Posted by Mark Hanson on his blog:
Most look to loan type and equity position as two of the most important factors when forecasting loan default. In fact, I believe that epidemic negative-equity is the overarching reason that the default, foreclosure and housing crisis remains in the early innings. But…negative-equity with a caveat.
While negative equity is a threat in and of itself, being in an over-leveraged household debt position is the true default catalyst for most in a negative-equity position. And being over-leveraged is also the primary default catalyst for those is a positive equity position. Being in a negative-equity position with lots of top line and disposable income each month is generally more of a mental burden than a reason to fly the coop.
How many homeowners are over-levered and at eminent risk of default? This answer is…a lot more than most think, especially those who got a loan from 2003-2007 due to a radical, yet subtle shift in loan guidelines across the mortgage spectrum that kicked-off the bubble-years.
Yes, even Prime full-doc borrowers in 30-year fixed mortgages with 20% equity who got their purchase or refi from 03-07 are at much greater risk than most think. Being over-levered was condoned – all the lenders, investors and loan programs operated in the same manner.
In my research, I often assume that everybody knows the subtle idiosyncrasies of how loans are really structured. I understand this is not the case. So, in an attempt to highlight why the total residential mortgage risk exposure is so much greater than anybody’s expectations, this report drills down on Prime, Alt-A and Subprime allowable debt-to-income (DTI) ratios that were made ridiculously lax relative to pre and post 2003 – 2007. This, in my opinion, is the real tempest in the mortgage teapot that buckets millions more loans that are still in existence today across all loan types, as risky.
- Time-Tested DTI Standards Thrown out the Window
A long time ago in a mortgage market far, far away (circa-2000 and before!) there was responsibility in lending. Age-old underwriting standards only allowed fully-documented debt-to-income ratios of 28% for housing and 36% for total debt (referred to as front and back DTI). On Jumbo loans, the ratios were 33/38 because Jumbo borrowers typically have more disposable income. On occasion, banks would make exceptions to this rule if the borrower had a large equity position or liquid reserves. At 28/36, homeowners can pay debt, shop, take their annual vacation, and even save money. At 28/36 DTI a house is a place to live first and an investment, second.
Bubble year’s loan guidelines not only pushed the boundaries of risk by exotic loan structure but also income leverage. Circa-2002, time-tested DTI standards went out the window. Allowable DTI ratios on Prime loans rose to 50% and much higher when considering that so many loans were made with limited or no income documentation. Alt-A and Subprime full-doc loans would routinely go to 55% DTI…and full-doc are supposed to be the safe loans. Given that full-doc only represented 50% of Subprime and 25% of Alt-A loans it is understandable why these two loan types are experiencing the worst trouble, even though across the Alt-A universe the average FICO was above 700 at the time of origination.
Around this same time, the investment bank’s participation and non-Agency lending and securitization began to really heat up. Guidelines expended further…hey, if the loan was going to be off the books in a few months, who cares how over-leveraged the borrower is.
- Going Exotic in Plain Sight
Before too long — circa-2003 — lending guidelines were fundamentally changing with many lenders allowing leverage through increased DTI ratios never seen before. Obviously, this expanded affordability sharply. When all of a sudden you can spend 50% of your gross income on debt vs 36% before, you can afford to buy much more house or take a much larger cash-out refi.
Subtly changing loan guidelines by raising the allowable DTI on traditional loans, such as a 30-year fixed, was a more sneaky way of easing credit and going exotic than blatantly advertising for ‘no doc’. In fact, 30-year fixed loans and the borrowers that chose them were deemed to be so safe, the underwriting was much more lax than on an exotic structured loan, such as a Pay Option ARM.
By 2004, as property values pushed house prices to levels that were unaffordable and stated income was not the norm yet, the new-normal in mortgage lending was allowing up to 50% of gross income to go to total debt. The mortgage obviously was the largest chunk.
And remember, the 50% is only mortgage PITI and other debt listed on the credit report. It does not include income taxes, auto insurance, food or all the other things that individuals spend money on over the period of a month.
And it didn’t stop there. As the mortgage credit strengthened the borrower’s credit profile, other credit was made available, including second mortgages, that could take total DTI far above 50%. Nevertheless, at 50% DTI, the house becomes the largest investment of a person’s life because there is no way for most to put out half of their gross income to debt each month and invest elsewhere.
