NAR speaks on LTV vs CLTV

[zk]

The NAR provides an email address on the website with their “anti-bubble” reports, inviting us to ask them questions or make comments. They promise a NAR research staff person will reply promptly. I emailed them a week ago, and got their response today. I specifically asked about the flaw pointed out by the professor. Their response indicates that they are perhaps partaking Professor Piggington’s peculiarly potent potions.

Following are the email I sent, their response, some further comment from me, and the email I’ll be sending back to them.
—————–
Hello,

I am a prospective home buyer in San Diego, CA. In deciding whether to buy a home now or to wait, I have tried to educate myself on the San Diego housing market. I read your “Home Price Analysis for San Diego-Carlsbad-San Marcos” as part of that effort. I have a few questions about that analysis, and I thank you in advance for the promised prompt response.

In your “Home Price Analysis for San Diego-Carlsbad-San Marcos,” you state that “only 3% of loans have loan-to-value ratios above 90%, so the foreclosure risk is minimal. (That is, prices would have to decline by more than 10% to have a measureable impact on foreclosure rates.)”

The paragraph immediately preceding the paragraph in question discusses loans made in 2004, so I’ll assume that’s the year during which the loans in question were made. The statement that only 3% of loans have LTV ratios of 90% or more, combined with the following statement that therefore prices would have to decline by more than 10% to have a measurable impact on foreclosure rates, seems to imply that 97% of homeowners have borrowed less than 90% of the value of their house. But according to DataQuick, 38% of San Diego home purchaces in 2004 were made with a down payment of less than 10%, and 27% of purchaces were made with zero down payment. It seems obvious that the discrepancy between your figure of 3% and their figure of 38% lies in the fact that many homeowners used piggyback loans to supplement their first mortgage. So, while the statement that only 3% of loans have LTVs of 90% or less may be technically accurate, the statement that, “prices would have to decline by more than 10% to have a measureable impact on foreclosure rates” is the part I don’t understand. Your logic seems to be: “Only 3% of loans have LTVs of 90% or more, therefore only 3% of homeowners will owe more than their house is worth (and therefore be at risk of foreclosure) if prices decline less than 10%.” But that logic seems faulty, because, while only 3% of loans have LTVs of 90% or more, 38% of home buyers have put down less than 10%. It seems to me that the LTV of a particular loan wouldn’t be what determines how much at risk for foreclosure a homeowner is, but rather the total amount a homeowner owes in relation to the value of his property. And, as the statistics from DataQuick show, almost 4 in 10 San Diego buyers borrowed more than 90% of what they paid for their house in 2004.

Please enlighten me as to whether I’ve made some poor assumptions, missed a point, used faulty logic, or otherwise misinterpreted what your report states. I’m kind of on the bubble as to whether to buy a house or not. I’d heard about large numbers of people putting very little money down on their houses, and that made me worry about large numbers of foreclosures (and a possible decline in prices as a result). Your report lessened my concern somewhat, but I’m still not sure how I should interpret the paragraph I’ve asked about. Some of the things I read in an effort to educate myself are a bit hard to understand, and I appreciate the opportunity to ask for clarification.
Thank you,

(Let’s call me Mr. Steele)

————————————
Their response:
————————————

Mr.(Steele),

The data for LTV that we use comes from the Federal Housing Finance Board. It has been published for years and is widely used in the academic and housing community. Furthermore, when the data is surveyed from 75+ companies by the FHFB, the instructions for collection are to explicitly leave out
piggy back loans.

I am not familiar with Dataquick’s products or their survey methodology, but the FHFB data is the standard. If you have questions regarding their data, you should contact them directly.

On a related note, we wrote the local price analysis reports to refute many of the erroneous theories that have been circulating in the media by pundits and economists. However, while we feel strongly about our economic arguments, it would be inappropriate for us to provide you with purchase advise either one way or the other. I hope you understand the sensitivity of this issue.

Good luck with your decision and have a nice weekend.

