How to Fix Non-agency Lending

Part I: Hedging and Execution

Part II: Underwriting/AUS

Part III: Loan delivery/Data Integrity

Part IV: The Role of TPR Firms

Part V: The Missing Link (AKA What the Industry Can Do Now)

In this five part series I’ll be focusing on how non-agency vs agency loans work. This is my attempt to break down the four key components that allow agency loans to scale, in spite of mortgage manufacturing plants NOT because of them. The purpose isn't to, necessarily, show how the sausage is made but more to inform the market about why its able to be made, at scale, in spite of all the shortcomings that exist (lack of lender expertise, technology, etc.). This guide is intended to shed light on the weak points of agency loans in order to, hopefully, educate investors, technology experts and the industry in an effort to come up with universal?solutions to reduce taxpayer backed mortgage lending while increasing stability and quality in both agency and non-agency loans. The fifth part will be recommendations for how to solve for the issues exposed in parts I-IV.

Comment, share, poke holes, improve. Ignore at our peril (March 2020, March 2022….)

Part I: Hedging and Execution

Most mortgage companies are run by executives with strong sales backgrounds not by those with deep understanding of the nuts and bolts of capital markets. The reasons a mortgage company can be successful, despite this lack of expertise, is due to the solid backbone provided by the agency MBS (mortgage backed securities) forward market (TBA market), the various tools provided by the GSEs (Cash window execution, for example) and a healthy secondary market for MSRs (mortgage servicing rights) where trades happen frequently enough to get solid color on where servicing values are. All of these pieces combine to allow either the mortgage company itself to create stability in future loan sale prices or allow larger mortgage companies (correspondent lenders/aggregators) to buy from the minnows using the future price stability provided by the aforementioned tools. Let’s take an example. If an organization understands their fixed and variable costs, they can reasonably calculate their required margin (profits) to build into a loan. The amount of margin “buffer” (additional spread needed to make sure the target margin is hit) needed to hit that target margin is a function of the certainty of the future pricing the lender will receive and the certainty the loan will be purchased, if funded. There are multiple factors that influence this margin number including: competition, efficiency, scale, cash balances, net interest income (or expense) for loans dwelling on warehouse lines, etc. For the purposes of this article, I want to simply things and focus on the basic mechanics so we can expose the major differences between agency and non-agency execution.

For agency loans, there are four major factors to determine a rate sheet price (a price offered to a borrower, broker or another lender for a lock which may eventually become a loan).

1.) Forward execution price

2.) MSR value

3.) Margin

4.) Lock Term

In order to put a rate (or price) out to a customer, the math is as simple as (1+2-3) adjusted for the time value of 4 (i.e. diff lock terms get different pricing). Additionally, the forward execution price has several public, easily accessible, fungible pricing discovery methods through 3rd party pricing providers (Optimal Blue etc.) as well as TBA forward pricing, cash window execution (pricing provided directly by Fannie and Freddie updated by the minute) to name a few. The fact that virtually all GSE loans can be sold to virtually all agency loan investors (aggregators) and the pricing is readily available, regardless of how turbulent the market gets, allows for stability in execution and thus stability in the product itself. The pricing certainty, if hedged appropriately, allows for tighter margins (less buffer needed due to pricing uncertainty) and a consistent offering (the market never disappears).The TBA forward market is similar to the oil futures market in the sense that it allows someone today to lock in a price in the future for a particular range of note rates (this range can be solved for by using a combination of MSR values, TBA prices, guarantee fees [insurance fee paid to Fannie/Freddie to guarantee the loan] and buy-up buy-down grids provided by the GSEs) to be delivered at a specific date in the future. This forward market has strict delivery eligibility rules and extreme amounts of transparency which translates into pricing certainty and liquidity (because investors know with reasonable certainty what to expect). Additionally, the market has “specified pools” which can offer even more certainty (at a price). Specified pools are similar to TBAs except they are stipped for certain characteristics an investor may find appealing for various reasons (Investor only loan pools, maximum loan size, etc.). Since the forward market for MBS is so strong and liquid, and since there is no credit risk involved in these products (they are all government guaranteed), the only “spread” risks agency lenders have to deal with are MBS vs Treasury spreads (investor appetite for MBS) and Primary Secondary spreads (an indicator of competition in the lending market). Both of those spread risks can easily be traded through a combination of interest rate instruments (Treasury notes, futures, options, swaps) and TBAs. Lastly, since there are various layers to the agency market (big fish eating smaller fish loans), the GSEs buying loans directly from lenders via the cash window (taking the complexity of MBS out of the lenders hands and executing on their behalf) and selling loans with or without the servicing, it allows small, unsophisticated lenders to compete with the bigger more sophisticated shops.

