Monday, September 9, 2013

Broker v. Lender: Lender's Perspective - Trying to Make Sense of QM Pricing Model

The Qualified Mortgage (QM) points and fees and Higher Priced Mortgage Loan rule (HPML) threshold calculation looks like an MC Escher drawing where you go around an infinite staircase.  While brokers face QM comp issues, lenders have an even more daunting challenge: trying to calculate allowable discount points, LLPAs, MI, and LO Comp, and still keep loans below HPML thresholds.  In anticipation of the January, 2014 QM and Loan Originator Compensation Rule (LO Comp) implementation dates, the industry as a whole struggles to make sense of the regulatory scheme.

Lenders face even greater challenges with the conundrums and complexities of the rule in its implementation. The HPML adds wrinkles to the equation.  Even with a clear understanding, a practical solution to meeting both a borrower's financing needs and maximizing lender compensation may represent two incompatible objectives.

The Objective:  A Qualified Mortgage

Remember, the whole point behind this set of pricing exercises is to limit lender liability.  QM removes the borrower's ability to pursue a claim that the lender did not consider his or her ability to repay under the Qualified Mortgage Ability to Repay Rule (ATR).  As long as the loan meets Appendix Q requirements a court necessarily might dismiss the action. However, with an HPML, the lender loses rebuttal presumption exposing himself to risk if a borrower chooses legal recourse to challenge the mortgage.  Even if all other elements of ATR fall into line, the lender still takes risk when exceeding HPML thresholds because of the borrower's right to challenge.  Of course, the lender can still rebut the borrower's assertion but must prove ATR conclusively. (** See Blog Comment Below - attorneys may find a way around rebuttal presumption.)

When trying to parse out the various permutations of QM and HPML, it helps to look at specific examples. The biggest concern for many is the effect of the Loan Level Price Adjustments (LLPA) on the cost and rate structure. Since LLPAs are secondary market loan sale eligibility fees, and not specifically profit or pricing gain, it seems unfair, to lenders and borrower alike, to have had them included in the calculation of points and fees. One potential solution: "adjust out" the loan level pricing by adding to the coupon rate - effectively buying out the LLPA upfront cost.  While for brokers the regulatory scheme may simply limit fees and create triggers for HPMLs, for lenders the rules create a minefield of quandaries.

Quandary 1:  What is the price?

If you cannot see the embedded spreadsheet above, then click on the link to the right.

To determine whether a loan will trigger HPML and how much discount you can charge to offset LLPAs, you need to work backwards into a price in a calculation similar to borrower pre-qualification.

You may download a copy of this spreadsheet to work on by clicking here.  If you have found another resource, feel free to share it and we will post it on our updates page.

Until you have loan level price adjustments factored in you have no way of knowing the final price of the loan, and no way of knowing whether you can charge an adequate discount to offset the LLPA. Naturally, you can always increase the interest rate, tantamount to buying out the LLPA fees.  But this may be of little comfort if it triggers HPML.

Quandary 2:  Limits to Buy-ups and Prepays?

If the most you can charge, net of compensation, is two discount points (provided you keep your rate within 1% of the Average Prime Offered Rate (APOR)), a higher LTV, lower credit score borrower could potentially price himself out of the lenders rebuttal presumption for QM.  It may not be possible to charge enough fees to cover all of the upfront cost of the LLPAs.  To deal with this the lender may solve this problem by "buying up" the rate to "zero out" or minimize the upfront LLPA.

The problem: The buy-up/buy-down cost formulation has always had an options seller's position built in. From par, buying down the rate with points costs progressively more and buying up the rate yields progressively less.  The more you are out of the money, the less that option is worth (or the more the insurance costs).  You may essentially "run out of rope" as you move up coupon rates to offset the cost.  This is particularly true on ARM loans, which offer a much more restricted rate "buyup".

Buy-Up Rate to Reduce LLPA Fees - it gets expensive and you eventually run out of rope.

Quandary 3:  Bought up the rate, now the loan is HPML

Secondary has always experienced the added prepayment risk from loans priced with high yield spreads. The higher the coupon relative to the market, the higher the prepay rate. Lenders offset this risk by adding prepayment penalties.  However, if the coupon is high enough to make the loan an HPML you may not be able to offset with a prepayment penalty because the Higher Priced Mortgage Loan rule prohibits these features.

This calculation shows how compensation, when combined with LLPAs for even moderately priced loans can create HPML Status.  In addition, high LLPAs can create HPML loans by their very nature.  Will the market accept the risk of not having presumptive rebuttal status?

Quandary 4:  You can't get there from here - APOR's artificiality

The regulatory scheme seems to imagine the mortgage market as some sleepy backwater drifting along oblivious to the machinations of global credit markets.  A loan with steep LLPAs already looks like a candidate for triggering HPML status.  The features of the APOR as an index calculation can make avoiding this impossible.

The APOR index looks at last week's data.  Because of this, depending on when an originator looks to lock a loan, APOR's relation to current market can be as much as two weeks old.  In addition, because this survey represents a range of lender pricing we don't know the extent to which the surveyed rate has been impacted by hedging strategies and pricing concessions.  One only needs to look at the rate surveys posted in local daily newspapers to see how bottom fishing originators may skew this data.

Because we confer QM status at the time of lock-in, the APOR index, with market conditions moving as they have in the past year, could potentially create a pipeline of HPMLs for loans with even moderate LLPAs.

Brokering Doesn't Sound so Bad Now...

Considering this complexity, the broker's limit of 3% compensation seems like an attractive option for avoiding issues in the regulatory scheme instead of the inverse.  Again, we ask the question:  Does mini-correspondent solve the broker's problem, or just add a set of additional problems.  For lenders, who carry the risk of liability, the questions are more dire than compensation.

When we overlay the pricing reality against the regulatory construct of QM, it seems clear that many more loans will achieve HPML status than originally conceived.  Will the market for these transactions shut down and add further inertia to a fledgling economic recovery?  Or will lenders accept the risk of no rebuttal presumption and continue make HPMLs.  During the 80's lenders originated 95% LTV loans using qualifying ratios of 25/33.  Certainly, if there is a place for HPMLs, they will have very restrictive guidelines.  


  1. Great explanation and helpful spreadsheet Thomas, great job!

  2. When will the lending industry unite and just say.....No More!

  3. Thomas,
    You got some good stuff here. Not sure whether you have any legal background or not, but as an attorney I can tell you that you should refrain from making statements pertaining when a court will or will not dismiss an action (see 4th sentence in The Objective: A Qualified Mortgage). There are many factors that will go into play before a judge takes a case of the jury's hands. While the lender may have used Appendix Q, it does not mean it used it correctly or whether the point and fee calculations were correct. A Safe Harbor is just a cushion. A good attorney would simply focus his plan of attack on establishing that the proposed QM had a defect. Once established, he/she can allege non-compliance with ATR. Note that I said allege. Litigation is long and windy road with many turns and twists. As attorney, we don't always know how things will turn out but how they most likely may. Thus we never make statements like "a court necessarily would dismiss the action." You should not either.

    1. That is an excellent point "anonymous". This is a point we have been trying to make - there is no limit to potential litigation loss, even under a putative "safe harbor". The phrase you reference is a paraphrase of the agency commentary, not my own invention. It is completely true that, once litigation starts, there is no telling what a judge will do.

    2. Fair enough sir. Just trying to help.

    3. I did change the phrase to "may" as you made such an excellent point on this matter. Do you really believe that the "Safe Harbor", as it applies in a litigation context, doesn't really exist?