HousingWire Propaganda Not To Be Believed, Part 1: Re-Analyzing the Data

By | December 3, 2012

Friday HousingWire ran a six-and-a-half page big bank/mortgage servicer propaganda piece called “Living Large“, by Tom Showalter. The article, subtitled “A person’s lifestyle plays into whether they will pay their mortgage after a loan modification”, purports to explain why people default on loan modifications. Instead, it spins a bank-exonerating morality play not justified by the data supposedly being interpreted.

I’ll get to the morality play and the “irresponsible borrower” propaganda it represents in my next post to keep this one to readable length. First, to clearly show the wrongness of the bank-serving mythology being sold as its interpretation, I’m going recap the data the ‘article’ presents to answer the questions the underlying study apparently aimed at: why did so many people with mortgage mods made in 2009 default on those mods by 2011? And what needs to be done to make mods more successful going forward?

Note: I can’t assess the data quality because I don’t have access to the underlying tables and sourcing info; I am working off HousingWire’s/Showalter’s analysis of it. I just take his numbers at face value, though as I discuss below, however, something is screwy either in the some of the data or Showalter’s reporting of it.

Regardless, after looking at the data as presented by Showalter, the answer’s clear: People defaulted because the loan mods weren’t steep enough to make them solvent. To get successful mods in the future, payment reduction needs to be enough for borrowers to be solvent again. That’s all. Importantly, a post-mod solvent borrower is what would’ve happened if the loan restructuring were allowed to happen in bankruptcy; we’ve long understood what dealing with debt crises takes.

The Data: Tremendous Financial Stress, Inadequate Modification

So let’s look at the data as described by Showalter. Here’s the setup:

A control group of a random sample of 1 million people with mortgages in 2009, skewed to be a little more than 50% subprime, Alt-A and jumbo prime, was compared to a random sample of 55,000 loans modified in 2009 and tracked through 2011. (Showalter does not say if the modified loan sample is a subset of the control group or not.) The debt loads, bill paying performance and other characteristics of the people in both groups, and of people within the modified group, were compared. In fact, the people in the modified group were broken out into seven different groups, A through G.

So what do we learn about these various groups? The metric most discussed is  who was seriously delinquent–60-days or more past due–on credit lines from 2007-2009. This data point is fixated on because, according to the analysis, it is the overwhelmingly strongest predictor of defaulting on a 2009 modification by 2011.

Turns out the baseline borrower–the average borrower in the million person sample of people with mortgages in 2009–was seriously delinquent with 16% (1 in 6) of their “active credit lines” sometime in 2007-09. Think about that: the average borrower with a mortgage in 2009 became seriously delinquent on one or more accounts in 2007-09. That is quite a backdrop of financial stress.

So how do the people who qualified for a mortgage modification compare? Unsurprisingly, mod recipients were even more debt stressed: 30% of their accounts had been seriously delinquent during 2007-09. Mind you, they’re not deadbeats; they were keeping 70% of their accounts current. Still, they were struggling much more than the average mortgage holder.

What about the people who defaulted on their loan mod by 2011? Turns out they were even worse off, having been on average behind on half (51%) of their credit lines during 2007-2009. No surprise there–people with the most debt stress at the start continued to have the most trouble paying their bills.  Still, that’s not the whole story.

See, the data show that the tipping point at which “loan modification redefault performance erodes substantially” is serious delinquency on 25% of accounts in 2007-09. But wait–the average for all modified borrowers is 30%! There’s only conclusion possible: the mods were systematically inadequate, unable to address the financial crisis faced by the average person they purported to help.

Redefaulting Data Screwy as Reported

How likely to redefault? Depends on the subtype. At best one in ten redefaulted (11%, subtype F, representing 2% of the sample), and then one in four (26%, subtype G, 3% of the sample). That leaves 95% of the modified borrowers redefaulting at a higher rate, and also signals that the data are screwy, at least as presented by Showalter.

See, he says the average redefault rate was 30%. Group A, which came in at 30% redefault rate, represents 22% of the sample. If 5% of the sample is below the average, and 73% of the sample above, shouldn’t the average be higher? It’s not like the other rates were very close to the average. Subtype C comes in at 34%, D  at 37%, and E and B, which combined account for 56% of the loan mods, redefaulted at 41% and 51% respectively. No way that average is 30%; either Showalter or the data are just wrong.

UPDATE: The sentence Showalter writes is “These borrowers…redefaulted at below average rates (average redefault rate <30%).”   I read that as meaning the average redefault rate is 30%. Otherwise the 30% is a meaningless, arbitrary number. But considering that a 30% average doesn’t make sense, per the above, and given how incoherent the data presentation was generally, maybe Showalter doesn’t mean the average is 30%. Maybe he picked 30% just because it was the first round number above 26% and 11%, the redefault rates of the groups in question, and the average is in fact higher. Doesn’t really matter.

