And I still say that painting this issue as a "sub-prime" (or undocumented, non english speaking) is WRONG.

The issue is that home prices went too-high because a real estate craze that took the whole country. A country that became addicted to credit instead of getting:
1. education,
2. fair wages and
3. fiscal resposibilty starting with savings.

Saving is what you have in your saving account (this is how I started) - most people think "saving" is when WalMart is having a sale and they can load up on made in China junk they don't really need.

"You just eliminated the entire 20-something .."

Not True.

I came to the US as an immigrant when I was 20 something.

I could not even rent a car because no one would rent me one without a credit card. You have no idea how hard it is to "get into the system"

But I worked hard. With the few dollars I had when I came into the country I bought a $900 Ford Pinto (never understood why everyone was laughing at me) used it for several years until I could afford something else. That ford pinto allowed me to move around. I did not look for anyone to make any favors. I found a job, I rented for years, first at $900 a month later at $1200, later at $1400 and by the time my rent got higher I bought an apartment with 50% down payment. My mortgage is now paid in full.

What is so wrong with saving ?
With working hard ?

Above 620 is not sub prime.

So here is a post BK 656 FICO:

Mortgage Grapevine: UT, $175,000, 656 FICO, 83.33% LTV, Refinance, SF, full, OO

It was already two years since his BK so I guess he is credit worthey once again...... with 50 base point for the added risk he should find any loan he needs and even get some cash back. Lovely.

and another:

Senator Dodd,

There are five important steps you can take to make housing affordable again in America:
1) Appraisal process: All appraisals must be ordered thru a third party agency, removing appraisers from the control of lenders, agents, buyers & sellers. There is too much pressure and too much fraud.
2) Audit “Owner Occupied” loans. Speculators are buying multiple houses at at time, blatantly lying abut occupying the homes. This increases risk, drives up prices, and forces true homeowners to compete with investors. Get FNMA and FMAC back to their roots.
3) Use qualifying incomes for purchases based solely on using tax returns. You will raise tax revenue and eliminate the cheaters. If people want to buy a home, they must prove they can carry their own fair share of the public infrastructure which supports housing.
4) Enforce mortgage fraud laws. Fraud is so blatant today, it is laughable. People believe they will not get caught. Ten “auditors” in every city could each identify 20 cases/day, or 1,000/week, 50,000/year. Well placed inquires to Realtors, mortgage brokers, sellers and buyers would put a little fear of enforcement into the air and end 90% of the scams immediately. This type of action and publicity has a huge cost benefit ratio.
5) Provide more balanced laws supporting the home building industry to produce more housing. The bureaucratic nightmare involved in building housing units is out of hand. Increasing supply will do more for affordability than any subsidy or Fed funds policies.

The worst thing you can do, Senator Dodd, is provide a bailout for lenders and/or irresponsible buyers who created the current housing problems. You will perpetuate the problem and you will not be re-elected. The majority of responsible people in the U.S. will find it reprehensible for you to facilitate placing the burden of this mess back on the taxpayer.

By the way, I have voted in every election since I turned 18.

I argue for having more considerations in the mortgage credit process than mere FICO, but I also want to keep it. In part, because it becomes a useful thing to stop certain kinds of discrimination in their tracks: if you want to argue that A just "isn't the kind" to "deserve" a mortgage, you have to confront the fact that A has the same FICO as B. If you cannot find and document objective measurements of creditworthiness that would show a true difference between the borrowers, you are going to get regulators waving a copy of the Fair Lending regs coming down on you, and you would deserve it.

That word "deserve," used in terms of mortgage credit, makes my skin crawl. You qualify or you don't. You show a lender you have repayment willingness and ability, or you don't. Mortgage credit is a business, not a church, and the day we let ourselves believe it's about rewarding people we approve of and punishing people we don't is the day we really did become a doomed economy. I do not consider myself an arbiter of human worthiness or the expert on acceptable aspirations; I'm just a mortgage underwriter. I might have colleagues who are convinced that they should be able to judge everyone else's motives, deserts, and worth, but I avoid those people like the plague and if you all want to see them in charge, good luck to you. They will not stop with mortgage credit; they will want to tell you where you should live, what job you should have, what you should spend your money on. Then they'll get into who you should sleep with, how you should reproduce, and where you've been on Sunday morning. Haven't we had enough of the moralists in charge?

