We recently had a Home Equity closed-end loan that was approved, the borrower signed the loan documents, then withdrew because he changed his mind. The loan officer marked the loan as withdrawn within our loan system for the final credit (Reg. B) decision. The loan was reported on the HMDA LAR as withdrawn based on this information.

The question the loan department is trying to resolve or seek best practices on is the gathering and reporting of HMDA data and the difference in definition for Reg B “loan decisioning” purposes versus HMDA reporting purposes for approved loans not completed by borrowers. Our loan syatem does not code for HMDA and all LAr data is enetered manually based on the loan officers worksheet.

I know this is confusing, but we were wondering what other banks do.

Based on what you have stated this should have been reported as Approved but not accepted on your HMDA LAR. Once the bank approves the loan, then withdrawn can no longer be used.

If you presented a counteroffer to the customer's original request and they did not accept it then it would be reported as denied.

Added: What loan documents are you referring to that they signed?

You should have very few HMDA files marked "withdrawn." Bankers Compliance Consulting published a very good article on this subject....can't share because I think you have to have a subscription (paid for) to access the articles....though there ARE a multitude of free stuff on the website. But anyway, the article argues that if the bank issues early disclosures, the loan is in a temporary "conditional" state of approval. Because to have provided "meaningful" disclosures, the bank will have had to analyze all the facts as presented (though unverified) and at that point, have no reason to deny...thus a "conditional" approval. If withdrawn at that point, the application would be considered "approved but not accepted." If the bank has accessed the applicant's credit report and it's bad, but before the bank can issue an adverse action notice, the customer withdraws, it's still a denial and an adverse action must be provided. If the customer withdraws before the bank has a chance to analyze anything (1st three days)....only then would the application be coded as "withdrawn" for HMDA.

This is just my opinion. But in any event, if you utilize the above method as your procedure, the analyzing process is way narrowed down so that similar applications will be treated and reported in the same manner.

If the borrower signed loan documents, wouldn't this be a rescinded loan?

From Appendix D of GIR:

2.
Action taken—rescinded transactions. If a borrower rescinds a transaction after closing, the institution may report the transaction either as an origination or as an application that was approved but not accepted

But anyway, the article argues that if the bank issues early disclosures, the loan is in a temporary "conditional" state of approval. Because to have provided "meaningful" disclosures, the bank will have had to analyze all the facts as presented (though unverified) and at that point, have no reason to deny...thus a "conditional" approval. If withdrawn at that point, the application would be considered "approved but not accepted."

Just be aware that this is a minority opinion in the compliance world and many HMDA and other compliance experts do not agree with BCC on this stance.

So, how would you define it? At my prior bank, a loan wasn't a go until the FINAL secondary market underwriter approved it....which was usually only a week or so prior to closing. Until then, it wasn't DEFINITELY approved. So that bank defined approved but not accepted as only occurring subsequent to that event. And since the possibility of someone withdrawing at that point is slim next to none, the bank had zero approved but not accepted and scores of withdrawn applications. Both the internal auditors and the firm that assessed fair lending frowned on the numbers.

At my new bank, we haven't adopted BCC's method but are thinking about it. When you audit, what do you look for to confirm withdrawal or approved but not accepted?

For a withdrawal, I look for a documented express withdrawal of the application by the applicant prior to loan approval.

For approved but not accepted, I follow the guidance in the HMDA FAQs, which includes an approval subject to customary loan-commitment or loan-closing conditions.

Customary loan-commitment or loan-closing conditions include clear-title requirements, acceptable property survey, acceptable title insurance binder, clear termite inspection, and, where the applicant plans to use the proceeds from the sale of one home to purchase another, a settlement statement showing adequate proceeds from the sale. See comments 2(b)-3 and 4(a)(8)-4. An applicant's failure to meet one of those conditions, or an analogous condition, causes the application to be coded "approved but not accepted." Customary loan-commitment and loan-closing conditions do not include (1) conditions that constitute a counter-offer, such as a demand for a higher down-payment; (2) underwriting conditions concerning the borrower's creditworthiness, including satisfactory debt-to-income and loan-to-value ratios; or (3) verification or confirmation, in whatever form the lender ordinarily requires, that the borrower meets underwriting conditions concerning borrower creditworthiness.

I'm not going to start this argument again, there is a very long series of posts regarding this issue. This clearly states that an applicant cannot be "approved but not accepted" if the lender still requires: "verification or confirmation, in whatever form the lender ordinarily requires, that the borrower meets underwriting conditions concerning borrower creditworthiness."

I would not adopt the BCC approach; if you really are considering it, have a long heart to heart with legal counsel about what issuing disclosures means vs "approving" an application.

We recently had a Home Equity closed-end loan that was approved, the borrower signed the loan documents, then withdrew because he changed his mind.

So, how would you define it?

I have to assume that since this was a HE loan the borrower exercised their ROR. Hopefully the bank properly treated the "withdrawal" as a rescission since the loan documents were executed.

The FI has the option to report a rescinded loan either as an approved but not accepted or as an origination. Withdrawn is not an option.

Thank you both for your valuable input! This will help me outline definitions for my bank. Based on the above, an approved but not accepted applications will be few and far between, but I guess it is what it is. File documentation will support the labels.

I will search for the other posts.

GIR - Page D-8:

2. Action taken—rescinded transactions. If a borrower rescinds a transaction after closing, the institution may report the transaction either as an origination or as an application that was approved but not accepted.

I've been gone from BOL for the past 3 weeks so I'm catching up past threads. As swiggles reports, we had an article in our July newsletter that basically stemmed off of several BOL discussions from March & April about early disclosures and the correlation to making a conditional approval. The article is 7 pages long, so I won't post it here.

Randy states Just be aware that this is a minority opinion in the compliance world and many HMDA and other compliance experts do not agree with BCC on this stance. and KB states I would not adopt the BCC approach; if you really are considering it, have a long heart to heart with legal counsel about what issuing disclosures means vs "approving" an application.

While I'm not certain I want to open this can of worms again, I will tell you that we ran this article by numerous compliance professionals, legal experts, the ABA and lots of regulators before issuing it to our clients. While it is true that some disagreed (Randy was one), most did not. Therefore, I don't think it's accurate to say this is a "minority opinion". Also, not one regulator disagreed. However, several said they wouldn't cite a bank for a different stance, if they could support it.

Here's how the article closed. If you disagree, you need to consider this:
Our position is really a “bright line test”. If you accept this position, you can’t go wrong. An examiner is not going to cite a bank for saying, “when we issue preliminary disclosures, we believe we have made a conditional approval. Therefore, if the applicant doesn’t proceed or wants to withdraw after this action but prior to an intent to proceed, we code it ‘approved, not accepted”. We believe it is a clear and consistent approach and is well documented and supported.

If you don't believe issuing preliminary disclosures requires a decision, ask yourself “when do we make a decision and how do we document when that decision is made?” HMDA requires all information on the LAR to be supported. How would you support the date of “withdrawal” or “approved not accepted” consistently if you don’t have a way to clearly substantiate when that occurs? Examiners could (and many have) cite your bank for incorrect action codes.

If you accept this position, you’ll have fewer withdrawals on the HMDA-LAR and it’s common knowledge that withdrawals are a “red flag” for fair lending exams. To say you approved an application and then the applicant withdrew shows a proactive approach from the lender. Numerous withdrawals make examiners dig to find out if the bank is discouraging applicants.

While many may see this as a “gray” area, there’s no risk to accepting this position. We feel there is risk to not accepting it.

Yup - makes sense to me. I apply for a mortgage loan, you give me the early disclosures, I don't like your pricing and say I am going to go down the street and you code the application as approved not accepted.

Those that adopt this approach, I wish the best of luck.

Manipulation of whatever logic you are using to reduce the number of withdrawn applications because "withdrawals are a "red flag" for fair lending exams" is a much more dangerous practice (the CFPB stated that withdrawals made up 7 percent of all applications in 2000 in the latest HMDA proposal). The disposition of applications are what they are. If you have a higher number of withdrawals than the national average, then I suggest you figure out why and not try to hide them.

I have yet to see one other person put the support for this approach in writing for all to see. The only support that I would accept prior to informing my clients that they should think about adopting this approach is a written opinion from the CFPB.

That is likely not to happen as the CFPB has already stated their position quite clearly in the proposal of the new HMDA regulations which exactly mirrors the FFIEC FAQs which have been out there for years:

"If the applicant expressly withdraws before satisfying all underwriting or creditworthiness conditions and before the institution denies the application or closes the file for incompleteness, the institution reports the action taken as application withdrawn. If all underwriting and creditworthiness conditions have been met, and the conditions are solely customary commitment or closing conditions and the applicant expressly withdraws before the covered loan is originated, the institution would report the action taken as application approved but not accepted."

So what you are really saying is that by issuing the early disclosures, the applicant has met "all underwriting and creditworthiness conditions" of the creditor. In that case, can we also eliminate the "file closed for incompleteness" category since they have met all of the documentation requirements for meeting "all underwriting and creditworthiness conditions"?

There are plenty of other examples in the proposal. If the CFPB felt that issuing early disclosures were the equivalent of saying the application was approved but not accepted they sure wasted a lot of time discussing how to the report on withdrawn applications.

Basically, what you have proposed is that an application would have to be expressly withdrawn by the applicant prior to the creditor getting around to issuing early disclosures but possibly after obtaining a credit score and calculating the applicant's DTI, which I believe is part of your "conditional approval" argument? If that is the case, then why did the CFPB bother with the following?

Regarding credit scores: "They also include a code to use if more than one credit scoring model was used in developing the credit score, as well as a code for any other credit scoring model that is not listed, a code for purchased loans, and a code for use if the financial institution did not rely on a credit score in making the credit decision or if a file was closed for incompleteness or an application was withdrawn before a credit decision was made."

"Similarly, if an application was expressly withdrawn by the applicant before a credit decision was made, the financial institution complies with § 1003.4(a)(23) by reporting that no credit decision was made, even if the financial institution had calculated the applicant’s DTI ratio."

But you are correct that neither your or my opinion matters much, it is up the CFPB and the prudential regulators. All I am saying is that if someone wants to take this approach, then they need to be absolutely sure that it is correct.

David, I have to agree with Randy. You are taking a specific situation that happened with banks in your area who were taking specific actions and trying to extrapolate that to the entire industry.

Issuing early disclosures is NOT a credit decision unless in the three days that banks has actually underwritten the application. In many many (if not most) situations, no one with any authority to make a credit decision is even aware that an application is in house.

All that is in good faith is that you are providing fee estimates with a level of due diligence behind those fees, not fees that have been pulled out of the air to complete a form. That has nothing to do with whether or not a credit decision has been made.

I think it's best not to open this topic again. Therefore, I'm not going to write a response to each of your points. We obviously don't agree. I'm not trying to force anything. I've written a position paper that is well supported. If you don't agree, I recommend you determine when your institution does make a conditional approval. My experience with this is that most banks don't have a definitive point in the application process that they can point to. That's a problem.

The point I was trying to make in my last point is that this isn't a "minority" view. Many people have signed off on this approach and agree with it including Carl Pry, Patti Blenden, several at the American Bankers Association & numerous regulators. In fact, the only people to disagree have been the vocal majority here on the BOL threads. Dan can speak for himself, but he told me his regulators agreed with my position too, so this isn't just regulators in my area. I've also spoken with regulators from around the country about this issue. At the ABA Regulatory Compliance Conference in June, this topic came up in the HMDA session I taught (not by me), another session on Fair Lending and at the FDIC break out session I attended. Rick Freer (from the ABA) said they've had several banks report being cited for not taking the position I represent. Mark Kruhm (who answers most calls on HMDA for the ABA hotline) also said this has been a highly debated topic and has lit up his phone in the past few months. During the FDIC break out session, a FDIC representative made a response that supported exactly our position and added (paraphrasing) "if you're issuing early disclosures without underwriting the loan previously, you're risking UDAAP issues as you'd better be able to live by everything you stated in your disclosures."

My hope is that our position paper will drive the examiners to take an official stance on this confusing area. Several of our local examiners asked me to put this in writing so they could formally push it up the chain to D.C. level. I hope that occurs so we can put this to rest.

"if you're issuing early disclosures without underwriting the loan previously, you're risking UDAAP issues as you'd better be able to live by everything you stated in your disclosures."


From the GFE:

This GFE gives you an estimate of your settlement charges and loan terms if you are approved for
this loan.

I say 90% or more of ours are nowhere near approved within 3 days of walking through the door.

UDAAP? Let the arguments begin...that's plain crazy.

Exactly. There are so many variances between A LOT of issues from application package to application package. If it's hard for me as a compliance professional, to draw the line between approved-but-not-accepted and withdrawn, how in the world can I train a bunch of lenders located all across the state, in different markets, how to tell the difference?

Is it when we're confident enough of an approval to order an appraisal? Is it when we have verification of financials and employment (when, according to lenders, some applicant bring financial information with them to initial application ). As with my former bank, is it when the secondary market underwriter examines the entire package and says, "yes, we can close next week?" It just seems like there are so many variables, it's hard to make sure that all aps are reported in a like manner.

I have always struggled with this and so have enjoyed this discussion.....though, the discussion still does not provide me with a definitive answer. David's method makes more sense in that using that method, it's easy to decide how to report each application.

And I don't think it's "hiding" withdrawn applications. Our policy would state what we do....how we decide....and then the files are there for examiners/auditors, etc to look at.

In any event, I am not the Compliance Officer here, so it's not up to me to chose a method. I don't know what we will do.

If we see a file marked 'withdrawn', we question it for HMDA purposes. Out of the few of those we have, most are actually approved-not accepted.

