Why CRA Is Not to Blame for the Housing Crisis

Today in The Gartman Letter, which I highly respect, I came across the following regarding the $700B government bailout and the housing crisis:

"Senator Dodd... upon whom we pin much of the blame for the current circumstances, for it was he and others on the Left who pushed the nation's banks, savings & Loans and mortgage brokers to expanding home lending operations into minority areas that were clearly incapable of handling large debt loads... seems intent upon adding various additions to the law as presented, sufficient to slow passage, and perhaps even to have it defeated." [Emphasis mine]

This isn't the first time I've seen this argument; it not exactly new. The basic idea is that back in the day, banks were sticking with lending to their bigger depositors, who were rather obviously middle and upper-class. CRA is the Community Reinvestment Act and it requires that FDIC institutions lend to all of the people in the community they serve without prejudice. In 1995, the Act was expanded to increase the number of loans being made to lower-income areas.

Now the argument goes that the banks were 'forced' by this regulation to loan money to people who really couldn't afford a house to begin with. Now I must clear something up right from the start: loans to low-income borrowers to meet CRA obligations and subprime lending are NOT the same things.

When the CRA requirement first started, banks did most of their lending to these lower-income borrowers through FHA (or VA) loans, which the government-sponsored entities, Fannie Mae and Freddie Mac, will buy off the originating bank. That off-loads the risk, and since the number of these borrowers was never that high, and the houses they were buying were cheap, the risk was reasonable. The loan terms were also very conventional; generally fixed rate 30-year products.

After 1995, CRA lending did indeed increase, and this marked in some ways the beginning of subprime, but not by FDIC banks! Instead, those institutions continued lending prime products to fufill their CRA obligations, or simply avoided opening branches in places where they felt uncomfortable lending.

So just looking at the chart, it becomes clear that FHA picked up after 1995, as did subprime, but then both cooled down in 2001. However, subprime then exploded in 2002 onwards while FHA loans slumped. Its backed up by data from the Treasury that the CRA-affected institutions continued meeting their obligations with prime loans during this period.

So what exactly *is* subprime? Well it can be a lot of things, but basically a 'prime' product is one that is considered either 'conforming', which means that the GSE's will buy it (which banks really like, because it allows them to get rid of the loan). Or it means the borrower is extremely well-qualified for the loan and is considered low-risk. Let me make a point here: banks hate to hold loans themselves; they'd much rather sell them off, either to the GSE's, or to private buyers in the secondary market who will package them up into Mortgage Backed Securities.

Some people don't seem to understand that fact, that banks don't actually loan their own (or depositors) money for mortgages; they originate the mortgage and sell it to an investor, usually to one of the GSE's, which off-loads the risk and free's up lots of capital while still collecting almost all of profit (and usually retaining the 'servicing' portion of the loan, which is the billing, etc).

The GSE's get to define what a 'conforming' product is, and there are certain things an originator must do to meet these guidelines. However, if you do, the GSE's guarantee they will buy the loan from you, which is great. So basically if it wasn't for the idea of a conforming product, most people would be considered too high of risk to lend to. Remember, before the GSE's, most mortgages from private banks were 5-year loans with a balloon (a really big) payment at the end.

Subprime was for borrowers that 1) didn't fit a Prime product and/or 2) couldn't get a FHA/VA loan (low-income). But subprime was more than that; it was all kinds of exotic loans; everything you could dream of beyond the 'conforming' standard.

As someone who worked in the mortgage industry for a number of years (before running away), I can tell you that the low-income type of borrower is one problem. But another, often much bigger problem, are the loans like NINJA's, Options ARM's, non-Amortizing ARM's, often of Jumbo size, often with 95% or greater LTV, sometimes with over 100% LTV! These loans did not go to the demographics targeted by CRA.

By the time sub-prime was at its height, institutions were over-originating 4-5 times their CRA requirements, in additions to all of these other toxic subprime loans that to not serve low-income borrowers. Remember, subprime was not just for low-income borrowers; it was offered as a solution for anyone who needed a unique, flexible, or unconventional product.

At the institution where I used to work, we used to do a loan called Stated Income Stated Asset, but we called it the Mobster loan, because there was really no reason to use the product unless you had no way to justify how you had come across a giant pile of money. Basically we said 'we won't ask questions, we'll loan you a ridiculous amount of money, secured at say 95% LTV, and we'll slap a pretty ridiculous rate on it, but as long as you keep paying, we'll never ask any questions about your personal details....' We even did loans to illegal immigrants. And this was an institution that did not get heavily involved in subprime.

So I think to blame CRA entirely is unfair at best. Yes, it made some contribution, but had HUD-type lending been constrained to a reasonable number of borrowers, the risk is entirely mitigatable, particularly if you keep those borrowers in non-toxic loan structures, ie. fixed-products with reasonable LTV's. I do not think 100% LTV's are completely unreasonable in all cases (I got my first house that way). But consider that NINJA loan for 800k. Its going to take a lot of HUD loans to add up to 1 of those, and that spreads out the default risk, and in general, those lower-priced loans probably have a better shot of being modified to something reasonable.

In case your wondering why I keep emphasizing fixed rates over ARM's, is that the ARM rates often reset to higher rates, but appear to be lower payments on paper when you ask "what's my monthly payment going to be?" Check out the default rates:

As you can see, the ARM's have not done well, and the interest-only loans will start to be even worse as those are really 'speculation' loans (originally meant as Prime products for really really rich people who could buy a house with cash, but have better investments they could do in the meantime).

Some people want to blame the GSE's for the mess, but this is also wrong. Yes, they were allowed to purchase some subprime loans, but the amount they could buy was very very small. Instead they lobbied for relaxed restrictions, the "Alt-A" loans, which are basically "less than prime" loans, but what's critical is that they are largely of non-toxic structures, meaning they have fixed rates, or they are fully amortizing ARM's. But for the most part, the GSE's could only buy conforming loans, which is why their market share shrank badly from 2002+ as the private sector expanded its operations.

So back to the original point, are low-income minorities at fault here? Well, the data seems to say otherwise. They benefited the most in terms of increasing home-ownership, but all groups really benefited quite comparably.

Finally, these last three charts have all been from Harvard's Joint Center for Housing Studies, and the last is the most depressing. Even college-educated people are losing income over the last several years.

But if you want to read a real treatise on the subject, you can go here. So hopefully I've dispelled the myth that CRA somehow 'forced' banks to make bad loans. No way, no how. And I'm not the only one making such an assertion.

There are always two sides to every transaction, and with subprime, there is no doubt that many borrowers got in over their heads, or speculated, or knowingly bought houses they couldn't afford. But at the same time, there was a party on the other side willing to lend them money against collateral (the house) that they assumed would always increase in value (and thus cover their losses), and that the risk could be diversified enough.

But without enough history of these exotic products, no one could correctly model the default rates, the risk, etc. And that was the real problem; everyone rushed for profits in what was a very opaque house-of-cards built on one assumption: that real-estate would always go up.