We should remember that we started somewhere

[drwex]

Someone sent this around the office recently.

See http://query.nytimes.com/gst/fullpage.html?res=9c0DE7DB153EF933A0575AC0A96F958260&sec=&spon=&pagewanted=all for the full thing.

September 30, 1999
In a move that could help increase home ownership rates among minorities and low-income consumers, the Fannie Mae Corporation is easing the credit requirements on loans that it will purchase from banks and other lenders.

The action [...] will encourage those banks to extend home mortgages to individuals whose credit is generally not good enough to qualify for conventional loans. Fannie Mae officials say they hope to make it a nationwide program by next spring.

[...]

In addition, banks, thrift institutions and mortgage companies have been pressing Fannie Mae to help them make more loans to so-called subprime borrowers. These borrowers whose incomes, credit ratings and savings are not good enough to qualify for conventional loans, can only get loans from finance companies that charge much higher interest rates -- anywhere from three to four percentage points higher than conventional loans.

[...]

In moving, even tentatively, into this new area of lending, Fannie Mae is taking on significantly more risk, which may not pose any difficulties during flush economic times. But the government-subsidized corporation may run into trouble in an economic downturn, prompting a government rescue similar to that of the savings and loan industry in the 1980's.

You think, maybe?

Comments

[whuffle.livejournal.com]

Very good point.

[wotw]

Thank you for posting this.

[drwex]

You're welcome. If you find other examples of people saying back then "this is going to involve a really huge amount of risk and maybe a big-ass bailout" I'd love to read them. I personally believe that this entire mess was 100% predictable to anyone who looked at the subprime concept objectively.

[wotw]

I can convey a first-hand report that ten years ago, talk about exactly
how this subprime market was going to play out---including accurate estimates
of the size of the disaster---were standard fare around the lunch table
at at least one Federal Reserve bank.

There was also a report from the Shadow Open Market Committee maybe
five years ago saying exactly the same thing.

[drwex]

Interesting. Is that report online?

My first exposure to this came when I started trying to understand fixed-income securities, only about 4 years ago. Within about a month of starting to learn about it I was personally convinced that it made no sense. But four years isn't particularly prescient.

[pierceheart.livejournal.com]

well, there's also Warren Buffet calling derivatives financial WMDs.

[quietann.livejournal.com]

Yes but... This is being used over on the conservative side to justify racism. That if banks had not been forced to lend to "those people" there would be no economic crisis (never mind that plenty of white people got in over their heads and used their equity like an ATM).

Remember, until not long ago at all, "those people" included Jews. The neighborhood just north of where I grew up had a no-Jews covenant until the early 1960s. One of my father's friends, not a Jew, was responsible for getting rid of it.

The change in the law was simply the first step. What sent things over the edge was unscrupulous mortgage agents and all sorts of legal gambling (via Wall Street) on the value of mortgages. The amount of foreclosures seems inversely related to the legal oversight required of any given institution.

[drwex]

I think this is an example both of the Law of Unintended Consequences and people using things for their own purposes.

It's pretty clear that one of the intentions here was to make mortgages available to people who couldn't usually qualify, particularly people of color. It's also clear that a lot of banks and mortgage companies took this opening and piled on like crazy, regardless of whom they were shoving untenable mortgage contracts at.

So the unintended consequence was the piling on. And of course, people who want to make a racial issue of this are using it for their own purposes.

[wotw]

I believe that the foreclosure rate among blacks is considerably higher
than the foreclosure rate among whites right now, which suggests that
standards were lowered more for blacks than for whites (it does not
*prove* this, because of average-versus-marginal issues, but it does
suggest it).

As far as I'm aware, the only evidence for discrimination *against* blacks
in this market was a single paper from the Boston Fed that seems to have
pretty much fallen apart under scrutiny (though I haven't carefully
followed that discussion, so there might be something I don't know about).

[drwex]

I don't have data on foreclosure rates by ethnic group, but it seems likely that if more untenable loan arrangements were extended to blacks than to whites it would likely follow that there would be a higher foreclosure rate for that group, just because they're overrepresented in the population "people who hold untenable mortgages".

[wotw]

In the absence of discrimination, you'd expect *marginal* black mortgages
(i.e. those that just got in under the wire) to have the same default
rate as marginal white mortgages. In the presence of pro-black
discrimination, you'd expect higher default rates for black mortgages
at the margin.

