Earlier this week I noted that I had changed my mind on the Community Reinvestment Act.
Contrary to my initial conclusion, the evidence is overwhelming that the CRA played a significant role in creating lax lending standards that fueled the housing bubble. Once I realized this, I had to abandon my suspicion that the anti-CRA case was a figment of the rhetoric of Republicans attempting to distract attention from their own role in the mortgage mess .
So I laid out the facts and arguments that had convinced me to switch sides in the CRA debate. It was a long series of posts that generated hundreds of responses and counter-arguments. Felix Salmon’s response is here. Barry Ritholtz’s here. Mike Rorty's here. Ryan Chitum’s here. and Matthew Wurtzel’s here. All of my posts are here. Henry Blodget's earlier post on the CRA, with which I largely agreed until recently, is here. If you carefully run through these posts and the accompanying comments, I think you'll see that every argument raised by the "Defend CRA at all costs" crowd has been refuted.
For people with less time on their hands, here's a quick guide to the main points raised by the CRA defenders and the arguments that refute them. If I’ve left out any salient points, please let me know and I’ll add them to the list.
- How could a piece of 1977 legislation be significant to the deterioration of mortgage standards 25 years later?
The CRA was not a static piece of legislation. It evolved over the years from a relatively hands-off law focused on process into one that focused on outcomes. Regulators, beginning in the mid-nineties, began to hold banks accountable in serious ways. Banks responded to this new accountability by increasing the CRA loans they made, a move that entailed relaxing their lending standards.
- That’s still too early. Why would changes in the mid-nineties result in a mortgage boom a decade later?
Throughout the nineties banks, as banks lowered their mortgage standards, mortgage rates remained high. The laxity was spreading but the incentives for borrowers to re-finance even under relaxed standards remained low. New buyers often still didn’t know that some of the loosey-goosey mortgages existed. Speculators had an internet bubble, so they weren’t yet attracted to real-estate. Treasury rates were not yet so low that investors seeking yield would pour into mortgage backed securities. Securitization levels were low enough that banks weren’t yet willing to fully embrace the loose standards. The historical data on default and loss rates from the lax lending were not yet available, so they weren’t embraced by banks or the broader market.
But as the years went by, these factors changed. The Fed pushed interest rates down. This made refinancing more attractive, and created an investor demand for yield. Fannie and Freddie popularized low-income securitization. Low defaults and loss rates from lax loans made them seem not as risky as previously expected. A shrinking consumer asset base thanks to the dot com bust created a demand for home-equity loans and high loan-to-value loans, as consumers exchanged high-interest credit card debt for low interest home debt. Speculators seeking higher returns and ordinary home buyers became aware that lax lending standards would allow them to buy bigger homes with little or no money down.
In short, the lax lending standards created in response to the CRA had dug a pit that was waiting to get filled when the circumstances were right.
- Ah ha! So it wasn’t the CRA that caused the mess. It was everything else!
Of course it wasn’t the CRA that caused everything. The CRA was a factor in lowering lending standards. This was a necessary, although not sufficient, cause for the mortgage mess.
- Wait a minute! Paul Krugman told me the CRA was relaxed during the Bush administration. What about that bit of evolution, buster?
It’s true that the CRA requirements were relaxed during the Bush administration. But at this point the lax lending standards were already in place. In any case, the relaxation took a peculiar form that actually made CRA lending more important rather than less. You see, the government let banks drop things like putting in ATMs in rural areas in favor of letting their compliance be judged entirely on CRA loans. This means the CRA had more of an influence on home lending after the requirements were relaxed, not less.
What's more, George W. Bush was a major proponent of the kind of mortgages that banks had started making under the CRA. He urged low-to-no doc mortgages and the elimination of downpayments, just like the CRA regulators had long done. “We certainly don't want there to be a fine print preventing people from owning their home,” the President said in a 2002 speech. “We can change the print, and we've got to.”
- What about "No Money Down" Mortgages? Were they required by the CRA?
