In the WSJ, FNC reported that jumbo mortgages saw a larger year-over-year gain than conforming:
Home sales using a jumbo mortgage had year-over-year growth of 7.9% through September, compared with 2.7% for nonjumbo sales, according to an analysis for The Wall Street Journal by mortgage-technology company FNC.
Could this be a sign that credit is thawing a bit more quickly at upper end of the market?
Capital Economics Ltd., in a recent research note, found that jumbo loans are going to borrowers with credit scores as low as 700, compared with 720 or higher previously, and that financing has generally reached $2 million from a previous upper limit of $1.5 million.
Anecdotal sure, but when looking at the actual jumbo mortgage data, it appears that from 2005 to 2012, mortgage volume for jumbo fell 83.1% and non jumbo fell 46.9%. In other words, jumbo mortgage volume fell 2x further than non jumbo from peak. Also, a number of the high cost markets had their base level lowered expanded what is now considered a “jumbo loan”:
Also, the floor for a jumbo loan fell in some high-cost areas last fall. In Los Angeles and New York, for example, the definition of a jumbo dropped to $625,500 and up from $729,750 and up. With the lower floor, a loan of $700,000 would now be a jumbo loan.
So the fact that jumbo volume is up 7.9% versus 2.7% reflects it being calculated from a much lower base number, and with lower jumbo thresholds, more loans are being classified as jumbo. This likely resulted in a somewhat larger jumbo market share, reaching 5.5% of total mortgages in 2012 compared to 5.2% in 2011.
My takeaway here is that jumbos are not growing at 3x the rate of conforming as FNC seems to be suggesting, but jumbos (5.5% of the first mortgage market) are more likely consistent with the balance of the mortgage market.
Since jumbos have no real secondary market to allow mortgage lenders to free up their capital to lend more, jumbos are actually performing amazingly well however you slice it, just not better than conforming mortgages.