Are credit risk transfers fatally flawed?
“One part of GSE reform is already working,” writes Mark Zandi of Moody’s Analytics. He refers to credit risk transfers, wherein Fannie or Freddie pay outsiders to assume certain levels of credit risk in various static mortgage pools. CRTs are, he says, “an unheralded success story.
That assessment is premature, because the viability of CRTs can only be tested during a cyclical downturn, which seems far off.
Indeed, CRTs have a serious flaw that is best illustrated in Fannie Mae's latest chart on, Cumulative Default Rates of Single-Family Conventional Guaranty Book of Business. Loans originated over a 15-year time span, from 2002 up through the second quarter of 2017, are segmented by year. As we can see, most vintages—2002, 2003 and 2009-2017— were highly profitable because of nominal default rates over time.
Default rates for years preceding and following the mid-2006 peak of housing prices, vintages 2005 through 2008, are of an entirely different order of magnitude. Without the defaults and consequent losses suffered by loans booked in 2005 through 2008, Fannie would not have faced financial duress. The credit risk transfers for vintages 2005 - 2008 would have reduced the losses borne by Fannie Mae.
But the CRTs sold before and after that four-year period—2002-2003, 2009-2017—would have done nothing to reduce Fannie's financial distress. In fact, Fannie would have been better off had it not sold CRTs during those years and simply kept the interest and fee income. Up until now, the GSE model diversified of credit risk of its all of its mortgages in a single pool. The profitable vintages offset the losses from the unprofitable vintages. CRTs upend this tried-and-true system for diversifying market timing risk.
Zandi discusses how CRTs would protect Fannie and Freddie if they faced another severe downturn:
If a similarly severe downturn occurred today, the agencies would suffer smaller losses given that the mortgage loans and securities they own are of much higher quality. But more important, because of the capital market risk transfers now in place, approximately two-thirds of the losses would be borne by private investors, not Fannie and Freddie. And given how much capital the agencies would be holding if they were private institutions, they would avoid insolvency and another government takeover.Perhaps. It all depends on whether the GSEs would be able to sell credit risk on loans originated in 2006 through 2008. In early to mid-2006, California delinquencies spiked and prices began to level off. This would have signaled to anyone who is familiar with housing cycles that prices were very close to a cyclical peak. Remember, residential mortgages represent the only major credit market that relies on asset appreciation. And every housing boom has been followed by a multi-year period of stagnant or slumping prices. So investors, anticipating an end to the boom, would simply stop buying CRTs until housing began to emerge out of a cyclical trough.
Which is why an assumption, that Fannie and Freddie can sell CRTs during a downturn like that of 2007, should be viewed with skepticism.
3 comments - Are credit risk transfers fatally flawed?
How can any one trust reform plans proposed by such experts?
Thanks, for easy to understand great analysis with visual presentation.
This proves the point JTIMOTHYHOWARD has been making in his abstract financial analysis.
When pro financial establishment analysts like Mark Zandi like something and recommend it, then it must be highly beneficial to oneself and his clients at the cost of FnF. It is self revealing for Mark Zandi.