Servicers Matter for Compare Ratios; How will the FHA Insurance Premium Change Impact Production?

As has been said, there is plenty of blame to go around. Sometimes the press even notices that not all borrowers are above culpability.

Texas is hoppin'! Pacific Union Financial is looking to fill the following positions for its new Dallas/Fort Worth Wholesale Fulfillment Center: Operations Manager, Underwriting Manager, Funding Managers, DE Underwriters, Account Managers, Closers and Setup Personnel. Candidates must have extensive experience with FHA loans and they would prefer applicants to have a minimum of 5 years in recent wholesale experience. Candidates must reside in the greater Dallas/Fort Worth Area or be willing to relocate to the region. If you're interested contact Darius Mirshahzadeh at darius@loanpacific.com.

(By the way, yesterday the commentary mentioned Houston's First Continental Mortgage looking for retail originators, branches and company acquisitions in the South, Southeast, and Texas. The company was actually founded in 1992, not 2002, as noted. Contact Paul Peters at ppeters@fcmchou.com to discuss opportunities.)

Also, yesterday the commentary mentioned Grand Bank's OCC consent decree, along with the bank owning ICON Residential. As a clarification, the order has been in place since October - there is no new order that I am aware of, and the commentary did indeed say "rumored." "Nothing new" is a good thing in today's investor environment, and hopefully there was no confusion.

Lastly, I would like to clarify some news about MGIC that the commentary published Sunday. The pricing actually is effective March 12th, not last week. In addition, "borrowers with credit scores over 760 (not 660!) can take advantage of reduced monthly, annual and single premium rates on their loans."

It seems that lenders are watching their compare ratios more than their waistlines. One vet wrote, "I have read, with middling interest, the thoughts and commentary put forth by your readers regarding the FHA compare ratio.  For me, the thing that is striking is how much of it is servicer dependent.  Like many, I sold our FHA production to various conduits and ultimately have to live with how effective, or ineffective, they are at servicing the loans we sell to them.  As an organization, our FHA compare ratio is in the mid-70's on a combined basis. However, if you break that down into the various branch IDs that feed into the overall score, I have one branch at 154%, one at a 125% and on down from there. If I then break the loans in each branch down by my eight largest servicers, I note that one servicer accounts for 54% of my "seriously delinquent" and "claims" population but only 26% of the loans being included in the calculation of the compare ratio."

The vet continued, "Digging deeper, that same entity had priced itself to a commanding market share in the two localities that are included in the 154% and 125% compare ratios.  And in fact, although this servicer has about a 60% share with us in these two localities, they have all the reported seriously delinquent and claim loans versus everyone else with no claims or serious delinquencies.  Bottom line, I'm unconvinced that the poor FHA compare ratio that is being reported in those two locales is attributable to the credit decision made by my organization. Instead, it looks to me like poor selection with respect to which entity ended up with our servicing is the sole driver of our compare ratio in these two localities in particular but for my organization in general.  So, I think I'll stop worrying about FICO scores, or credit overlays, or streamline refinance with or without an appraisal and instead focus on who is servicing my loans as the best and strongest predictor of FHA compare ratio."

Of course, many delinquencies are caused by job losses, and another wrote, "The argument about lots of people losing their jobs seems like a classic Type II error.  Job losses should be evenly dispersed, unless smaller lenders are 1) not adequately confirming stability of employment (I believe three years is the relevant FHA standard) on its borrowers, or 2) they were really concentrated in lending to employees of the local factory that unexpectedly shut down.  #1 is an error attributable to the lender and should be rightly reflected in their compare ratio, #2 is a statistical aberration that should be easy to explain.  #3 is that the entity servicing their loans is using a predictive dialer 30 days after delinquency where another servicer might be calling, emailing  and writing borrowers feverishly starting the 16^th of the month and would apply in those situations where job losses in an MSA are evenly dispersed across loans made by all lenders in said MSA."

Over the last week, there have been numerous news articles and official releases on FHA premium changes. Investors are ruminating on how expectations for FHA premium changes will impact the refinancing incentive for GNMA pools issued at different periods of time, especially streamline refinances. It is a "slicing and dicing numbers" game: the impact of changes to the annual MIP structure is contingent on the loan origination date, the current MIP structure, the future MIP structure, the current rate, and even the changes to the upfront MIP. Some borrowers would actually have a "disincentive" to refinance given their current MIP versus what it will be starting in April and then July. For loans originated prior to May 2009, the disincentive to refinance goes down from 60bp to 0bp. In other words, these borrowers will see their incentive to refinance increase by 60bp once these changes become effective, all else being equal. On the other hand, the incentive to refinance for all other borrowers actually declines by 10bp. Based on the prevailing and future MIP structures, many investors are now interested in owning Ginnie II 4.5% pools issued prior to October 2010 and after June of 2009 - the thinking being that these borrowers will leave well enough alone.