All of my years as a lender, we never had a hard time knowing what stage we were at. We were never confused....whether in commercial lending, or consumer.

So what were your guidelines....what did you typically see in the file?

The fact is, many are unable to define when they've given conditional approval. So here's the bottom line question KB: When does a lender make a conditional approval?

If you recall, that's what started this entire discussion.

Not KB but I'll give you our definitions as hammered out with our FRB examiners.

Up to the time a loan is underwritten AND the decision is communicated to the borrower via a commitment to lend letter, the loan can be withdrawn by the borrower.

Once that commitment to lend is delivered, if the borrower says 'no thanks' it's approved but not accepted.

Seeing the discussion here, I'm truly grateful that we have no 'conditional approval' situation here.

At my former bank, we called it approved but not accepted when the file documentation contained documents that cost money such as an appraisal and a flood cert....because why would a lender order those if the loan wasn't in at least a conditional approval status? We also looked for tax returns, financial statements, etc because this meant that the lender had more than likely analyzed the financial portion of the file and saw no reason to deny. Still....though.....if the final underwriter nixed the deal, it could still be denied....which would mean it could not have been fully approved.....so confusing.

So such thing as commitment letter here......just file documentation to look at.

What happens when a borrower walks in, gives you all 6 points of information, you pull credit and disclose. The only thing you know is their credit score qualifies. How can that be a conditional approval if nothing is verified?

In this situation, "all underwriting or creditworthiness conditions" have not been satisfied but you have disclosed. So based on the FAQs Randy posted above, this cannot be approved/not accepted.

We have numerous applications where the credit scores are right on the borderline of acceptable, so we disclose and have to do more research into program requirements. If disclosing equates to "all underwriting and creditworthiness conditions" being satisfied, can we deny based on credit later?

BCC's approach makes no sense to me, as it seems to fly in the face of the FAQs.

Here's the only guidance HMDA offers about conditional approvals:
http://www.ffiec.gov/hmda/faqreg.htm#action

You'll see nothing about the decision being communicated to the borrower, commitment letters, etc. In fact, the 2nd FAQ states:
"If a credit decision has not been made and the borrower has expressly withdrawn, use the code for "application withdrawn." That code is not otherwise available."

swiggles stated: why would a lender order those if the loan wasn't in at least a conditional approval status? Exactly.

I'm not KB, but all I can say it is not contingent on issuing early or any other disclosures.

Every lending shop is going to be a little different depending on the structure of the credit underwriting function. If you are having a hard time determining a bright line test in your organization, then you have an operational problem and not a HMDA problem. It is pretty simple, the customer has met ALL underwriting or creditworthiness conditions. Not some, not most, but ALL. If you can't determine when that happens in your organization, then how do you ever decide when a loan is approved?

If you have a central underwriting department, unless the loan has made it that far, I can tell you your approval hasn't happened. If you have a central underwriting department and the front line or another centralized function issuing early disclosures, I can tell you approval has definitely not happened by the mere fact that you issued disclosures.

If you happen to get a complete loan package on day one for your underwriters to issue an opinion on, then you might on a rare occasional be able to say an applicant has satisfied ALL underwriting or creditworthiness conditions within the first three business days, but I would expect that would be a somewhat rare occasion.

If you are have problems figuring out a bright line test, then I suggest sitting down and having a chat with your underwriters and figure out when the "ALL underwriting or creditworthiness conditions" are met happens in your organization.

Just because your organization or someone's understanding of the organization has no logical or standardized process to make this determination, picking the issuance of early disclosures as the bright line test is not the answer.

Oh, and if these are all approved but not accepted at the time of the issuance of early disclosures, you better not have any adverse action notices in file for these applications unless they are only referring to customary commitment or closing conditions as defined by the CFPB.

The approval is conditioned upon all the information provided in the application being true, once verified.

You guys have totally misconstrued the conditional part of the requirement. The only conditional thing about it is the customary commitment or closing conditions. You are not allowed any leeway on the underwriting or creditworthiness conditions. Those ALL have to be satisfied.

Back to my second post: "Customary loan-commitment and loan-closing conditions do not include (3) verification or confirmation, in whatever form the lender ordinarily requires, that the borrower meets underwriting conditions concerning borrower creditworthiness."

If an underwriting condition has not been verified and you require it before closing the loan, it is not an approval under HMDA.

I don't understand this statement. An application might be motoring right along towards closing, and at any time (possibly even a day before closing), a piece of information might surface, causing a decline.

A loan is in a semi-state of approval until the customer either withdraws the request or the bank finds a reason to deny it.

I issue early disclosures - if the applicant withdraws sometime before loan closing, I code all such applications as approved but not accepted.

Well, if every application is approved but not accepted once the early disclosures are issued, you must have considered that at the time of issuing the early disclosures that they met all underwriting or creditworthiness conditions. So the loan is already sitting in the "approved but not accepted" bucket. How can you then turn around and say, for example, once you get the tax returns and can't justify the income listed on the application, deny the loan for that underwriting condition - failure to met required DTI?????

Verification documents are not considered customary loan-commitment and loan-closing conditions.

Because the bank orders this as soon as they have intent to proceed, the customer is paying (and the bank collected the appraisal fee or an application fee to cover the cost), and the bank wants to know if there is anything worth talking about.

KB: You still haven't answered my question about when you believe conditional approval is given. Please do.

Randy, the issuance of preliminary disclosures is not a commitment to make a loan. In fact, §1024.7(g) states the GFE is not a loan commitment. A similar reference is found in the Section-by-Section Analysis of the Integrated Disclosures Rule regarding the Loan Estimate. A lender is never committed until the loan is consummated and always has the option to deny an application up until the loan is closed. However, that doesn’t mean the lender hasn’t made a conditional approval when the GFE/LE and Preliminary TIL disclosures are issued. We believe a conditional approval must be made in order to provide binding and meaningful disclosures.

The RESPA final rule in the November 17, 2008 Federal Register states:
HUD has adopted a single application process for the final rule. Under this approach, at the time of application, the loan originator will decide what application information it needs to collect from a borrower, and which of that collected application information it will use, in order to issue a meaningful GFE. However, before providing the GFE, the loan originator will be assumed to have collected at least the following six items of information: the borrower’s name, Social Security Number, and gross monthly income; the property address; an estimate of the value of the property; and the amount of the mortgage loan sought. The borrower’s Social Security Number would be collected for purposes of obtaining a credit report.

Furthermore, the loan originator is presumed to have relied on the borrower’s name, the borrower’s monthly income, the property address, an estimate of the value of the property, the mortgage loan amount sought, and any information contained in any credit report obtained by the loan originator before providing the GFE.

Additionally, the RESPA section by section analysis specifically states [i]“HUD determined that this approach provides the flexibility originators need to properly underwrite, while limiting bait-and-switch methods whereby the originator uses the GFE to draw in a borrower and, after a significant application fee is paid or burdensome documentation demands are made, claims that a material change has resulted in a more expensive loan offering”.

This tells us that there is an underwriting process that occurs BEFORE the GFE is issued. No one says it's the final underwriting. If they pass the lender's underwriting, a GFE/P-TIL is issued. Therefore, some sort of approval is given at this point.

I don't see why this is such a difficult concept.

David,

Are we talking about RESPA compliance or are we talking about HMDA compliance. The two do not equate. One is a reporting regulation and one (as it relates to the GFE) is a disclosure regulation. Each have their own rules. So let's just pick one regulation to argue about.

Your stance is that by issuing early disclosures under RESPA that you have met the conditions to report the loan as approved but not accepted under HMDA.

That is simply just not the case. A GFE is almost always issued long prior to the applicant meeting all underwriting or creditworthiness conditions required by the lender including the receipt of all verifications or confirmations, in whatever form the lender ordinarily requires.

If you want to talk about the actions that a bank should take and what information should be relied upon prior to issuing a GFE in good faith, then that is a totally separate subject and should be left for another discussion.

Whatever those actions might be under RESPA or the upcoming TRID rules does not normally rise to the level of determining that an applicant meets all underwriting or creditworthiness conditions required by the lender including the receipt of all verifications or confirmations, in whatever form the lender ordinarily requires.

This is not a "point in time" determination. It is not a moving scale. You don't get to reset the underwriting or creditworthiness conditions as the application moves thorough different stages of the application process. There is only one point of time in which this happens.

We need to totally stick to the definitions and regulatory guidance as given for compliance with the HMDA requirements.

You keep saying that "We believe a conditional approval must be made in order to provide binding and meaningful disclosures." Well that may be under RESPA, but as I said, that should be left to another discussion. However, your definition of a "conditional approval" at the time of issuing a GFE does not equate to the definition of "approved but not accepted" under the current FFIEC guidance and the CFPB guidance in the new proposed HMDA regulation. The only thing an "approved but not accepted" under HMDA can be "conditioned" on are "customary loan-commitment and loan-closing conditions".

P.S. I'm done

I am heading out of town to a wedding. If I come back to this it will be next week.

Randy: That's why I said in my original post I didn't think it was good to open this can of worms again. We've discussed this pretty thoroughly in March/April. You give lengthy responses, but never seem to address the regulatory comments and citations I bring up, like the requirement to underwrite a loan BEFORE a GFE is issued.

Yes, I'm talking about HMDA Action Codes, but I'm looking at RESPA too. I agree you can't take a term from one regulation to use in another, but all regulations must work together. You can't violate one reg to comply with another.

It's also interesting to me that you make long comments and then say "I'm done".

KB: You also don't seem to want to address the questions I've asked of you, but you've made two posts since. Odd to me.

I'm not trying to be rude to anyone. I'm trying to point out something that a vocal few here at BOL don't seem to want to acknowledge, but my research on this topic for several months resulted in most everyone agreeing. Again, Carl Pry, Patti Blenden, ABA personnel, several regulators, etc. Just because a few here don't agree, doesn't make it right.

Swiggles presented the logic very well. Again, odd to me that others don't acknowledge it and don't seem to want to respond to her and my questions. We don't need to debate this, but a few of you seem to want to but then won't address the issues presented. So if you'd like to continue to discuss this, please reply to the questions we've asked about when conditional approval happens in your opinion. Then we can have a healthy discussion instead of blasting people and running away.

Then David, I won't "run away". However, I think I have already stated my case plainly and succinctly.

But I only have one question, since you seem to be in the mood of calling people out on this subject.

Why don't you tell everyone the whole story behind why some of your clients were originally criticized and questioned about reporting the HMDA code wrong on their LAR.

You previously told me privately that: "Some of our clients issue letters with the GFE/P-TIL that say “You’ve been approved”. It’s pretty tough to say there’s no decision."

And my response to you was: "Well if they are issuing letters to the applicants saying that they are approved for the loan at the time they are delivering the GFE – then they have more problems than their HMDA reporting."

So, then you turn around and use these isolated findings and try to make this argument that in all cases when you issue a GFE, you have an "approved but not accepted" loan under HMDA, when the fact of the matter is, that is not the point the regulators were making or implying at all.

The regulators were stating that you sent a letter to the customer that they were approved for the loan and since a decision on the loan had already been made and communicated to the customer, the loan could no longer be withdrawn by the applicant.

As far as addressing your regulatory comments, they just don't make any sense. One can't use some interpretation one has made in one regulation (RESPA) to override the clear and bright line test put forth by the FFIEC and CFPB in HMDA, no matter how much one tries.

There is only one fact and one fact alone to examine in this whole convoluted mess and we can just let each individual reading this make their own determination.

When a bank issues a GFE can they say that the loan application is fully approved outside of customary loan-commitment and loan-closing conditions, taking in consideration of the follow:

Customary loan-commitment and loan-closing conditions do not include (1) conditions that constitute a counter-offer, such as a demand for a higher down-payment; (2) underwriting conditions concerning the borrower's creditworthiness, including satisfactory debt-to-income and loan-to-value ratios; or (3) verification or confirmation, in whatever form the lender ordinarily requires, that the borrower meets underwriting conditions concerning borrower creditworthiness.

If the bank answers no, then the application cannot be coded as approved but not accepted on their HMDA LAR.

There is no RESPA argument here. What you do or not do from a RESPA standpoint is absolutely irrelevant when to comes to HMDA reporting and there is nothing in RESPA that requires you to get an application to this point of the approval process before issuing a GFE.

If you care to continue this discussion using only HMDA guidance and definitions, I look forward to it.

So....why does everyone holler about too many withdrawns raising fair lending red flags. If we wait until everything is verified and we are ready to close, to use the approved but not accepted code, we won't have any. Because it's rare for a borrower to withdraw at that point. Our LAR will only have withdrawn entries.

I should think examiners would have an issue with you approving a bunch of loans that weren't accepted. And my question to you, swiggles, remains: How can you approve a loan before everything is verified?

We have more withdrawns than approved but not accepted and it's never been an issue in all the exams I've been through. As long as the file is properly handled and things like the notice of incomplete sent, then you should have some of everything on your LAR without a real preponderance of anything, except originations, of course.

To make any reasonable argument that "withdrawn" must happen within the 3-day window of issuing disclosures, your action date for withdrawn applications would always have to happen within that time period, which makes zero sense.

For an application that was expressly withdrawn by the applicant, you may enter either the date shown on the applicant’s letter or the date that you received the letter or notice.

Nowhere does it state that this could only occur within the initial 3 days. What about the millions of loans not subject to RESPA to begin with?

5 Action taken date—approved but not accepted. For a loan approved by an institution but not accepted by the applicant, the institution reports any reasonable date, such as the approval date, the deadline for accepting the offer, or the date the file was closed. Although an institution need not choose the same approach for its entire HMDA sub mission, it should be generally consistent (such as by routinely using one approach within a particular division of the institution or for a category of loans).

