Everyone knows that real estate markets are in recovery. Right? Everyone except me – the one holdout.
For five years, I’ve shown readers that the so-called real-estate recovery is only a mirage that disappears when you approach it. Wall Street and most pundits disregard the data, charts and graphs I have published. They stick to their familiar sources regardless of whether they are accurate or reveal anything useful.
Let’s look at some recent data that shows the extent of the delinquencies in these mortgages with a focus on the most exposed large metro areas. We’ll then examine the implications for advisors whose clients own funds, ETFs or REITs that hold securities with underlying RMBS mortgages.
The clear failure of mortgage modifications
The Obama administration unveiled its Home Affordable Modification Program (HAMP) in the spring of 2009. Since then, “do something” advocates have pinned their hopes on mortgage modification as a solution to the mortgage delinquency crisis that erupted out of the housing crash. Let's see how it has fared.
According to HOPE NOW, an alliance among companies in the mortgage industry, roughly 7.5 million permanent mortgage modifications have been granted by mortgage servicers since late 2007. This includes both the government-sponsored HAMP modifications and what became known as “proprietary modifications” offered by the mortgage servicers.
What you are probably unaware of is that another 14.1 million “workout plans” have also been provided to borrowers who claimed a hardship. Some of these plans include repayments, payment reductions and forbearances. Spokespersons for Fannie Mae have told me the purpose of these plans is to enable delinquent borrowers to keep their homes.
How successful have all these plans been to date? Not very. In April 2013, the Treasury Department's overseer of the HAMP program reported that more than half of these modifications had been canceled because the borrower failed to meet all the requirements for maintaining the modification. Modifications from the peak year of 2010 were showing re-default rates of more than 30%. One of the main reasons for this high failure rate is that many of the HAMP participants still had very high total debt-to income ratios even after modification.
A year earlier, Trans Union, a credit-reporting firm, issued the results of a study based on an examination of roughly 600,000 borrowers from its database who had received a mortgage modification sometime between January 2008 and July 2011. It found that nearly six out of every 10 borrowers had re-defaulted within 18 months of receiving the modification.
The most recent research from the Federal Reserve Board is an article abstract that was posted in December 2014. It took a detailed look at 60,000 private securitized loans in the database of Loan Performance Securities (a subsidiary of Core Logic). These loans had been modified after January 2008. As of the end of 2013, only 38% of the loans modified in 2010 were still current. The year 2010 was when the number of modifications peaked.
Modifications from 2011 fared little better. Roughly 54% of these modified loans were current at the end of 2013.
Why is the re-default rate so shockingly high? One important factor is that the vast majority of modification recipients had either purchased a home or refinanced one during the bubble years of 2004-2007. Hence most of them owned properties that were still underwater. Underwater homeowners default at a much higher rate than those with equity in their property.
Another key reason is that there are concrete benefits for defaulting and even re-defaulting. It is the delinquent homeowner who receives a modification. According to TCW's latest Mortgage Market Monitor, servicers wait an average of 15 months before modifying a delinquent non-agency prime jumbo mortgage.
Most discouraging to owners of the non-agency RMBS holding these loans is that the re-default rate is getting worse. According to Black Knight Financial Services' April 2015 Mortgage Monitor, the recipients of roughly 70% of all new trial modifications and repayment plans had already been through one or more home retention actions. That number was only 45% in 2011.
Black Knight also showed this deterioration in a different way. The percentage of properties in active foreclosure where the borrower had already been through at least one home retention action has risen steadily over the last five years. That number climbed from 26% in 2010 to 53% in 2015.
Deadbeat borrowers have not paid for years
For several years, I have documented the incredible number of delinquent borrowers in New York City and Long Island who have received pre-foreclosure notices from mortgage servicers warning them of the possible loss of their homes. The figures are so high that many readers probably dismissed them as exaggerated … or even made up. I can understand this skepticism.
Fitch Ratings – one of the three major credit rating firms – has recently released figures that confirm what I have been asserting since 2011. Take a good look at this table, which shows how long the deadbeats have been getting away with no longer paying their mortgage.
Serious Delinquency of Non-Agency RMBS Loans
Ten Worst States
Hawaii62%53.9 months
| State |
Per Cent of Loans Delinquent for More Than 4 Years |
Average Number of Months Loans are Delinquent |
| US Overall |
35% |
37.5 months |
| Hawaii |
62% |
53.9 months |
| New Jersey |
59% |
52.3 months |
| Florida |
57% |
51.6 months |
| New York |
56% |
52.5 months |
| Washington, D.C. |
52% |
46.7 months |
| Vermont |
41% |
41.5 months |
| Nevada |
40% |
41.4 months |
| Massachusetts |
39% |
41.3 months |
| Maryland |
39% |
39.4 months |
| Oregon |
39% |
40.2 months |
Source: Fitch Ratings using Loan Performance Securities data; graphics by A.J.Kaps Group
More than one-third of all the delinquent non-agency mortgages have been seriously delinquent for more than four years. Yikes! Moreover, 26% of them have been delinquent for more than five years.
