QRM - tomorrow; More on Fingerprints; Changing LO Comp; Purchases Worth More than Refis?

By: Rob Chrisman

I am spending some time in Kansas, and some of the discussion is about how hiring is so much more interesting and fun than eliminating staff. Obviously the aggregators have cut back, but in the small and mid-size tiers "right-sizing" seems to be harder to do. There are lots of shops out there that are holding off making staff reductions for various reasons, and I hope that they're correct. Margins have indeed contracted, and no lender wants to lose the savings they built up after a great 2012 and first half of 2013. But other P&L changes are in the works - head count is just one thing. And it is impacting other industries. For example, starting January 1 UPS will no longer offer health benefits to working spouses of its employees who are eligible for insurance with their own employer. UPS, the 4th largest employer in the nation, said its healthcare costs are expected to rise 60% (to 11.25% in 2014 vs. the usual 7.00%) as the full impact of the Affordable Care Act is felt.

Is the economy strong or not? Yesterday a story in Mortgage News Daily reported, "Home prices keep edging closer to pre-crash price levels and today's Home Price Index report from Lender Processing Services (LPS) indicates that national prices are now back within 15.2 percent of that peak.  The index for June rose to $229,000 from 226,000 in May, an increase of 1.2 percent and is up 6.9 percent from the end of last year.  The peak, in June 2006, was $270,000. LPS used its loan-level databases and June 2013 residential real estate transactions to conduct a repeat sales analysis of home prices.  The LPS HPI represents the price of non-distressed sales by taking into account price discounts for bank-owned real estate (REO) and short sales."

Contrast that versus the news last week that Jeb Hensarling (R-TX), Chairman of the House Financial Services Committee, issued a statement last week in response to President Obama's meeting with financial regulators to discuss the implementation of the Dodd-Frank Act. Congressman Hensarling called the bill "Harmful to our floundering economy and in dire need of repeal." But is the economy really currently floundering?

"Did the CFPB or any other regulator announce the QRM news yet?" No - it is expected tomorrow from the FDIC, and the several other organizations that participated in formulating the "skin in the game rules." All indications point toward equating it with the QM rules (which is a good thing) with the possible addition of a 30% down payment twist (perhaps any loan with 30% will fit inside the QRM box and other attributes won't matter).

Yesterday the commentary described that while fingerprints don't expire in three years, if there are state-level issues, or a complication in the application process, an LO needs to have it done again. I received this note from Matt Ewald, a mortgage consultant with 1st Advantage: "Regarding the NMLS fingerprints, I've recently experience this expiring fingerprint issue and thought I'd shed some light on it for you. Where this 3 year rule comes into play is certain states (Florida comes to mind in my case) require the background check every year at renewal, so after 3 years in the system you can no longer have them use the old prints and you're forced to go get prints again of your same fingers and pay additional fees. The other time this will come up as an extra fee is when you add states, which many people are doing with the recent UST test implementation. In this case if you got licensed in your home state at the beginning of the NMLS you're prints have expired or will expire soon, so when you go add a new state instead of authorizing a background check for that state using your old prints you'll be told you need to go get new prints. It seems to be all about the money for NMLS and the states since I don't think your fingerprints change over time and they are taken electronically so it's not like the card has deteriorated." Thank you Matt!

While we're on some compliance, there seems to be continued confusion over LO compensation and whether or not a lender could reduce it if the deal with the original investor didn't work out. Most compliance folks know this already, but the CFPB offered up a supplement, more specifically Comment 36(d)(1)-7, which reads as follows:
 
"7. Permitted decreases in loan originator compensation. Notwithstanding comment 36(d)(1)-5, § 1026.36(d)(1) does not prohibit a loan originator from decreasing its compensation to defray the cost, in whole or part, of an unforeseen increase in an actual settlement cost over an estimated settlement cost disclosed to the consumer pursuant to section 5(c) of RESPA or an unforeseen actual settlement cost not disclosed to the consumer pursuant to section 5(c) of RESPA. For purposes of comment 36(d)(1)-7, an increase in an actual settlement cost over an estimated settlement cost or a cost not disclosed is unforeseen if the increase occurs even though the estimate provided to the consumer is consistent with the best information reasonably available to the disclosing person at the time of the estimate. For example: i. Assume that a consumer agrees to lock an interest rate with a creditor in connection with the financing of a purchase-money transaction. A title issue with the property being purchased delays closing by one week, which in turn causes the rate lock to expire. The consumer desires to re-lock the interest rate. Provided that the title issue was unforeseen, the loan originator may decrease the loan originator's compensation to pay for all or part of the rate-lock extension fee.
 
ii. Assume that when applying the tolerance requirements under the regulations implementing RESPA sections 4 and 5(c), there is a tolerance violation of $70 that must be cured. Provided the violation was unforeseen, the rule is not violated if the individual loan originator's compensation decreases to pay for all or part of the amount required curing the tolerance violation."