I’ve purposely waited a few days to respond to this post. I’m happy to have a healthy discussion about this topic. I’m not willing to argue and sling mud at each other. In that light:
Randy, you said:

I told you about the experience of ONE client but we’ve also had several that have been cited for this issue (calling an application withdrawn when the regulators believe a conditional approval has been given because disclosures have been provided) that didn’t issue “you’re approved” letters. We’ve also had several more clients that have had discussions with their regulators about this – but not cited.

Further, while teaching HMDA at the ABA School in March, a student brought up this topic. I asked the audience of approximately 340 students if they have had similar discussions with their regulators. About 1/3 of the students raised their hands. Many wanted me to discuss it in detail, but there wasn’t sufficient time there.

As I stated in a previous post, this was brought up at 3 different sessions at the ABA Compliance Conference in June – once by the FDIC during their breakout session.

Your comments make it sound like I’m trying to do something unethical (hiding the full truth, etc.). This topic is not something I created, but rather something I’m trying to react to. We’re in the business of making life easier for compliance officers. That’s what I’m trying to do whether you agree with it or not.

So there’s the “whole story”. We’ve been researching this topic for months, talking to regulators, bankers from around the country and discussing this in depth in our team of 9 consultants. This isn’t something that I dreamt up.

Second, you say:

HMDA clearly says that if a “conditional approval” is granted, you can’t call an application “withdrawn”. HMDA never defines what a “conditional approval” is. However, the FFIEC FAQs do give two answers that illustrate it. The 2nd one states:
If a credit decision has not been made and the borrower has expressly withdrawn, use the code for "application withdrawn." That code is not otherwise available.

Notice this doesn’t say “FINAL DECISION”. It says “a credit decision”. I can easily interpret this to say “ANY credit decision.”

In your last post, Randy, you state:


Nowhere does HMDA say “fully approved”, “final decision”, “preliminary decision” or any other qualifier. I don’t know why you are adding those words (fully approved). It seems VERY clear to me that if you issue disclosures, you have evaluated information about the applicant and made some sort of decision – at a minimum, not to deny the request. Therefore, it would be difficult to argue no decision of any sort has been made. The FFIEC FAQ goes on to say you can’t use “withdrawn” once A decision has been made.

Just to be clear – this is from the HMDA FAQs. We’re not talking about RESPA, TIL or crossing any other lines into other regulations. However, all of these regulations work together. You can’t get in a HMDA silo and just comply with it and then move to the RESPA silo. They work together. I’m not using RESPA to define HMDA’s terms (as explained above with the HMDA FAQ only), but our position paper does support this logic by referencing RESPA.

Where you seem to be the most concerned is the RESPA requirement to underwrite loans BEFORE you issue GFE’s (and there is such a requirement as I quoted 2 posts of mine above). If a bank is required to underwrite a loan to issue “meaningful and binding” disclosures, I once again find it very difficult to argue a bank hasn’t given some sort of conditional approval (the HMDA term).

Now I’m going to stop here. What’s most difficult is trying to respond many of you at the same time. I’m not wishing to ignore anyone’s position or responses. Trying to do so, would require lengthy posts and a lot of time, however. I also won’t continue to dialogue with people that won’t answer questions I pose.

Joker: I'm curious how you make this statement in your last post, but it appears to contract your response this week to the following in a RESPA forum post. The question asked was:
What if the applicant never returns the Intent to Proceed and you would have otherwise approved the loan- do you consider withdrawn for Reg B purpose and "approved not accepted" for HMDA?
You replied to code the loan Approved, Not Accepted" for HMDA. That seems to follow my position.

Here's the post:
http://www.bankersonline.com/forum/ubbthreads.php?ubb=showflat&Number=1997191#Post1997191

Truffle stated:
I should think examiners would have an issue with you approving a bunch of loans that weren't accepted.
Why would examiners have an issue with approving a bunch of loans that weren't accepted? The lender is demonstrating they are willing to offer credit, the borrower shopped around and didn't choose the bank. That's what RESPA/TIL and the new TRID rules were intended to do - inform people so they can shop. That doesn't mean you'll get 100% acceptance. The lender also demonstrates they aren't dragging their feet (discouragement under Reg B).

And my question to you, swiggles, remains: How can you approve a loan before everything is verified?
That's exactly what RESPA and TRID require! The lender is supposed to think "IF this is true, then we could offer this loan at this rate and fees." The TRID rules even require the lender to make assumptions if there's incomplete information. Of course, you verify things - that's prudent underwriting, but it's not compliance and it happens AFTER you issue disclosures.

Again, no one is saying "final approval". HMDA asks for "conditional approval". Conditioned on what? Verification! The FFIEC FAQ on Action Code clearly demonstrate this.

Conditional approvals---failure to satisfy creditworthiness conditions. How should a lender code "action taken" where the borrower does not satisfy conditions concerning creditworthiness?

Answer: If a credit decision has not been made and the borrower has expressly withdrawn, use the code for "application withdrawn." That code is not otherwise available. See Appendix A, I.B.1.d. If the condition involves submitting additional information about creditworthiness the lender needs to make a credit decision and the applicant has not responded to a request for the additional information in the time allowed, use the code for "file closed for incompleteness." See Appendix A, I.B.1.e. If the borrower has supplied the information the lender requires for a credit decision and the lender denies the application or extends a counter-offer that the borrower does not accept, use the code for "application denied." If the borrower has satisfied the underwriting conditions of the lender and the lender agrees to extend credit but the loan is not consummated, then use the code for "application approved but not accepted."

For example, if approval is conditioned on a satisfactory appraisal and, despite notice of the need for an appraisal, the applicant declines to obtain an appraisal or does not respond to the lender's notice, then the application should be coded "file closed for incompleteness." If, on the other hand, the applicant obtains an appraisal but the appraisal does not support the assumed loan-to-value ratio and the lender is therefore not willing to extend the loan amount sought, then the lender must use the code for "application denied."


I've read this section you referenced above, David, at least a dozen times. In my mind, and in my world, there's no way that any kind of approval can occur in the same day to three day time period between taking an application and giving the GFE. You haven't convinced me otherwise. I'm sticking with what my examiners have verified as the way to report. But I appreciate you taking the time and sharing your sources.

Good discussion Truffle and I absolutely see your point. My logic (and that of those that "forced" us to look at this from a different angle) was some sort of credit decision has to be made initially to provide early disclosures. If that's true, then "Withdrawn" is no longer an option.

I appreciate your input. I might be overly optimistic, but our position paper has now been sent to the CFPB and to numerous field examiners in hopes that it works it way to the top. Hopefully, we can get some sort of uniform, regulatory stance on this issue.

Amen, brother! That would be awesome. I'm sure you will keep us posted....just in case the CFPB doesn't.

One other aspect that I don't believe we've touched on in this lengthy discussion is online applications.
I'm reviewing a file now that has the following at the bottom of the early disclosures.


No human has set eyes on the information submitted. The disclosures were computer generated. Yet if I'm understanding David's contention correctly, this loan would be coded approved but not accepted if nothing further happened and the borrower withdrew it. The thought of computers giving conditional approvals is mind boggling to me.

How? Because they stated they would have otherwise approved the loan. The prospective customer never showed back up and didn't send a notice to withdraw (before an apparent decision had been made)...so since an apparent decision had been made, then it was approved, but was never accepted (originated).

Can I sidetrack slightly? How does everyone document a withdrawal? Do you try to get the borrower to send something in writing? If by phone, does the officer just document that the applicant withdrew by phone? Does anyone document the reason for withdrawal? If the applicant has provided all pieces of info and early disclosures have been provided and then the lender can't get the applicant to respond to phone calls etc....just document that fact and call it withdrawn (or possibly approved but not accepted)?

If this needs to be a separate discussion, I will start one.

I'll say this again too. What about the gagillions of HMDA reportable loans that are not RESPA loans. Are you sticking to a 3-day rule on those too? Last time I checked Reg B, you had 30 days to make a decision.

I'm not trying to argue, but I think the 'logic' is completely 'illogical' and one-sided/narrow in scope.

Swiggles, our files are typically marked withdrawn(unless done by a more savvy HMDA-oriented person) and usually because the customer called and said 'nevermind' or "i'm just going to pay cash/pay it off/whatever". All of our non-originated loans are looked at to verify HMDA reportable (or not). If they are HMDA reportable and marked ANA, we go no further. If they are marked "Withdrawn" we look deeper and discuss with the originating officer, if needed, to clarify and document whether they are actually W/D or ANA.

Any written communication is placed in file. If via email, a copy of the email is placed in the file. Telephone withdrawals are documented by the LO as to when received.

Awesome...thanks. I want to put in place a system whereby the officer chooses WD or ABNA and then compliance looks at the selection and file documentation to see if it looks feasible......thus my need for a guideline or facts which will help loan officers decide and compliance to agree.

I find that officers' perception of WD and ABNA are all over the board. And I agree regarding investigating WD files as some officers tend to mark their files withdrawn, automatically, as though ABNA doesn't exist.

And thanks, Truffle.

Yes, swiggles...and it's likely a result of Reg B rules being older than HMDA. !!

Absolutely not! No where have I ever stated (or implied) a bank has 3 days to make a decision. Non-RESPA loans don't have this requirement and therefore wouldn't be subject to this.


What I am saying is RESPA requires a bank to give a "binding and meaningful" GFE. RESPA also requires a bank to underwrite a loan PRIOR to issue a GFE (as I have quoted from the Federal Register). Therefore, RESPA applicable loans require underwriting (a conditional approval) when GFE's are issued.

When a lender provides a GFE & P-TIL, aren't they saying they would approve the loan with those terms and fees (assuming everything is verified and checks out)?

David, I do agree that I would not deliver a GFE on a loan product/terms that doesn't exist and couldn't be made. I simply do not agree that a loan must be underwritten and conditionally or otherwise approved before issuing disclosures.

It's often not even humanly possible!

You will find absolute no disagreement from me on this statement. But as I have previously pointed out, that is not the issue.

The issue remains the specific definition of an "approved but not accepted" loan application under the HMDA guidelines, which continues to be sidestepped by discussions regarding RESPA requirements.

David, please address what you think these statements from the CFPB mean:

1. The CFPB has stated: "The commentary generally provides that financial institutions should report loans approved subject to underwriting conditions which are not met should be reported as a denial, but it also provides that certain customary loan commitment or loan-closing conditions are not underwriting conditions. Additional guidance on this topic had been published in the FFIEC FAQs."

2. The CFPB has stated: "If the applicant expressly withdraws before satisfying all underwriting or creditworthiness conditions and before the institution denies the application or closes the file for incompleteness, the institution reports the action taken as application withdrawn. If all underwriting and creditworthiness conditions have been met, and the conditions are solely customary commitment or closing conditions and the applicant expressly withdraws before the covered loan is originated, the institution would report the action taken as application approved but not accepted."

3.The CFPB provided additional examples of customary commitment or closing conditions in the HMDA proposal: "These examples include: acceptable title insurance binder; clear termite inspection; a subordination agreement from another lienholder; and where the applicant plans to use the proceeds from the sale of one home to purchase another; a settlement statement showing adequate proceeds from the sale. The existing examples of a clear-title requirement and acceptable property survey are retained.

4. The CFPB also provided examples of underwriting or creditworthiness conditions which mirror the current statements from the FFIEC: "These examples include: conditions that constitute a counter-offer; satisfactory debt-to-income ratio or loan-to-value ratio; determination of the need for private mortgage insurance; satisfactory appraisal requirement; or verification or confirmation that an applicant meets underwriting conditions; including documentation of income or assets.


Didn't your last statement ("assuming everything is verified and checks out)?") not just confirm that all underwriting or creditworthiness conditions have not been met when you provide a GFE & P-TIL??

If not, I continue to be really confused.

Randy, that is crystal clear to me. I don't see how on earth it can be interpreted any other way.

Now see, I do disagree with David's statement.
When we give a GFE all we're saying is 'Here are the estimated costs for the loan you applied for.'

In my world, the only thing that gets looked at prior to issuing the GFE is the actual credit score. If it's too low for the program applied for, we deny. Otherwise, it goes straight to get the GFE done and out the door within the three days.

TR - Well, you have to admit that you can only disagree with that statement to the extent that once you issue the GFE and P-TIL, if you end up actually making the loan, those terms and fees will be correct (barring allowable tolerances) or unless a valid change in circumstances happens.

Agree...agree...agree. That's all it's saying. It has no bearing on anything else other than a commitment to fees disclosed (outside of a COC later on)

You're actually supporting my case, Randy.
I'm saying 'here are the estimated costs for the loan you applied for.'
You're correct that if all goes right, those estimated costs will end up on the HUD.

There is, however, no way that I'm saying what David says and you quote above that (I) would approve the loan with those terms and fees.

To me the distinction between these two things is crystal clear.

This is from the 1998 Federal Register announcing the RESPA rules that went into affect in Jan 2010. This requires lenders to underwrite a loan PRIOR to issuing a GFE so they don't bait and switch:
“HUD determined that this approach provides the flexibility originators need to properly underwrite, while limiting bait-and-switch methods whereby the originator uses the GFE to draw in a borrower and, after a significant application fee is paid or burdensome documentation demands are made, claims that a material change has resulted in a more expensive loan offering”.

The regulation requires lenders to underwrite the loan and then live by what they disclose - unless there's a changed circumstance. If there is no changed circumstance, your final disclosures better match your GFE (within tolerances).