- Borrower’s Always Borrowed the Max
When buying or refinancing, most got a purchase or refi loan for as much as their banker or Realtor said they could, which was what 50% of their gross income paid for in most cases. Most borrowers don’t say “we know we qualify for $500k but just to make sure we have some wiggle room in our budget, let’s stick to a $400k loan”. Bottom Line – everybody borrowed too much because all of the lenders and loans — from the safest full-doc Prime loans to Subprime trash — allowed it. And after the fact, most expanded their credit portfolio because all credit was so easily attained until a couple of years ago.
- GSE Loans – A Culture of Fraud
During the bubble years the GSE’s looked at DTI secondarily to credit score, LTV, and cash reserves as measured by liquid cash and 70% of retirement. Both Fannie and Freddie have automated underwriting systems called DU and LP respectively. During the bubble years, if the LTV was low enough and/or score and cash reserves high enough, the system would approve virtually anything.
Many lenders, especially the big banks, had in-house DU and LP underwriting ‘trainers’ that would go around to the various mortgage branches and teach underwriters how to ‘trip’ the systems in order to achieve automated loan approvals when a declination was certain, or simply get fewer approval conditions on a loan that was borderline. Getting a loan approval out of DU/LP on a borrower with a 100% DTI — with limited documentation required on the automated findings — was not uncommon.
In fact, many that needed to pump up a borrower’s strength who was light on income — instead of lying about the income — would pump up another aspect of the loan. The most common was to increase the borrower’s cash reserves, particularly retirement. This way, the retail sales worker buying a house well beyond their means would not need an obviously fraudulent income level, rather a believable household retirement total of maybe $100k. Doing it this way simply raised fewer red flags for the underwriter and investor.
Few Loans Were Ever Denied at First Pass
During the bubble years, very few loans were ever denied. Denying loans was not ‘production oriented’. The culture across all lenders was to ‘approve everything subject to’. If you did not do it this way, your competitors would get all of the business.
The approval process was for the underwriter to run the loan through DU/LP and if the system did not issue an approval (or an approval the borrower and the loan officer were happy with) to go back into the input file and edit the income, assets, retirement (or all three) until the system approved it. Some loans were edited 30 or 40 times until the GSE system issued an approval.
Then, the approval was sent out to the borrower and loan officer even if it required them to verify $100k more in retirement reserves than the borrower had per the original loan application. Within a few days, a new back-dated loan application and a retirement account statement reflecting adequate reserves would arrive, the underwriter would sign it off and the loan would be on its way to the doc department. There was no way to verify if the document was a fake, unless it physically looked altered. In many cases the borrower never even knew this was happening.
Note – this process was not GSE exclusive…this is just how it was done across all lenders.
- Affordability out of Control
Then circa late-2004, as affordability declined sharply even with 50% DTI the norm, stated income came into play in a big way. This super-charged affordability and house prices in ways we will never see again in our lifetime.
Stated income was around for years prior but limited to verified self-employed borrowers. The new-era Stated income loan allowed anyone with a two year job history to get a loan. All of a sudden, everybody earned $150k a year. From then on, the housing market had no shortage of purchases, cash-out refinances or HELOCs and house prices never looked back…well, until the exotic loan programs went away in late 2007.
Circa early-2006 when it became obvious that Stated income was being abused because everybody (hair dressers, public sector workers, and anyone that said they were self-employed for two-years and could provide a fraudulent CPA letter that the lender never verified) suddenly was earning $12k a month, lenders became more cautious.
What was the answer? Begin to push Pay Option ARMs with low teaser rates and payments. This way the borrowers could earn less so their fake income looked more believable. In addition, this is about the time that No Doc and No Ratio doc type options began to show up on every lender’s rate sheet, which provided the ultimate in plausible deniability.
Bottom line - 80% of all Alt-A (including Pay Options), 50% of Subprime, 50% of Jumbo Prime and 30% of Prime loans from 2003-2007 were limited documentation loans for a reason – because the borrowers didn’t even have the 50% DTI needed for full doc.
- How Big is the Total At-Risk Mortgage Universe?
Of the loans in existence today at least 75% were refinanced or attained through a purchase from 2003-2007 – the bubble years. On several occasions the past couple of years, Jim Cramer has quantified the at-risk loan universe as being around 14 million, which represents everyone who purchased a home between 2005-2007. But then he says ‘”here is no way everybody who bought a house from 2005-2007 will ever default”. So, he pairs it back to 20% or 25% of 14 million – whatever. He is incorrect on a number of levels.
