"I think there are at least three arguments for the canary theory and against the domino theory."

Here's a post on The Baseline Scenario about Subprime Lending and its role in the crisis by James Kwak:

"Asking whether subprime lending caused the crisis raises all the questions about agency and causality that I’ve raised before. On the agency question, insofar as there was a problem in the subprime lending sector - and few would deny that there was - does the fault lie with borrowers who took on loans they had no chance of repaying, perhaps sometimes without understanding the terms; with the mortgage lenders who lent them the money without doing any due diligence to determine if they could pay them back; with the investment bankers who told the mortgage lenders what kinds of loans they needed to package into securities; with the bond rating agencies who blessed those securities while taking fees from the investment banks; with the investors who bought those securities without analyzing the risk involved; or with the regulators who sat on their hands through the entire process? Note in passing that it may have been perfectly rational, as well as legal, for an investor to by an MBS even knowing that the loans backing it were going to default, but making a bet that he could resell the MBS before the price fell, under the “greater fool” theory of investing. (It may have been rational for an investment bank to do the same, but not necessarily legal, given the disclosure requirements relating to securities. Goldman Sachs is being sued over precisely this question.) Readers of this blog know that my opinion is that, although there is blame to be shared along the chain, the greatest fault lies with the regulators, for a few reasons. First, although the desire to make money may cause problems, it can be no more be said to be a cause of anything than gravity can be said to be the cause of a landslide; second, bubbles are inevitable, at least in an unregulated market; and third, there is a difference in kind between the mistake made by an investor, who is foolish and loses some money, and the mistake made by a regulator (or a legislator who votes to reduce funding for regulators), whose job is to serve the public interest."

I don't agree with this.
1) The desire to make money is always there, but there isn't always Fraud, Negligence, Fiduciary Mismanagement, and Collusion. That is an essential element of this crisis.
2) Bubbles are not inevitable.
3) The markets were not unregulated. Indeed, one also has to remember that there are professional codes that might have also been violated here as well as criminal ones.
4) There's plenty of research that Regulators act as Private Actors do. ( I'm thinking of James Buchanan )

I believe that the Banks and other Financial concerns in general are the most responsible, including the Credit Rating Agencies and Mortgage Brokers, for example.

"But that was all the preamble, because today I want to talk about the question of causality.

I think it’s generally accepted that the crisis we know today first appeared in the subprime lending market, where an increase in delinquency rates triggered a fall in asset values. Those problems were clearly visible early in 2007 (it’s impossible to say exactly when they were first visible, because some people had been warning of the problem for years, to little effect), and over the next year the main entertainment in the financial sector was watching banks and hedge funds suddenly realize they had large subprime exposures and either take writedowns or fold. But I think there are three ways to understand the relationship of subprime and the current crisis:

1. Subprime was the first place where various structural problems appeared, but those problems existed elsewhere, where they only appeared later. If the subprime lending boom had never happened, we would still be roughly where we are today. Call this the “canary in the coal mine” theory.
2. Subprime was the first place where various structural problems appeared, and the subprime crisis generated additional pressure that exposed those problems in other areas. For example, subprime concerns caused a pullback in lending, which caused a leveling off in home prices, which caused a reduction in housing construction, which slowed economic growth, etc. Call this the “domino” theory.
3. Subprime was a necessary cause of the crisis. Without subprime, the levels of housing prices, indebtedness, and risk in the system would have been sustainable indefinitely. Call this the “prime mover” theory.

Only under the prime mover theory can subprime truly be said to have caused the crisis. Under the domino theory it played the role of a precipitating but unnecessary cause. Under the canary theory it is just a leading indicator.

In my opinion, subprime was probably the canary, and possibly the first domino. There are various arguments against the prime mover theory:

