What Caused the Crisis?

Rounding up some suspects:

* Don Boudreaux on monetary policy.

* Frank Shostak on Fannie and Freddie.

* Randal O'Toole on land use regulations.

* Chris Dillow on the principal agent problem.

* Russ Roberts on perverse tax incentives.

Meanwhile, Robert Higgs disputes the idea that the credit market is "frozen," and David Cay Johnston suggests some questions for skeptical business journalists. Ilya Somin sees a slippery slope ahead.

Comments

[Tulpa]

David Cay Johnston suggests some questions for skeptical business journalists.

The Lonewacko Gambit!

[JMR]

This graph

http://www.nytimes.com/imagepages/2008/02/17/business/20080217_SWAP_2_GRAPHIC.html

suggests "only" about a Trillion(!) probably WON'T be enough when credit derivative markets go south. The pigs are now feeding...

[PC]

The taxpayers will assume the risk, that way after this "investment" goes broke they will assume the losses. Oh by the way we will be paying above the market value for these items and put it all on the credit card. Why do they go after internet gambling, claiming addicts will fund risky bets on the credit card, but they have no problem doing that with the US economy? At least the gambler will use to market to get the least juice on his bet, the Federal gamblers are not even trying to get an adequate price, they will be paying a premium. This is ridiculous, the people deserve to get screwed because they continue to let this nonsense happen. The snake oil salesmen whine that "the markets are irrational" because nobody will buy this garbage [except the biggest suckers out there].

[Jesse Walker]

The Lonewacko Gambit!

No, no. These are good questions.

[robc]

JMR,

That graph is out of date, the number I saw earlier this week is $62 trillion or something like that. However, that is the total face value, not the amount of loss if those were all unwound. Still a lot but less than a trillion.

Dont fall for the scare tactics.

[Episiarch]

No, no. These are good questions

No, no, Jesse. They are GoodQuestions.

[Gilbert Martin]

There should be another section in the roundup about how the "mark to market" rules contributed to the financial crisis.

And another one about how the credit rating agencies like Moody's and Standard & Poors did as well.

[JMR]

Yeah, I should have mentioned the graph's out of date, but doesn't that mean the gambling market it describes is even MORE dangerous than the graph depicts?? We're in a heroin addict economy (and e-gold & the Ron Paul Liberty Dollar were the problem -- sigh) but if just a few of those bets go wrong, the cascading counterparty defaults our government tried to avoid by all this recent BS will have to happen anyway, right? This means the whistle hard money "kooks" like me have been blowing, loudly, was worth listening to, so eventually -- as society collapses -- at some point my side gets to say "I told you so" as we grab for our ammo, MREs, and water. IMO.

[Dr. Kenneth Noisewater]

Gilbert:

The claim that "mark to market" accounting rules have played a large role in this financial crisis is something I've been hearing elsewhere from some conservative accounting types who have a history of constructing elaborate apologias for CEO malfeasance. I don't know much about accounting so I'm wondering what's so evil about these rules? From what I understand, it basically instantiates into accouting rules the idea that an asset is only worth what someone else is willing to pay for it.

Isn't that theory of value pretty much a bedrock principle of the free markets much beloved by us libertarians and some brands of conservative? Aren't other theories of value useful tools for tyrants and scam artists?

[Gilbert Martin]

Dr. KN

Yes something is generally only worth what somebody is willing to pay for it.

The problem is that when markets freeze up because of fear and panic, there are a lot of securities that don't trade at all. And then the "mark to market" adjustment is not based on any actual trade but on models, guesses and presumptions.

Not all the debt instruments out there affected by this trading freeze are totally worthless. Some are backed by pools of mortgages that are not in foreclosure or default. Long term assets are being artifically drastically marked down as if they are nearly worthless.

It is akin to saying your car is totally worthless if you cannot find somebody to buy in within the next hour.

These are non cash losses but trigger all sorts of regulatory capital requirements that treat them as if they were cash losses. Then institutions have to attempt to raise new capital under the worst possible market circumstances. The interconnectedness of all these financial institutions exacerbates the problem as well. Markdowns at one can trigger more markdowns at others and it creates a cascade effect.

[Jesse Walker]

There should be another section in the roundup about how the "mark to market" rules contributed to the financial crisis.

It wasn't supposed to be an exhaustive list. Anthony Randazzo had a concise explanation of what's wrong with the mark-to-market regs in his Reason piece yesterday.

[Dr. Kenneth Noisewater]

Got it. It sounds to me like this sort of downward spiral due to market seizure is a risk that the investment banks should have known about. If they knew this was a risk, shouldn't that have created an incentive to stop investing in shitty mortgage securities and their derivatives? Or did they simply assume that any such risk could be discounted because they could count on their friends in the US Treasury to bail them out in the event of financial Armageddon?

If "mark to market" isn't such a hot idea, what are the other alternatives and what shortcomings do they have?

[R C Dean]

From what I understand, it basically instantiates into accouting rules the idea that an asset is only worth what someone else is willing to pay for it.

The problem with mark-to-market is that it substitutes a hypothetical market value for the last real market transaction for the asset. Under traditional accounting, you carry financial assets at their cost - what you paid for them. What you paid for them is the last real market value assigned to those assets. Everything else is speculative.

Mark-to-market is also very prone to abuse. Its what drove the Enron scam, after all. And it introduces a lot of volatility into balance sheets, which is one of the drivers (I think) of the current liquidity problem.