First off, the bubble years were really 2003-2007. But aside from that, the number of people who purchased a home is only a small piece of the entire pie. The bubble years was not about purchases, rather refi’s. During the bubble years, cash-out refi’s and HELOCs were at least 5:1 over purchases. A purchase is no more risky than an existing homeowner with a great payment history who pulled out 90% or 100% of their equity at a 50% DTI. In fact, the latter are more risky…purchases in general are always considered the safest loans.
This means the true potential at-risk loan universe is any Prime, Alt-A, or Subprime borrower that did a purchase or refi from 2003-2007. Obviously, not every single borrower is at-risk but we have no way of really knowing how many of the 43 million + loans from that period still in existence today are destined for trouble. This is especially true when even borrowers with 800 scores and 70% LTV’s are at risk of default because their DTI started out at 50% and after the fact, they expanded their credit portfolio because all credit was so easily attained until a couple of years ago.
- 13 to 15 Million Loans at Eminent Risk of Default
- Potentially, 20 Million Homeowners over the Next Few Years
The chart below breaks out all of the loans in existence by loan type. Of the loans originated during the trouble years, the far right columns show the conservative number of loans in which the borrowers either borrowed at 50% DTI or went Limited Doc (stated income, light doc, no doc, no ratio). The two columns are not mutually exclusive.
The last Mortgage Bankers Association report estimates that the total number of loans in some sort of delinquency, default, or foreclosure status to be about 8.2 million, or 14.41% of all loans. If the true number of eminently at-risk loans is somewhere between 13 and 15 million, the default and foreclosure crisis is about 60% over.
The problem with the final 40% is that it crushes everyone other than Subprime households and likely happens over a longer period of time than the two-year Subprime Implosion.
In addition to the eminent defaulters, a large percentage will default for various unforeseen reasons tied to the macro. Throw in top strategic defaulters and we could easily see a situation over the next few years in which 20 MILLION homeowners are either delinquent, defaulted, or in the foreclosure pipeline.
- What a 50% DTI Really Means
- Time-tested 36% DTI Means 60% MORE Disposable Income Each Month
1) What a 50% DTI Really Means?
Borrower Earnings: $100k per year
50% Total DTI: $50,000 per year to housing PITI & all other debt on credit report
25% Fed & State Taxes: $25,000 per year
Disposable income: $25,000 per year, or $2,083 per month
How does this well-above average household SAVE MONEY AND pay for utilities (power, water, cable, garbage, insurance (car, life, health), gas, food, car payment, fuel, clothes, household maintenance and more on $2,083 per month? How do they save an emergency fund or take even a drive-away trip for the weekend?
How do they shop this holiday season when over a trillion dollar in consumer credit was taken away in the past year?
A 50% housing DTI turns the house into the largest investment of your life and ruins most household’s balance sheet at the same time unless the gross income — and disposable income — is much larger.
For most in a serious negative equity position, it is better to walk away. Earning your way out of a $200k hole is impossible with disposable income of $2,083 per month less expenses. Why not walk – the borrower’s credit will be trashed for a few years but as long as they maintain their credit rating on all other credit, their overall rating will not be damaged for as long as their house remains underwater.
2) Now, let’s look at this with 28/36 time-tested debt-to-income ratios.
Bottom Line - 60% MORE disposable income each month.
Borrower Earnings: $100k per year
36% Total DTI: $36,000 per year per to housing PITI & all other debt on credit report
25% Fed and State Taxes: $25,000 per year
Disposable income: $39,000 per year or $3,250 per month
With $3,250 per month, a $100k household can likely save $20k per year. Still, this is not enough to make a real dent in a $200k neg-equity position. But, with this much disposable income the homeowner is not missing out on much and they are saving money, meaning their house is a place to live.
What do households spend money in every year? The U.S. Census bureau provides the answers:
• $200 billion on furniture, appliances ($1,900 per household annually)
• $400 billion on vehicle purchases ($3,800 per household annually)
• $425 billion at restaurants ($4,000 per household annually)
• $9 billion at Starbucks ($85 per household annually)
• $250 billion on clothing ($2,400 per household annually)
• $100 billion on electronics ($950 per household annually)
• $60 billion on lottery tickets ($600 per household annually)
• $100 billion at gambling casinos ($950 per household annually)
• $60 billion on alcohol ($600 per household annually)
• $40 billion on smoking ($400 per household annually)
• $32 billion on spectator sports ($300 per household annually)
• $150 billion on entertainment ($1,400 per household annually)
• $100 billion on education ($950 per household annually)
• $300 billion to charity ($2,900 per household annually)
The average homeowner household spends $22,785 per year, or $1900 per month on the above. When making an allowance for some of the items that are typically financed, the outgo is still roughly $1500 per month.