There's a difference between moral and moralistic business practices. It may be a hard difference to maintain; that's why we need to keep having the conversations over and over, and testing our assumptions. So I think the issues need to be raised. But as soon as I hear it implied that I, as underwriter, ought to be doling out credit to "reward good people" and denying it to "discipline bad people," I am convinced that we really crossed the line.

simply brilliant.

Exactly, Max.

Furthermore, I have no problems with the idea that down payment should be relative to the actual specific borrower. For instance, here's something you see a lot of in the more modest end of the conforming loan world:

A works for $15 dollars an hour. A has managed to save up a 5% down payment out of income over three years on a modestly-enough priced home that A can easily carry the mortgage payment including the MI that will be charged because the LTV is 95%.

B just took a $20/hour job out of college. B is making a 20% down payment on a fairly expensive home; B can carry the payments, but they'll be tight. The 20% down came from a gift from the parents.

Who has more "skin in the game"?

I consider down payment to be a measurement of the borrower's willingness to repay, not just his or her ability to repay. A $10,000 down payment coming out of a savings account that somebody funded by putting aside $20 a week out of the paycheck can say a lot more to me than a $50,000 down payment that came from sale proceeds in a hot market or a gift from parents. My own experience tells me that, all other things being equal, A will fight like a lion to keep that house. B will fold more quickly. To an underwriter, the source of the down payment is just as important as its size relative to the deal at hand, not just to some abstract rule where "20%" is a magic number.

Great post. Before getting to the math part, I'm really pleased that in the comments Tanta discussed the use of FICO to introduce objectivity to the loans process. Discrimination, especially in residential real estate transactions has gone on for a long long time. I've read of realtors, in the past, not even showing homes in certain areas to black people. An African-American told me this happened to her in Pasadena, CA. I'm east-asian in origin and I distinctly recall suspecting this happening to me in England in the late 70s, early 80s as I joined the real estate ladder.
So FICO scores have brought in a much needed objectivity to the loans process at least.

On the math side:
I've done predictive modeling ( heuristic learning ) for a decade so Tanta's comment on the Expected Losses modeling and how FICO can be such a dominant risk factor for mortgages immediately got my interest.

An essential for predictive analtics is that the training set you learn from is representative of the population you are predicting about. Did this stop being true during the subprime boom ?

A more subtle point that always concerns me is reflexivity - better put as "gaming the system" - When I did heuristic learning about machining tools, predicting recalibration intervals based on out-of-tolerance occurences - I never worried about the machining tool learning about me spying on them and in some sense altering their behavior so the risk factors I relied on disappeared BUT still carrying on with their old out-of-tolerance actions, just no longer predictable in advance.

In human action though, I wonder if this occurred - if so, those MBS ratings are going to be seriously out of whack with reality.

-K

Pay No Attention to that Man Behind the Curtain;

Man. Man. OK. That picture was too much.

Either Greenspan is a manipulative genius, able to swing the economy with a mere twist of the tongue, or he's an incompetent boob, blundering along while dragging our economy behind him.

Does anybody (besides Mish) really believe that ordinary people base their borrowing decisions on what Alan Greenspan says? Really?

I think you give the guy too much credit.

IMHO

Easy debt drives asset prices up and beyond the ability of the average joe/jill to save up for it. Sorta of the old saying that bad money drives out good.

Easy debt drives down the need for labor action. Just get the money from the cc and mortgage atm. Lately in France, there have been riots over wages. Chinese workers riot frequently over wages.

US workers just vote for the party of more Jesus and War, watch on their indebted TV, drive their indebted SUV, and live in an indebted home enslaved to the powers that be.

My political agenda is a bit puritanical, eliminate easy debt, make the US citizens suffer in one mad rush the sins of the past 30 years, with the hope that the children will elect better leaders. I prefer lefties, but a good conservative not whoring after corporate dollars would be aok.

"Will the Fed ease soon enough and by enough to bail these borrowers out? If it does, it effectively bails out feckless speculators, just as it did after the tech debacle."

