Will HELOCs Trigger a Banking Crisis?
by Keith Jurow, Capital Preservation Real Estate Report
After disregarding the looming home equity line of credit (HELOC) disaster for several years, Wall Street and media pundits have finally taken notice. Homeowners with HELOCs will soon see them convert to fully amortizing loans and will face a huge increase in their monthly payment. Banks have not set aside adequate reserves to cover a HELOC-driven crisis, which would impact even those investors in broadly diversified index funds.
Here is the problem in a nutshell: Over the next three years, roughly three million homeowners who either could not pay off their bubble-era HELOCs or were unable to refinance into a more affordable one will face soaring payments on these second liens.
Let's see how this problem could impact your clients.
How the HELOC bubble developed
In 2004, housing markets in major metros were on fire, fueled largely by swarms of speculators out to make their fortune. As underwriting standards collapsed, homeowners saw the value of their homes soar. Refinancing to cash-out the growing equity in their homes became irresistible. Banks were eager to accommodate.
This was done in one of three ways: refinance the first mortgage with a larger one; take out a HELOC; or refinance an existing HELOC with a bigger one. In each case, the borrower had money from the bank usable for any purpose.
Take a good look at this chart from Equifax that shows the number of HELOCs that were originated each quarter during the bubble years.
From the beginning of 2005 to the end of 2007, roughly 10.8 million HELOCs were originated. In the fourth quarter of 2005 alone, nearly $130 billion of HELOCs were issued. The average amount of a HELOC loan in those three months was more than $130,000.
When the sub-prime market collapsed in the spring of 2007, originations of HELOCs continued unabated for more than a year through the middle of 2008. According to Equifax, the total of outstanding HELOCs did not peak until late 2009 at $672 billion of cumulative issuance.
Some of these bubble-era HELOCs were taken out by homebuyers as so-called “piggy-back” second mortgages at the time of purchase. This enabled them to purchase a home with little or no down payment. However, the vast majority were originated after the home purchase – usually within two years.
Millions of homeowners were not content with taking out a single HELOC. Many refinanced their HELOCs – some more than once – to take advantage of rising home values and pry still more cash out of the house. California was the center of this HELOC madness. Between 2004 and 2006, roughly six million HELOCs were refinanced around the country.
HELOCs were encouraged and spurred on by lenders. In 2004, the FDIC reported that banks were charging non-use fees on HELOCs that were open but inactive. Some actually penalized homeowners for not borrowing more.
How a HELOC recast works
In the midst of this housing bubble, no one saw the potential danger posed by interest-only HELOCs, which would convert to a fully amortizing loan ten years later. After all, nearly everyone was convinced that home prices would continue rising.
To understand the danger of millions of bubble-era HELOCs recasting to a fully amortizing loan, I need to briefly explain HELOCs.
A HELOC is similar to a business line of credit. Using the residence as security, a homeowner was usually given a line of credit with a prescribed limit upon which the borrower could draw at any time. During the bubble years, some banks offered HELOCs where the available credit increased automatically as the equity in the house rose along with the home’s value.
For HELOCs originated during the bubble years, the homeowner had a period of anywhere from five to 10 years when funds could be drawn. During this time, the borrower was usually required to make only interest payments. The rate was adjusted monthly and was pegged to the prime rate.
At the end of the 10-year draw period, the loan converted to a fully amortizing one. The repayment period was typically between 10 and 20 years – usually 15 – at the end of which the HELOC had to be fully repaid.
HELOCs lured millions of homeowners between 2004 and 2007 because the interest-only monthly payment was not very much – often only a few hundred dollars. Why worry about the fact that in 10 years the loan would become fully amortizing when home values were soaring by double digits?
How many HELOCs will be recasting to fully amortizing loans?
Take another look at the Equifax chart showing quarterly HELOC originations. The annual origination figures look like this:
2005 – 4.5 million
2006 – 3.4 million
2007 – 2.9 million
That is a total of 10.8 million HELOC originations during the three peak bubble years. If you add the 2008 originations, that would increase the total to 12.5 million. Nearly 40% of these bubble era HELOCs were opened in California.