If you aren't underwriting the loan prior to issuing the GFE, how do you know what to disclose? How can your GFE be meaningful and binding. You can't issue garbage. Why wouldn't you want to underwrite (to the extent possible) so you don't back yourself into a corner?

Randy & Truffle: I'm pretty busy today - Tuesday. I'm not ignoring you, but I'm not certain when I will give you a complete reply. Be patient with me and I'll get back to you.

Besides the fact that you are again arguing a HMDA definition by using something from RESPA, that is not what HUD said at all. That sentence is only part of the paragraph in which it appeared. The sentence before the one you quoted in that paragraph says this:

"Furthermore, the loan originator is presumed to have relied on the borrower’s name, the borrower’s monthly income, the property address, an estimate of the value of the property, the mortgage loan amount sought, and any information contained in any credit report obtained by the loan originator before providing the GFE. The loan originator cannot base a revision of the GFE on this information, unless it changes or is later found to be inaccurate."

In the paragraph previous to this, HUD stated: "None of the information collected by the originator prior to issuing the GFE may later become the basis for a ‘‘changed circumstance’’ upon which a loan originator may offer a revised GFE, unless the loan originator can demonstrate that there was a change in the particular information or that it was inaccurate, or that the loan originator did not rely on that particular information in issuing the GFE."

It is giving the creditor the flexibility to actually underwrite the loan in any manner it may choose, including at a later time - it does not say that they must underwrite the loan before issuing the GFE.

HUD only indicated that the fees and charges quoted to the applicant cannot change based on a change in the factors that the applicant has already told the creditor under which the creditor has used to determine whether or not a GFE application had occurred.

What a creditor does with the application to ensure their fees are accurate is up to the creditor - but it does not "require" any underwriting. It only prevents the creditor from changing the quoted fees unless there was a change in the particular information or that it was inaccurate.

This discussion is all in the lead up to what would and would not constitute a change in circumstance allowing a creditor to send a revised GFE.

What RESPA really boils down to is:

1. You were required to issue a GFE when you hit the GFE application threshold.

2. HUD didn't really care how or what you thought you needed to rely on to come up with the numbers on the GFE. They really left that decision up to you.

3. Once you quoted them to the borrower, you could only change them under certain conditions (defined change in circumstance).

4. If you other wise changed them, you have a violation of RESPA.

5. If you abide by the change in circumstance rules and still close the loan with fees outside of tolerance, you pay the customer.

None of this has anything to do with underwriting the loan.

Randy's post two above this saved me having to write out the same thing. The quote David keeps falling back on does NOT "... require(s) lenders to underwrite a loan PRIOR to issuing a GFE so they don't bait and switch:...." As Randy said, you're leaving out the meat of the citation in the sentence immediately preceding it.

All of these pages of discussion aside, I stick to Reg C when it comes to reporting action codes on my LAR. Specifically "If a credit decision has not been made and the borrower has expressly withdrawn, use the code for "application withdrawn."

I think we have found the underlying problem with our disunity on this topic! This is actually exciting to me because I know you have all been frustrated with me and it's been frustrating for me to understand why you all don't understand.

Randy said: "None of this has anything to do with underwriting the loan. Truffle said: The quote David keeps falling back on does NOT "... require(s) lenders to underwrite a loan PRIOR to issuing a GFE so they don't bait and switch:...." And Joker said "I don't need to verify anything else, but I've not underwritten nor can I approve the loan until I do!"

I DO believe a lender is required to underwrite what they have available before issuing a GFE. It's not full underwriting. It's not the final underwriting. It's not complete underwriting. It's analyzing and reviewing EVERYTHING you have available at this point. It's making a decision on what you know and what you believe to be true at this point.

Let me make this clear:
The issuance of preliminary disclosures is not a commitment to make a loan. In fact, §1024.7(g) states the GFE is not a loan commitment. A similar reference is found in the Section-by-Section Analysis of the Integrated Disclosures Rule regarding the Loan Estimate. A lender is never committed until the loan is consummated and always has the option to deny an application up until the loan is closed. However, that doesn’t mean the lender hasn’t made a conditional approval when the GFE/LE and Preliminary TIL disclosures are issued. We believe a conditional approval must be made in order to provide binding and meaningful disclosures.

Here's where I pull this from. Once a lender has accepted an application (as defined in RESPA), the lender is required to issue a binding and meaningful GFE [1024.7(f)].

Binding GFE. The loan originator is bound, within the tolerances provided in paragraph (e) of this section, to the settlement charges and terms listed on the GFE provided to the borrower, unless a revised GFE is provided prior to settlement consistent with this paragraph (f) or the GFE expires in accordance with paragraph (f)(4) of this section. If a loan originator provides a revised GFE consistent with this paragraph, the loan originator must document the reason that a revised GFE was provided. Loan originators must retain documentation of any reason for providing a revised GFE for no less than 3 years after settlement.

Beginning in October 2015, the lender will need to issue a LE made in “good faith”. This is supported by the following regulatory citations:

Loan Estimate must contain a good faith estimate of credit costs and transaction terms. If any information necessary for an accurate disclosure is unknown, the creditor must make the disclosure based on the best information reasonably available at the time the disclosure is provided to the consumer, and use due diligence in obtaining the information. [§ 1026.19(e)(1)(i); Comment 19(e)(1)(i)-1]

Creditors are required to act in good faith and exercise due diligence in obtaining information necessary to complete the Loan Estimate. [The Small Entity Guide in reference to the Commentary to §1026.17(c)(2)(i)]

For example, the creditor might look to the consumer for the time of consummation, to insurance companies for the cost of insurance, or to realtors for taxes and escrow fees. [Commentary to §1026.17(c)(2)(i)-1].

Normally creditors may rely on the representations of other parties in obtaining information.
 [§ 1026.17(c)(2)(i)]

However, there may be some information that is unknown (i.e., not reasonably available to the creditor at the time the Loan Estimate is made). In these instances, the creditor may use estimates even though it knows that more precise information will be available by the point of consummation. However, new disclosures may be required under §1026.17(c) or §1026.19. [Comment 17(c)(2)(i)-1]

The RESPA and TRID rules require lenders to review all of the information they have available before they issue a GFE/LE. The TRID rules indicate there is a due diligence standard placed on lenders to get information they need to issue disclosures that are meaningful and represent terms you are willing to offer the applicant. If these disclosures are not as accurate as possible then the lender is risking UDAAP issues (bait and switch tactics).

The RESPA final rule in the November 17, 2008 Federal Register states:
HUD has adopted a single application process for the final rule. Under this approach, at the time of application, the loan originator will decide what application information it needs to collect from a borrower, and which of that collected application information it will use, in order to issue a meaningful GFE. However, before providing the GFE, the loan originator will be assumed to have collected at least the following six items of information: the borrower’s name, Social Security Number, and gross monthly income; the property address; an estimate of the value of the property; and the amount of the mortgage loan sought. The borrower’s Social Security Number would be collected for purposes of obtaining a credit report.

Furthermore, the loan originator is presumed to have relied on the borrower’s name, the borrower’s monthly income, the property address, an estimate of the value of the property, the mortgage loan amount sought, and any information contained in any credit report obtained by the loan originator before providing the GFE.

Additionally, the RESPA section by section analysis specifically states “HUD determined that this approach provides the flexibility originators need to properly underwrite, while limiting bait-and-switch methods whereby the originator uses the GFE to draw in a borrower and, after a significant application fee is paid or burdensome documentation demands are made, claims that a material change has resulted in a more expensive loan offering”.

The same premise will hold true under the Integrated Disclosure requirements, as the Commentary to §1026.19(e)(3)(iv)(A), #3 states “a creditor is presumed to have collected these six pieces of information . . .. ” In a separate discussion on the extent of credit information to be obtained, the Final Integrated Disclosures Rule in the December 31, 2013 Federal Register states, creditors should be able to receive sufficient information from the credit report…to develop a reasonably accurate estimate of costs….

Creditors may only use revised or corrected Loan Estimates when specific requirements are met. Creditors generally may not issue revisions to Loan Estimates because they later discover technical errors, miscalculations, or underestimations of charges. Creditors are permitted to issue revised Loan Estimates only in certain situations such as when changed circumstances result in increased charges. [§1026.19(e)3)(iv)]

This tells us that both rules expect lenders to USE and RELY ON the 6 items and the credit report before issuing the GFE/LE. How? The income and credit report are used to calculate a debt-to-income ratio. The estimate of value and loan amount is used to calculate a loan-to-value ratio. The score from the credit report can also be evaluated. All of these things are used/relied on to give the best GFE/LE possible – a meaningful and binding GFE/LE. Lenders are not to issue a “standard estimate of charges” as it would not meet the “good faith” requirement.

Additionally, Regulation B says lenders are to be diligent in obtaining whatever information they need: The creditor shall exercise reasonable diligence in obtaining such information. [§1002.2(f)] The Commentary to this section even specifically mentions obtaining a credit report diligently.

The regulation defines a completed application in terms that give a creditor the latitude to establish its own information requirements. Nevertheless, the creditor must act with reasonable diligence to collect information needed to complete the application. For example, the creditor should request information from third parties, such as a credit report, promptly after receiving the application . . . [Commentary to §1002.2(f)(6)].

Many applicants voluntarily provide verification documents at the time of application (tax returns, pay stubs, etc.) If a lender has these documents available prior to issuing the GFE/LE, they should review them and prepare a binding GFE/LE in good faith based on all information available. If a lender doesn’t review all of the information they have available, they can’t later use any of that information as a Changed Circumstance.

The loan originator cannot base a revision of the GFE on this information, unless it changes or is later found to be inaccurate. HUD determined that this approach provides the flexibility originators need to properly underwrite, while limiting bait-and-switch methods whereby the originator uses the GFE to draw in a borrower and, after a significant application fee is paid or burdensome documentation demands are made, claims that a material change has resulted in a more expensive loan offering.

None of the information collected by the originator prior to issuing the GFE may later become the basis for a ‘‘changed circumstance’’ upon which a loan originator may offer a revised GFE . . . [Federal Register 11-17-08].

This requirement is also supported in numerous places in the TRID rules.

So back to the original questions:
1. Does the lender make a decision in order to issue the GFE?
We think they have to because the lender is supposed to be reviewing all of the information possible prior to issuing the GFE/LE. They then issue a proposal to the applicant through the P-TIL and GFE / LE. These disclosures aren’t hypothetical. They represent what the lender will do (binding and meaningful), unless later verifications demonstrate the information is inaccurate or new information is provided that allows for a changed circumstance. If everything goes as planned, the P-TIL and GFE / LE will look just like the final TIL and Settlement Statement / Closing Document. There are tolerances and cure provisions, if the early disclosures are not accurate.

2. If the lender doesn’t evaluate what’s been provided and make a decision based on this information, how can the lender issue early disclosures that may represent how the final documents will be designed?
To issue a meaningful and binding GFE/LE in good faith, a decision must have been made about what program, rate & fees are used for the applicant. That IS a decision, although it may be conditional and a preliminary one. The point is, the lender must make a decision and be willing to live with it if there are no changed circumstances.

3. How can we make a decision if we haven’t underwritten the loan yet?
Remember the RESPA final rule in the November 17, 2008 Federal Register states:
“HUD determined that this approach provides the flexibility originators need to properly underwrite, while limiting bait-and-switch methods whereby the originator uses the GFE to draw in a borrower and, after a significant application fee is paid or burdensome documentation demands are made, claims that a material change has resulted in a more expensive loan offering”.

If you believe you haven’t made a decision yet, how do you comply with this requirement to have already underwritten the loan request to avoid bait and switch tactics? The term “underwrite” is not defined, but it’s clear it happens before issuing preliminary disclosures. While it may not be the final underwriting, as used within our industry, the regulation clearly uses this term and requires a decision to offer the best information possible to the applicant before preliminary disclosures are issued.

Now let’s go back to HMDA. The Commentary to §1003.4(a)(7) #4 states:
Action taken—conditional approvals. If an institution issues a loan approval subject to the applicant's meeting underwriting conditions (other than customary loan commitment or loan-closing conditions, such as a clear-title requirement or an acceptable property survey) and the applicant does not meet them, the institution reports the action taken as a denial.

In other words, once the lender receives an application (as defined by RESPA or TRID) the lender underwrites the application request, makes a decision to approve the application and THEN verification occurs. Therefore, when you combine RESPA, Truth in Lending and HMDA together:
1. A lender must underwrite what they know after receiving an application and either denies the request, makes a counter-offer or makes a conditional approval;
2. If the request is conditionally approved, the lender is to deliver meaningful GFE/P-TIL or LE demonstrating what you can offer the applicant;
3. The lender waits to receive an Intent to Proceed from the applicant;
4. Orders verifications to confirm all information, and
5. Makes further decisions and a final decision as verifications are received.

Many of you seem to be saying “That’s not how we do it at our bank” or “Our procedure is to make a decision after we receive all verifications” or other variations of these. We acknowledge that lenders can have different procedures. However, a lender’s procedures must conform to the regulation.

While these are all different regulations, they must still all work together. We’re not saying you can take a term from one regulation and apply it similarly to another. We are saying in order to comply with one regulation you can’t violate another. You won’t find the terms “conditional approval” in RESPA or TRID. However, HMDA uses them and clearly explains a withdrawal can’t occur after a conditional approval is granted (see the FAQs). [I absolutely agree with the last statement in Truffle's last post!]

If you don't believe issuing preliminary disclosures requires a decision, ask yourself “when do we make a decision and how do we document when that decision is made?” HMDA requires all information on the LAR to be supported. How would you support the date of “withdrawal” or “approved not accepted” consistently if you don’t have a way to clearly substantiate when that occurs? Examiners could (and many have) cite your bank for incorrect action codes.