Over the years the FED has repeatedly said, asset prices aren't relevant (especially when they head up). Let's see if they change their opinion when real estate prices start heading down.

Professor, certainly the 20% guideline comes from many decades of experience.

However, what it comes from is, largely, an "individual risk" model rather than a "group insurance policy" model. In other words, to make every one of your borrowers conform to the 20% rule is to make each loan "stand on its own," more or less. To work on a pool model, where you can in some sense "insure" the whole pool by having some 60% LTVs and some 95% LTVs, is not, to me, obviously a bad idea. It's just only as good as your underwriting.

Compare this to what we're seeing in health insurance, where you have insurers wanting to use these "pre-existing conditions" and qualifying medical exams and so on to keep older and sicker people out of the insurance pool. What's the point of limiting health insurance to the youngest and healthiest (who thus have the least need for it)? Profit. Certainly not any worthwhile understanding of the whole point of group insurance.

So go ahead, put young people into a situation where they fund their own retirements with 401(k) contributions, they fund their own healthcare with "health savings accounts," and they have to make 20% down payments in order to buy houses. Let's just decide that the little bastards can pull themselves up by their own bootstraps. Meanwhile, let's certainly not use this pool of older homeonwers with their 50% or 60% LTVs to function as a "risk pool" that could counterbalance the increased risk presented by the young 'uns.

Yes, that's what it sounds like to me. I want an average LTV in my portfolio of no more than 80%. That doesn't mean I'll never do an individual loan for more than that.

I'm not sure everyone here knows why people can't document their income.
People cannot document income because they, in every case, cheat on their taxes.
Stated income loans used to be for people with retail stores, landscaping and other cash businesses who are the most common cheats who could afford a house. Now that everyone is entitled to a large house with granite countertops, the tax cheats who clean houses or do other manual labor who are either self-employed or work for other tax cheats have been added to the existing pool of tax cheats.
The loans for tax cheats used to carry a fairly high rate to accomodate risk. Now that there is no risk priced in, there is little interest premium for these loans.

Tanta,

I have to disagree about the errors in the credit scoring data with regard to Expected Losses modeling, especially from a predictive analytics perspective. There are very good mathematical techniques with statistical proof backing them for handling noisy data and a colleague of mine who worked in analytics at Fair Isaac ( the FI in FICO score ) told me they routinely used them.
If a risk factor data is too noisy it simply loses it predictive power and is discarded..
But I'll get off my professional hobbyhorse now.

Of course, from the individual human perspective none of the above is any consolation if THEIR data about an included predictive risk factor is erroneous - I do "get that".

-K

Clarifying my own post:

"If a risk factor data is too noisy it simply loses it predictive power and is discarded.."

I mean the risk factor is discarded, i.e. excluded from the predictive model. I don't mean that they throw out the noisy data; well apart from "outliers" and using capping and fancy scaling techniques.

-K

Lama, as always, hits the nail on the head. The great All-American hobby of cheating on your taxes turns around and bites you on the ass when you want a mortgage loan. Forgive me if I don't weep.

Another way to put the point I'm trying to make is that we can consider "affordable housing" lending to be some kind of allocation of available credit. I used the example of balancing out a portfolio so that my WA LTV is 80% or less, but the individual loans might be under or over that. Another way to look at it is that I have X dollars--a percent of my lending capital, or my security pools, or whatever--to "spend" on high-LTV lending. It can be spent to put a third-time homebuyer into some overpriced monstrosity with granite countertops so that person can live the middle-class dream of more toilets per person, or it can be spent on getting FTBs into starter homes, and force the later stages in the chain to "move up" with actual equity.

It's a question, at some level, of where we want to take the risk. It's just like some stupid locality giving tax breaks to developers of McMansions, without figuring out how the schools are going to get funded or the roads paved, because teachers and road workers have been driven into the next town over where they can afford to live.

Professor, that is the thing, isn't it? The borrower has to ask how much risk he is really ready to take. Lately, the answer has been something like, "what risk?", which makes my hair stand on end and my blood run cold.

