Should we be excited about the July new home sales number?: The short answer is no. The July number was the worst July number since 1982; it just wasn't as bad as the June number, which wasn't as bad as the May number.
Everybody wants to know if we have hit bottom. There are three indicators suggesting we have--and three suggesting not. The good: prices in many markets have fallen below replacement cost (which is a pretty robust fundamental in the absence of population declines). Morris Davis at Wisconsin has shown that rent to price ratios have returned to be more in line with long term ratios, and given how low mortgage rates are, this is comforting. And resale inventories in California have dropped to under 4 months.
On the down side, we may have a lot of foreclosed houses coming at us in the next year. The employment picture is still atrocious. And if rents keep falling, prices will follow.
I would also guess that the first-time homebuyer tax credit is time-shifting sales, rather than raising them for the long term, but we shall see. On the other hand, the nature of investor sales is actually a positive indicator: investors are buying with cash and renting out units at decent rates of return. This is very different from the borrow, buy and flip model from the earlier part of this decade.
FWIW, I would assign a subjective probability of .7 that we are at bottom. On the other hand, around 2005, I assigned a .35 probability that we were about to face serious trouble.