Actually, yes they were. The regulators charged with enforcing the CRA praised the lowering of down payments and even their elimination. They told banks that lending standards that exceeded that of regulators would be considered evidence of unfair lending. This effectively meant that no money down mortgages were required. A Treasury Department study published in 2000 found that the CRA had successfully lowered down payments not just for CRA loans, but for all mortgages.
- Explain the shift in loan to value away from the traditional lending requirement of 80%.
- What about the elimination of payment history? How about income requirements?
Similarly, banks were expected by regulators to relax income requirements. Day labors and others often lack reportable income. Stated-income was a way of resolving the gap between actual income of borrowers and reported income. The problem, of course, comes when the con-artists and liars come into the game.
- Did the CRA require banks to develop automated underwriting systems that emphasized speed rather than accuracy in order to process the greatest number of mortgage apps as quickly as possible?
This was another lending innovation praised by regulators to the point that it became mandatory for banks. Those who were not employing automated underwriting would be putting their CRA ratings at risk. Automated underwriting was seen as a way of eliminating bias in lending.
- Point out to me where in the CRA or any regulation that any of this is required.
I cannot. But this kind of legislative fundamentalism misconstrues the way laws constrain business activity. An unenforced law exercises little constraint, regardless of how onerous it is worded. Think of the way anti-trust enforcement changes from presidential administration to presidential administration.
In the case of the CRA, it was the activity of the regulators that matters. And each of these credit innovations described above was put into place to satisfy the CRA regulators.
- Wouldn’t lending standards have been lax during the boom even if we didn’t have a CRA?
That’s an interesting question to which we’ll never have a satisfactory answer. It’s possible. But this kind of counter-factual is just a game. We know that we had the CRA and that it caused relaxed lending standards in the reality we actually live. In another universe, another reality? Well, if you get there send us a post-card and let us know how it turns out.
- Couldn’t the increase in CRA loans have been accomplished without these lax lending standards?
This is another interesting question about an alternate universe. It is possible that banks may have been able to meet their CRA obligations with tighter, more traditional lending standards. But we’ll never know. We know
that they thought they needed to employ lax standards, and so did the regulators. A banker who refused to relax standards risked the wrath of regulators, and—more importantly—if his unorthodox, novel attempts failed he’d be found to be in non-compliance with the CRA. In the real world, lax lending standards were the only known way of satisfying the CRA regulators.
- Isn’t it really securitization that is the culprit?
The government pushed for greater mortgage securitization in an effort to increase CRA lending. At the behest of HUD Secretary Andrew Cuomo, Fannie and Freddie promised to buy $2 trillion of “affordable” mortgages. The government was intentionally decreasing the risks to the original lenders in order to increase loans to low-income borrowers, and minorities in particular. In short, you can’t blame securitization without coming back around to the CRA.
- Weren’t the majority of the subprime loans made by mortgage service companies not subject to the CRA?
This is true. But it is largely beside the point. A huge driver of the demand for subprime loans was the demand for CRA bonds. Banks operating under the CRA could meet their obligations by buying up CRA loans or MBS built from CRA loans. The CRA created a demand that the mortgage servicers were meeting.
What's more, many smaller mortage service companies hoped to be acquired by larger banks. Increasing their CRA lending made them more attractive take-over targets.
A study put out by the Treasury Department in 2000 found that the CRA was encouraging the mortgage servicers to provide loans to low-income borrowers, in part because the CRA loans had been so successful.
Finally, the Clinton adminstration threatened to subject the mortgage companies to the CRA if they didn't comply voluntarily. They promptly agreed to increase their CRA-type lending in order to escape the kind of public scrutiny that comes with official CRA regulated status.
- If the CRA was forcing all this lax lending, why weren't bankers objecting?
Are you really in the dark about why the leaders of large public corporations wouldn't publicly object to a piece of civil rights legislation? Fine. I'll be totally open with you: this would have been career suicide and an open invitation to bias litigation and increased scrutiny from regulators. In this case, silence is misleading.
What's more, no one said the bankers hated the lax lending the CRA was requiring. Sure, some did. But those people were quickly shown the door, while the enthusiasts were promoted. The regulations themselves selected for enthusiasts for the program of lax lending.
