. If prices fall by 10 percent, then Cagan projects that the default rate would be more than 70 percent higher and the losses to the financial system would more than double.
How plausible is it that house prices don’t fall from their December 2006 level? Well, the National Association of Realtors data show that the median house prices has fallen 3.1 percent over the last year…
Does anyone think that house prices won’t fall in 2007?
I’d point out that the study is dated March 19, it’s not three months old. And the assumption that house prices do not fall is not an assumption that they do not fall in 2007: it’s an assumption that they do not fall over the next six years. Which, frankly, is an eminently reasonable assumption. Median house prices are volatile in the short term, and might well go down in 2007. But the chances of a nominal fall in house prices over the next five years are, I think, de minimis.
Baker also ignores a lot of pretty good news in Cagan’s report. First, he shows how limited the reset problem is:
The analysis anticipates a total of 1.1 million foreclosures with losses of about $112 billion, spread over six years or more. Our mortgage lending industry extends about $2 trillion of loans yearly, so these losses represent less than one percent of total lending. This will clearly not break the mortgage industry.
Even if prices fall by 10% and foreclosures reach 1.86 million — over six years, remember — we’re still not talking major systemic impact.
What’s more, the problem of negative equity seems to be getting better, not worse:
As of December 2006, 6.9 percent of the properties had negative equity, which is better than the 9.4 percent having negative equity (as of September 2005) found in the previous study published in February 2006. Although values have fallen in some markets during the intervening time, they have risen in other places, particularly on a year-to-year basis.
There’s another interesting datapoint, too: in the last study, 29% of the mortgages originated in 2005 had negative equity. In this study, just 9% of the mortgages originated in 2005 had negative equity, and less than 18% of the mortgages originated in 2006 had negative equity. Why is that? It’s because, contra Baker, prices did actually rise in 2006. If you look at the NAR report he’s citing, the 3.2% drop is from January 2006 to January 2007. But the same report says that the median sales price in 2006 was actually 0.6% higher than the median sales price in 2005, and there was even a 0.8% rise between January 2006 and December 2006. As prices rise, both foreclosure rates and negative equity rates fall.
(For much more on the old study and negative equity, take a look at this old post of mine from February.)
There’s much more in the study, which I would highly recommend to anybody who’s seriously interested in the relationship between interest rates, adjustable-rate mortgages, and foreclosures. Here’s Cagan’s conclusion, which I think is decidedly solid:
In conclusion, this study shows that the time-honored truths of economics are true after all. Risky investments are in fact less secure than traditional investments, and should be valued as such. This sounds too obvious to mention, but it should be disregarded at one’s peril. For short periods of time, under special market conditions such as the surging prices of 2004 and 2005, the old adages stood away from the market in temporary abeyance; but as time went on the time-tested fundamentals of economics have reasserted themselves.
Another time-honored economic truth is that residential real estate is properly a matter for the long term. Traditionally, people purchase their homes to live in and enjoy for a period of several years or even for a lifetime. Strategies with a short time horizon, such as flipping, can produce great profits but also carry risks. People who undertake fixed rate loans that they are able to pay, and hold their homes for several years or more, can in most cases expect to do quite well.
It is legitimate to enter a home with adjustable-rate financing, but borrowers and lenders should consider what the payments are likely to be in a few years as well as initially. If the future payments can be sustained and not only the initial payments, and if significant negative amortization is avoided, buyers can use adjustable-rate loans with confidence. Those who purchase, borrow, lend and invest following basic and proven virtues are likely to be rewarded for their wisdom.
The article did not clearly identify the new study, which had not been formally released at the time it appeared on the web, so I had wrongly assumed that it was referring to the 3-month old study.
As far as the more fundamental issue, the question is whether prices will decline at a point in time, not some arbitrary future point 5 to 10 years out, that will determine whether homeowners have positive equity that gives them both a strong reason to pay off their mortgage and also the means to borrow against equity, if necessary.
You can try to bury a decline in house prices by taking year round averages (yes, they were still rising in 2005, so year-round 2006 is above year-round 2005), but this doesn’t help all the people who bought in oct, nov, dec of 2005, or in 2006. Also, you know that the month to month changes are virtually meaningless. You have to beat up the data pretty good to come up with a story that prices are rising.
Oh, I’m not denying that house prices are falling. (I’m not necessarily admitting it, either — that’s a question for another blog.) I’m just saying that Cagan’s conclusions are perfectly commensurate with a situation where house prices are falling: you can believe that house prices are falling and also believe Cagan’s conclusions at the same time.
I don’t think UK house prices had regained their 1990 peak by 1996, and Japanese real estate certainly didn’t get back to the high water mark in six years. I wouldn’t call this risk “de minimis”.
These are the historical home prices for Southern California (hope you had a good trip)
As you can see, the (nominal) median home price of 1991 wasn’t reached until 9 years later. I’m afraid your calling the risk of flat home prices for five or six years ‘de minimis’ is nothing than your gut feeling. Further, the fundamental is affordability. Affordability today (on monthly payment basis) is worse than in 1991 both for Southern Cal
as well as the national market
Maybe the bulls are right on the risk minimization. Nevertheless, here’s Stephen Roach on the dotcom bust:
“Sub-prime is today’s dot-com — the pin that pricks a much larger bubble. Seven years ago, the optimists argued that equities as a broad asset class were in reasonably good shape — that any excesses were concentrated in about 350 of the so-called Internet pure-plays that collectively accounted for only about 6% of the total capitalization of the US equity market at year-end 1999. That view turned out to be dead wrong.”