Earlier I discussed the tendency for mortgage issuers to "discover" new categories of nonagency loans that can be securitized and sold into the capital markets. However, one trend from the past several years runs in the opposite direction. The agencies have expanded their underwriting guidelines so that mortgage originators can include in agency pass-through pools many loans previously not eligible for agency pooling. In particular, agency underwriting guidelines are now broad enough to include a sizable percentage of alt-A and subprime loans. However, because no data have been released on this topic, it is hard to estimate how many previous alt-A and subprime loans are now being securitized as agency pass-throughs, but I assume that it is sizable.
The agencies also have become involved in the nonagency sector in other ways. In 2001-2002, both agencies wrapped a large number of subprime deals. One of the stated objectives of this policy was to learn more about the subprime sector. Beginning in 2003, both agencies became active buyers of subprime deals, focusing on the AAA floaters, and they deemphasized their wrap programs. During 2003 and 2004, the agencies at times purchased as much as 50% of many subprime deals. After the accounting problems at Fannie Mae, there has been some speculation that the agencies, especially Fannie Mae, might withdraw from some MBS markets, including subprime. However, I suspect that they will remain active buyers of subprime securities for the foreseeable future. I say this because subprime loans have several attractive features for the agencies. They provide a yield pickup over competing securities, they do not require the expensive hedging that fixed-rate securities require, and subprime securities count toward the agencies' affordable housing goals, which are being increased over the next several years. Having said this, agency activity can fluctuate significantly from month to month because they are opportunistic buyers (i.e., buying more when spreads are wide than when they are narrow).
I have shown how the characteristics that issuers use to define their different shelves determine, to a large extent, the prepayment and credit behavior of the various parts of the nonagency market. These definitions have evolved over time and sometimes differ from issuer to issuer. Even with these changes, though, the definitions presented here are still useful. By examining the characteristics of individual deals, investors can make reasonable estimates of the future prepayment and credit performance of the nonagency MBS they purchase.
Was this article helpful?