That's certainly enough to chew on for a while. I believe I've supported well why we believe this to be true. I'll be interested in your comments. But as I said in the beginning of this post, I think you can see where we aren't lining up now.

One regulation at a time:

RESPA

Step 1. I do not underwrite a loan and I issue a GFE.

Step 2. I abide by the estimates that I gave the customer on the GFE and only make any changes allowed if I have a change in circumstance.

Step 3. I wait until I get everything I think I need to underwrite the loan and make a deny/approve decision.

Step 4. If approved, I close the loan with no tolerance violation.

Tell me where I am required to underwrite the loan in this scenario and how my GFE would have not been issued in good faith? I can give you estimated fee and charges and abide by them (all my banks do without underwriting) and as long as I close the loan with the fees as disclosed - tell me again how RESPA forces me to underwrite??

But again, that is a RESPA argument and it has no place in the HMDA discussion.

HMDA:

You still have not addressed the issue that even if you thought you needed to do some preliminary underwriting to come up with your GFE estimate, how that underwriting gets you to the point that the only thing left is "solely customary commitment or closing conditions".

In theory: If you issue a GFE without evaluating (underwriting) what the applicant presented (your step #1), and you are able to abide by the estimates that you gave (your step #2), and you close the loan with no tolerance violations (your step #3), then you're fine. But you know that's not prudent. I would think that you of all people - being a former banker for numerous years - know that this is not how it happens in the real world. If you were able to live by what you said even though you didn't underwrite the loan information you had before issuing the GFE, I don't think you would be in violation.

You continue to say: Tell me where I am required to underwrite the loan in this scenario . . .
GFE's are not to be hypothetical. RESPA requires GFE's to BINDING and MEANINGFUL - specific to the applicant. HUD even knows this and said they believe the 6 items and credit report would be enough for "originators need to properly underwrite" an application. There it is again. You continue to ignore this from the section by section analysis that I continue to quote.

As I said in my last post (and supported with regulatory references):
The income and credit report are used to calculate a debt-to-income ratio. The estimate of value and loan amount is used to calculate a loan-to-value ratio. The score from the credit report can also be evaluated. All of these things are used/relied on to give the best GFE/LE possible – a meaningful and binding GFE/LE. Lenders are not to issue a “standard estimate of charges” as it would not meet the “good faith” requirement.

So if an applicant had a credit score of X and a DTI ratio of XX% and a LTV ratio of XX%, the lender offers rates and fees based on this. But if the applicants creditor score was Y, DTI of YY% and LTV of YY%, they offer a different rate/fees on the GFE/P-TIL. How is that not underwriting?

You might have a different definition of "underwriting" in your mind, but why would you not evaluate all of that (plus anything else the applicant provides) prior to issuing the GFE? If you are, I think you're underwriting the loan request. If you don't think I am, show me where it says that this is not underwriting and where it says underwriting occurs.

Then you say: But again, that is a RESPA argument and it has no place in the HMDA discussion.
It absolutely does. Since when do you work in a vacuum? The point is that HMDA references conditional approvals. If you make one, you can't say the application is withdrawn after that. Agree?

I'm pointing to RESPA to show that when you issue a GFE, you have made a conditional approval. How is that not a valid point in this process? In the real world, lenders work with all regulations simultaneously. In the real world Congress and the regulators wrote laws and regulations that must coincide. You can't copy with one law and violate another.

I don't understand what you mean by this statement:
You still have not addressed the issue that even if you thought you needed to do some preliminary underwriting to come up with your GFE estimate, how that underwriting gets you to the point that the only thing left is "solely customary commitment or closing conditions".
I recognize that wording from the HMDA FFIEC FAQs, but I don't follow what you are asking.

You also continue to ignore the question that I've asked 3 times in this string:
If you don't believe issuing preliminary disclosures requires a decision, ask yourself “when do we make a decision and how do we document when that decision is made?” HMDA requires all information on the LAR to be supported. How would you support the date of “withdrawal” or “approved not accepted” consistently if you don’t have a way to clearly substantiate when that occurs?

I have a few comments and then I will retire (again) from this thread.

1. Why would a bank go to the time and expense of underwriting applications for which it has not received an intent to proceed? That could become exceedingly expensive and is certainly not prudent.

2. It is also not prudent to allow those processing an application just received, who are not experienced or authorized to make credit decisions, to determine if an application can go forward within the first three days. Even if there is a low score, there may very well be mitigating factors that an experienced person who will see the application further in the pipeline will notice and factor into the decision.

3. Documenting a file with memos/notes in the system is (or should be if a bank does not do so) standard to memorialize when a customer contacts the bank or when the bank contacts the customer. Calls or emails/letters from the customer withdrawing the application (or any such action) need to be documented...a note written, a copy in the file. The same is true for any communications out to the customer. To not do so is foolish. The bank/financial institution needs to do so to protect itself. Banks I have worked at over the years required this to be done. Properly documented, there will always be "proof" of what happened when. A bank without those processes in place is taking unwarranted risk.

"1. Why would a bank go to the time and expense of underwriting applications for which it has not received an intent to proceed? That could become exceedingly expensive and is certainly not prudent."

What is time-consuming and expensive about calculating DTI and LTV based upon the information the applicant has provided and information from the credit report? Most software systems do that automatically once the information has been entered. At that point, it's either "this looks good, let's go forward" (conditional approval) or "the numbers don't work, so at this point it a denial."

And then, David's comment.....which is currently my problem.



In the real world of banking...and I have been in it for 38 years....getting lenders to completely and uniformly document files with complex stuff such as this, is all but impossible. To add to that....I scrub the LAR. And I am one of those "inexperienced" people. I have never been a lender, but I know that underwriting is quite complicated. It's up to me to verify whether the decision WD or ABNA is appropriate. Most of the time, I really have no idea.

swiggles, speaking for my world only, Kathleen's second point answers your question about her first point. Granted, we don't have LOs generating GFEs or underwriting loans.

As Randy said so well back on the first page of this thread: We have very few files that are not properly documented. Thankfully, quarterly scrubs bring them to light while they're still fresh and easy to fix.

Wow. Wouldn't it be just that simple to say all HMDA loans, if passed the 3 day test, will be originated or ANA. End of discussion. Forget RESPA loans...ALL HMDA loans would have to be treated the same way.

Forget that Reg C states:

Action taken date—approved but not accepted. For a loan approved by an institution but not accepted by the applicant, the institution reports any reasonable date, such as the approval date, the deadline for accepting the offer, or the date the file was closed. Although an institution need not choose the same approach for its entire HMDA sub mission, it should be generally consistent (such as by routinely using one approach within a particular division of the institution or for a category of loans).

This right here gives you the OPTION of using the date the file was closed.

Now, if you don't know whether or not the loan has been approved (underwritten fully or not) then you have a documentation (training) problem.

If you don't believe issuing preliminary disclosures requires a decision, ask yourself “when do we make a decision and how do we document when that decision is made?” HMDA requires all information on the LAR to be supported. How would you support the date of “withdrawal” or “approved not accepted” consistently if you don’t have a way to clearly substantiate when that occurs? Examiners could (and many have) cite your bank for incorrect action codes.

Please refer to Reg C above. Any reasonable means. Now...if my file is properly documented (and they are) no freaking examiner is going to get away with citing me over this issue.

ETA: Yes, you will briefly look over a RESPA application to make sure it appears, on the surface to pass...otherwise you would want to go ahead and deny and not waste further time and effort. But to suggest it's even conditionally approved is not at all accurate...neither is it conveyed or suggested to an applicant.

Yes.

Yes.

And let me give one example. Very high end private banking. An ap might "on the surface" seem not to pass. Further review could uncover all sorts of mitigating factors.

As I said what now seems to be months ago, the brush being used here is too broad. There is no one size fits all.

But to paint the industry one color because some banks can't document files is just too much to take.

Yes.

KB wrote:

1. Because the regulation requires it as I have quoted and supported numerous times.

2. Why would you allow someone who is "not experienced or authorized to make credit decisions" to issue a GFE that is binding and meaningful?

3. I agree. I'm not sure how this doesn't line up with anything I've said.

As I said in my post 2 above, you don't get to make the regulation fit your procedures. A lender’s procedures must conform to the regulation.

Truffle said:

So you let people that don't know loans issue GFE's? That's the problem that HUD was trying to tackle with the 2010 RESPA rules (and now the CFPB with the TRID rules). So you issue disclosures that really don't mean anything?

Joker said:
I can't agree with the statement that it's not a conditional approval. You're not denying it because you decided to approve the application. But that's not a conditional approval? Then what is it?

Joker also said:
Do you know when this was written? PRIOR to the 2010 RESPA rules. When RESPA 2010 was written, it lined up a lot of these things dates.
How can you say it's approved for RESPA (when you issue a GFE), but not approved for HMDA?
"The deadline for accepting the offer" lines up with RESPA's intent to proceed.

No and you don't get to take it to that extreme to support your argument, David.
I'll put my processors up against anyone you chose to give a meaningful GFE.
But they are not underwriters and in my shop, only underwriters have the ability to approve or deny an application.

Again, the brush is too big, David.
Further, the section you keep quoting about 'this approach' refers to a single application process, which does not include underwriting prior to GFE delivery.

And I, for one, would like to apologize to the original poster who has yet to get a good answer to their question. From the sounds of this, you'd best pick the side that works best in your organization and be ready to defend it because I'm sure examiners have gotten wind of this and will be looking for it.

From the original poster.

I really have no idea.

The fact that examiners have never criticized, or really deeply scrutinized whether an application is approved but not accepted or withdrawn (maybe I'm just lucky).....because I don't think examiners have any guidelines either.....makes me want to just continue on the same road.....let the lender tell me what it is and go with it......meaning our entire LAR will consist of withdrawn applications. It is completely rare for an applicant to withdraw subsequent to the final underwriter placing a stamp of approval on the ap.

David....if you get a cfpb response, I would sure like to read it.

No, people who don't know loans do not issue GFEs. The bank has (should have) set up a process to issue accurate estimates based upon the loan applied for. The bank also has (should have) a process that only authorized staff can deny an application.

If you have clients that can't document files, can issue accurate GFEs, allows anyone to deny (or conditionally approve) a loan, then I feel sorry for those banks.

These are all internal control issues, risk management issues. If management continues to reward staff that does not document loan files, then management gets what it deserves when the regulators have findings. Compliance or internal/external auditors should be checking these things (documentation to support coding) and reporting on errors and weaknesses. That should be followed by training that improves the processes.

Truffle: I'm not trying to attack you or your people. You said "we don't have LOs generating GFEs or underwriting loans." That made it sound like you had non-lending specialist preparing the GFE's. That's all I was addressing. Whatever you want to call them, the people that prepare the GFE's are supposed to issue disclosures that are meaningful and binding on your institution. They must be accurate - which I believe requires some sort of underwriting - to ensure that what is offered fits the applicant and their request. This seems to be a mix of terminology as much as anything.


Does HUD offer any other type of "approach"? The approach I describe (and outline in my long post) is the ONLY approach HUD provides to properly comply with their regulation.

I remember in a post a few months ago, where you said you don't allow lenders to evaluate any extra information brought in voluntarily by an applicant (you said you tell them to throw it in a drawer). That is clearly outside of the regulation as it says you MUST consider EVERYTHING available. You don't get that option just like you don't get the option of making the regulation fit your procedures. Your procedures must conform to the regulation.


Most certainly the examiners have gotten wind of this. That's what caused me to research this extensively and write our position paper. The OP's question was answered in the first reply by NSFW. Swiggles then agreed and brought up the position paper that we wrote in our July newsletter - which opened up this discussion.

"Worth going forward" is not a legal term, a regulatory term and not a HMDA Action Code. I think you're making the regulation fit your procedure instead of your procedures fitting the regulatory requirements.

I agree (and have stated many times), future verifications can allow you to change your mind from "conditionally approved" to "denied" or making a counter-offer. But based on what you know, the bank is conditionally approving a loan when they issue early disclosures. Why else would they give the disclosures?

Please, David. It did no such thing. But moving on...

David, when we talk GFEs, extra information is not required nor can it be requested. TRID changes much of that. Assuming that we're talking today's RESPA, not the 10/3 changes, I stand by my statement that we don't look at anything else that comes in nor are we required to. We're giving meaningful GFEs based on the loan product requested and the six pieces of information that trigger an application.

Again I say that the citation you use over and over states in part "...this approach provides the flexibility originators need to properly underwrite,..." but no where does it state that the originator MUST underwrite at this point. That's what we all are saying is our sticking point. If you're going to use RESPA to determine HMDA action codes, then show me where it says I MUST UNDERWRITE prior to delivering the GFE thereby making it impossible for every application I do disclosures on to be marked withdrawn. Frankly, as Randy outlined earlier, there's much more specifically defined in Reg C/HMDA to support my stance than I find in RESPA to support yours.

I'm confident I'm not making regulations fit my procedures at all so I will continue to mark withdrawn if the borrower contacts me to do so prior to an underwriting decision being made, even after disclosures are given. Of course, if the CFPB (not HUD anymore) comes forward with something dictating otherwise, I'll have to change.

"If the applicant expressly withdraws before satisfying all underwriting or creditworthiness conditions and before the institution denies the application or closes the file for incompleteness, the institution reports the action taken as application withdrawn. If all underwriting and creditworthiness conditions have been met, and the conditions are solely customary commitment or closing conditions and the applicant expressly withdraws before the covered loan is originated, the institution would report the action taken as application approved but not accepted."