——————-
My comments:
——————-

I suppose it’s conceivable that they actually didn’t understand what I was saying. Quite possibly, unbiased reasoning does not register at the NAR. But I think it’s probable that they knew exactly what I was talking about, tried to obfuscate the issue by implying that possibly DataQuick’s products were flawed, then hoped that somehow I wouldn’t notice that they completely failed to address the issue at hand. Maybe they thought I’d just up and believe that all those bubble theories out there are “erroneous” because the “National Association of Realtors” (speak like Mr. Howell when you say it; they probably do) says they are, and then panic and run out and buy a house right quick before I got priced out of the market. I might not be Ben Bernanke (or Professor Piggington, either), but I’m not falling for THAT.

The haughty tone of the email is mildly amusing, but their “we wrote the local price analysis reports to refute many of the erroneous theories that have been circulating in the media by pundits and economists” line actually made me laugh out loud. I can see the blank stares at the NAR when they read non-NAR economists’ theories. I can picture them trying to jam square logic into their misshapen logic receptacles: THEORY INDICATES COMING DOWNTURN. DOES NOT COMPUTE. OBVIOUSLY ERRONEOUS. DISCARD. Their lack of a coherent response to my email makes it obvious that they either can’t follow fairly simple logic (albeit logic that doesn’t fit into their receptacles), or they are purposely trying to mislead me (and, of course, the general public). Either way, how amazingly arrogant of them to flatly call the theories of others “erroneous.” Especially when their own “theories” are suspect in their logic, selective in which facts they use and, of course, outrageously biased.

Their “while we feel strongly about our economic arguments, it would be inappropriate for us to provide you with purchase advise either one way or the other,” line is kinda funny, too. I didn’t ask for advice (or “advise,” for that matter). I merely asked them to answer whether I misunderstood what they were saying or not. I guess they didn’t see that question as the attempt at tact that it was, and they were hoping that I really didn’t know whether I misunderstood them. Well, what they don’t know is that I know that I didn’t misunderstand them (understand?). Of course, I did understand them, and it’ll be interesting to see how they respond when they see that I understand all too well. I’m guessing that once they see they can’t fool me, they’ll ignore me. Just like they ignored our favorite Pundit. I mean Professor.

Following is the email I sent them in response. They took a week to answer my last one. We’ll see what happens with this one.

——————–
I’ll be responding tomorrow. Following is a draft of the email I’ll be sending back. Any suggestions would be appreciated. Professor, they ignored you as “Mr. Toscano;” if you want to sneak in a few words as Mr. Steele, I’d be happy to throw them in there.
——————–

Sir or Madam,

I appreciate your response to my email, but the question that I had wasn’t cleared up. Perhaps I didn’t state my question clearly enough, so I’ll try again. Also, you implied in your response that I was asking for purchase advice. I’m not asking for purchase advice. I merely want to educate myself as much as possible, and in that endeavor I am trying to understand your Home Price Analysis for the area in which I am trying to decide whether to buy a home. Clarification of your report is all I’m asking for.

Below in this email are the email I sent you and your response, if you need to refer to them. If you’re not familiar with the issue I’ve raised, I believe you’ll need to read those emails before reading my questions.

You imply that possibly DataQuick’s information is inaccurate, but that isn’t the issue. The issue is that you seem to be assuming (and for this assumption it doesn’t matter whose data you use) that if 97% of first mortgages made were for less than 90% of the home’s value, then 97% of homeowners owe less than 90% of their home’s value (and thus are not at risk for foreclosure unless prices drop by more than 10%). But if some of those homeowners also got piggyback loans, then that assumption is not true.

You state in your response that, “the instructions for collection are to explicitly leave out piggy back loans.” I understand that if you take all the first mortgages and all the piggyback loans, you will get a greater percentage of loans whose LTVs are less than 90% than if you just take the firsts. And apparently, that was what you thought my problem with your data and your reasoning was. But that was not the issue/question that I have. The thing that I want to be clear on is the conclusion that you come to based on the fact that 97% of first mortgages have an LTV of less than 90%. It seems to me that if you are trying to come to conclusions about how at risk of foreclosure homeowners are, then you’d have to include all of what they owe, and not just what they owe on their first mortgage. This was the point of my first email, and the basis for my questions (in that email and this one).

