BANKTHINK
RISK DOCTOR

Let the FHA Charge for Risk, Like Fannie and Freddie Do

Print
Email
Reprints
Comment (1)
Twitter
LinkedIn
Facebook
Google+

The mortgage securities business has become a shell game. Consider the disconnect in pricing of credit guarantees between Fannie Mae and Freddie Mac, on the one hand, and the Federal Housing Administration on the other.

One of the important differences in pricing between the government-sponsored enterprises and FHA is that Fannie and Freddie adjust their prices to some degree for risk while FHA largely charges fees that are flat across risk attributes. And although FHA has raised upfront and annual mortgage insurance premiums since the crisis, the all-in pricing for a GSE mortgage-backed security and a corresponding Ginnie Mae MBS (which is where most FHA loans end up) leaves FHA the clear winner for the riskiest pools. In that regard, we are merely shifting risk from one set of federally insured entities (the GSEs as wards of the state) to another. There does not seem to be any apparent rhyme or reason to this pricing — one more piece of evidence of the lack of cohesion in pricing in federally dominated mortgage markets.

The issue is far more than academic. The latest actuary report estimates that the FHA’s Mutual Mortgage Insurance Fund is underwater by more than $16 billion although the Obama administration’s estimate released Wednesday places it now at just under $1 billion. If FHA employed a more granular pricing approach, charging much higher premiums for riskier attribute combinations and less for lower-risk attributes, it would reduce the risk of the agency’s troubled MMIF over time.

Look at the pricing decision between a pool of 95% loan-to-value, 660 FICO score fixed-rate 30-year mortgages going into a 3% coupon Fannie Mae or a comparable GNMA security. The Fannie bond was recently quoted at $104.09 per $100 of principal, the GNMA at $105.56. For this pool, Fannie would charge an “adverse” delivery fee of 25 basis points, a loan-level pricing adjustment of 225 basis points. And since the loans have such little equity and low credit scores, a private mortgage insurer would charge 350 basis points for the risk that it bears.

For the same pool, FHA would charge an upfront premium of 175 basis points and an annual premium of 130 basis points. If we further assume that the pools have an effective life of five years, then that annual premium takes 650 basis points off the dollar price of the bond.

Taking into consideration origination and differential servicing costs that also affect the bond coupon, the net price of the Fannie MBS is 95.24 per $100 of principal while the GNMA MBS is $97.57. In this example, the higher-priced GNMA MBS is the “best execution” for delivery of 95% LTV, 660 mortgage pools.

This simple illustration of relative mortgage pricing underscores the need to revisit FHA’s approach to pricing and it has implications for the extent of the federal guarantee as well.

The lack of any risk-sharing between FHA and other suitable counterparties such as private mortgage insurance companies amplifies the contingent liability of the MMIF. Although the GSEs have their own problems (which they appear to be working off), the ability to share risk at least for loans with LTVs above 80% provides an effective mechanism for distributing risk. And this year’s credit enhancement pilot with the GSEs initiated by the FHFA will further test the boundaries of viable credit structures, including deeper use of private mortgage insurance.

FHA should be permitted to follow suit and test the waters for risk-sharing arrangements with private mortgage insurance providers.

It is hard to imagine how the pricing for a fund the size of the MMIF ($1 trillion plus and counting) could be handled in such a simplified fashion. Still, easy as it is to take aim at FHA given the state of the MMIF, remember that FHA has served this country well for nearly 80 years, however. Like other institutions that struggle with the pace of change in their markets, it has been unable to adapt to significant structural changes in the mortgage capital markets. The advent of sophisticated risk transfer mechanisms and pricing capabilities of other market participants has left FHA at competitive disadvantage.

The time is now for FHA to adopt risk-sharing and risk-based pricing practices in order to put the MMIF on a secure financial footing that will ensure FHA’s important legacy for borrowers for the next 80 years.

Clifford V. Rossi is the Executive-in-Residence and Tyser Teaching Fellow at the Robert H. Smith School of Business at the University of Maryland. 

JOIN THE DISCUSSION

(1) Comment

SEE MORE IN

Who's Who in Auto Lending Investigations
As U.S. auto lending has boomed, the industry especially its subprime sector has become a growing target for a slew of prosecutors and regulators. Here are seven government agencies to keep an eye on.

(Image: Bloomberg News)

Comments (1)
Having introduced the FHA and FHFA to the concept of risk sharing during a series of 2010/11 presentations I was pleased to see that the concept had taken root in FHFA's strategic goals & scorecard. Risk sharing is a critical response tool necessary for the execution of efficient long-short strategies. However, I will caution that reliance on pre-crisis sharing strategies will, holding all else constant, create the pre-crisis environment I believe we would all prefer to avoid. The FHA, as an insurance provider, should be equipped to manage the risk they insure, including access to tools and methods for doing so, but without the expertise to wield properly we will simply end up with one mortgage insurance company promising to protect another mortgage insurance company...and we saw where that series of domino's leads.
Posted by RBirtel | Thursday, April 11 2013 at 2:39PM ET
Add Your Comments:
Not Registered?
You must be registered to post a comment. Click here to register.
Already registered? Log in here
Please note you must now log in with your email address and password.
Already a subscriber? Log in here
Please note you must now log in with your email address and password.