Randy has pointed out this FAQ a few times, but I haven't seen anything from David to dispute it. In no way does issuing disclosures qualify as "satisfying all underwriting and credit conditions". Isn't a satisfactory appraisal an underwriting condition? And you can't have that info prior to issuing disclosures.

TMatt: What Randy has quoted is from a proposal. It's not here, nor is it final. If it comes - without change - we'll need to evaluate the position we've taken and may be able to use that information to debate with examiners. But you truly want me to argue a position based on a proposed comment?

Truffle: You said "I stand by my statement that we don't look at anything else that comes in nor are we required to. So please tell me what this statement from the 2008 Federal Register means:

None of the information collected by the originator prior to issuing the GFE may later become the basis for a ‘‘changed circumstance’’ upon which a loan originator may offer a revised GFE . . . [Federal Register 11-17-08].

We don't collect anything beyond the six items that make an application, David. If we did, we wouldn't use it as a basis for a 'changed circumstance'. In fact, I'd go so far as to say that the vast majority of any redisclosures we do are at the request of the borrower, not something we as the loan originator initiates.

David, I suggest you read the FFIEC FAQs one more time. THe CFPB has said nothing more than what is and has already been there for years. What the CFPB did in their HMDA proposal was to expand on that definition.

Action Taken

Conditional approvals---customary loan-commitment or loan-closing conditions. The commentary indicates that an institution reports a "denial" if an institution approves a loan subject to underwriting conditions (other than customary loan-commitment or loan-closing conditions) and the applicant does not meet them. See comment 4(a)(8)-4. What are customary loan-commitment or loan-closing conditions?

Answer: Customary loan-commitment or loan-closing conditions include clear-title requirements, acceptable property survey, acceptable title insurance binder, clear termite inspection, and, where the applicant plans to use the proceeds from the sale of one home to purchase another, a settlement statement showing adequate proceeds from the sale. See comments 2(b)-3 and 4(a)(8)-4. An applicant's failure to meet one of those conditions, or an analogous condition, causes the application to be coded "approved but not accepted." Customary loan-commitment and loan-closing conditions do not include (1) conditions that constitute a counter-offer, such as a demand for a higher down-payment; (2) underwriting conditions concerning the borrower's creditworthiness, including satisfactory debt-to-income and loan-to-value ratios; or (3) verification or confirmation, in whatever form the lender ordinarily requires, that the borrower meets underwriting conditions concerning borrower creditworthiness.

http://www.ffiec.gov/hmda/faqreg.htm#action

^^^This^^^

Truffle: In an 4/30/15 post on this same topic I posted 4 questions. #3 was:
3. Sometimes applicants bring in verification documents (tax returns, W-2's, etc.) at the time of application. Do loan officers review these documents before preparing the GFE?

You responded:
3. absolutely not. I tell my LOs to stick it in a drawer until the GFE is delivered and intent received.

Here's a link to the discussion if you want to verify:
http://www.bankersonline.com/forum/ubbthreads.php?ubb=showflat&Number=2010268&page=2

In the post you made yesterday afternoon, you stated: "I stand by my statement that we don't look at anything else that comes in nor are we required to." and I'm trying to point out that you MUST look at everything you have avaialbe before issuing the GFE. The RESPA rules state "None of the information collected by the originator prior to issuing the GFE may later become the basis for a ‘‘changed circumstance’’ upon which a loan originator may offer a revised GFE . . ." [Federal Register 11-17-08].

Randy: First, my apologies for not responding to your questions previously. I hope you can understand that this whole discussion can be very time consuming. Here's my response concerning the FFIEC FAQs and how I interpret them:


This FAQ you quote, uses the term “underwriting conditions” but then separates them into two groups. Let’s separate the terminology used in this FAQ:

1. Loan-commitment / loan–closing conditions
These include things like clear title requirements, surveys, and inspections. They are “property conditions” that are not typically ordered before issuing a GFE/LE. If one of these specific conditions becomes a factor that does not allow the loan to be closed, this HMDA FAQ makes it clear to report the application as “approved not accepted”.

These do NOT include “creditworthiness conditions” (see below) as stated in the last sentence of the FAQ.

2. Creditworthiness conditions
These are applicant conditions and include things such as loan-to-value, debt-to-income, etc. Due to the applications requirements found in RESPA (6 required pieces of information) and the allowance to order a credit report, the lender can review and evaluate these conditions (although unverified) prior to issuing the GFE/LE. For example, RESPA requires the loan amount and an estimate of the value of the property; thus, a lender can determine a loan-to-value.

This FAQ says a lender has made a “conditional approval” subject to loan-commitment or loan-closing conditions. It also implies creditworthiness conditions (LTV, DTI) should have already been completed and therefore a credit decision (approval) was made. That’s also why the Commentary states you deny a request when a loan is approved and then the verification of creditworthiness conditions don’t match what was previously relied upon (via W-2, tax return, appraisal value, etc.). If the consumer’s creditworthiness doesn’t “check out”, the application is coded as a denial.

Now, let’s look at the second FAQ.

Conditional approvals---failure to satisfy creditworthiness conditions. How should a lender code "action taken" where the borrower does not satisfy conditions concerning creditworthiness?

Answer: If a credit decision has not been made and the borrower has expressly withdrawn, use the code for "application withdrawn." That code is not otherwise available. See Appendix A, I.B.1.d. If the condition involves submitting additional information about creditworthiness the lender needs to make a credit decision and the applicant has not responded to a request for the additional information in the time allowed, use the code for "file closed for incompleteness." See Appendix A, I.B.1.e. If the borrower has supplied the information the lender requires for a credit decision and the lender denies the application or extends a counter-offer that the borrower does not accept, use the code for "application denied." If the borrower has satisfied the underwriting conditions of the lender and the lender agrees to extend credit but the loan is not consummated, then use the code for "application approved but not accepted."

For example, if approval is conditioned on a satisfactory appraisal and, despite notice of the need for an appraisal, the applicant declines to obtain an appraisal or does not respond to the lender's notice, then the application should be coded "file closed for incompleteness." If, on the other hand, the applicant obtains an appraisal but the appraisal does not support the assumed loan-to-value ratio and the lender is therefore not willing to extend the loan amount sought, then the lender must use the code for "application denied."

The example given in the last paragraph is outdated (customers can't order their own appraisals anymore), but the point being made is the same. If the lender has not made a decision and the borrower withdraws, it is coded “application withdrawn”. But once a decision is made (even a conditional one) that action code (withdrawn) can’t be used. This FAQ clearly states a withdrawal cannot be made after a conditional approval has been made, regardless of the applicant’s actions. It can either be "denied", “closed for incompleteness”, the lender can issue a counter-offer, or it could be coded "approved, not accepted."
-----------------

I'll admit I haven't studied the comments you post from the HMDA proposal. Why? Because they are only proposed and even if accepted are at least 17 months away from being effective. I will agree with you that there are many things in the HMDA proposal that really aren't changed, but clarifications that I think can be used in today's interpretations. I'll look at those comments more in light of my position.

I suggest that unless somebody has something new to add, maybe we should just table this topic. There's no doubt in my mind that we're all smart people and passionate about our industry. What I'm seeing lately is the same topics being brought up with the same responses. I feel like some of this has become personal and argumentative instead of just a healthy discussion. I'm not trying to run from anything but don't think I have anything more to add.

As I stated before, we have submitted our position paper to the CFPB. IF we get a response, I will let everyone know. I'm not optimistic we will however. We tried to get their input before writing the position paper and basically were told by the HMDA people "we're not RESPA or TRID experts" and we were told by the TRID experts "That's a HMDA question."

Not that any of you care for my opinion....or really anything else I have to say.....but I appreciate this discussion and have enjoyed reading it. I'm glad all of you provide your opinions on BOL as the opinions are very valuable to me.

I do have 2 straight-forward questions, David. Since you covered Q2 above, I'll skip that...but what was your point or reasoning for this wording [see below]:

Here's how the article closed. If you disagree, you need to consider this:
Our position is really a “bright line test”. If you accept this position, you can’t go wrong. An examiner is not going to cite a bank for saying, “when we issue preliminary disclosures, we believe we have made a conditional approval. Therefore, if the applicant doesn’t proceed or wants to withdraw after this action but prior to an intent to proceed, we code it ‘approved, not accepted”. We believe it is a clear and consistent approach and is well documented and supported.


Your stance has been that you have a "conditional approval" before you issue your GFE. What would it matter then, about the time period between issuance of disclosures and 'intent to proceed'...that would be moot, would it not? Why is it referenced that way?

And one last question I asked earlier that I don't believe was ever addressed. This entire position is based on RESPA. What about all of the HMDA reportable loans that do not fall under RESPA? If a bank's problem is documenting when an approval is actually done/happens/whatever, how are they handling these deals?

One more comment, and I'll be done (I think!). In order for a borrower to know what conditions are being made, you must communicate with them so that they either 1)don't comply and become denied or become and incomplete file or 2) comply and get approved or denied.

Without a written (or otherwise communicated and documented) commitment, conditional or otherwise, you really don't have one. Not really.

And yes, Swiggles...it's been a good debate!

I think I know what you're asking (but not certain). If I miss it with my response, please ask again.

If an applicant doesn't give an intent to proceed, it can easily be seen as the applicant withdrawing the request. Either in this string (or the previous similar discussion in March/April), some were saying they consider it "withdrawn" until the borrower gives the intent to proceed. I'm trying to emphasize that approval is not in the hands of the borrower. It's in the hands of the lender. It occurs when the lender (or decision maker) decides they can make this loan and prepares a GFE/P-TIL to represent the "offer".

Your point (I think) that it could continue to be "approved, not accepted" after intent to proceed is received is true. It depends on what happens. It could also be a denial, closed for incompleteness or counter-offered (which may result in a denial, as well). But if you issue a GFE/P-TIL & don't get Intent to Proceed from the applicant, I think the only action code is "approved not accepted". The lender shouldn't be issuing a GFE and then denying, counter-offering or saying "closed for incompleteness" during that window of time.


(Again, I THINK I know what you're saying - but please reply if I'm missing this)

Most banks issue a cover letter with the early disclosures. The letter explains what's enclosed (early disclosures) and what needs to happen next (borrower provide a deposit, verifications, etc.). I don't think you have to call that a written commitment, but it serves as the written communication you describe.

This is the area that I think most people are getting hung up on. Some say "that's not how we do it at our bank". While that may be true, my point has been I don't think you have to communicate "approved" to the applicant. You have internally given a conditional approval when you issue the disclosures.

I'll close with this statement again:
If you accept this position, you can’t go wrong. An examiner is not going to cite a bank for saying, “when we issue preliminary disclosures, we believe we have made a conditional approval." What's the risk of taking this position? it is a clear and consistent approach and is well documented and supported. The problem with any other position seems to be "when do you consider the application "approved"? That's the whole crutch of our position.

Good discussion!

Thanks David, I think you interpreted what I was asking correctly.

I can see how your closing argument would be hard for an examiner to take issue with simply because the bank would be stating 'we believe we have made some sort of 'approval'. They really cannot argue that....even if it's not true.

We report based on facts. You are either 1) denied 2) approved or 3) in process and yet to be determined.

I still would like for you to opine on the following question, based on your position paper:

And one last question I asked earlier that I don't believe was ever addressed. This entire position is based on RESPA. What about all of the HMDA reportable loans that do not fall under RESPA? If a bank's problem is documenting when an approval is actually done/happens/whatever, how are they handling these deals?

Sorry Joker. I didn't mean to ignore your last question. I got caught up in the other responses and forgot to include a response to this. I did, however, answer this (or so I thought) on a 7/30th post (#2030961) where I said:

No where have I ever stated (or implied) a bank has 3 days to make a decision. Non-RESPA loans don't have this requirement and therefore wouldn't be subject to this.

What I am saying is RESPA requires a bank to give a "binding and meaningful" GFE. RESPA also requires a bank to underwrite a loan PRIOR to issue a GFE (as I have quoted from the Federal Register). Therefore, RESPA applicable loans require underwriting (a conditional approval) when GFE's are issued.

You are correct with your last sentence about documenting when an approval is actually done - no matter if it is subject to RESPA or not. Banks will need to have a way to demonstrate this in all loans subject to HMDA.

"RESPA also requires a bank to underwrite a loan PRIOR to issue a GFE (as I have quoted from the Federal Register). "

You have not quoted that from the Federal Register, so I really wish you would quit saying that. It may be how you have misinterpreted one single sentence in a 86 page Federal Register release, but HUD did not and has never said that and neither has the CFPB. Please stick with the facts.

Agree. Totally taken out of context, which may be why the Bureau took away that 'extra flexibility' when drafting TRID. We get limited, basic info but yet must still produce meaningful disclosures. It's on us to be able to do that and it has nothing to do with a time frame. Disclosure are due out in 3 days. We can't do much else until Intent is received. Unless it's a slam dunk denial, it's just a loan in process.

Wow Randy! We can have a healthy discussion (and thought we were), but when you say “you have not quoted the Federal Register”, I interpret that as calling me a liar. You can have your opinion and don't have to agree with me (I've made that clear). Once again, I will quote from the FR and NOT from just one single sentence, so that you can see exactly what it says. You can look all these up yourself here:
http://www.gpo.gov/fdsys/pkg/FR-2008-11-17/pdf/E8-27070.pdf

From page 28211 of the 11/17/08 FR:
Industry representatives questioned HUD’s legal authority to: limit information originators can request to underwrite a loan;.
This is referring to the information (limited) prior to issuing the GFE.