In the non-agency market, virtually none of those tools exist. Additionally, the largest players in the non-agency market tend to be the least sophisticated lenders (not in all instances and this isn’t meant to be disrespectful). The reality is, the margins in agency loans are, generally, not very big and thus require a lot of capital (because cash margin is generally negative, though not always, or in other words even if there is a financial gain reported it is cash intensive to sell the loan servicing retained). Since the margins are low, in tougher markets, the least efficient lenders are priced out first. Additionally, large lenders generally avoid this market for various factors such as: liability for non-agency loans is not as well understood, the expertise required for non-agency is hard to find, the custom changes required for an LOS are intensive, the market is not stable enough for lenders to originate at scale with minimal risk, etc. Thus when the agency market gets tough, lenders that can’t effectively compete in the agency product begin to migrate to other products (FHA/VA/USDA loans, Non-agency, etc.). This leads to a market where the least efficient and least sophisticated players participate on the lending side. Keep in mind, the non-agency market also has the least pricing transparency because it lacks liquidity. Why? There really is no forward market for pricing (no TBA market), the GSEs will not buy these loans and do securitization on small lenders behalf, guidelines are different among virtually every lender, product features are different, and investors enter/exit the market at unpredictable intervals. Additionally, these product have inherent credit risk. If the borrower defaults Uncle Sam isn’t going to pay to make you whole (like it will on agency/Govt loans). Lastly, there are various middlemen in the process that take a “piece” of the transaction and the execution is different depending on who the end buyer is.

Once a loan is funded, an agency loan can be off a lender’s warehouse line in under five days. This means from the time the loan disburses, until the time the warehouse line is paid off is less than a full business week. In non-agency, this process can take 30+ days depending on the execution (who the loans are sold to and the type of execution). The execution types can be best-efforts or mandatory (single loan or bulk). Where best-efforts means the buyer of the potential loan takes the hedge risk (and timing risk) vs mandatory where the seller takes the hedge/timing risk. In agency loans, this spread is generally 25 bps for single loan mandatory (one loan at a time) but can be in the 50+ bps range for bulk mandatory (selling many loans at once).?Additionally, on agency loans, when a pool of loans is sold the loan level execution price is explicitly known and the pool composition does not materially affect the execution price. On non-agency loans, the pool characteristics are extremely important and have a very material impact on pricing. Because of the lack of certainty around future pricing in non-agency vs agency and the importance of the pool characteristics being known at the time a pool is bid, the non-agency market can experience mandatory bulk pricing sometimes POINTS higher than best-efforts pricing. This carrot is many times too enticing to pass up in good markets and impossible for some to pass up in tough markets (to stay competitive). Thus, ironically, the least efficient lenders which, generally, lack the capital markets acumen of their larger counterparts, end up taking on the most risk and they happen to, generally, also be the least well capitalized (having enough money to withstand shocks/unexpected market circumstances). In order to hedge non-agency loans, the market turns to either forward trades (attempting to specify a forward pool of loans to be sold with some characteristics agreed to that will result in price adjustments if missed), or to some “credit” instruments to hedge these products (equity vol, swaps, etc.). The average non-qm lender has very little expertise or ability to trade in the latter, so they engage in the former (or they just aggregate loans and hope for the best). The problem with forward trades, which we saw first-hand in March 2020, is that even the most credit-worthy counterparties can/will fail to fulfill those obligations with little to no recourse for the lender (and if the lender is racing against the clock because warehouse lenders are breathing down their neck even if they have a case, the legal system takes too long to help). We saw many big non-qm takeouts abandon their obligations for one reason or another in 2020, leaving lenders and warehouse banks with large losses.