I would never object to anyone counseling a borrower that a 20% down payment is a good idea. But all you're saying is that, if the property value drops significantly, the borrower's 20% will get spent first. Hopefully, it will be enough to cover the loss, so you won't see a situation like we had on an earlier post where the borrower had to bring cash to the closing in order to sell. But how, exactly, are you protecting a borrower from excessive RE price risk by just telling the borrower to risk his own money instead of mine?

The answer, of course, is that you're hoping that the borrower with his own money at risk will be more prudent about getting into such a deal in the first place. I agree. But that's where I part company on the across-the-board 20% requirement. I used an example above of someone for whom a 5% down payment might well be a much bigger gamble than another borrower's 20%. I am not willing to assume that all borrowers think in terms of percent of value. Some think in terms of, "that's my $10,000 that it took me three years to scrape up by eating a lot of peanut butter. I'm not going to risk losing it in an RE deal unless it's really a good deal." I can respect that thinking all day long.

You know, you really ought to consider writing a book aimed at serious consumers and investors. Your style makes it entertaining.

Oh well, there I go again. I'll be quiet now.

Go Tanta go

For whatever reason, we as a society, have substituted credit for cash. I see it as a vast right wing conspiracy to limit unions, worker's rights and 'fair wages'. That said, even it if just happened because our leadership found it expedient to let it happen because it put off hard choices, the result is the same. Assets replace income and credit replace cash.

So we have a new entitlement, instead to a 'fair wage' it is instead to its substitute credit. Rather than a career because careers are a thing of the past, it is a house you are entitled to as part of the American Dream.

We are drinking, downing, drowning and deluging the kool aid. We are the kool aid.

A Britsh news paper did a wonderfull study few month ago.

They calculated CPI for:
1. Youngsters (20s)that still live with their parents
2. adults
3. Senior citizens

It turned-out that for very young people CPI is very low (1%-2%) while for senior citizens it was more like 10%

Tanta:FICOs are at least an attempt at a consistent, efficient means to communicate credit quality.

We have secretly replaced reality with an over-simplification. Let's see if anybody notices...

The best part of waking up
Is the Folgers in your cup!
(oh, the irony)

Efficiency and larger scale usually means bigger messes to clean up once the bugs in the plans finally come to fruition. With great throughput comes great prudence or something. Hopefully, with a little bailout "innovation", Wall Street can come out of this fire-sale making a profit.

You down with RTC (Yeah you know me)
You down with RTC (Yeah you know me)
You down with RTC (Yeah you know me)
Who's down with RTC (Every last homie)

Can they accellerate loan that are being paid on time because their margin is perhap unattractive?

Ab-so-lute-ly not. Residential real estate mortgages have no call option on them; the lender cannot accelerate because years later (or any time later) the lender has decided it wasn't priced properly up front. You can call a mortgage only on default or sale of the collateral. You gotta be careful when you make a 30-year loan to people. Remember, this is why most lenders sell their fixed rate loans (and ARMs with long fixed periods, too).

If a federally insured lender goes down, the FDIC will find another institution to assume that lender's assets. The reason why we had an RTC back in the ugly days (Resolution Trust Corp) was that those assets in some cases weren't very attractive. Let's just say the regulators might have to tie a pork chop around their necks to get the other banks to play with them. But someone has to.

Lurker, neg ams can be and are securitized, just not as often as other ARMs. They obviously have to be put in custom security structures--because their cash flow is weird. But if you think investors have been requiring actual cash flow these days . . . um, no.

Really, it's not all that different from any "zero coupon" or strip-type bond, from that point of view; the investor gets some of its principal and interest periodically, and some of it eventually. Assuming the loans don't all go into foreclosure and lose too much money, but you'd want to worry about that with any "Alt" backed bond.

Tanta,

I'm late to this discussion, but I'm wondering why no one is calling for a credit scoring metric specifically for mortgage credit. As you pointed out, the FICO score was built for installment credit. The algorithm is heavily weighted the quality of payment history and the length of payment history (roughly 50% based on information released by Fair Isaac) while debt capacity is weighted only at about a third.

For installment credit the low debt capacity weighing probably makes sense since in the normal course of things people don't usually get enough installment credit to really hang themselves (not universally true I know, but probably generally the case outside of divorces, big medical debts, etc).