So, to put my questions more directly, and hopefully more clearly:

Is it your position that 97% of 2004 San Diego-area buyers have borrowed (counting all of their home loans) less than 90% of the value of their homes?

If that is your position, then how do you come to that conclusion if you have explicitly instructed the 75+ companies who collected the data to leave out piggyback loans? If you don’t have data on the piggyback loans, then how can you add the first to the piggyback(s) and come up with a ratio of what a homeowner owes to the value of his house?

If it is not your position that 97% of buyers have borrowed less than 90% of the value of their home, then how do you come to the conclusion that “only 3% of loans have loan-to-value ratios above 90%, so the foreclosure risk is minimal. (That is, prices would have to decline by more than 10% to have a measurable impact on foreclosure rates.)?” You seem to have directly concluded that the LTV ratio of the first mortgage is the important factor in determining the risk of foreclosure, without taking into account the total ratio of what a homeowner owes to what his house is worth.

Is it your position that the LTV ratio of the first mortgage is the important factor in determining the risk of foreclosure, and not the ratio of the total amount that a homeowner owes on his house to what his house is worth? If so, then please explain the logic behind that position.

Whether to buy a home now or to wait is a very important decision for me and my family, and I want to make sure I am as informed as possible prior to making that decision.

Thank you for helping me to inform myself. I look forward to your response.

Sincerely,

(Mr. Steele)

[zk]

I said I’d be responding tomorrow; actually, I’ll be responding to the NAR by email on Monday. Suggestions for the email appreciated. Thanks,

Mr. Steele.

[barnaby33]

I would shorten it if I were you. You want to be explicit but you also need to be concise.

Josh

[bailedintime]

Agreed. Edit.

[Bugs]

I’m betting they go from restating the original theme (which is a favorite trick of politicians) to outright stonewalling – no response at all to your follow up.

[powayseller]

Great job! My first question is: Why is the NAR and the FHFB benefitting from reporting only the LTV, and not the CLTV?

I spent an hour reading some very lengthy documents (while the kids are cooking dinner), and came up with this: The GSEs purchase the first lien in structured jumbo or piggyback loans!! Thus, by excluding the piggyback portion, they can keep lower capital requirements because they are not exposed to the total amount of the loan, just the first lien portion!! Clever, right? And the NAR would just report the first lien data, because it is also in their benefit to downplay the risks.

For further info, read on. The risk of GSEs comes from their new foray into investing. As of 2000, they had a $900 billion portfolio of mortgage loans and MBS, which exposes them to interest rate and credit risks, and operational risks. A significant percentage of their revenue comes from their investments. I bet that figure has multiplied in the 6 years since the report was written.

The “model” (I didn’t read enough to get whose model it is) allows Fannie Mae to purchase the 80% portion of an 80/10/10 loan, and no credit enhancement (i.e. PMI insurance) is requried by their charters under this structuring. This is also done for Jumbo loans. The model understates the credit risk and required capital for these structured loans. The problem is that one risky loans, by simple division, becomes two less risky loans on the books. GE’s Finance Division, which wrote this critique (obviously because the GSEs are a competitor they’d rather expose), pointed out that an 80/15 loan has the risk of a 95% LTV loan, and should be treated as a 95% LTV loan, and not an 80% LTV loan in the GSE risk profile.

The GSEs aggregate all loans > 95% LTV. However, risk increases in steps. The default risk for a 97% LTV and a 100% LTV loan ia 34% and 75% greater, respectively, than a 95% LTV loan!

The OFHEO does not consider CLTV, and underestimates the capital needed by the GSEs. For every $1 billion in an 80/10 or 80/15 combo loan, the capital should be increased by $48.9 million and $100.1 million, respectively! Just consider how much Fannie Mae can scrimp on their capital requirements by not counting the piggyback portion of their loans.

GSEs are allowed to purchase either the 1st or 2nd lien in a structured deal, so either the 80% or 20% portion of an 80/20, subjecting themselves to a 100% LTV default risk, which is, as you recall, 75% greater than the default risk on a 95% LTV.

GSEs encourage piggyback or structured loans, and it is a growing part of their business (again as of 2000). They make a Jumbo loan into a GSE-eligible mortgage.