The buyers of non-qm loans are much more sophisticated in terms of hedging ability, however, just like with prime jumbo, non-agency hedges can not perform as hoped/intended. Unlike the TBA market which guarantees a lender a price in the future, the non-agency credit hedges provide no guarantee. While they may perform directionally, as intended, they generally do not perform as needed to guarantee pricing. On top of that risk, non-qm issuers (those that do securitizations) rely on other buyers to be interested in the bonds they create (which can take weeks or months to finalize) at consistent spreads (prices). This delay to create these bonds injects additional uncertainty as the more time passes, the higher the chance something happens in the market before the bonds can be sold. This exposes securitization to getting “hung up” and these issuers are using warehouse lines, just like lenders do (they are different in important ways but outside the scope of this article). This can cause margin calls on the non-qm loan buyers, the inability of the market to buy loans and thus downstream impact lenders (think March 2020). This can cause the market to literally disappear “overnight”. On the other hand, there are real money (unleveraged) buyers in non-agency. These buyers buy loans for their portfolio (balance sheet) and are thus not exposed, directly, to the issues the securitization market has. However, since a single balance sheet is buying loans, there is a limit to how much they can buy and when they can buy. Another important difference is that since these buyers are holding the loans, they tend to not want to buy certain loan characteristics that can easily be “tucked away” in non-qm securitizations. So if the market struggles with securitization, many loans are not eligible for real money buyers regardless of rate. Lastly, these types of buyers also see what happens when securitization markets get stuck and respond by, generally, worsening their prices to take advantage of distressed sellers or to protect against increasing volatility and spread widening.

In order to make non-agency more stable and liquid, better hedges need to be introduced or better execution options need to be developed otherwise, market turbulence will continue to kill off lenders and loan buyers at various, unpredictable intervals. We will propose solutions for this in Part V of this series.

Part II: Underwriting/AUS

When you hear the word “underwriting” or “underwriter” do you think of someone reviewing a loan file from start to finish and making a credit decision on their own? If you do, then you won’t understand why its so difficult to transition agency underwriters to non-agency. Let’s dive in.

What is an AUS? An AUS is an “automated underwriting system”. The name itself implies that the file is being underwritten by a machine and not a person. To simplify this conversation, I am going to explicitly discuss Fannie Mae’s AUS which is called Desktop Underwriter (DU) and I am going to leave out a lot of the minutiae in an effort to keep this essay short, but explain the main issue with non-agency as it pertains to underwriting.

In the world of agency loans, by the time a file reaches a human underwriter the file has essentially already been “underwritten” by DU. The “underwriter” in an agency loan shop is really more of a “processor” than a true underwriter (vs a manually underwritten file). Again, if you are an agency underwriter, this guide is not an attempt to undermine the importance of your role nor is it to downplay the work involved, it is simply to explain how different agency vs non-agency underwriting is and to get this point across to the reader so they can understand why a solution is necessary.

In order to get a file submitted to an underwriter, a loan application must be filled out (URLA/1003) by a borrower. This data is placed into a standardized dataset called a 3.4 file and this information is fed into the AUS, in this case DU along with a credit report. When this is done, a printout will be created by DU which will essentially tell the lender whether the loan is approved or not. It will capture relevant file information such as loan characteristics and property details. A sample snapshot is below:

Additionally, based on various datapoints the DU may also grant an appraisal waiver meaning the file doesn’t need an appraisal ordered and thus no one will even need to review the collateral on the file. The DU will also run loan eligibility and give the lender a message discussing potential items that need to be looked into along with rules for how to handle such items. For example, here is a sample of one such message:

In other sections, the DU will provide additional guideline information including when a file must close by, how to handle HELOCs, it will provide feedback on the credit report (which is read automatically by the DU), it will tell the lender which accounts were not considered, which debts require evidence of payoffs, it will review employment details and let the lender know what else is required to prove income (paystub, w-2, VOE, etc.) along with requirements such as “be dated no earlier than 30 days prior to the initial application date”, etc.

At this point, it should be clear that the agency AUS system is substantially more robust than the non-agency “AUS” 3rd party vendors have tried to sell the non-qm world. Fannie Mae has over 2,000 pages of guidelines, which an underwriter must be familiar with. However, the DU provides substantial feedback and, essentially, tells the lender that a file is approved as long as documentation is collected that “proves” the details of the file submitted in the 3.4 file were accurate. So essentially, and this is oversimplifying a bit, the DU underwrites a file, underwrites the collateral (home value), determines eligibility, does a black box credit qualifying to determine approval, and then calculates pertinent information such as LTV, DTI, etc. and displays that, along with any additional messages, to the underwriter. From that point, if the correct documents are collected, reviewed and shown to “prove” the data in the file, then the loan will be approved by the underwriter.

Essentially, 90%+ of the heavy lifting is done by the AUS system in agency and it provides significant guardrails for a lender. This makes it virtually impossible for a file to be “accidentally approved” by the underwriter if the AUS does not provide an “approval”. The collateral doesn’t need to be reviewed in many cases (if a PIW is granted), the credit report is already read and reviewed by the AUS, the eligibility is already determined based on the Fannie Mae guidelines, and the file is partially conditioned by the AUS (this is a fancy way of saying your loan is approved subject to conditions, such as provide a W-2 to prove you make what you said you did on the loan application).