I don't know what upsets me more: me typing a song into YouTube and then having Google (which owns YouTube) in its Big Brotherish way, suggest a dozen songs that I might like. Or that Google is often correct. The securities industry is keenly interested in Richmond, California's eminent domain lawsuit, and the latest news is that "Richmond says lawsuit premature, sees no irreparable harm." Yup - let's see what happens to its market if Fannie & Freddie stop buying loans, or the FHA stops insuring, loans from that municipality. Richmond recently inquired about buying more than 600 troubled mortgages and threatened to use eminent domain to seize the loans. But the banks that provide customer service on the mortgages notified the city that the loans aren't for sale or that they don't "have the contractual authority to sell the loans."

The market for servicing is alive and well. Yesterday word went out that the MountainView Servicing Group has been retained on an exclusive basis to offer for sale $1.4 billion of bulk Freddie Mac and Fannie Mae MSRs. Bids are due on Thursday, August 29 at 12:00 pm Eastern Time. "We continue to see upward price pressure on conventional and Ginnie Mae servicing," said Matt Maurer, Managing Director at MountainView Servicing Group. "And this $1.4 billion conventional offering should be no different, given the portfolio's low weighted average interest rate and stellar loan performance."

Speaking of servicing, I received this note. "Rob, I understand that the price of a loan is based on the security price (in the secondary market) and the value of the servicing income over the life of the loan. Do investors pay more for refis than for purchases?" Good question. While the value difference is rarely a game changer/deal breaker, yes, purchase MSRs are generally valued higher than refis. There are many models behind it, but the best rationale I've heard is that it's a borrower behavior play - specifically with refis you know all the borrowers are actively rate seekers, but with purchases, there is at least some subset % that's passive... So therefore longer duration begets higher IO value. And Art Yeend with Mountain View writes, "It is not a simple answer since values are influenced by a number of factors. Although MSR valuation models can be very complex, value is a lot about determining life expectancy, cost to service, and loss exposure.  Given a deep dive through a number of attributes and factors, purchase vs. refinance normally isn't a significant driver of value, at least on most conventional products.  It can be more of a factor on streamline refinances because of LTV uncertainty and on government loans because of delinquency advance requirements.  And is a real issue with VA IRRRL's because of the fuzzy LTV's and VA guarantee which can require significantly more advances. So, yes in many cases they are valued basically the same, but not all!"

Which is interesting since it puts the residential mortgage lender in an odd spot. Purchase loans cost more to originate than refis, for basically two reasons. First, purchase loans are viewed as tougher to do. There is more communication with more parties (Realtor, borrower, inspections, appraisers, etc.), and more documentation and underwriting scrutiny. And second, purchase loans are typically closed during the last week of the month (refinance loans can usually be scheduled to suit the convenience and capacity of the lender). So many lenders find that they are paying overtime for staff near the end of the month in order to process purchase loans, which always take priority over refis. And as STRATMOR points out, "This means that back office staffing will not decline in direct proportion to volume declines if most of the decline is in refinance volume. In effect, then, as mix shifts to purchase loans, unit fulfillment costs will rise."

Rates improved a little yesterday, and although it was due to a weak Durable Goods number, we'll take it! Durable Goods is always a volatile number, and after three months of increases, the unexpected July drop of 7.3% should be taken with a grain of salt. "The report shows struggling overseas markets and the effects of federal government spending cuts are lingering and holding back manufacturing, which accounts for about 12 percent of the economy." MBS prices on 30-year product opened a tick higher along 3.5s through 4.5s and then quickly rallied .375 in price. Price is certainly a factor of supply and demand, and trading volumes suggested that lock volume and hedging is waaaaay down (a technical term).

Today we'll have the 9AM S&P/Case-Shiller home price indexes for June (June? It is expected +1.0) and an hour later both the August Richmond Fed and Consumer Confidence (expected moderately lower). Lastly, Treasury auctions off $34 billion of 2-yr notes (5- and 7-year auctions Wednesday and Thursday). Yesterday the 10-yr ended at a yield of 2.80%, and in the early going it is down to 2.78% - look for agency MBS prices to be a shade better.