While shocking, these numbers really don't surprise me. Delinquent homeowners with underwater properties realize that it might be four or five years before their servicer moves to foreclose on their house.
Servicing Management, a mortgage servicing trade magazine, described horror stories of maintenance contractors and inspectors entering some of these seriously delinquent properties in a July 2015 article entitled “Tales from the Field.” They discover dead animals, garbage and debris, attacks by vicious pets and ferociously angry occupants. One vacant property had been rigged to explode when anyone flipped the light switch. You can only imagine the condition of these vacant properties.
I just received updated figures on the non-agency RMBS universe from Black Box Logic, a provider of loan-level RMBS data. As of July 2015, their database had just under 3.8 million active loans. Of these, roughly 667,000 (17.6%) were either in foreclosure, seriously delinquent, in bankruptcy or had already been repossessed.
This percentage is for the entire nation. Figures for the worst major metros are much higher. Furthermore, one-third of all the active loans have been previously modified. The re-default rate for these modified non-agency loans has been 30-60%; a similar re-default rate for active loans will ensue.
The big problem is in roughly 20 major metros
While the pundits point to the declining nationwide delinquency rate published by the Mortgage Bankers Association, a trade organization representing the real-estate finance industry, that figure hides the location of the coming delinquency disaster.
In June, my housing-market article focused on the very high delinquency rate of the residential mortgage portfolio of the “too-big-to-fail” banks. Now I want to focus on a problem at least as large: the high delinquency rate of non-agency mortgages in the largest major metros.
I have an excellent data source for these non-guaranteed loans – Black Box Logic. As of July, the total outstanding balance of these mortgages was roughly $701 billion. This table shows the delinquency rate of the 10 worst major metros.
Serious Delinquencies of Non-Agency Mortgages
Ten Worst Major Metros
| MAJOR METRO |
OUTSTANDING MORTGAGE BALANCE |
SERIOUS DELINQUENCY RATE |
| NEW YORK CITY |
$74.5 Billion |
37.70% |
| MIAMI |
$26.0 Billion |
33.80% |
| TAMPA |
$7.3 Billion |
32.80% |
| ORLANDO |
$6.6 Billion |
30.20% |
| HONOLULU |
$3.0 Billion |
29.80% |
| PHILADELPHIA |
$9.6 Billion |
29.50% |
| PROVIDENCE |
$3.2 Billion |
28.00% |
| LAS VEGAS |
$8.8 Billion |
27.70% |
| CHICAGO |
$17.8 Billion |
26.30% |
| BOSTON |
$10.2 Billion |
25.00% |
Source: Black Box Logic, graphics by A.J. Kaps Group
The total mortgage balance outstanding for all of these metros is $167 billion. Nearly half of that lies in the New York City metro – with roughly 19 million residents. Nearly 24% of the entire nationwide balance of non-agency loans was originated in these 10 large metros. This is where the worst of the delinquency problem is situated. If you extend it a little a wider, the worst 25 metros hold just shy of 56% of all non-agency securitized loans.
You may be surprised that no California metros appear on this list. All the major California metros have delinquency rates much lower than the 10 worst. For example, the delinquency rate for the Los Angeles metro is only 13.7%. Yet, California led the nation in non-agency mortgages originated between 2005 and 2007. Underwriting standards collapsed during these three bubble years. Why is the California’s delinquency rate so low?
According to Black Box Logic, as of July 2015, 40.2% of all active California non-agency mortgages had been modified. That percentage has risen steadily from January 2011 when it was only 16.9%. Delinquent loans that are modified become “current,” which explains California’s low delinquency rate. A growing number of these modified California first liens will re-default.
This mortgage modification “game” cannot go on indefinitely.
The housing bubble was very concentrated in roughly 50 major markets. That is where an extremely high percentage of the bubble-era buying and refinancing took place. According to Black Box Logic, nearly two-thirds of all the non-agency securitized mortgages were originated in these metros. It is where the greatest price increase took place during the 2004-2007 housing bubble. It is also where the bulk of the bubble-era jumbo mortgages on the books of the too-big-to-fail banks are situated.
As I have shown in this article, millions of mortgage modifications and other loan workouts have done nothing to solve the deep-rooted problem of defaulting borrowers. The 10 worst major metros have serious delinquency rates of 25-38%. In several articles, I have explained that Standard & Poor's now expects 100% of these seriously delinquent loans to default.
What will happen to the one-third of all these mortgages that have been modified? Or the 40% of California loans that have been modified and are considered current? Past history tells us the answer: Roughly 30-60% of them will re-default.
Is your fund or ETF exposed to these risks?
I have analyzed the mortgage delinquency mess in several articles. But how does it impact you and your clients?