The GSEs do classify and track liens by CLTV. But they don’t have to meet capital requirements by it, or report it.

So there you have it. Now we know why only LTV is reported. The government has not required the GSEs to update their risk profiling by CLTV.

I know why GSEs benefit from this, but why hasn’t the government required them to update their capital requirements for >80% CLTVs?

I wonder how many of you reading this are exposed to the risks at the GSEs. It’s even worse than I thought. They are holding and investing in loans based on the wrong risk profiles.

Ready to get out, yet???

[bailedintime]

Wow! Do you think REIT money has much exposure to this risk? Nice research, I hope dinner was good!

[RightSide]

Here is another stock that is going to ZERO. NYSE:PMI . They sell the insurance that people need to buy when they finance with less then 20% equity (the real at risk group of people). They are insuring BILLIONS of loans, yet guess how much cash on their balance sheet? 700mm.

They will be hosed.

You want a good trade if you believe this market is tanking? Go short PMI.

I’m not short yet, but I’m lining up my trades. Forget about investing in gold. If you want to get rich, then you’ve got to actively participate in the decline. There are numerous opportunities to do so.

Did you know that there is a swap market that allows you to short MBS? The risk of MBS has been spread to many people…will this pose a systematic risk or is this what the new global markets are all about? We shall find out…

[powayseller]

How can you short safely, i.e. without getting a short squeeze? With all the automated computer trades, the first sign of loss causes other sell trades to execute, and then you can get stuck w/ a margin call, right? I’m just thinking this could be way over my head. I can identify the companies which will go down, but I’m not sure I can win my bets against the sophisticated traders and computers. That’s why I was thinking of which companies to go Long.

As far as REITs, I don’t know. Read the prospectus, and if you’re still unsure, e-mail the company and get their holdings in writing. Don’t REITs hold real estate in physical form, such as shopping centers, apartment buildings, commercial buildings? GSE investments are held by banks, insurance companies, banks and hedge and pension funds. They buy Fannie Mae and Freddie Mac bonds. I don’t know if REITs hold any bonds.

[RightSide]

Shorting is very easy and the risks are far overblown…but you do need to be smart and educated to be successfull, but that goes with most things. You are not going to get squeezed out (forced to give you your short postion) in any of the stocks I’ve talked about.

[powayseller]

How do you short MBS, and how do I know I won’t get a short squeeze, since I only like the squeeze I get from my husband:)

[Jim Brubaker]

They’re going out of business for another reason.

The 80/20 loan is an end run on PMI. The PMI drops out of a 80/20 no money down loan. Plus if your house has appreciated to an 80% LTV, you can drop the PMI. In essence, the 20% second is the looser on the default.

The “Second Trust Deed” is going to have the same ring as “Peruvian Bonds” in the near future.

[powayseller]

I checked w/ a mortgage broker (SoCalMtgGuy), and he added that the GSEs don’t know when someone takes out a HELOC.

[Jim Brubaker]

When the GSE’s (Government Sponsored Enterprise ie Fannie Mae or Freddie Mac) take in a first trust deed loan, they are oblivious to a HELOC. The HELOC (home equity line of credit) has to stand on its own two feet. Its a second trust deed. In a default, the HELOC would have to cure any deficiency’s on the first in order to foreclose on the property.

If its an 80% with a 20% HELOC and the property drops in value by 20%, the bank holding the HELOC is not going to cure the first trust deed its going to walk with a loss.

The HELOC is not going to go away from what I understand. It might drop off of the Trust Deed when the first forecloses but the amount owed is still owed by the originator of the loan, at that point it is an unsecured loan. I could be wrong on this–some of the stuff I read, I haven’t had the time to research

[privatebanker]

I believe your right on if a property drops 20% on an 80/20, the heloc essentially becomes an unsecured loan. However, they still have (on paper) a lien on the property. It’s almost like comparing preferred stock to common stock in the event of a liquidation of a company. I do know that, for example, if a borrower has an option arm and the loan balance exceeds a certain LTV from the originating value, the borrower will need to pay down some of the loan. Almost like a margin call. I’m not sure if all lenders require this but I know my bank (that I work for) requires this.

[Author]

Posts