What are the consequences for mortgage lenders in getting it wrong?

Factual Questions
1

My understanding is that one contributing factor to the subprime mortgage crisis was that since mortgage lenders were selling virtually all the loans that they underwrote anyway, they had less incentive to do sound underwriting. Someone else was going to take the loss if things went wrong. But I can’t imagine it was as simple as that, or that it still is if it ever was. I would have to think there must be ways that the mortgage purchasers (e.g. Fannie, Freddie, et al) make at least some effort to make sure they’re not stuck with deadbeat borrowers.

I can think of two possibilities. 1) the purchasers track the percentage of loans that default by underwriting lender, and if a particular lender has too loose standards then their default rates will be unacceptably high and they won’t be able to sell them (or sell them at face value, anyway). 2) The purchasers have their own standards which they impose on the sellers, and in the event of default (or perhaps even otherwise) they look to make sure that the underwriting lenders adhered to those standards, and if otherwise, they can penalize the lenders.

I believe the second is true in fact, but I wouldn’t think that can be the extent of it, because if it were so then all the underwriting requirements would be completely consistent for different lenders, when this is not the case. So there must be some incentive for the underwriting lenders to keep the creditworthiness high independent of the standards of the downstream purchasers.

2

Former mortgage post-closer here. My job actually used to be to sell loans to Freddie and Fannie.

Few things in play here:

  1. Yes, every mortgage purchaser has their own guidelines for what loans they’ll purchase. Usually they start with Fannie/Freddie guidelines and add on their own overlays.
  2. Every loan goes through a pre-purchase audit. We send docs over to the purchaser, and they make sure we have everything in a row (paystubs, bank statements, etc.) If we don’t they won’t buy it. There’s a whole department devoted to getting what we need to get to get a loan sold.
  3. Most investors will audit defaults. Sometimes everything in underwriting goes right and the borrower still defaults. Nothing they can do there. But if we were to make some sort of misrepresentation of the borrower’s situation, they can make us buy it back. Again, another department devoted just to that.
  4. And yes, if we were to consistently sell bad loans, we would get caught off and/or get our license revoked.
3

Thank you very much. That’s the best of this MB - a guy who actually knows what he’s talking about, answering the question directly and comprehensively.

4

I feel like we’d really need to understand what % of defaults were sold as conforming in the standard secondary market versus what % were sliced, diced, and repackaged into instruments like Mortgage-Backed Securities and Collateralized Debt Obligations.

In the latter, a car theft ring stole thousands of cars, shipped them over the border to chop shops in Tijuana, and sold them off as (maybe nearly or maybe entirely) unidentifiable bits and pieces.

In the complicated debt securities end of the equation, much of which was sold off to foreign investors, I wonder with what ease individual loans could be identified, much less tracked back to their originators.

The executive summary of the Inquiry from the US Financial Crisis Inquiry Commission (Warning: BIG honking PDF) goes into this at some length.

5

This piece from the report I linked to may directly address your OP:

In theory, every participant along the securitization pipeline should have had an interest in the quality of every underlying mortgage. In practice, their interests were often not aligned. Two New York Fed economists have pointed out the “seven deadly frictions” in mortgage securitization—places along the pipeline where one party knew more than the other, creating opportunities to take advantage. For example, the lender who originated the mortgage for sale, earning a commission, knew a great deal about the loan and the borrower but had no long-term stake in whether the mortgage was paid, beyond the lender’s own business reputation. The securitizer who packaged mortgages into mortgage-backed securities, similarly, was less likely to retain a stake in those securities.

In theory, the rating agencies were important watchdogs over the securitization process. They described their role as being “an umpire in the market.” But they did not review the quality of individual mortgages in a mortgage-backed security, nor did they check to see that the mortgages were what the securitizers said they were.