In Non-agency, there are AUS systems BUT they are not the same. Many 3rd party vendors will claim it is, but it is not. The vast majority of these systems are simply eligibility engines, so they can read things like the LTV is too high relative to the credit score, etc. Most don’t even integrate credit reports into them, so they can’t read liabilities, bankruptcies, etc. They also don’t automatically condition files in almost all cases and not a single one conditions a file fully. On top of that, there are so many different types of products with differing income calculations, and differing eligibility depending on what “matrix” these files fall under. Additionally, every non-qm guideline set has some form of the following language “where these guidelines are silent, fall back on Fannie Mae’s guidelines”.

For non-agency underwriters, they must be familiar with at least one set of non-qm guidelines (which means somewhere between 50-250+ pages) and also must, by definition, be familiar with Fannie’s 2,000+ pages of guidelines and then must realize where to apply the Fannie guides when the non-agency guides are “silent”. Additionally, many lenders offer different classifications of non-agency. To make things simple, I’m going to use Deephaven’s original concept of: Expanded Prime, Non-Prime & DSCR (some have even more). Each of these classifications has separate eligibility requirements, different pricing (which affects Debt-to-income calculations), and DIFFERENT guidelines. Also, keep in mind that a lender may be selling loans to Deephaven and other lenders and, potentially, the underwriter will need to review a file against another aggregators guidelines (Verus for example).

At this point, it should be clear how confusing and how many layers there are to this. Not to mention, that before a file even gets to the underwriter it must be submitted to a particular program (in other words someone has to determine if the file is Expanded Prime or Non-Prime, for example). In order to determine what program to select, a salesperson must be well trained to help a borrower or loan officer find the program with the best pricing. Keep in mind that one lender’s Expanded Prime guideline/eligibility might be another lender’s Non-Prime program (or a completely different name altogether).

Let’s pause for a second and get our bearings. In agency loans, a file goes through a loan application, credit is pulled and the file is run through the agency’s AUS (DU in our example). From there the file is essentially approved and the lender must now just prove (via documentation) that what the borrower stated on their application was accurate. On non-agency loans, a file goes through a loan application, credit is pulled and someone must make a determination of what program they think the borrower fits best. The non-agency AUS (if anyone is even using it), may/may not provide a little bit of help here, but it does not do anything remotely close to what DU is doing. Additionally, a smart lender is going to first run a file through the DU to see if its eligible for Fannie (before trying to fit it into a non-agency product). Many lenders do not do a good job of this, especially on DSCR loans (because they don’t collect income and other information required to make sure the file doesn’t fit agency guides). Then from here, an underwriter must be familiar with what guidelines to use (expanded prime, non-prime or DSCR in our example) and then know when to apply the 2,000+ pages of Fannie Mae’s guidelines, as applicable.

If your head is spinning it means you are keeping up, and we haven’t even talked about income. One of the major factors in proving ability-to-repay (on non-exempt loans) is income. In non-agency, income is calculated differently by many different lenders. One popular way of documenting income is via “bank statements”. This is a fancy way of saying deposits. How bank statements are analyzed (what can be included/excluded as income) and what type of expense factor is applied vary across job types AND lenders. Thus not even income calculations are consistent across different lenders/different guidelines. Also, keep in mind that in order to complete a loan application, income must be filled in at the time of the application. So this analysis of bank statements must be done either by a non-underwriter or by a desk staffed with underwriters who do “bank statement analysis” upfront for their customers, then later when an underwriter re-calculates income there’s a chance the original income calculation doesn’t match this one, leading to upset customers, sales people, etc.

In order to avoid this essay being over 100 pages long, I have left out a significant amount of detail. However, it should be clear that unlike agency loans where the guidelines and eligibility are exactly the same, regardless of the lender (overlays aside); non-agency loans do not have this “luxury”. Even if you have five years of underwriting experience at a certain non-qm lender, if you move to another lender you will have to learn all new guides or at least various nuances. This means mistakes, errors and, at the very least, a steep learning curve. It also means that someone with agency experience is not going to be able to “plug” into a non-agency lender as they are used to the safety and guardrails that the DU (AUS) provides. Since the market has been essentially conditioned to let an AUS underwrite a file, the manual underwriting skillsets of the mortgage industry have atrophied significantly.

As you can see, there are dramatic differences in agency vs non-agency loan underwriting. Agency loans are essentially underwritten by an AUS while non-agency must be manually underwritten to multiple sets of guidelines and eligibility criteria. This seriously complicates the process and makes scale very difficult to achieve (since underwriters with specific guidelines knowledge is scarce). In Part V, I will be discussing solutions for this element and make recommendations to improve the quality and scalability of non-agency underwriting.