Moreover, it appears that FICO scores have little prospective risk assessment beyond the total credit available on credit cards. It appears there is little to no weighting given to how the terms of the financing will change at some point in the future.

These two characteristics (low debt capacity weighting and no prospective view of debt terms) would appear to be the achillies heel of FICO scores as it relates to assessing mortgage credit risk. As we are currently seeing in all the vivid ugliness before us, mortgage debt gives consumers plenty of rope to hang themselves if it has terms that become very onerous down the road.

So my question is, why isn't anyone calling for an end to this state of affairs. I realize there are all the inertial excuses about FICO being the only thing out there. However, this is an important enough issue in a big enough market that there is certainly an economic return to devising a better mortgage credit metric. Yes it will add a little more friction to the home buying process, but even if it cost you $50 to get a "M-FICO" score as part of your application process that is a modest price to pay compared to watching a whole segment of the mortgage market disappear in 60 days due to flawed underwriting metrics. The data to create such a thing is already being collected and the Barney Franks of the world could surely find the will to mandate yet another form that would neatly capture all the data needed to make it work.

I have not heard one person suggest or proposed such a thing. Is it just to early in the scandal cycle for that or am I missing something?

Subprime lending is about people with credit depth that have just naturally spotty histories or single life events that have kicked them out of the prime world. Folks with little or no credit histories have never been the target of subprime, that is where FHA comes in...

So why has the subprime share of origination been growing exponentially over the years and the FHA share shrinking?

Your view of subprime, producer, is I think a bit dated. Both subprime and Alt-A have become first-time homebuyer programs, and they're taking borrowers who once would have gone into FHA. This particularly outrages me because FHA loan programs are safe structures: your choices are a fixed rate or a 1-year ARM with 1.00% annual adustment caps (the overwhelming majority are fixed rate). The subprimers put these people into 2/28s and 80/20s with a variable rate second lien and other high-risk products.

Besides, my point was about how people get "credit depth" as much as how they can lack tradeline credit. If your very first credit experience is with high-rate, high-fee subprime credit cards and car dealer loans, your odds are very good of stumbling with it (because it's too expensive or because you didn't understand the complex rules of the payment structure), and so you fall into the "once subprime, always subprime" trap. I am proposing that if entry level consumer credit to people with no existing credit history were made on safer and sounder terms, they might not so often end up with a "subprime" credit history.

Calmo, the flip side on the OER is thatif one assumes no appreciation in house prices for, oh say the next decade, which at these levels in many markets seems to be a reasonable ssumption, one has to be a blithering idiot to buy rather than rent if s/he has to move. As the realization of low to negative house price appreciation becomes a more common perception, more people will, at the margin, decide to rent rather than buy. The stock of rental properties has not increased nearly as much as the stock of owned properties. For the last decade or more Condo conversions have been very common in major metro areas. Reverse condo conversions have been exceedingly rare. While we may see more of them in the future, it will take time to catch up. Over all this process should force rents up and housing asset prices down. Over the long term, the value of an asset is dependent on its ability to generate cash. While housing serves many functions beyond its role as a cash generating asset, over the long term its price should still reflect its ability to generate cash (namely in the form of rent). Put another way, use of OER substantially understated inflation on the way up (rising house prices were never directly incorporated into inflation measures). Now it is likely to play a bit of catch up.

Tanta,
A little bit of data on the question of the 20 somethings in the market. A big part of the overall increase in the home ownership rate has been a stunning rise in the ownership rate of those below the age of 25. While the total ownership rate rose from 65.4% of households in 1996 to 68.0% in 2001 to the current 68.9%, the rate among households headed by someone under age 25 rose from 18.0% in 1996, to 22.5% in 2001 to the current 24.8%. Thus in a cecade where the total ownership rate rose by 5.4%, the young ownership rate rose by 37.8%. These kids have to be sub prime borrowers. After all, a kid just out of college, no matter how good a job he just landed, and how mature, honest and mature he seems just has not had the time to develop a really good credit history. Let alone save up enough of his own money while actually renting an apt. to make a significant downpayment.