- How do you explain the fact that CRA loans historically had low levels of default. Doesn’t this mean that loan standards were not relaxed?
Actually, no. We know that lending standards were relaxed under the CRA. The fact that they had relatively low default levels was initially a surprise but it isn’t an indicator that lending standards were secretly high. There are plenty of explanations for this, including the lack of borrower ruthlessness in unsophisticated home owners and a tendency of delinquent low-income borrowers to sell the home and prepay the mortgage rather than default. Especially during a period of rising home prices, the default option was not heavily exercised.
- But even during the crisis, CRA loans didn’t default at higher rates than other mortgages. CRA banks aren’t failing more than other financial institutions, CRA areas aren’t hotspots of defaults. What about that?
In part, this is evidence that the lending standards of the CRA had spread to the rest of the mortgage market.
- I thought you said CRA loans caused this crisis.
Nope. It isn’t losses from CRA loans that drove the crisis (although they are disproportionately responsible for losses at some banks ). Instead, the CRA required lax lending standards that spread to the rest of the mortgage market. That fueled the mortgage boom and bust.
- So how and why did the CRA lax lending spread to the rest of the mortgage market?
The structure of the CRA regulations encouraged the spread. Banks that were the best at making CRA loans were allowed to grow by making acquisitions and opening new branches. This created a kind of political-financial Darwinism that reward the biggest enthusiasts for lax CRA lending standards. Of course, the most successful people under this regime were not the types who needed their arms twisted to make loose loans. They were who were predisposed to engage in loosey-goosey finance, who discovered that the CRA had made the world their oyster.
As for the “why” part of your question, the answer is a bit ironic. Banks making CRA loans initially expected that defaults would be higher due to lax lending standards. When they discovered the low-income borrowers had an unexpected propensity to pay their mortgages. After years of data poured in showing that borrowers were paying mortgages despite high LTVs, low down payments and unconventional income measures, bankers began to believe that many of the traditional measure of credit worthiness were overly conservative. Recall what I said earlier about how mortgage service providers started pursuing low-income borrowers in part because of the CRA.
What they didn’t take into account was that different types of borrowers may behave differently, and that much of the data on those lax lending mortgages was warped by increasing home prices. Wealthier, more sophisticated borrowers ruthlessly default when their mortgage goes underwater, for example. What’s more, the reversal of housing prices meant that defaults across all borrower classes increased.
Making matters worse, President Bush pushed hard for lax lending standards. He wanted to expand minority and low income home ownership far beyond what the CRA required. So he pushed even harder for the broadening of these lending standards.
- Wait. So you’re saying that the CRA's lax lending to low income borrowers became dangerous when it was extended more broadly?
In part. Ironically, the low income lending practices, particularly when undertaken on a limited basis in a low-securitization and rising house price period, seemed safe. This led to the practices being spread across the much broader category of loans. In a sense, you can look at the mortgage mess as what happens when CRA lending was applied much more broadly. If we'd confined the lax lending standards to just the geographical areas and low-income borrowers directly targeted by the CRA, our mortgage crisis would be far more manageable.
Studies have suggested that only 6 percent of subprime loans were extended by CRA-regulated lenders to either lower-income borrowers or neighborhoods in the lenders' CRA assessment areas. Since these loans suffer from outsized losses (for reasons not yet clear), we'd still have a major problem. But it would probably be only about 1/4 of the size of the current mortgage mess.
- That doesn’t sound very Republican of you.
I told you from the start this wasn’t some plot to make the GOP look good.
But don't get carried away with this irony. Now that we are in this crisis, loans located in low income areas are almost twice as likely to be in foreclosure as other loans. There's also an unfortunate racial angle, with African Americans being 1.8 as likely to be in foreclosure as whites, and Latinos being 1.4 likely to be in foreclosure.
What's more, an enormous amount of subprime loans were made to lower-income borrowers target by the CRA. Forty-five percent of subprime loan originations went to lower-income borrowers or borrowers in lowerr-income neighborhoods in 2005 and 2006, where the foreclosures are almost twice as likely. This suggests that the kind of low income borrowers targeted by the CRA are likely to be responsible for the majority of subprime foreclosures.