Dots, Connected

“No good deed goes unpunished” is a maxim that pretty much everyone except the irredeemably foolish learn at some point, if only temporarily.  “No good intention goes unperverted” would be the corresponding maxim of government.

The Community Reinvestment Act was passed in 1977, back in the waning days of blind faith in government “initiatives” to accomplish just about anything, with the stated goals of increasing the number of Americans who lived in their own house.  On its face this seems to be not just a morally commendable goal — “landlord” comes in neck-and-neck with “tax collector” and “kulak” for most-reviled occupation — but Good Policy as well.  There are hosts of desirable socio-economic data which are positively correlated with home ownership, and equal numbers of undesirable data which are negatively correlated.  So by increasing home ownership we increase the Good Things and decrease the Bad Things.  All at the margins, of course, but hey!, it’s still a twofer and every step forward is a step forward.  And all that, right?  So let’s throw it against the wall and see how much closer we can get.

How the statute worked was that it loosened the loan underwriting requirements that banks were otherwise required to observe in making residential loans.  “Loan underwriting” is the technical term for “those criteria by which a bank determines whether any particular loan is an acceptable risk.”  Much shorter and easier to remember, in other words.  Generally banks, the deposits of which are insured by the FDIC or its S&L counterpart, FSLIC, or those which are to be guaranteed by an agency of the federal government, such as the VA or the FHA, or which are to be purchased by an entity which is backed by the federal government, are required to make some sort of effort to figure out ahead of time whether any particular loan is likely to be paid back, and not to make the loan at all if the answer is too close to “no.”  There are a whole host of things that banks look at, such as income-to-fixed expense ratios, the amount of the loan expressed as a percentage of the collateral’s value, the prospective borrower’s track record in paying his other creditors, whether the borrower has, within a certain period of time before, sought the protection of the bankruptcy system, the prospective borrower’s other unencumbered assets expressed as some percentage of the prospective loan, etc. etc. etc.  Generally those lenders are required to verify what their borrowers tell them.  They do that by requiring the production of tax returns, bank or investment account statements, or obtaining information from unrelated third parties like appraisers of the proposed collateral, or credit reporting agencies which collate information about the proposed borrower.  All well and good, and proper when you consider that, however imperfect it may be in that function, the past is in fact the only thing we have to rely on in figuring out what the future is likely to look like.

The CRA explicitly provided that for certain borrowers, in certain locations, and under certain circumstances, a lender was permitted to look at a loan application that fell into the “don’t make this loan” territory and ignore the warning signs.

A quick aside:  Those folks who think that banks just made up all this stuff about loan underwriting requirements in order to deny loans to people actually able to pay the money back, just because . . . well, just because, are fools.  Banks do not make money if they do not make loans.  OK?  They owe the money on their deposits, including interest; those deposits are obligations on the bank’s books.  Banks cannot get money to loan without themselves borrowing it, from their depositors, or from the Fed, or from other banks.  The only way they can pay the interest is by charging interest and then actually collecting it.  If they make good loans, they stay in business.  If they make bad loans, they cannot pay their debts and, once they’ve exhausted their own capital, they go out of business.  Period.  Banks which willingly pass up good loans are leaving money on the table.  Banks which by bigotry exclude entire classes of borrowers or depositors because they’re . . . well, not people like us, create business opportunities for others who are savvy enough to court those good loan and deposit customers.  Like the Italian-American Immigrant Bank, which loaned to that group because they understood that Italian immigrants paid their bills.  You may have heard of them; they’re now known as Bank of America.  Don’t think I’m just making this up, either; Gary Becker won a Nobel in economics in part for his study of the economics of unjustified invidious discrimination.  He showed that actors who indulged their bigotry under circumstances where there was not some concrete basis for doing so (such as, for example, not serving a drunk stinking of his own piss and who also happened to be black, versus someone who just decided he wasn’t going to do business with the 20% of the town’s population that was black) paid a price for it, and that in a free market, without direct or indirect government subsidy of such behavior, it tended to go away because the people who paid the price realized they were paying it.

The mildly-common-sensical reader will immediately note a few problems with the thinking behind the CRA.  For starts, it treats the genuine economic considerations behind loan underwriting as if they in fact did not exist.  Telling Lender X that it may ignore adverse loan underwriting results for a specific loan application does not mean that the loan is now less likely to prove to be a bad loan.  Secondly, it assumed that because A and B are positively correlated, it must be because A causes B, where A is home ownership and B is positive socio-economic attributes.  This of course not only confuses correlation with causation (the first is a statistical phenomenon that can be determined based upon counting instances of A and B within a population; the second is a logical relationship the existence  and directionality of which can only be determined by experiment and examination of the specifics of the population in question), but it made the assumption that the direction of causality was from A to B.  Of course, what if it’s B that causes A, or what if A and B have no direct causal relationship but rather are both results of (that is, both are caused by, rather than cause) some other factor C?  Artificially increasing A is simply not going to produce more of B in either such case.