• The U.S. subprime sector is simply not big enough. Although the numbers have been shifting in the last couple of years, roughly 80% of outstanding residential mortgages in the U.S. are prime; the other 20% is split between subprime and Alt-A. About 50 million homeowners have a mortgage, of which about 7 million have subprime mortgages. The idea that an increase in the delinquency percentage among 7 million U.S. homeowners (total mortgage value about $1-2 trillion, so losses on foreclosure - assuming a 100% foreclosure rate - about $0.5-1 trillion) could have by itself caused the largest economic downturn in the world since the 1930s is hard to credit.
• In absolute terms, losses in the subprime sector will be dwarfed by losses in the prime sector. Credit Suisse is now forecasting 8.1 million foreclosures by 2012, over 5 million of those outside of subprime. Current-month foreclosures among prime mortgages have already caught up to and passed (see chart on p. 4) foreclosures among subprime mortgages.
• The U.S. and global economies bumped along passably for over a year from the beginning of the subprime crisis. The U.S. recession did begin in December 2007 (Econbrowser for a good post on recession dating), but most of the numbers don’t start falling off cliffs until the second half of 2008. By the time Lehman went bankrupt in September, it’s probably true that all of the bad news about subprime was already priced into the various markets. What’s happened since then is new bad news about every other market.

Deciding between the canary and domino theories is tougher. The canary theory is that there were lots of boulders perched precariously on a cliff and subprime was just the first one to fall. The domino theory is that the subprime boulder knocked into a lot of much bigger boulders and knocked them off, but something else could have knocked them off just as easily. The domino theory could go something like this: Subprime caused writedowns and instability in the financial sector and nervousness in the housing market; nervousness in the housing market caused housing prices to start to fall, making it harder to refinance and increasing delinquencies on all kinds of mortgages; expanding writedowns caused a liquidity run on banks such as Bear Stearns and eventually Lehman; falling house prices and the consequent wealth effect reduced U.S. personal consumption, slowing economic growth; reduced consumption had the usual multiplier effect, reducing incomes and creating a recessionary cycle; the the recession hurt the value of every other type of debt (commercial mortgages, credit cards, etc.), triggering a full-scale banking crisis; and the fear created by the banking crisis led to the sharp downturn in credit and in consumption that put us where we are today.

I think there are at least three arguments for the canary theory and against the domino theory.

First, there is the issue of timing. The subprime crisis took an awfully long time to blossom into a full-fledged global recession and, as I said above, by the time the latter occurred the full scale of the subprime problem was more or less known to everyone. On that principle, the other boulders withstood the bump they got from the subprime boulder.

Second, once we had a housing bubble, it was inevitable that it was going to pop one way or another. So one question to ask is whether subprime lending was the reason for the housing bubble. Even at the peak of housing prices in 2006, subprime loans only made up about 20% of total mortgage origination volume. (Everyone cites Inside Mortgage Finance, but you have to pay for their data; here’s an NPR primer on subprime with a chart.) Could that 20% have have been solely responsible for the bubble? I suppose it’s possible, depending on the shape of the supply curve, but count me as skeptical.

Third, there is another good explanation for what pushed all those boulders down. James Hamilton thinks that the economy was structurally fragile, and the shock that knocked the boulders down was the oil price spike.

My view is that we were teetering on the edge of a cliff last summer, and the oil price shock may have been just enough to tip us over the edge. As we did so, the financial disaster that had always been a potential became a reality.

The trouble is, now that the economy is in free fall, it’s going to take more than $2 gasoline to pull us back up.

Ultimately, I think this question (canary or domino) is not definitively answerable, like many historical counterfactual questions, but I’m on the side of the canary.

One final note: Blaming subprime can have a disturbing overtone of blaming poor people for reaching beyond their means. First of all, it’s not true that subprime has more than a vague correlation with income. In the words of the late Tanta:

The capacity C of traditional underwriting was, of course, always relative to the proposed transaction. A lower-income person buying a lower-priced property was, you see, not a case of subprime lending; assuming a reasonable credit history, it was a prime loan. People with quite good incomes and stellar credit histories who tried to buy way too much house got turned down by the prime lenders.

More often, however, people in gentle society realize it’s not proper to blame poor people, so they take aim instead at the Community Reinvestment Act and liberal politicians generally for attempting to extend homeownership to people who couldn’t afford it. This line of attack was most recently exhibited on the New York Times op-ed page. I will leave the rebuttals to the experts:

• Mark Thoma (2 separate posts, with additional links)
• Barry Ritholtz (2 separate posts)
• Randall Kroszner (cited by Ritholtz)"

I suppose I also accept the Canary Theory, in that there were other investments that could have triggered this crisis. The lowering of capital requirements infected a number of investments. The complexity of the investments also contributed as well. In general, there was a mountain of Wishful Thinking that infected a whole range of investments.