Part III. Loan delivery/Data Integrity

Note: This will be the most ignored essay (of the five I write) and I would argue it is the most important of all if you want to fix both agency and non-agency operations.

In Part II we discussed underwriting and AUS, in order to fully comprehend how important the AUS is, please re-read the AUS component of this essay; it will help you grasp the magic.

When you think about “shipping” or “loan delivery” functions inside a large agency lender, do you think of someone creating a digital loan file (with all the documents), sending a data tape to an investor and the investor doing due-diligence on the file (reviewing the data, reviewing the document file, etc.) and then, only once everything is aligned, they purchase the loan? If you do, then you won’t understand why it is so difficult to transition agency lenders to non-agency. I’ll explain why.

First, we need to revisit the AUS (automated underwriting system). In part II, I discussed:

“In agency loans, a file goes through a loan application, credit is pulled and the file is run through the agency’s AUS (DU in our example). From there the file is essentially approved and the lender must now just prove (via documentation) that what the borrower stated on their application was accurate.?Essentially, 90%+ of the heavy lifting is done by the AUS system in agency and it provides significant guardrails for a lender. This makes it virtually impossible for a file to be ‘accidentally approved’ by the underwriter if the AUS does not provide an ‘approval’. The collateral doesn’t need to be reviewed in many cases (if a PIW is granted), the credit report is already read and reviewed by the AUS, the eligibility is already determined based on the Fannie Mae guidelines, and the file is partially conditioned by the AUS (this is a fancy way of saying your loan is approved subject to conditions, such as provide a W-2 to prove you make what you said you did on the loan application).”

The first step in making the above happen requires sending a uniform dataset to the GSEs called the MISMO standard 3.4 file. Once this file is read by the AUS, it is stored along with additional expected future details based on AUS responses. Secret tip: its this step that “allows” the GSE to have an “expected” result in terms of data. What do I mean? Based on this AUS approval, without an underwriter ever having looked at the loan, Fannie already knows what to expect when post-closing uploads the file data (ULDD dataset) after the loan is funded.

To say this another way, to get the GSE to “accept” the loan file (purchase the loan) the lender needs to, after a loan is funded, upload a standardized data tape to the GSE. That data (ULDD) will be compared to the expected data (based on the AUS run) and will either be rejected or accepted. You read that right. There is no file review, there is no making sure loan documents match the data (or vice versa), there is no human being reviewing anything to make the purchase decision. If a loan is sold to the GSE (either via cash-window or via MBS), the only factor required is the data matches. That’s it. Period.

Now here is the ultimate issue with mortgage lending, if it could be simply boiled down to one thing, operations works off documents (paper or digital) while Sales, Capital Markets, Post-Closing, Investors, and the GSE’s themselves work off Data. Despite all the money thrown at “fintech” and “mortgage tech” there simply has been no solid solution to marrying data to documents (or vice versa). In fact, there has been little implementation of (already existing tech) marrying expected data results (from the GSE tools) to the actual data (inside the Lender’s loan origination system). How do I know that? The GSEs share information based on loan delivery data quality relative to quartiles. Did you know that the top 25% of lenders that ?deliver loans to one of the GSEs have their file rejected one out of every two times (at first delivery attempt)? Let me rephrase. Despite the fact that the tools exist to mimic loan delivery throughout the loan origination process, despite 30+ days spent inside a lender gathering documents, inputting information, going through random quality control audits, etc. after the file is already funded, after the borrower has already closed on the loan and after the operations function has long forgot about the file, when the lender extracts the information from their LOS and inputs it into the GSE loan delivery system, a machine (not a person, not someone thoroughly reviewing a loan file, not a chief credit officer digging through a file with a fine tooth comb) simply says “hey the information you just uploaded, doesn’t match what we approved”. And this happens 50% of the time, at the “best” companies in lending. 50%!

Now at this point, when the loan delivery team gets this error message (which takes place in the GSEs website not their LOS) they have two choices:

1. Take this information back to operations, have operations update the LOS and then reimport the information to the GSE

2. Type in the expected information into the loan delivery system directly.

If they do number 1, it basically delays the loan delivery process (think Lucy on the chocolate factory line trying to eat all the chocolates) but the data in their LOS will match the loan delivery system. If they do number 2, it will speed up the process and get the loan purchased faster, it won’t “bug” the operations people upstream BUT the data in the LOS won’t match the loan delivery system. Which one do you think they do 90% of the time? That’s right, number 2.