Do we know how many bubble-era HELOCs will be recasting into fully amortizing loans? The estimates of the major data firms are all over the place. Last March, RealtyTrac released a HELOC Resetting Report that found that roughly 3.26 million HELOCs with a remaining balance of $158 billion were still active and scheduled to recast between 2015 and 2018. They also claimed that 56% of these properties were seriously underwater.
Two months later, credit-reporting firm Experian published a study that found that a much-higher $265 billion in bubble-era HELOCs would recast during this same four-year period.
According to Equifax, there were 11.2 million HELOCs outstanding in August 2015 with a remaining balance of roughly $497 billion. We also learn from Equifax that a total of just under six million were originated between 2009 and 2014 and another one million in 2015. That leaves roughly 4.2 million HELOCs still outstanding from the four peak-bubble years.
As I noted earlier, roughly six million HELOCs were refinanced between 2004 and 2006. This means that nearly all the pre-2004 HELOCs no longer exist. Furthermore, the HELOC refinancings in 2005 and 2006 wiped out most of the 2004 HELOCs as well. Hence what remains from the bubble-era are HELOCs originated during the peak home price years of 2005 through 2008.
How much of that $497 billion in outstanding HELOCs are from the 2005-2008 bubble era? You would think it would be less than half given the number of remaining bubble era HELOCs. Don't be so sure. During those three years, HELOCs with enormous credit lines were originated -- in California especially. Here is a typical example from the Irvine Housing Blog, now called the OC Housing News:
- On 3/11/2004 the wife appears alone on the title, and the first mortgage is $999,800.
- On 8/30/2004 she refinanced with a $1,000,000 first mortgage.
- On 12/28/2005 she refinanced with a $2,170,000 first mortgage.
- On 2/1/2006 she got a HELOC for $250,000.
- On 8/22/2006 she refinanced with an Option ARM for $2,500,000.
- On 11/15/2006 she opened a HELOC for $490,000.
- On 8/1/2007 she refinanced with another Option ARM for $3,225,000.
- On 10/22/2007 she opened a HELOC for $500,000.
- Total property debt is $3,725,000.
- Total mortgage equity withdrawal is $2,725,200 during a four-year stretch.
As I have reported in several articles over the past five years, HELOCs of $200,000 to $500,000 were quite common in California. Thus, it is safe to assume that more than half of the $500 billion of outstanding HELOCs are from 2005 through 2008. Experian's estimate of $265 billion may be fairly accurate and possibly too low.
How ugly can it get?
Let’s examine the situation of a real borrower whose plight was discussed in an August 2014 article in the Los Angeles Times. The homeowner took out a $167,000 HELOC in 2006 on his Huntington Beach, California condo. He received a notice from his mortgage servicer that the HELOC would convert to a fully amortizing loan in July 2016, and his monthly payment would soar from $400 to more than $1,100.
The shocked homeowner said, “We both now live on fixed income and will not be able to make the payment.”
In this LA Times article, the authors tried to minimize the concern over the fallout from these HELOC resets. They concluded by citing Dean Baker, the co-director of the Center for Economic and Policy Research. Baker said, “I don’t see it shaking the financial system....This isn’t anything like we saw in 2008 and 2009.”
How reasonable is this assertion?
Pundits who tell us not to worry often say that lenders have already written off many, if not most, of the bubble-era HELOCs. Is this true? Take a look at this graph from Equifax.
Since 2009, the annualized write-off rate for all outstanding HELOCs has been very low. Even in the two years immediately following the implosion of the financial system, 2009 and 2010, the rate never exceeded 4%. Before you breathe a sigh of relief at the declining write-off rate, understand that this was partly due to the origination of post-bubble HELOCs where the delinquency rate has been very low.
This next graph from Equifax shows the cumulative dollar write-off rate broken down by when the HELOC was originated. Even for the worst vintages in 2007, the overwhelming majority of HELOC loans have not been written off and are still outstanding.
The total balance of outstanding HELOCs declined from $672 billion in the summer of 2009 to roughly $497 billion as of August 2015. Isn't that a good sign?
My answer is no; Not all of this decline was due to the write-off of delinquent or defaulted HELOCs. Many bubble-era HELOCs were rolled into a refinanced first lien.