The Mods Were Inadequate

It’s not just the pre-mod financial stress data that shows mods simply weren’t substantial enough to work. It’s the other data Showalter fixates on, namely what people did with the cash freed up by the mod. Given the “Living Large” headline, I expected to read that redefaulting, ‘bad’ borrowers blew their mod-freed up cash on shopping sprees, decking out their lives with flat screen tv’s, jewelry and vacations. Living Large is about self-indulgent pleasure chasing, isn’t it? But here’s the “lifestyle” choice that Showalter says make some borrowers good, and some bad: whether the borrowers decided to use the cash to pay Peter (the mortgage bill) or Paul (the revolving debt/credit card bill.) Seriously.

Showalter writes:

“the borrowers in the high redefault group applied their discretionary cash to their distressed revolving debt, significantly reducing the proportion that rolled to a more serious level of delinquency. …The low redefault group behaved much differently. By 2011 they were letting their revolving debt role to more than 60 days overdue at a much higher rate.”

Since when is someone “Living Large” by paying down some bills while falling behind again on their mortgage?  Snark aside, here’s the key point: both groups–the high redefaulters and low redefaulters–are still insolvent after mortgage modification. They still can’t pay all their bills even though they are trying to.

Showalter talks about high redefaulters paying down distressed revolving debt. He doesn’t claim the high redefaulters rang up large new debts with the cash, just that they chose to pay existing creditors other than the mortgagee.

Mods Should Make Borrowers Solvent

Let’s remember the context for these mods: our nation faced, and faces, a housing and foreclosure crisis triggered by the collapse of a lender-inflated, fraud filled housing and housing-related securities bubble. The crisis has been exacerbated by the unemployment crisis of the Great Recession. Our government stepped in to help, and responded quickly to save the bankers and Wall Street. The big help for Main Street was supposed to be loan modifications/refinancings.

Seen through the lens of public policy aimed at avoiding foreclosure, the inadequacy of the mods is even more obvious. Why did the government let the banks do mods that left people still insolvent?

But wait, banks might say: it’s not fair to put the homeowner debt reduction burden solely on the mortgage; why should our payment be cut so much that the borrower can service all their debts post-mod? Why not force the other debt lines to take a hit too? A variation on this whine is: but wait, by reducing the mortgage payment so much, you’re hurting pension funds and other investors, folk we don’t want to hurt.

Here are my responses:

Dear Bankers:

If you hadn’t falsely inflated principal balances by suborning appraisal fraud and adding fake demand to the marketplace by abandoning underwriting, thus directly and fraudulently increasing the homebuyer’s debt load;

If you hadn’t engaged in widespread predatory lending including steering people into more expensive mortgages when they qualified for cheaper ones, thus directly and fraudulently increasing the homebuyer’s debt load;

If you hadn’t engaged in predatory servicing, misapplying payments, forcing borrowers to purchase grotesquely priced insurance policies (even when they were maintaining insurance of their own), and charging rolling late and other junk fees, thus directly and fraudulently increasing the homebuyer’s debt load;

If you hadn’t worked so hard to kill mortgage-loan restructuring in bankruptcy, which would have led more people into bankruptcy and thus forced other creditors to take a hit too;

If you hadn’t defrauded homebuyers-as-taxpayers by making false mortgage insurance claims;

If you hadn’t defrauded homebuyers-as-pension-fund-participants by lying about mortgage-backed securities to the investing pension funds;

If you hadn’t criminally manufactured documents to foreclose on people’s homes;

It would have been enough to justify you lowering the mortgage payment to make borrowers solvent that you were bailed out with tax dollars and not held accountable for any of your crisis-causing misdeeds.

Dear Investor Protectors:

The issue isn’t whether the mortgage payment should be modified sufficiently to allow borrowers to become solvent. It’s who should pay for that modification. I agree with you: the banks should pay. There’s nothing about modifying the amount of the payment due you that comes out of the borrower’s pocket that requires investors to take the hit for the difference, except investor passivity and comparatively (to the banks) weak political power.

The Bottom Line

Insolvent people will default on debts owed, making the choice that seems most rational to them about whether to pay Peter or Paul. Public policy aimed at keeping people in their homes based on loan mods is a total failure if the loan mods leave borrowers insolvent. And apparently the 2009 loan mods did just that. And that failure is unconscionable against the backdrop of banker wrongdoing and government solicitousness of bankers.


Leave Your Comment

Your email will not be published or shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>