In May, new home sales sunk to the lowest level since at least 1963, when the Census Department began keeping records. They fell to an annualized rate of 300,000, down 33% compared to April. This number is much worse than the 405,000 new sales economists expected. Unlike existing home sales, new home sales more fully take into account the loss of the home buyer credit. Clearly, the housing market has taken note.

Today's report is bad in pretty much every way imaginable. Not only did May's sales set a new all-time low, but March and April were also revised downward significantly. March fell from 439,000 to 347,000 sales, and April fell from 504,000 to 446,000. If you use the previous estimate for April, home sales would actually have fallen by 40%. So really, May's numbers showed twice the 20% decline economists expected. Given this trend in revisions, it wouldn't be surprising if we learn next month that May's new home sales are even lower than today's first estimate indicates.

Here's a historical chart of new home sales, since 2000:

new home sales 2010-05.PNG

You can see how far we've fallen. Sales peaked at 1,389,000 in July 2005. The annualized rate of sales has fallen by more than one million since then.

What's an even scarier prospect for the housing market is that this might be where sales are settling. The Mortgage Bankers Association also released their weekly mortgage application data today. According to that, applications for new purchases were down last week by 1.2%. That's after the first rise in six weeks the week prior. At this point, it looks like the market might have found a new equilibrium. Here's a chart showing the applications for new purchases since the year began:

mba purchases 2010-06-23 - detail.PNG

It's pretty clear that the level is beginning to stabilize around the green line -- last week's level. This is based on the MBA's index, so it's hard to know precisely what this number is. But that rate is extremely low, as this historical chart shows:

mba purchases 2010-06-23.PNG

Applications for purchases are down 39% since they peaked at the end of April. Now they're somewhere near the mid-1997 level.

The gut reaction to this kind of lackluster demand for new homes is to worry about jobs. After all, construction will certainly suffer if fewer new homes are being built. But the hope is that renovations on foreclosures and short sales could pick up some of that slack.

It's actually better for the housing market if buyers focus on draining the inventory of existing homes before building more new ones. Yet, the MBA data casts doubt on whether there's much demand in that aspect of the market either. Although existing homes sales only fell slightly in May, by 2%, that was much worse than the 5% rise in sales economists expected. It also doesn't fully take the housing credit's absence into account. We won't see how demand for existing home purchases has changed without the credit until July.