Many advisors have put their clients into what are often called total-return bond funds. Let's take a look at one of the largest and most popular – DoubleLine's Total Return Bond Fund (DBLTX), managed by Jeffrey Gundlach. I’ll use this fund as an example because of its size and popularity, but my analysis applies to any fund that owns securities with similar real-estate exposure.
Gundlach started the fund in 2010 and has overseen its growth to roughly $47 billion in total assets. He has owned non-agency mortgage-backed securities (RMBS) for quite some time.
As of the end of July, 22.3% of the fund's assets were invested in these non-agency RMBS. That comes to roughly $10.5 billion. These RMBS tranches are neither guaranteed by the government (i.e., by the GSEs) nor insured by the FHA.
Nearly all of these non-agency mortgages were originated during the bubble years of 2005-2007 when underwriting standards collapsed. That is why nearly 18% of the fund's assets are below investment grade (i.e., junk) and another 5.1% no longer have any credit rating. Without a doubt, these are the non-agency RMBS.
Isn't having such a large percentage of the portfolio invested in the worst mortgages dangerous for a bond fund? Some will insist that this is not a big concern, because the fund managers have the credit risks of the non-agency portfolio under control because the tranches were bought at a substantial discount. That is true. But, there are two key questions to consider.
Have the fund's managers underestimated the credit risks of these bubble-era mortgages?
Did the fund purchase the non-agency portfolio at a deep enough discount to cover the full risks of defaulting mortgages?
Let's start with the first question. The price for a given tranche depends primarily on the ultimate default rate during the life of the security. The key word is “expect.” It is the best estimate of what will happen over the next several years.
Over the last few years, I have written several articles on the RMBS market. I described in detail how Standard & Poors (S&P) has had to revise its default projections several times because it had seriously underestimated the problem.
My most recent article on RMBS – published in February – discussed the latest revision that S&P had to make. An S&P report published in April 2013 revealed that of the 7,111 non-agency tranches still rated AAA in 2012, only 1,119 remained there a year later. Over 85% had seen their rating downgraded and more than 10% of them had lost their rating completely.
The question investors must ask is if your fund manager’s estimate of the total default rate on a particular tranche uses S&P's assumption that 100% of all seriously delinquent mortgages will ultimately default and will have to be liquidated. Remember that for the 10 worst metros, the delinquency rates are 25-38%.
Keep in mind that one-third of all the active non-agency mortgages have been modified and that the figure is 40% for California loans. Do the fund managers assume (as I do) that 30-60% of all these modified mortgages will also end up re-defaulting?
Let's stay with the California loans. These are almost entirely bubble-era mortgages with the worst underwriting standards. The vast majority of the prime loans are jumbo mortgages with interest-only payments. Hence a majority of these prime loans are still underwater.
The great majority of all California non-agency mortgages underwritten in 2006 and 2007 were “stated income” loans where the borrower did not have to verify his or her income. It is well documented that fraud was widespread with these loans. Roughly 20% of all non-agency loans were given to borrowers who lied about intending to live in the property. They were speculators, not owner-occupants. Does the funds’ team build all of this into their assumptions? I have my doubts.
I probably disagree with most fund managers on where housing markets are headed. For four years, I have been presenting evidence that the housing crash was merely interrupted in 2009 and that it will soon resume. That is not consistent with the views of Gundlach or most fund managers.
If housing prices start heading south again, my assumption is that underwater borrowers will resume what used to be called “strategic defaults.” They will walk away from their mortgages rather than wait to regain equity.
Investors need to assess whether those assumptions are consistent with those of their fund’s manager.
A final consideration is what the so-called “loss severity” on repossessed properties will be. As I stated earlier, many of the long-term seriously delinquent properties around the country are in abysmal condition. That is also the case with bank-owned properties (REO). Many are currently uninhabitable. This number will grow. Has your fund manager’s team factored in worst-case loss severities that take all this into account?
Mortgage REITs
What I have said about Gundlach's non-agency RMBS portfolio also applies to some of the larger mortgage REITs with substantial holdings of non-agency tranches. I have discussed several of them in an earlier article – including Annaly (NLY), American Capital (AGNC) and Two Harbors (TWO).
It is very likely that portfolio managers for these mortgage REITs invested in a manner similar to Gundlach's Total Return Fund – and investors must ask if their credit risk assumptions were too optimistic and they are overpaying for their non-agency tranches.
My conclusion
My analysis of where non-agency securitized mortgages are headed argues for lightening or liquidating long positions in funds, ETFs and mortgage REITs with substantial holdings of non-agency RMBS.
A skeptic might respond with this question: What if your assessment is wrong, Keith?
My reply is simple: What if I am right? Given what is now occurring in U.S. stock markets, the prudent course for an investor is to sell these assets before we see a repeat of 2008-2009.
Keith Jurow is a real estate analyst and former author of Minyanville’s Housing Market Report. His new report – Capital Preservation Real Estate Report – launched in 2013.
Read more articles by Keith Jurow
