The Washington Post had a piece explaining the seeming anomaly that cash-out refinancing is still well below bubble levels even though house prices have recovered much of their ground. The piece explains this gap by the fact that homeowners are a wealthier group on average than they were in the bubble years and therefore less likely to tap equity for spending.
While this is in part true, the more obvious explanation that is that inflation adjusted house prices are still almost 30 percent below the peak of the bubble, which is a good thing. It is also likely that homeowners do not expect continually rising house prices as they did in the bubble years which would make them less likely to withdraw equity from their homes.
As a practical matter, it is not plausible that the decline in moderate income homeowners could explain much of the $200 billion drop in annual cash-out refinancing between the bubble peaks and 2013. The homeownership rate has fallen by a bit less than 4 percentage points from its bubble peaks, translating into roughly 4 million fewer homeowners. For a $200 billion drop in annual cash-out financing to be explained by this loss of moderate income homeowners, we would have to believe that these moderate income homeowners had been cashing out $50,000 a year on average from their homes.
Given that this group would have had limited equity in their homes and their homes would have been lower priced than the average, it is implausible that on average they would have been cashing out even one-fifth of this amount. But hey, why mess up a good story with arithmetic?