From the same page of the FR:
In order to prevent over burdensome documentation demands on mortgage applicants, and to facilitate shopping by borrowers, the final rule specifically prohibits the loan originator from requiring an applicant, as a condition for providing a GFE, to submit supplemental documentation to verify the information provided by the applicant on the application. Loan originators, however, can require applicants to provide such verification information after the GFE has been provided, in order to complete final underwriting.

So there's underwriting prior to issuing the GFE, AND there's final underwriting. The terms “underwriting”, “preliminary underwriting” and “final underwriting” are used numerous times in the final rule. They are clearly distinct and different if you’ll read and try to understand it.

Also from page 68212: This is discussing FINAL underwriting after the GFE is issued. It supports the idea that there is preliminary underwriting before the GFE is issued.
After the GFE has been received, the loan originator may collect additional fees needed to proceed to final underwriting for borrowers who decide to proceed with a loan from that originator. As noted, at that time, verification information or any other information could be required from the applicant, such as bank statements and W–2 forms, to confirm representations made by the applicant in the application.

This next paragraph is on the same page of the FR and discusses information collected prior to issuing the GFE and how it can and can't be used as a changed circumstance. This also indicates an originator is to have done underwriting prior to issuing a GFE and then discusses final underwriting when verifications are received:
None of the information collected by the originator prior to issuing the GFE may later become the basis for a ‘‘changed circumstance’’ upon which a loan originator may offer a revised GFE, unless the loan originator can demonstrate that there was a change in the particular information or that it was inaccurate, or that the loan originator did not rely on that particular information in issuing the GFE. A loan originator would have the burden of demonstrating nonreliance on the collected information, but may do so by various means, including through, for example, a documented record in the underwriting file or an established policy of relying on a more limited set of information in providing GFEs.

And again from the same page of the FR. The context is about underwriting the loan prior to issuing the GFE so that the originator doesn't bait-and switch later.
HUD determined that this approach provides the flexibility originators need to properly underwrite, while limiting bait-and-switch methods whereby the originator uses the GFE to draw in a borrower and, after a significant application fee is paid or burdensome documentation demands are made, claims that a material change has resulted in a more expensive loan offering.

If that doesn't line up with your interpretation, read this from page 68265 in the same FR. This CLEARY says the originator has to underwrite a GFE:
• HUD has adopted a streamlined single application process for the final rule. The new definition will allow loan originators more flexibility in determining the information they need to underwrite a GFE.

From page 68264. This is discussing the compliance costs to applicants and originators by adopting this new rule.
It would be reasonable to assume that in the high-application scenario, where there is an increase in preliminary underwriting costs, that the cost of an initial credit report would be passed on to all applicants.

From page 68266. This quote not only supports our position that the originator must conduct a preliminary underwriting, it also explains how the fees on the GFE must be "firm".
Some lenders were concerned about their ability to produce firm cost estimates (even of their own fees) within a three-day period, given the complexity of the mortgage process. Lenders wanted clarification on their ability to make cost adjustments as a result of information they gain during the full underwriting process.

From page 68269 – a discussion about underwriting costs to prepare the GFE:
A second criticism of the analysis of the compliance cost of the GFE is that HUD does not consider the possibility that the rule could increase the administrative costs to loan originators by generating a greater demand for GFEs. Although HUD believes that it is just as likely that applications do not increase, HUD has included a sensitivity analysis of compliance costs by the number of applications. (See Section VII.D.2.)
The NAR points to another cost not included in the IRFA: the cost of preliminary underwriting.
You’ll find many more references to the term “preliminary underwriting” and the costs of preparing the GFE in the Compliance Costs Analysis on page 68276.

If you still don’t get it, turn to page 68278 of the FR. There’s a heading entitled “Multiple Preliminary Underwriting”. It states:
Comment. Every application under the new rule requires preliminary underwriting. Since borrowers who shop may seek out multiple GFEs, there will be multiple underwritings. Commenters said this will add to the underwriting burden firms incur today. The National Association of Realtors calculated an additional cost of multiple underwriting at $30 per loan for an application per loan ratio of 2.7.

Response.Every application under the final rule that generates a GFE will require preliminary underwriting in order to come up with an early offer for the borrower. Originators can charge a fee for issuing a new GFE limited to the cost of a credit report. It 
is hoped that the charge for this, if any, would be small enough so that it is not a significant deterrent to effective shopping. But whether or not there is a charge, there are real resource costs associated with preliminary underwriting. The additional cost generated depends on the number of applicants and the number of GFEs they receive. Since every completed loan eventually gets underwritten in full, the additional cost of preliminary underwriting depends mainly on the number of additional times that preliminary underwriting occurs beyond the one associated with the full underwriting that would have occurred under the existing scheme.

Read the first sentence of the Comment again. The new rule “REQUIRES preliminary underwriting”. (Still think I’m misquoting this?) Then read the 2nd sentence of the Comment. This clearly states that borrowers that shop will be making multiple banks underwrite. If underwriting doesn’t occur until after the GFE is issued and Intent to Proceed is given, why would this be? It wouldn’t! HUD is saying they realize every creditor will be underwriting the application to issue a GFE. Therefore, there could be “multiple underwritings” for the same borrowers that go to multiple creditors.

Read the first sentence of the response again. In order to produce a GFE, the lender is REQUIRED to conduct PRELIMINARY UNDERWRITING. Then it says (in the 6th sentence), there is “full underwriting”. This paragraph makes it very clear that the current regulation (prior to the 2010 change didn’t require preliminary underwriting (see the last sentence). The new RESPA does require underwriting prior to issuing the GFE.

Want more proof? Page 68279:
Under the final rule, preliminary underwriting should decrease the number of applications that go to full underwriting (e.g., an applicant may be denied during the preliminary without having been charged for an appraisal); that is, some of the 8.75 million that are not originated may be disapproved at the preliminary stage rather than going through full underwriting (as they might today). This savings in appraisal, verification, and other incremental underwriting costs that are avoided would tend to offset the increase in cost resulting from the extra preliminary underwriting noted in the above paragraph.

There are 10 places on this page alone where the terms “preliminary underwriting” or “additional underwriting” are used referencing evaluations before the GFE is issued. They also use the term “final underwriting” or “full underwriting” 6 times on this page of the Federal Register to refer to evaluations after the GFE is issued.

So Randy & Joker. Do you think I’m still missing it? Do you think I’ve taken this out of context? Do you really think RESPA doesn’t say originators need to underwrite before issuing a GFE? If so, what does all of these mean? Please reply and let me know your thoughts.

Moving to TRID: This only gets more concrete. As Joker said, the CFPB is taking away even more flexibility, which makes my position even stronger. Originators will need to assume things and underwrite based on what they have and their assumptions prior to issuing the LE.

I can't wait to hear from you.

Yes, I do still think you are missing it. Here is a blurb that to me, explains the difference fairly well. It comes from 68211 and differentiates between underwriting an 'application' vs underwriting a 'mortgage'. A preliminarily approved 'application' simply means the type of loan requested is something that can be provided and disclosed...and then once intent is received can be moved into the 'mortgage/loan' end of things. That is the difference if you ask me.

Industry representatives also
requested that HUD permit borrowers to
expedite the application process and
proceed to the mortgage application
stage, when the borrower so desires due
to timing or other concerns.
Industry representatives stated that
the new application definitions in the
March 2008 proposed rule would
present uncertainty in complying with
other mortgage-related statutes and
regulations. They commented that
compliance with other statutes and
regulations is triggered by a mortgage
‘‘application.’’ Because HUD’s proposal
included both a ‘‘GFE Application’’ and
a ‘‘Mortgage Application,’’ they
commented that it is not clear which
one is the ‘‘application’’ for purposes of
compliance with other regulations. In
particular, lenders expressed concern
with the possibility that the ‘‘GFE
Application’’ would trigger compliance
obligations under FCRA, ECOA, HMDA,
and the TILA requirements. They
requested that ambiguities surrounding
compliance with these statutes and
other laws be addressed to provide
clarity and mitigate litigation exposure.
For example, one lender noted that to
calculate the spread for high-cost loans
under Regulation Z and many state
predatory lending laws, the index used
is based on the month in which the
‘‘application’’ for credit is received by
the creditor. This lender stated that it
was not clear from the proposed rule
whether the GFE application is an
application for purposes of Regulation
Z.

Missed part!

Industry commenters expressed
confusion about preamble statements
regarding whether HMDA or ECOA is
triggered by the GFE Application. They
indicated that the preamble stated that
whether HMDA or ECOA is triggered by
the GFE Application should be
determined under Regulations C and B,
as interpreted by the Board. They noted,
however, that the preamble stated that
based on consultations with the Federal
Reserve Board, TILA disclosures would
be provided within 3 days of a written
GFE application unless the creditor
determines that the application cannot
be approved on the terms requested.
The commenters further noted that the
Regulatory Impact Analysis states ‘‘[t]he
proposed rule clarifies that only the
mortgage application would be subject
to Regulations B (ECOA) and C (HMDA),
which is the current situation today.’’
These commenters requested
clarification of this matter.
Industry representatives questioned
HUD’s legal authority to: limit
information originators can request to
underwrite a loan; require that
originators accept an abbreviated
application from which to complete a
GFE; require a new GFE when a
counteroffer is made; and require a
consumer to be notified within one
business day of a lender’s decision to
reject an application, among other
concerns.
Additionally, one lender commented
that under HUD’s March 2008 proposed
rule, lenders would be required

and this may be further clarification.

Once the applicant submits to the
loan originator all the mortgage
application information deemed
necessary by the loan originator to
process the GFE, the originator will be
required to deliver or mail a GFE to the
applicant within 3 business days. HUD
is now also limiting the fee that may be
charged for providing the GFE,
consistent with the Federal Reserve
Board’s recently finalized rule limiting
the fees that consumers can be charged
for the delivery of TILA disclosures.
After the GFE has been received, the
loan originator may collect additional
fees needed to proceed to final
underwriting for borrowers who decide
to proceed with a loan from that
originator. As noted, at that time,
verification information or any other
information could be required from the
applicant, such as bank statements and
W–2 forms, to confirm representations
made by the applicant in the
application.

Joker: What you are quoting from (page 68211) in your first paragraph of your first post is addressing the proposed rule. The FR is saying they didn't accept this because of the issues raised. Then the FR goes on to state what they did put in the final rule. Read the section called "HUD Determination" on the bottom of the 2nd column of the FR on page 68211.

Did you read the quotes I gave from pages 68278 and 68279? Where it states "Every application under the new rule requires preliminary underwriting."?

There's a lot on the table right now. Let's discuss one thing at a time. I'm simply trying to show you where it says the originator must underwrite what they have before issuing the GFE - that's why you and Randy challenged me today. We can get to how this applies to HMDA later.

Your second post references "final underwriting". I'm not disputing that. I'm stating that the originator must do preliminary underwriting before issuing the GFE. THEN they do final underwriting when verifications are received.

I get it David, but the concept truly didn't change. Preliminary underwriting of multiple applications as compared to loans is discussed heavily in the section you refer to. Obviously you are going to weed out some in that stage due to poor credit or terms that are not available. However, that is not the same as the credit "full" underwriting process to arrive at an 'approved' status, subject to nothing more than closing conditions.

I will end this with the following from the final rule section:

As noted, this public policy is further
supported by the Federal Reserve Board
through its recently issued final rule
limiting fees that can be charged for the
delivery of the TILA disclosure. Under
this rule, borrowers must receive the
TILA disclosure before paying or
incurring any fee imposed by a creditor
or other person in connection with the
consumer’s application for a closed-end
mortgage, except that creditors may
charge a bona fide and reasonable fee for
obtaining the consumer’s credit history.
Whether an application under a
particular set of facts triggers ECOA or
HMDA requirements must be
determined under Regulation B or
Regulation C, as interpreted by the
Federal Reserve Board’s Official Staff
Commentary.

Let's just say there is a world of difference between a 'preliminary approval' vs an 'approved loan' subject to only closing conditions.

that is not the same as the credit "full" underwriting process to arrive at an 'approved' status, subject to nothing more than closing conditions.
I've never said it was. In fact, I've stated this preliminary underwriting is a "CONDITIONAL approval" - not FINAL approval and that's the tie to HMDA.

The quote in your last post basically says an application can be different for Reg B & Reg C. I don't disagree. Again, my only point is that RESPA clearly states a lender must do a preliminary underwriting prior to issuing a GFE. If they do, I think it's very easy to say that they have given conditional approval. If they have, HMDA says, you can't use "withdrawn" anymore.

Let's just say there is a world of difference between a 'preliminary approval' vs an 'approved loan' subject to only closing conditions.
Let me chew on that. I'm not disagreeing. I don't know that I'll get back to this today.

Good discussion Joker!

I'm still interesting in hearing from Randy - whether he thinks the 2008 FR doesn't require underwriting prior to issuing a GFE.

Joker: Back to your comment saying: Let's just say there is a world of difference between a 'preliminary approval' vs an 'approved loan' subject to only closing conditions. I agree there is a world of difference between preliminary and final approval. Let me address that and the "closing conditions" part of your comment.

I hope you're seeing that RESPA does require a lender to do a preliminary underwriting of the loan request as I showed with the numerous quotes from the 2008 FR. Probably the strongest being:
Every application under the final rule that generates a GFE will require preliminary underwriting in order to come up with an early offer for the borrower.

If you agree with that, then I hope you can understand my interpretation that this is a conditional approval. IOW, the purpose of underwriting the request (as required) is to arrive at a decision to make or not make the loan. Again, I'm not saying it's a final approval - it's a preliminary (conditional) approval.