So to recap, if a loan is “sold” directly to the GSE, loan files are not reviewed (at time of delivery), data is compared to expected data and the purchase moves forward or doesn’t based on that information (data comparison result), this attempt to upload data is rejected 50%+ of the time, the data is generally updated in the GSE system and not the LOS, and thus at the end the loan is purchased with data discrepancies between the LOS and the GSE and these are virtually never reconciled.

Because of this, I always recommend every lender to sell at least some portion of their loans to an aggregator (a middle man/correspondent lender). Why? Because those lenders will actually review loan files and provide feedback. This helps the lender keep a good pulse on their file delivery quality, but many lenders don’t do this because it “delays” loan purchase timelines (well sure, especially if the data is wrong and/or the documents don’t match the data, or the lender’s stacking order is a mess [most are]).

So now that we’ve discussed how agency loan delivery works, let’s juxtapose that against non-agency. For non-agency, as discussed in Part II, there is little to no use for the AUS (its not going to capture this data and compare to loan delivery like Fannie/Freddie which, ironically, is one of the major points of having an AUS and, arguably, the most important from a scale perspective), the underwriter will manually underwrite the file, the operations team will still do their documents herding and reviewing and capital markets, investors and, to some extent, sales will still do their jobs using data. However, when it comes time to get a loan purchased, a lender will upload a non-standardized data tape (with additional fields required for non-agency loans that don’t exist for agency), they will upload a data file and all this will be reviewed by a TPR firm (we will discuss this in Part IV) and the firm that purchased the loan.

At this point, I could write several more pages on the topic, but what I want the reader to takeaway from this is that the entire process of delivering a loan to a non-agency investor exposes the parts of the process most lenders are the worst at AND it doesn’t include the feedback (data comparison) that a lender is used to getting from the GSEs AND it should give the buyer of the loan a bit of heartburn when they realize that if a lender can’t get the data right on an agency loan 1 out of 2 times, how will they get it right on a non-agency loan when all the tools an agency loan has (AUS, ULDD, Data Delivery Feedback, etc.) don’t exist? In other words, in agency loans if the lender says the data is X and an auditor says the data is Y the gold standard is the AUS/ULDD which may say the data needs to be Z. This standard allows for excellent audit results. However, on non-agency loans if a lender says the data is X and the auditor says the data is Y, how does anyone know if the data should have been Z?

In order to fully appreciate the brilliance of the AUS, you must start at the end of the loan origination process. Unfortunately, most of the investment in tech in mortgage is on the front end. If they started at the back, they’d solve more problems, increase tech ROI, reduce errors, cycle times AND create better solutions for the front end. If non-agency wants to scale, it needs to understand how the AUS and ULDD work hand-in-glove (and how it can replace/reduce reliance on TPR), and this will lead to an obvious conclusion which I will discuss in Part V as one of the solutions to fixing the non-agency market.

Pro-Tip (if you work at a lender you should advocate for this, no matter what your position in the firm, if the below is implemented it will improve your life):

Having been a capital markets professional in the agency mortgage space at multiple lenders, I cannot emphasize enough how important it is for lenders to grasp what I’m about to say. Despite numerous tools available (and integrated into most 3rd party LOS systems), free, readily available documentation online, and very responsive reps at the GSEs almost no lender does this right. If there was a single item that would materially improve cycle times, post closing backlogs, QC, cost per loan, etc. it would be doing what I’m about to say properly: understand and implement the “early delivery” toolset from Fannie Mae.

Here are required reading on the subject:

How the original AUS data gets to Fannie:

https://singlefamily.fanniemae.com/media/7571/display

How loans are delivered to Fannie (data not documents)

https://singlefamily.fanniemae.com/media/7571/display

How to make sure your data is correct (at any point in the origination process, not just after funding):

https://singlefamily.fanniemae.com/applications-technology/earlycheck

If you build the above into your origination process, you will be shocked at the improvements you see in customer experience, operational efficiency, and post-closing/secondary will thank you for saving them 50% of their time. Everywhere I’ve implemented this, we have achieved a 99% loan delivery success rate on attempt one. This project takes less than three months to complete and can be done in as little as 30 days, depending on how well the LOS is setup.

Part V: The Missing Link

Its been several months since we discussed Part III of this series. The reason for the delay was that we had to do a deep dive on compliance rules in order to set the stage for where Mortgage Lending may go in the future. Since the laws dictate what's possible, its important to understand both what they were and, since there was a major change to QM rules on 10/1/2022, what they are now. Under QM 1.0 rules there was a clear(er) line between what was QM vs what was NonQM. Add in the wrinkle that under the "QM Patch", any GSE loan was QM even if it didn't follow the rules and you'll understand why many people believed QM meant agency and, therefore, NonQM meant non-agency.

