The recent contraction in the spread between Jumbo- and conforming-balance mortgage loans has obscured continued problems in the market for non-agency loans. The lack of a securities market outlet for loan production means that virtually all loans ineligible for agency execution must be held in portfolio by the lender. In addition to requiring LTVs of 75% or lower, the pricing of non-agency loans is dependent on each lender's balance sheet and capital position, along with their appetite for credit and interest rate risk. Ultimately, the banking system's collective balance sheet is not large or strong enough to indefinitely support the upper tier of the housing market on its own.
In this light, a key question for the housing markets (especially in high-cost states such as California) is how soon the new-issue private-label market will recover and provide economical execution for prime jumbo-balance originations.
While capital market execution is theoretically available, it is currently uneconomical for newly-issued loans. Even if highly aggressive assumptions are used, securitized execution currently translates into par price for jumbo loans in excess of 9%, as compared to the Bankrate.com national average of 6.16%. The problem is two-fold: new-issue subordinates would trade at very low prices, and the subordination required by the rating agencies would be extremely large.
A number of potential solutions have been proposed to alleviate the problem posed by the high required credit support levels. One modification to traditional securitization practices would be for lenders to hold a first-loss piece on any loans they securitize, reducing the amount of subordinates that the underwriter would need to place. However, this would still require a substantial capital outlay by the originator; in fact, the capital necessary to hold a first-loss piece may be greater than that required to hold a loan package.
An alternative would be to finance new lending by issuing covered bonds. These are securities that are collateralized by a pool of loans but, unlike traditional MBS, don't represent an ownership interest in the underlying mortgages' cash flows. As with the holding of the first-loss piece, however, covered bond issuance would not alleviate balance sheet issues and reduce the need for capital, even if execution levels are economical.
Nonetheless, some observers expect an early recovery for the private-label market. They argue that investors view prime residential mortgages as fundamentally high-quality assets, based on their long history of steady credit performance; once there is evidence of improved performance for newly issued loans, they expect execution to improve quickly. In this view, the problems with loans originated between 2004 and 2007 resulted from a combination of sloppy underwriting (which theoretically has been corrected) and the collapse of a once-in-a-lifetime real estate bubble.
I suspect that recovery may be a longer process. The experience of the last few years has demonstrated the strong link between borrower performance and homeowner equity, which, in turn, is driven by the direction of home prices. I also expect real estate prices to remain under pressure. In addition to working out the flood of foreclosures, demographic trends (such as the aging of the Baby Boomer population) will begin to alter demand patterns in many communities. While credit performance will certainly improve, it won't be enough to sufficiently boost execution enough to make it an economical outlet for loan production.
In this case, the Jumbo market will remain dependent upon the ability of portfolio lenders to fund lending activities with reasonable rates and terms. This will leave the upper tier of the housing markets, vital to high-cost states such as California, vulnerable to the continued weakness of the banking industry.
Bill Berliner is a consultant based in Southern California. His Web site is www.berlinerconsulting.net
(c) 2009 Asset Securitization Report and SourceMedia, Inc. All Rights Reserved.