Most know that unlike closing costs, the upfront annual insurance premium can be rolled into the balance of the loan when the borrower goes through a streamline refinancing. Thus, the 75 basis point increase in upfront premiums should lead to a roughly 10 basis point disincentive for potential borrowers to refinance. This coupled with the 10 basis point increase in annual premiums, means that FHA borrowers will see a 20-25 basis point increase in their effective mortgage rate. Given that approximately 75-80% of GNMA MBS is FHA, this should lead to a 15-20 bp reduction in refinancing incentive, all else being equal.

Continuing, how do the changes impact originations from a practical viewpoint? I received this note from Dave Lewis: "HUD raised its UFMIP as well as its annual MIP, and instituting the majority of the rate hikes into case number assignments beginning April 1, 2012. Remember that when the current administration took office, HUD loans, in general, had a 2.25% UFMIP and .50% annual premium.  On April 1, 2012, HUD loans will cost 1.75% UFMIP and 1.25% in annual premium.  Folks need to keep things in perspective! As the month of February came to a close, we had a customer and her husband closing a conventional, 85% LTV, primary, SFD, cash out refinance.  The private mortgage insurer is charging them an annual premium of .57%.  Assume for one minute that this couple applied in mid-April of this year.  Assume that they receive their required 'alternate methods' disclosure, showing the differences between a conventional, FHA and let's say an ARM loan.  These customers have mid FICOs in the 729 range."

It goes on: "That simple disclosure will tell the story that you will write a few years from now.  HUD has managed, (again) to set the insurance fund up for adverse selection.  As its fees have risen, its credit quality will decline.  Any reader out there can do the quick analysis.  These anonymous borrowers are being dinged by the Agency LLPA and by the private MI company for requesting a loan for the purpose of getting cash out, and they are being walloped for having a mid FICO under 740.  That said, their rate is 4.375%, with an annual 57 basis points for MI coverage, their cost of money is 4.945%, (the loan is being done at zero points). By mid-April, the same request, using a HUD loan instead, will cost them the nominal 4.375%, plus 1.25% annual MIP  for five years, resulting in a whopping 5.625% annual cost, and they would pay HUD 1.75%, ($3,937 in this case, on a $225,000 loan) for the privilege of closing on a HUD insured loan.  Obviously, the side-by-side comparison will eliminate HUD as an option on this loan.

"But suppose the borrower's FICO was some number between, let's say 580 and 680.  At those levels, the HUD alternative begins to look a little better. The lower the FICO, the better the comparison gets in favor of the HUD alternative. So, in the average month, the Agency loan portfolio will include more loans made with scores over 700, and the HUD insurance fund will be insuring more loans, (comparatively) with scores under 700.  Is it hard to predict that three years down the road, the delinquency rate will be higher on the HUD sample than on the Agency sample?  The folks at HUD are expert marksmen; they never miss their foot when they shoot." Thank you Mr. Lewis.

Monday, was, for lack of a term, quiet. We've been in the same price range for what seems like months, and yesterday prices went up or down a shade, but by the end of the day things were nearly where they were at the end of Friday. Today might be the same, as the economic calendar is nil. The 10-yr T-note closed at 2.00% and stocks were down a shade, but this morning rates are down slightly (1.97% on the 10-yr; MBS prices better by .125) ahead of "Super Tuesday."


(Part 2 of 3)
After the budgie jumping disaster yesterday, Seamus arrives up at Connor Pass.
He's been to the pet shop too and walks up to the edge of the cliff carrying another cardboard box in one hand and a shotgun in the other.
"Hi, Paddy, watch dis," Seamus says.
He takes a parrot from the box and lets him fly free.
He then throws himself over the edge of the cliff with the gun.
Paddy watches as half way down, Seamus takes the gun and shoots the parrot.
Seamus continues to plummet down and down until he hits the bottom and breaks every bone in his body.
Paddy shakes his head and says, "And I'm never trying dat parrotshooting either!"


If you're interested, visit my twice-a-month blog at the STRATMOR Group web site located at www.stratmorgroup.com . The current blog discusses the possible future roll of Freddie Mac and Fannie Mae as the FHFA might model it. If you have both the time and inclination, make a comment on what I have written, or on other comments so that folks can learn what's going on out there from the other readers.