We've spent the last several months debunking the agency = QM theory and walked through the new QM 2.0 rules to show both why that was true and, why calling non-agency loans "nonQM" has been devastating to private lenders. If you fail to comprehend that QM 2.0 rules make any loan possible for QM, you'll fail to innovate and adapt.

The main rule we've been discussing is Dodd-Frank. Under Dodd-Frank there are two major pieces that almost work hand-in-glove: Regulation Z (QM rules) and Credit Risk Retention (QRM rules). Below is visual aide to assist in comprehension:

Hopefully that flow chart is easy to follow but I want to draw your attention specifically to the QM 1.0 vs 2.0 rules:

Under the QM 1.0 rules (officially ended 10/1/2022), loan programs had to follow specific underwriting criteria benchmarked to FHA underwriting standards (Appendix Q). Since GSE loans would NOT be QM under 1.0, they were exempted by the rules via the "QM Patch". This is why Government and GSE loans came to dominate 80-90% of all residential loans. To put this in perspective, during 2020, one single top 10 non-bank lender originated more loans in a quarter (some in only a month) than the entire Non-Agency ("NonQM") market originated in its best year.

After the mandatory compliance date for QM 2.0 (10/1/2022), three major things happened. First, Appendix Q was removed from the QM test. This means any set of guidelines can be QM. To avoid confusion, several sets of guidelines were included in the QM test that are "safe harbor" guides. The punchline to the first change is that all prior "NonQM" guidelines under 1.0 can technically be QM guidelines under 2.0. So if you have a covered loan (not exempt/business purpose), bank statement income, asset qualification, full doc + asset depletion, 1099, etc. ALL can technically be QM. Do you see why this is so hard to understand if everyone calls their products "NonQM"?

The second major change is the 43% DTI cap was removed and replaced by a Price Based Threshold (APOR test). This is designed to focus more on the interest rate than the percentage of income being used to qualify a borrower. While this is a much better approach, in my opinion, it means that if your rates are too "high" your loan will be NonQM.

The third change, which the GSEs adopted over a year ago, was that the QM Patch was retired and instead the GSE guidelines were added to the QM Safe Harbor set (once again making them all QM, unless exempt).

The key summary here is that product features and/or interest rates will be the main factor in why/if a loan is NonQM. So if you add interest only, go over a 30 year term, charge too many points or offer too "high" of an interest rate your loan is NonQM; your loan is not NonQM simply because it didn't use FHA guidelines anymore.

Now moving onto the right side of the flowchart, you'll see that nothing technically changed on the Credit Risk Retention Rule BUT since the QRM rule is linked to the QM rule, something material DID change. Under QM 1.0 rules, the GSE eligible loans were NonQM or exempt, however if the GSEs securitized the loan it was automatically QRM & QM (due to the QM Patch) yet if a private issuer securitized an exempt loan it was not QRM (even if it was agency eligible). Non QRM loans require 5% risk retention, GSE securitizations require 0%. Going forward, prior "Non QRM" loans like bank statements can be QRM under 2.0 rules and both products (agency eligible QM and agency ineligible QRM) can be securitized in the private markets with 0% risk retention. This allows originators to securitize their production with significantly less capital required (in theory).

If any of the above is still confusing, I would highly recommend you look through the last several months of articles and posts on my LinkedIn homepage. I have dedicated many pieces to these topics and have gone into great detail to help you understand where we are and how we got here. Assuming you are tracking, we will now go into "What next?".

What Next?

If my crystal ball is working (probably isn't), here's what I expect to happen.

Someone is going to start doing bank statement loans and full doc non-agency loans as QM. This is either going to be the GSEs or the private sector. The private sector is generally not incentivized to do things that lower interest rates or improve liquidity (as it hurts returns), but it may pull in new investors and capital that is more comfortable with QM loans and wants to arbitrage between agency rates and non-agency. If I had to bet, I would say the GSEs win this one. Why?

In Parts I-IV of this series we discussed the major "problems" with Non-agency: Hedging, a lack of underwriting standards, cost to produce, all the diligence needed, etc. The fastest way to solve all those is simply have the GSEs update their guidelines to include these types of qualification and voila! Safe Harbor QM. If this happens, small NonQM lenders will be left fighting over a very small population of loans (only the ones that fail the 3% points and fees test, can't qualify to the agency alt guides, have loan terms over 30 years, interest only, etc.). The problem here is that the most likely outcome, if GSEs get involved, is that the rates GSE will provide will be so much better than the NonQM rates that taking a 40 year term or an interest only (or both) likely won't reduce the payment vs the GSE option, so what's the point?