At 50% DTI, the $100k earner with a disposable income of $2083 per month will have extra monthly income of $583 based upon typical spending. That does not leave a lot for savings, or items not listed such as auto insurance, vacations, gas etc. That definitely is not enough to ‘earn their way out’ of their negative equity hole.
However, the 36% DTI borrower will have an extra $1750 month, which allows for living life and saving money, significantly reducing the chance of loan default due to negative-equity..
Bottom Line - This shows vividly why 50% DTI — even with borrowers making $100k a year and with 20% equity in their property — is in fact over-leveraged and a recipe for loan default for any number of reasons.
- HAMP — More Exotic than Bubble-Years Loans
Now you know why I have been calling HAMP “the most exotic loan ever created” since its inception.
But from the HAMP headlines you could not tell. All that is ever focused upon is the 31% DTI. But that is the front DTI…the housing-only DTI. If you read the guidelines, the back DTI (total debt) allows borrowers to go to 55%!
In fact, if the borrower’s DTI is over 55%, the borrowers are required to go to credit counseling. A little news for ya – a borrower paying out 55% of their gross income to debt does not have time for credit counseling because they have a second job.
Bottom Line: HAMP was designed to lower ‘payments’ for underwater borrowers, but also designed to suck every bit of disposable income every month to the bank. Being underwater in a high-DTI situation is the recipe for default, so it is no wonder the program is not performing as thought.
Borrower’s realize this and are simply using the HAMP multi-month processing and approval process as a way of staying in their home rent-free for a longer period of time. At the end of the day, those that do make it to a permanent mod — but have a high back DTI — will ultimately fail.
For small percentage of those that fit the HAMP sweet-spot, it is great and absolutely the right medicine. However, at what cost? For many that can technically afford the house and would have gone on paying for 30-years — but can’t qualify for a new-vintage refi — a pre-meditated loan default and subsequent HAMP mod is an easy route to a government subsidized no cost refi.
For all of these reasons and more, I believe HAMP will be fundamentally changed in 2010, perhaps to finally include principal balance reductions. Principal reductions are the only way modifications will stick. I hate the idea of any gov’t interference, but if they are going to be spending hundreds of billions anyway, they may as well target it.
I also believe that HAMP will be ultimately responsible for a sizable wave of foreclosures beginning in the near-term from those who do not make it through their trail period, which as of recent data, is most. With foreclosures averaging 80k a month for the past six months and 700k foreclosures held up in the pipeline due to HAMP, even a trial mod failure rate of 40k a month would increase foreclosures by 50%.
However, this is housing market bullish. The biggest threat to the housing market in 2010 is a lack of distress inventory, which is some states still makes up 70% of all sales. Foreclosures are what is in demand and the biggest unintended consequence of HAMP is to keep those who can’t afford their houses in them and others than can afford them away.
- Fannie Mae to Tighten DTI Guidelines
Lastly, the following story talks about a recent move by Fannie to raise minimum credit scores and to lower the max allowable DTI to 45%. Operating in a pro-cyclical manner like this only sucks more liquidity out of the mortgage and housing market, but does make for safer loans in the future. It is also validation that DTI and household leverage — something rarely focused upon any any analysis I have ever read — is beginning to get the attention it deserves.
Fannie Mae to Tighten Lending Standards: Report
Published: Thursday, 26 Nov 2009 | 6:40 AM ET
Fannie Mae plans to raise minimum credit score requirements next month and limit the amount of overall debt that borrowers can carry relative to their incomes, The Washington Post reported on Thursday.
Starting Dec. 12, the automated system that the government-controlled mortgage finance company uses to approve loans will reject borrowers who have at least a 20 percent down payment but whose credit scores fall below 620 out of 850, the newspaper reported. Previously, the cut-off was 580.
Also, for borrowers with a 20 percent down payment, no more than 45 percent of their gross monthly income can go toward paying debts, the newspaper said.
A Fannie Mae spokesman told the newspaper that the limits reflect the company’s recent experience.