I also agree that RESPA and HMDA have different definitions of application, as you were pointing out in your last post. I think it's very clear that if you have an application for RESPA, you most certainly have an application for HMDA (assuming it meets the collateral and purpose components of HMDA). I can't think of one scenario where I would say "I have an application for RESPA, but don't for HMDA", based on the information given by the applicant. Maybe you can. So let's look at the HMDA FAQs again:

Conditional approvals---failure to satisfy creditworthiness conditions. How should a lender code "action taken" where the borrower does not satisfy conditions concerning creditworthiness?

Answer: If a credit decision has not been made and the borrower has expressly withdrawn, use the code for "application withdrawn." That code is not otherwise available. See Appendix A, I.B.1.d. If the condition involves submitting additional information about creditworthiness the lender needs to make a credit decision and the applicant has not responded to a request for the additional information in the time allowed, use the code for "file closed for incompleteness." See Appendix A, I.B.1.e. If the borrower has supplied the information the lender requires for a credit decision and the lender denies the application or extends a counter-offer that the borrower does not accept, use the code for "application denied." If the borrower has satisfied the underwriting conditions of the lender and the lender agrees to extend credit but the loan is not consummated, then use the code for "application approved but not accepted."

For example, if approval is conditioned on a satisfactory appraisal and, despite notice of the need for an appraisal, the applicant declines to obtain an appraisal or does not respond to the lender's notice, then the application should be coded "file closed for incompleteness." If, on the other hand, the applicant obtains an appraisal but the appraisal does not support the assumed loan-to-value ratio and the lender is therefore not willing to extend the loan amount sought, then the lender must use the code for "application denied."

The example given in the last paragraph is outdated (customers can't order their own appraisals anymore), but the point is the same. If the lender has not made a decision and the borrower withdraws, it is coded “application withdrawn”. But once a decision is made (even conditional), that code can’t be used. It can be "denied", the lender can issue a counter-offer or it could be coded "approved, not accepted."

So let's put it all together:
1) You have to underwrite the request to issue the GFE,
2) If you decide to give a preliminary approval, I think that's equivalent to a conditional approval for HMDA.
3) HMDA says if you've given a conditional approval, you can't call the application "withdrawn" from that point on.

THAT is what I've been trying to say since day 1. If I understand you last few posts correctly, you're struggling with my logic of the term "conditional approval" as used in HMDA based on the preliminary underwriting requirement in RESPA. If this doesn't make sense to you now, please let me know why and where the logic breaks down.

Once again, I know this is a paradigm shift in thinking - it was for me too. But the more I look at it, the more it actually makes sense.

Randy: Still waiting to hear from you. Not trying to be argumentative, but your last post said RESPA doesn't require underwriting prior to issuing a GFE. Do you agree that it does?

Okay, I re-read the entire section entitled: B. Changes to Facilitate Shopping
1. New Definitions for ‘‘GFE
Application’’ and ‘‘Mortgage
Application’’ beginning on page 68210 of the 2008 FR.

I still do not draw the same conclusion that you have, David. It depends on just how far you go (or don't go). All they did was get rid of the bifurcated application process.

It all still goes back to the fact that you must use those few pieces of information along with (or not) a credit report in order to do the piece of underwriting that determines:

1) that the application is for a consumer closed-end loan secured by a 1-4 family residence
2) that the loan request will be for the purchase, refinance or home equity-type of loan

so that the lender can verify that the application is for a type of loan that the originator actually originates.

Once that is determined, along with the credit score/report, if needed, to determine appropriate rates/terms to be offered, you have enough information to generate a meaningful GFE of which cannot be changed later based on any of that early information, while at the same time not overly burdening the applicant while they are in the 'shopping' phase.

The "underwriting" could stop at that exact spot. There is no mandate that the loan be 'preliminarily approved at that point subject to verification.

You couldtake it to that level...but you certainly don't have to. You just have to make certain that you know what type of loan John Doe wants and that you have that product available in order to generate a GFE that is accurate.

Beyond that, you'd better live with what you offered unless it changes due to something later on that has nothing to do with the early required information.

We do it daily, but I can guarantee you...other than perhaps running a DTI off of stated income and calculating a probable LTV...that's all we do that early in the game. And all of that is in an effort to generate an accurate GFE...not to conditionally (or otherwise) approve a borrower.

However, we do strive to deny a borrower during that period if it's a request that just won't fly.

I truly believe it is 'lender specific' as to just how far you go, or need to go, during this early stage.

You are correct that they got rid of the bifurcated application process.

I'm confused about what you're not following in my conclusion. It sounds like you agree that a lender must conduct a preliminary underwriting but disagree about the level of underwriting that must be performed. Is that correct?

If so, here's a portion of the long post I made on 8/1st - earlier in the string:

The RESPA final rule in the November 17, 2008 Federal Register states:
HUD has adopted a single application process for the final rule. Under this approach, at the time of application, the loan originator will decide what application information it needs to collect from a borrower, and which of that collected application information it will use, in order to issue a meaningful GFE. However, before providing the GFE, the loan originator will be assumed to have collected at least the following six items of information: the borrower’s name, Social Security Number, and gross monthly income; the property address; an estimate of the value of the property; and the amount of the mortgage loan sought. The borrower’s Social Security Number would be collected for purposes of obtaining a credit report.

Furthermore, the loan originator is presumed to have relied on the borrower’s name, the borrower’s monthly income, the property address, an estimate of the value of the property, the mortgage loan amount sought, and any information contained in any credit report obtained by the loan originator before providing the GFE. [/i]

Let me stop and give you my opinion on this:
This tells us lenders are expected to USE and RELY ON the 6 items and the credit report before issuing the GFE. How? The income and credit report are used to calculate a debt-to-income ratio. The estimate of value and loan amount is used to calculate a loan-to-value ratio. The score from the credit report can also be evaluated. All of these things are used/relied on to give the best GFE possible – a meaningful and binding GFE.

Again, do you agree with this part so far?

The same premise will hold true under the Integrated Disclosure requirements, as the Commentary to §1026.19(e)(3)(iv)(A), #3 states “a creditor is presumed to have collected these six pieces of information . . .. ” In a separate discussion on the extent of credit information to be obtained, the Final Integrated Disclosures Rule in the December 31, 2013 Federal Register states, "creditors should be able to receive sufficient information from the credit report…to develop a reasonably accurate estimate of costs…."

From reading your post, I think we're in agreement so far, but it's the "preliminary approved" part that separates us. Is this right? Let me stop there and see if you agree with this part or where we might differ.

Pretty much, yes. I do agree. Where we are disagreeing has everything to do with actual context.


We are presumed to have the six pieces allowed and cannot later change anything that is based on those six pieces. It does not mean you are conditionally approving a loan. At best it means you are approving the application, itself, as a request for a product that you can disclose to a prospective borrower.

Great. I see where we are in agreement and not. That helps.

In your last post, you said "At best it means you are approving the application . . ."
If you agree you are making an "approval", then why is it wrong to call this a "conditional approval"? I agree you are approving the application. To me that's a CONDITIONAL approval. We're not saying "for sure". We're saying "looks good based on what we have, but there's a lot more to review, so it's approved conditionally." If it is a conditional approval, HMDA says you can't use the action code "withdrawn" anymore.

So it really boils down to the definition of "conditional approval", which is not defined in HMDA. If I understand correctly, we all agree a lender:
1. Must do a preliminary underwriting of the loan prior to issuing the GFE;
2. They use and rely on all of the information they have available at that point to conduct the preliminary underwriting; and
3. No changed circumstances can occur based on this information unless it is later found to be inaccurate.

You don't believe this action is a "conditional approval" and I believe it is. That's really the crux of this whole discussion. Right?

The Commentary to 1003.4(a)(8)-4 states:
4. Action taken--conditional approvals. If an institution issues a loan approval subject to the applicant's meeting underwriting conditions (other than customary loan commitment or loan-closing conditions, such as a clear-title requirement or an acceptable property survey) and the applicant does not meet them, the institution reports the action taken as a denial.

This commentary uses the words "A LOAN APPROVAL" (not final approval) and then says "subject to meeting underwriting conditions. So this approval decision came BEFORE the final underwriting. This ties right into the RESPA requirement to conduct a preliminary underwriting to issue a GFE. It's an approval before final underwriting occurs.

I also believe this meets what HMDA is referring to as a "conditional approval" because the FFIEC's HMDA Action Code FAQ describes what a conditional approval is, without technically defining it. Here is the FAQ:

Conditional approvals---customary loan-commitment or loan-closing conditions. The commentary indicates that an institution reports a "denial" if an institution approves a loan subject to underwriting conditions (other than customary loan-commitment or loan-closing conditions) and the applicant does not meet them. See comment 4(a)(8)-4. What are customary loan-commitment or loan-closing conditions?

Answer: Customary loan-commitment or loan-closing conditions include clear-title requirements, acceptable property survey, acceptable title insurance binder, clear termite inspection, and, where the applicant plans to use the proceeds from the sale of one home to purchase another, a settlement statement showing adequate proceeds from the sale. See comments 2(b)-3 and 4(a)(8)-4. An applicant's failure to meet one of those conditions, or an analogous condition, causes the application to be coded "approved but not accepted." Customary loan-commitment and loan-closing conditions do not include (1) conditions that constitute a counter-offer, such as a demand for a higher down-payment; (2) underwriting conditions concerning the borrower's creditworthiness, including satisfactory debt-to-income and loan-to-value ratios; or (3) verification or confirmation, in whatever form the lender ordinarily requires, that the borrower meets underwriting conditions concerning borrower creditworthiness.

This FAQ uses the term “underwriting conditions” but then separates them into two groups. Let’s separate the terminology used in this FAQ:

1. Loan-commitment / loan–closing conditions include things like clear title requirements, survey, and inspections. These are property conditions that are not typically ordered before issuing a GFE. These do NOT include “creditworthiness conditions” (LTV, DTI, etc.) as stated in the last sentence of the FAQ. If one of these specific conditions becomes a factor that does not allow the loan to be closed, report the application as “approved not accepted”.

2. Creditworthiness conditions are the consumer’s conditions. The lender can review and evaluate the loan-to-value and debt-to-income ratios from the application requirements of the lender (6 items and credit report). They have already been provided and should be reviewed and relied on prior to issuing the GFE - as RESPA requires this as we have already discussed in the exchange about the requirement to do preliminary underwriting.

This FAQ says a lender has made a “conditional approval” subject to loan-commitment or loan-closing conditions. The FAQ also implies creditworthy conditions (LTV, DTI) should have already been completed and therefore a credit decision (approval) was made. Isn't that what you are saying too? That’s also why the commentary states you deny a request when a loan is approved and then the creditworthy conditions don’t match what was previously relied upon. If the consumer’s creditworthiness doesn’t “check out”, the application is coded as a denial.

This has been a good discussion and as I started out saying - we're seeing where we line up and where we disagree. I'll admit, there isn't a clear cut answer. I'm benefitting from the discussion and I hope you (Joker) as well as others are too. Thanks!

Okay, I'm going to take this and break it out from how I see and how most banks I'm aware of operate at this early stage in the game. [Having been a Mortgage Originator for many years prior to the insanity of compliance, I'm also speaking from a hands-on POV].



I hope all that lined up properly and is easy to pick out. We may end up agreeing to disagree on this, but from a banker's standpoint, I stand firmly when I say few banks dole out conditional approvals during the GFE phase. It's just not smart business nor is it practical in that time period to get it to an authority who can make a real decision on the loan...the application's viability is another story.

^^^This and

^^this equals HMDA action taken code at this point will be denied or withdrawn or closed for incompleteness. Originated and approved but not accepted cannot happen yet.

TR, to be accurate, I did add to that last section quoted before I realized you had chimed in to:

Without a written or otherwise communicated commitment or some form of approved status, you do not have an approval in any shape..

Reason being, in a very informal shop, or just based on timing of events, you could have an approval before that approval is communicated to the borrower and if the borrower changes their mind between those two events, it would still be ANA.

Small detail, but an important one.

Our examiners helped us draw the line with the 'communicated' part. The definition we worked out with them came down to if the borrower doesn't know you approved it, they cannot opt to not accept it. So, no communication = withdrawn.

There's a lot to comment on from Joker's post (2 above). However, let me jump to the last comment by Truffle (and I believe Joker agrees with Truffle's comment):

"Our examiners helped us draw the line with the 'communicated' part. The definition we worked out with them came down to if the borrower doesn't know you approved it, they cannot opt to not accept it. So, no communication = withdrawn."

Both of the Action Code FAQs repeatedly state "a decision has been made". They never say "and it is communicated to the applicant". I think saying "we haven't communicated that to the applicant is a leap that isn't supported in any regulation, commentary, FAQ, etc.

Truffle says "if the borrower doesn't know you approved it, they cannot opt to not accept it. . .."
1) You say you APPROVED IT. How is that not an approval?
2) HMDA Action Codes aren't reliant on what the borrower knows and then dictates to you. The Commentary defines when to call it dependent on lender and borrower action. For instance, if the lender say "we can make this loan" and the borrower says "never mind, we're withdrawing", it's NOT a withdrawal. If the lender says "I can't do this, but I could do that" (counter-offer) and the borrower says "never mind", it's not a withdrawal, it's a denial. The borrower's words are not the test to determine regulatory reporting.

Can either of you show me where the communication part is supported in a regulation - not "our examiners told us."?

Interesting. Our stance at every bank I've worked at FRB or FDIC went with what the file status was at the time of the borrower changing their mind.

Anyway, it's never been an issue I've had to argue.

Our stance at every bank I've worked at FRB or FDIC went with what the file status was at the time of the borrower changing their mind.
Exactly! The file status is what determines the action code, not what is communicated to the borrower or what the borrower says.