The CFPB is talking about how to apply more technology to lending and the FHFA is worried about access to credit. Since both of those topics are getting a lot of attention, its easy to see how a case could be made either via "affordable" initiatives or simply adapting to the new workforce (self-employed, gig workers, etc.) could allow the GSEs to expand their guides to incorporate those types of borrowers. Additionally, since the "nonqm" lenders have simply refused to harness technology in any meaningful way (see sections III and IV), the ability for a lender to use the AUS, automated conditioning, and loan delivery functions would be an easy sell on the Tech + Access to Credit front. On top of this, Freddie recently announced how their AIM system will read digital asset statements (bank statements) to look at cash-flow and make better/more informed AUS decisions. They are using Plaid, Finicity, etc. to do this work so its easy to see how they are not far from simply doing bank statement income using this tech (all this has been available for the private lenders/issuers to take advantage of for many years yet most still have a "bank statement" desk staffed with humans to do income calculations).

The alternative would be for the Non-Agency universe to get together and follow some of the guidance I put out surrounding the Non-Agency Alliance concept. A self-policing/oversight board, outside of the MBA, which works to get a safe harbor non-agency guideline set incorporated into QM 2.0 and then utilize the AUS/ULDD concepts under a proprietary AUS (I've built one in the past or get one of the GSEs to do it for them, like they do for Ginnie loans). Once these pieces are in place, we simply have to realize that the GSEs are insurance companies. So an insurance company would provide a "wrap", re-insurance would step into the space as well, and alternative execution options could be developed from there. The added benefit here is that the ALM profile of this product fits with insurance money. For more details or ideas check out the ACIS side of the CRT playbook. For those of you selling whole loans to insurance companies, you'll want to root for this as well. If secondary mark-to-market pricing goes up (or rates go down), you'll be able to use this to negotiate better whole loan pricing.

In either case, the other piece of the QM test that needs to be addressed is the "Points & Fees" (P&F) test. I've written an entire article on why we are seeing this test keep borrowers from getting loans. The summary is that the P&F test was essentially built for a "high premium" environment. Since we've never had a mortgage market with large G-Fees, substantial LLPAs and little/no premium, something is going to have to change. Either the LLPAs will have to be reduced, the G-fees cut, or the Fed/GSEs/Some other quasi Government sponsor is going to have to be allowed to buy agency RMBS. We simply can't expect things to "go back to how they used to be pre-2008" when pre-2008 there were very few LLPAs, much lower G-fees, the GSEs bought MBS AND there was no P&F test (no QM test). If we expect to be able to harness the guideline side of QM (removing Appendix Q), the P&F will need to be addressed to keep loans from being NonQM simply because borrowers had to pay points up front, instead of financed for 30 years.

The last piece I'll get into is where the other asterisk is on my flow chart. The 5% risk retention. Note that while the QM side of GSE vs Private is now "aligned" with 0% risk retention, the exempt side is not. The GSEs can still securitize exempt loans with 0% while the private sector has to retain 5%. I dedicated an entire article on this subject earlier this month, so please take a look at how I used the regulation itself to make the case for changing this. The punchline is that the exempt underwriting quality would go up if the rule was changed and this is one of the carve-outs the rule provides for making such an exception.

If the exempt side changes to allow certain subsets of exempt loans to have 0% risk retention, many potential DSCR loans will go the 0% route (likely under more full doc style) and thus fewer DSCR loans will be left for the rest of the universe.

In summary, my hope is not that the GSEs swallow up more of the lending market. In fact for the last several years I have been pushing all of these changes in the non-agency market so that they would be ahead of the curve. If I am right, its highly possible the small, dedicated NonQM lenders won't make it. There simply won't be enough true "nonqm" production to survive and the loans that get done under QM 2.0 (bank statements, etc.) will simply come down to who has the most efficient and lowest cost to produce. This would be a small windfall for the big agency lenders (if the GSEs get involved) but would put tremendous stress on smaller non-agency shops left to fight over fewer and fewer loans. It is my hope that the industry comes together to collaborate on guides, tools, compliance rules and oversight to use technology to beat the Government at its own game and put some healthy competition in the market. Additionally, it would be great to get some insurance/reinsurance competition against the rating agency PLS models as well. There are alot of arbitrage opportunities between g-fees, LLPAs, and rating agency draconian credit enhancement/model assumptions.