So the integrity of the market depended on two critical checks. First, firms purchasing and securitizing the mortgages would conduct due diligence reviews of the mortgage pools, either using third-party firms or doing the reviews in-house. Second, following Securities and Exchange Commission rules, parties in the securitization process were expected to disclose what they were selling to investors. Neither of these checks performed as they should have.

ETA: Document page 165. It really gets even more interesting from there.

6

As I understood it from “Market Crash 101” in assorted books about what happened -

The problem was not that banks or assorted F’s bought mortgages. The problem was far deeper. Perhaps someone more expert can chime in (or someone who has read the books more recently).

First, the sellers took a huge assortment of mortgages and glommed them into one bucket. This consisted of a variety of mortgages of various risk levels.
Then, they sliced the loan bucket horizontally into investment bonds - the first X% of the mortgage payments for that bucket was the ground floor, then next was mezzanine, then succeeding “stories” of risk. So if a few borrowers defaulted, the higher floors had a higher risk of not getting paid completely. In return, those bonds paid a much higher rate of return.

Sometimes, they took a bunch of buckets and mixed them together and shared them out into different new buckets to spread the risk - after all, what are the odds an issue would cause defaults in New Jersey, Omaha, and Portland at the same time?

Logic was not bad… at first. OK, you have a collection of high risk borrowers in the bucket, but what are the odds that a lot of the borrowers would default? Safety in numbers. As a result, “more than friendly” rating services gave these bonds very solid ratings.

Trouble was that in the times of low low interest rates (something new at the time) the high rate but good rating bonds were gobbled up. This strong demand caused the suppliers - people who actually gave out mortgages - to increase the supply, as selling the mortgages and then the bond buckets was increasingly lucrative.

Basically, it got to the point where people were being sold mortgages they could not afford- usually on teaser rates (exploding ARM’s, or Adjustable Rate Mortgages). Buy a house now, pay a low rate for two or three years, then the interest goes sky high. This also fueled the house-buying frenzy, then the house building frenzy, then the continuously rising house price frenzy as all those easy mortgages chased fewer houses. People could buy a house they could not afford, but flip it for a profit when the unaffordable high interest rate kicked in a few years later.

Needless to say, the urge to generate more mortgages to feed this frenzy led to shady practices and outright fraud. paperwork was fudged, lies were accepted, to allow people to create mortgages for the purpose of aggregating them and slicing and dicing the aggregates.

To get back to the OP - the people buying these bonds were assured they were far more reliably rated than they were. Often the bonds were on collectives of mortgage aggregates that collected other mortgage aggregates to the point where tracing the original mortgages was impossible; and like other professions, investment managers took the easy way out and took the rating services’ word for it the bonds were reliable. Anyone who actually tried to trace what they were actually buying eventually found the trail too difficult to dive into.

As a result, we presume that there are far stricter controls that mortgages are only given out when the borrower can actually afford it, and that the house used as surety is reliably assessed correctly.

And in the adage of “once bitten, twice shy” anyone who purchases mortgage-backed securities today demands better accountability about the source and reliability of what they purchase. We hope…

7

I though the big driver of the problem was lending to people who were not just buying 1st time houses, but to people who wanted to invest in a 2nd property mostly for rental/investment purposes, and many of these people bought in areas that could not support a surplus of rental properties; in other words buying foolishly. When these foolishly bought properties didn’t move or appreciate in value, people started defaulting on those loans. Society gets stuck with a bunch of crap loans and a glut of houses in many areas with no one interested in them.

8

This is typically long after the original closing, right? We’re talking about the selling of a potentially years old mortgage? If so, these pay stubs and bank statements would no doubt be out of date and I assume you’re not running after the mortgagee to try get updated documents. Is this just a “did you do your due diligence many years back at closing? And did you preserve the records correctly?” check?

9

Different lenders work in different ways, but we start the selling process pretty much soon as the loan closes. But yes, it’s still a matter of confirming we did our due diligence at the time. If someone loses their job two years after they close and then default, that’s not really the fault of the lender.