In 1995 the enforcement mechanisms under the CRA were significantly sharpened.  Now your examiners were authorized to determine how many CRA loans you ought to be making, and if they found you weren’t making enough, they could fine you.  Really.  No kidding.  You exposed yourself to liability to the government if you did not make sufficient loans to people who could not reasonably be expected to pay them back.  At about the same time, Fannie Mae and Freddie Mac decided, on their own, that they were going to increase the percentage of their total loan portfolios represented by “sub-prime” loans (in other words, loans made to people who weren’t likely to pay them back, including specifically CRA loans) to up to 50%.  You read that number right:  Fannie Mae decided that it wanted to have half of its loans consist of paper that was more likely than not worthless.

Well now.  What happens next?  In point of fact sub-prime loans, as a percentage of the total loan market, went through the roof.  And the race was on.

Audit of a bank’s CRA compliance was no empty threat.  Bank of America within the past few months paid $25 million to settle a CRA enforcement action brought by the DOJ.  That’s a lot of money to pony up because you tried not to lose money on bad loans.  But that’s the perversity of the incentive system we created.

Bank of America is of course a large bank.  But what about smaller banks, community banks?  Well, let’s say that I work for Community Bank X, and I know that my bank has a corresponding relationship with Fannie Mae.  That means that I know, for a fact, that every loan which I make which fits within the loan standards established by Fannie Mae, whatever those standards may be, I can originate and have sold and off Community Bank X’s books before the ink is even dry on the closing documents.  In other words, I know that, so long as I have a loan applicant who fits within Fannie Mae’s underwriting requirements, I can make that loan with zero effective risk to my bank.  Community Bank X only has a re-purchase obligation in the event the loan didn’t comply, at the front end, with whatever requirements Fannie Mae had in place at the time.

Let’s see how that plays out:  I have a loan applicant come to see me.  I can look at their application and just about guarantee that they’ll never, ever be able to repay this money.  But wait:  This application in fact does fit within the You-Gotta-Be-Kidding-Me program recently rolled out by Freddie Mac or Fannie Mae.  So I can make this loan and in less than 24 hours Fannie or Freddie will have taken it off my hands and if three years later the whole thing blows up, it doesn’t do so on my desk.  Do I make the loan?  Well, if I do not make the loan, and the applicant is a member of a pet constituency, or the proposed property is located in a CRA area, what I am doing is risking tens if not hundreds of thousands of dollars in defending either a private civil action alleging illegal discrimination or a DOJ enforcement action for failing to comply with the CRA or Fair Housing Act, or whatever.  Those risks are uninsurable, meaning the bank gets to pay to defend them and then pay out of pocket any liability.  Even if the bank wins the discovery process will absorb hundreds of thousands of man-hours and attorney’s fees and forensic accountant’s fees.  So do I make the loan?  Hell yes I make the loan.  Have I done anyone any favors?  No.  Is this person any more likely not to lose his home to foreclosure at some point down the line?  Not in the slightest.  But I’ve protected my bank from liability; we can even brag on our website about how we aggressively support the Community Reinvestment Act.

Note, by the way, that even if my lender is for whatever reason not subject to the CRA, or the loan in question is not a CRA loan, so long as the prospective borrower fits within whatever cock-eyed Fannie Mae loan programs are within the scope of my lender’s corresponding relationship with Fannie, if I refuse the loan my lender is still subject to liability.  And in truth how do I justify refusing to make a loan that someone solvent has already promised to buy off me so that my risk in making the loan is zero or close to it?

And now some folks have come along and demonstrated by precise examination that yes, in fact the CRA and its enforcement did increase the loan risk accepted by large lenders in CRA areas.  They conclude pretty plainly that the Community Reinvestment Act in fact materially contributed to the subprime lending bubble and therefore to the subsequent crash.  In fact they allow that because of the constraints of their data sampling it’s likely that their study understates the impact of that misconceived statute.  They only looked at large lenders who were in the midst of CRA examinations.  They observe, “If adjustment costs in lending behavior are large and banks can’t easily tilt their loan portfolio toward greater CRA compliance, the full impact of the CRA is potentially much greater than that estimated by the change in lending behavior around CRA exams.”  In other words, the study’s authors admit that their sampled lenders might have been engaging in Potemkin lending, but that they can’t exclude that the observable behavior extended outside their window of observation.

The above link is to Reason magazine; the actual study is not downloadable for free (except to certain people), but here’s the link for those who wish to pay.

By the way, the Community Reinvestment Act remains out there, unchanged, over four years after the disaster it contributed to exploded into the worst economic downturn the country has experienced since the Great Depression.  Just like the Belgian farmers still plow up, and occasionally get blown up by, old artillery shells from World War I, the CRA still harbors its lethally defective assumptions beneath the American banking system.