What we're seeing is higher default rates for black mortgages on *average*.
If we knew this were true at the margin, we could infer pro-black
discrimination, but we don't know that. (I suspect it's likely, though.)

[darkrosetiger.livejournal.com]

"Pro-black discrimination??"

[drwex]

Hmm, I'm not sure whether or not we disagree, so I'll try to state things as I see them and you can tell me if we agree or not. I don't like the phrase "black mortgages" so I'm going to try to avoid that.

Let's divide mortgages into two categories: traditional, and risky. We know, statistically, that people of color were disproportionately denied traditional mortgages. This may be for various reasons, including discrimination or simply the fact that fewer black people occupy the higher socio-economic strata and didn't have the credit histories or present incomes to get these traditional mortgages.

So if the default rate on these mortgages is R_tm we would expect black people to have a lower percentage of defaults, due to their lower representation.

Now we have risky mortgages, which we also know were given to black people (or at least people in color generally whom I'm going to handwave into the 'black' camp) at a higher than normal rate. Further, we know that the default rate on these mortgages, R_risk, is much higher than R_tm.

To my knowledge, the national default rate isn't really separated this way. The mortage default rate is reported as one number, which is really R_tm+R_risk. However, since black people hold far more of the risky mortgages, they now become disproportionately (more) represented in the defaults.

But that's a side effect of holding the wrong type of mortgage, rather than any form of discrimination.

[drwex]

*sigh* for "people in color" I mean "people of color."

[wotw]

Right. The default rate is not separated this way. So this is not what I
was saying.

Let me try again:

Suppose each mortgage applicant has creditworthiness that we can rate on
a scale from 1 to 10.

To get a mortgage, you've got to be above a certain cutoff.

I take "discrimination" to mean that the cutoff for whites is different
than the cutoff for blacks.

So if the policy is "You have to be 3 or above to get a mortgage", I call
that non-discrimination. If the policy is "You have to be 2 or above if
you're white, but 4 or above if you're black", I call that pro-white
(or anti-black) discrimination. If the policy is "You have to be 2 or
above if you're black, but 4 or above if you're white", I call that
pro-black (or anti-white) discrimination.

It is an empirical fact that blacks are currently defaulting at a higher
rate than whites.

Here is a tempting but false argument:

1) Blacks are defaulting at a higher rate than whites.
2) Therefore, on average, blacks who received mortgages were probably
less creditworthy than whites.
3) Therefore, the black cutoff must have been lower than the white
cutoff.
4) Therefore there was pro-black discrimination.

To see that the argument is false, all it takes is a counterexample. Here
is a counterexample: Suppose that 90% of whites are 1's and 10% are 2's.
Suppose that 90% of blacks are 5's and 10% are 2's. Suppose the cutoff
(for everyone) is 3. Then there is no discrimination, but every white
who gets a mortgage is a 10, every black is a 5, and more blacks than
whites will default, despite the non-discrimination.

So my points were: 1) The higher default rate among blacks suggests a
tempting (but false) "proof" that there was pro-black discrimination in
the granting of mortgages. 2) That proof is in fact false, so we do
not know there to have been pro-black discrimination. 3) Nevertheless,
if I had to guess, based on further numerical patterns in the data, I'd
say there probably was.

My subsidiary point is that it is often claimed that prior to the wave
of alleged pro-black discrimination (fostered by, e.g. the Community
Reinvestment Act), there was significant pro-white discrimination. My
points regarding that were: 1) There is, to my knowledge, only one paper
that finds this, and 2) That paper has withered under scrutiny. Also
3) There might be more evidence I don't know about.

[pierceheart.livejournal.com]

there's also info that needs to be looked at about how CRA aligned banks didn't sell off mortgages as derivatives.

And how, of the Fannie and Freddie held subprimes that were checked (as of mid sept 08) 20% of them were made to people who would have qualified for prime mortgages.

http://www.traigerlaw.com/publications/traiger_hinckley_llp_cra_foreclosure_study_1-7-08.pdf

[taura-g.livejournal.com]

Yeah, just a little unfortunate foreshadowing going on there...

*just grumpy about the whole damn mess*

[gentlescholar.livejournal.com]

Only unfortunate in that it was ignored.