I am currently working on demand factors during the housing boom. I don't know how to proceed here with any description of the period that isn't going to sound crazy to most readers. Here's a taste of why.
First is the proportion of securitizations, by type. In 2003 and 2004, with the misdiagnosis of the housing problem already affecting policy, the Bush administration, supported by Congress and the Federal Reserve, started pushing hard against the GSEs. GSE securitizations, as a result of the pressure, both formal and informal, declined sharply. In effect, the government knocked out the conventional form of home financing without replacing it with anything. Since home prices were high because of supply problems, not demand problems, this did nothing to stop the trend in home prices. Notice in this graph that subprime didn't actually increase that much as a proportion of mortgages, and to the extent that they did, they replaced Ginnie Mae loans, which had similar characteristics. Alt-A loans replaced GSE loans.
There was a sharp decline in the proportion of homes funded by either Ginnie Mae or subprime loans between the 1990s and 2007 - a drop in market share of more than 10%. This is why, despite all the hand wringing about deteriorating borrower characteristics, it doesn't show up in national data.
The GSE's barely managed an 8% growth rate from 1998 to the height of the boom. The subprime and Alt-A loans were not excess - they were the first responses to demand deprivation. If you have to bid up the real estate in Los Angeles or New York City to get into the job market there, you're going to need a mortgage. That problem doesn't go away just because we pressure the GSEs.
During the period when the GSEs were being pushed around, banks were also filling in the gap. Mortgages were increasing as a portion of bank assets. They weren't suddenly engaged in a game of moral hazard, pushing off irresponsible mortgages to securitizations. As with so many issues here, the facts are in opposition to the normal story.
Part of what I see in the readings, and I'm not completely sure how to address it, is that every single review of the period basically just treats the entire mortgage market as a grab bag of scary statistics to relay about the excess that we all just know was happening, so it confirms our preconceptions. So, a reading will mention that half of some bank's loans were no doc, and another bank lowered its FICO scores on a certain type of product, and another bank increased the mortgages it issued with low down payments. But, you have to read carefully to notice that those no doc loans were for jumbo loans with 40% down payments and the low FICO scores were in a region with low home prices and the low down payments were on products with financially secure buyers, etc. Basically, every mortgage is a compromise among several factors, and it seems to be commonplace to just pick the worst sounding parts, because that's what everyone needs to hear.
So, you go look at national data, and it's like bizarro world. Nothing much happened with home owner incomes, down payments, mortgage payments, etc., in the aggregate. It really is just a good, old-fashioned moral panic that we all just had.
Remember, home prices rose until early 2006 but home ownership topped out in early 2004 - which just happens to be when the GSE's were cut down and Alt-A came in to replace them. The Bush administration basically just undercut the means that regular Americans used to buy homes. And, remember, Three-quarters of the country lived in places where mortgage affordability was quite good.
In the second graph, note that securitizations actually dropped from trend in 2004 and 2005 when home price increases were at their sharpest. Sure, it was a high trend (15% per year), but IW readers know that was due, in large part to rising prices in cities that won't build houses. It certainly wasn't due to prices in places like Dallas and Atlanta and (at that point) Phoenix, where most of the homes were being built. The private conduits were filling a gap. Imagine what disaster would have happened in 2004 if they hadn't. The government knocked out the longstanding method of safely funding mortgages, so it had to be replaced with something.
At every step, our attempts at solving this problem with demand deprivation just created dysfunction and dislocation, and this began as early as 2003. The reason we had a crisis is because the universal certainty felt about this error would not stop at anything else. A decade into it, that certainty remains strong.