As Sam Khater, chief economist for CoreLogic describes in a blog post in November 2014, there is an even worse problem of recasting HELOCs into amortizing loans. Khater pointed out that 27% of all outstanding HELOCs were taken out by borrowers who utilized only 10% of the line of credit and that some of these borrowers had pre-paid their loan. Thus a significant portion of the decline in outstanding HELOC balances was due to borrowers paying off their loans early.
This means that the least problematic HELOCs with the lowest balances have already been removed from the balance sheets of the large banks. That leaves roughly three million large HELOCs from the bubble years whose borrowers either could not pay off their loan or were unable to refinance into a more affordable one.
As the graph of write-offs by vintage year shows, approximately 85% of bubble-era HELOCs are outstanding and will be recasting over the next three years into fully amortizing loans.
Put yourself in the shoes of a borrower whose HELOC will soon recast. Assume that your HELOC servicer notifies you that the reset will occur this year, that it will fully amortize over 15 years and that the monthly payment will double.
Since you bought or refinanced close to the housing market peak, your home is almost certainly still underwater. Should you take a chance that your local housing market will strengthen and that eventually you will rebuild some equity?
Those lofty price expectations you had when you bought the house were only a fantasy. Unfortunately, you are stuck with the terrible burden of decisions which you made back then.
You have read that Wall Street is confident that the worst is over and that the housing market is recovering. That is not reassuring because your gut tells you otherwise. So you ask yourself these questions: What if Wall Street’s optimism is, again, misplaced? Do I really want to continue paying off my two burdensome mortgages for the next 15 years?
Millions of homeowners will be asking these questions in 2016, 2017 and 2018. Do Wall Street economists have any idea how many HELOC borrowers from the bubble years will decide to default? I doubt it.
Toward the end of 2013, Bank of America conceded that 9% of the outstanding HELOCs that had already converted to fully amortizing loans were no longer performing. That did not include any of the bubble-era HELOCs from 2005 through 2008. Around the same time, Equifax reported that 5.6% of HELOCs originated in 2003 that had recast were delinquent. Since they were taken out three years before the housing bubble peaked, many of these properties probably had some equity when they recast.
In the blog post referred to earlier, Sam Khater of CoreLogic published graphs showing that the delinquency rates of HELOCs originated from 2001 to 2004 quadrupled within a few months after recasting to amortizing loans.
The massive problem of underwater HELOC borrowers
Millions of bubble-era HELOC borrowers have properties that are severely underwater. The RealtyTrac HELOC report mentioned earlier confirms my assertions. According to this report, 56% of the bubble-era HELOCs that will recast are attached to underwater homes.
According to Black Knight Financial Services, roughly 45% of all 2006 home purchasers took out second liens either at the time of buying the house or within a few years after buying. A 2012 study by the New York Federal Reserve Bank found that the total loan-to-value (LTV) ratio on all properties bought throughout the country during the bubble-era that had a second lien was at least 95%. Thus it is safe to conclude that the vast majority of bubble-era borrowers who took out a HELOC ended up with equity of less than 5%.
A March 2013 study entitled “Equity Extraction and Mortgage Default” put out by the Federal Reserve Board clarified the issue. The author had access to CoreLogic's database of all mortgage liens for every Los Angeles County single-family home from 2000 through 2009.
He focused on purchases by owner-occupants between 2002 and 2004. He discovered that 50% of those who borrowed with a first lien also took out a piggy-back second mortgage at the time of purchase. Furthermore, 45% of these buyers subsequently extracted equity with a cash-out refinance.
What the author reported about mortgage defaults is extremely important. Of homeowners who purchased between 2000 and 2003, 25% of those who later defaulted had properties with LTVs of 140% and 10% of them had ratios of more than 170%. His conclusion was that equity extraction through a cash-out refinancing of either the first lien or the HELOC accounted for roughly 80% of all defaults between 2006 and 2009. Equity extraction was a far greater cause of default than job loss, reduced income or anything else.
Given what has happened to home prices after the bubble collapse, it is very reasonable to suggest that – at least in the major metros – 80-90% of these bubble-era HELOC borrowers are now underwater on their property.
RealtyTrac's report apparently confirms this conclusion as well. According to their report, the two states hit hardest by the housing collapse – Nevada and Arizona – showed underwater figures of 84% and 74%. Two other states where prices plunged – Florida and Illinois – reported underwater rates of 71%. California – the state with the most recasting HELOCs – had an underwater figure of 66%.
How does this HELOC mess impact investors?
How might the impending HELOC recast disaster affect your investor clients? Let's examine one of the “too big to fail” banks – Wells Fargo.
In its latest FDIC call report for the third quarter of 2015, Wells Fargo showed $67.3 billion in HELOCs on its balance sheet. That is down substantially from the $108 billion in its portfolio at the end of 2009. Clearly, the vast majority of HELOCs reported in 2009 were from the bubble era.
How did Wells Fargo manage to reduce its outstanding HELOC balance by more than $40 billion? Was it by writing down the delinquent ones? No way. Total charge-offs of HELOCs for the first nine months of 2015 were a mere $437 million. They weren't sold off to investors either. It is a mystery.
What is the bubble-era HELOC portfolio of Wells Fargo worth? Their call report for the third quarter of 2013 reveals a great deal. For the first nine months of 2013, Wells Fargo charged off slightly more than $1 billion of HELOCs and had recoveries of $147 million. That equates to roughly 15% of the outstanding balance.
Some of the HELOCs charged off may not yet have been sold to investors by the end of the third quarter. Nevertheless, these figures clearly show that the market value of the HELOCs on underwater properties is substantially less than the outstanding balance. The banking regulators have never forced Wells or any other “too big to fail” bank to mark down their HELOC portfolio to its market value.
What about the delinquent status of its HELOC portfolio as of the third quarter of 2015? The FDIC call report showed that $1.94 billion of these HELOCs was not accruing interest. Another $522 million was delinquent but still accruing interest. That is a total delinquency rate of 3.6%. Is that credible? Not to me.
Unfortunately, we know nothing about when the loans in the HELOC portfolio of Wells Fargo will recast into fully amortizing loans. Although that would be helpful to learn, the banks do not have to report this in their FDIC call report. However, it is safe to assume that the vast majority of its bubble-era HELOCs are 2005-2007 loans and – as I have shown – perhaps 60-80% of them are collateralizing underwater homes.
Because these are second liens, the HELOC on an underwater home becomes essentially worthless if the borrower defaults. The bank will eventually take a total loss on the loan.
A HELOC disaster for Wells Fargo would not include its enormous $255 billion first-lien portfolio composed largely of non-guaranteed jumbo mortgages. The delinquency rate for these first liens is nearly 11%.
This analysis applies to J.P. Morgan Chase and Bank of America, whose HELOC portfolios are very similar to those of Wells Fargo.
Although your clients probably do not hold any individual bank stocks, they may own financial ETFs. PowerShares' KBW Bank ETF (KBWB) is a pure play on banks. Forty per cent of its holdings are five large banks – J.P. Morgan Chase, Wells Fargo, Bank of America, Citigroup and U.S. Bancorp. Because the banks have been heavily hit by selling for months, this ETF is down 24% as of February 5 from its 52-week high. It is likely to face substantial further declines.
Another vulnerable ETF is State Street's S&P Bank ETF (KBE). Although the “too big to fail” banks do not make up as large a portion of the portfolio as in KBWB, the top 10 bank holdings make up more than 25% of the portfolio weighting. Nearly all of these ten stocks are down more than 25% from their 52-week highs.
Perhaps the greatest exposure facing your clients is with the largest ETF of them all – State Street's S&P 500 ETF (SPY). As of early 2016, roughly 16% of the entire weighting of the S&P 500 Index is the financial sector – dominated by the large banks.
Investors use this ETF as a way to gain broad exposure to equities. Daily volume of trading in SPY has been enormous – especially on down days – and has often exceeded $50 billion. It has become a favorite trading tool for speculators. In the decline last August, and again last month, volume on the market's down days has far exceeded that on up days. That has been a warning signal ignored by most investors.
Early February action in the stock market indicates that the market decline may be resuming in earnest. If this occurs, traders may dump the SPY in a big way and leave your investor clients holding the bag.
I strongly urge you to consider whether the risk of HELOC-driven problems is accurately priced into the ETFs you and your clients own.
This was published previously at Advisor Perspectives 22 February 2016.
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.p>