One knock on cost-based accounting is that, if your assets lose value, your balance sheet looks better than it really is. The reason the Wall Streeters don't like cost-based accounting is that it keeps them from trading on/leveraging unrealized gains. As we are seeing, leveraging your balance sheet with mark-to-market is a good way to realize short-term gains, but like all leveraging, it is risky.

[Secretary of Treasury, Mr. Han]

Let's have mark-to-market when the market is going up, and cost accounting when the market is dropping. And when we own all the mortgages, we can have mark-to-election-year-politics.

[Geotpf]

I hate zoning laws/land use regulations. I absolutely agree that restrictive zoning and other government meddling into what people can and can't build on their property is partially responsible for the current house price bubble, not to mention increasing things like pollution and traffic.

As for cost based vs. mark to market accounting, I think the second is much more accurate. If I paid $10,000 for a Ford Escort, or $3,000 for an Apple IIE, in 1983, but both are near worthless today, they shouldn't be listed as assets at the price I paid for them. Now, those are equipment and not land, but one can also buy land that used to have great value that is now worthless (say it was found to have toxic waste on it after I bought it). If nobody is willing to purchase something, at any price, it's value should be zero, even if I paid lots of money for it originally.

Now, I can also see how mark to market can (and has) cause a feedback loop that causes companies to go into bankruptcy. Companies need to have x dollars worth of assets to maintain their bond rating, etc., but the price of those assets drops, so their bond rating falls, and so does everybody else's, so nobody wants to buy those assets, so the price of the assets drops, so their rating drops, so nobody wants to buy those assets, etc. The proposal mentioned by Newt Gingrich (a three-year mark-to-market rolling average) seems like a good compromise to minimize this effect.

[joe]

The land-use argument is laughable.

In fast-growing places with no such regulation, such as Dallas, Houston, and Raleigh, housing prices did not bubble and they are not declining today.

Captain Cato needs to visit the Inland Empire in California, or the gulf coast of Florida, or newer suburbs around Phoenix, and count the empty homes. Sure, the bubble only happened in low-growth areas. Tell me another one.

[Gilbert Martin]

"As for cost based vs. mark to market accounting, I think the second is much more accurate. If I paid $10,000 for a Ford Escort, or $3,000 for an Apple IIE, in 1983, but both are near worthless today, they shouldn't be listed as assets at the price I paid for them."

A morgage backed security that is still generating incoming cash flow from principal and interest payments is not analgous to the depreciated value of an old car or computer.

It may be worth less than the purchase cost but it is not "near worthless".

[joe]

By definition, a bubble is created when speculative investment raises the price of a good far above what can be sustained by actual demand. That's why they pop - because there honestly isn't the underlying demand to keep the prices that high.

Obviously, the housing bubble is yet another consequence of invading Iraq, failing to provide universal health care, and emitting too much carbon dioxide into the atmosphere. I've been trying to warn you people about this for years, and now, this crisis proves that I'm right about everything!

[Rationalitate]

O'Toole truly is a tool. Where the hell did he get the idea that "Dallas, Houston, and Raleigh" didn't have land use regulations?? I guess it's understandable when you see the world through O'Toole's eyes - that is, minimum parking regulations, zoning, and other land use policies that force sprawl where the market doesn't want it are not regulations; but the same bodies enacting opposite legislation (maximum parking requirements and other New Urbanist kind of policies) are regulating. Do yourself a favor, Reason Magazine: don't pay attention to the idiotic statists behind land use policy at the Reason Foundation. They're ignorant at best, and shills at worst.

[robc]

joe,

Category A: fast growth, low reg
Category B: fast growth, hi reg
Category C: slow growth, low reg
Category D: slow growth, hi reg

In fast-growing places with no such regulation, such as Dallas, Houston, and Raleigh, housing prices did not bubble and they are not declining today.

This is a reference to Category A.

Captain Cato needs to visit the Inland Empire in California, or the gulf coast of Florida, or newer suburbs around Phoenix

You reference Category B.

He wasnt referring to Hi vs Lo growth, but A vs B/C/D.

[Jesse Walker]

I'm surprised to see more comments about O'Toole's sloppily phrased comment about land-use regulations than Dillow's assertion about the relative dangers of national and dispersed private ownership. Anyway, don't take the links in this post for unqualified endorsements. I think O'Toole makes a good case that land-use regs played a role in the housing bubble. I'm not persuaded that they played the central role.

[joe]

robc,

If you read the report, the author claims that the cause of the bubble was a restriction on supply produced by regulation. Since there are fast-growing places with heavy regulation, his theory makes no sense. Since fast-growing places also experienced a bubble and a pop, his theory that supply restrictions caused the bubble makes no sense.

Absent the link between regulation and supply, and absent the link between growth and bubbles, his argument is magical thinking.

In addition to the fact that bubbles are, by definition, spikes in price caused by speculative purchases unmoored from supply, so any theory of bubbles being caused by unmet demand for the underlying goods is nonsense. If there really was unmet demand substantial enough to cause prices to rise as much as they did during this bubble, prices wouldn't have declined like this. I mean, come on, Demand Kurv!

[joe]

Let me add, I agree with the point that land use restrictions - some of them, anyway - have caused housing prices to rise. Snob zoning is a big problem, and the suburbs that enact it need to open up their zoning.

But that can't explain a spike like this, and it is wholly incompatible with the collapse in real estate prices we've seen over the past year or two.

[joe]

I'm surprised to see more comments about O'Toole's sloppily phrased comment about land-use regulations...

Hi, Jesse. Have we met?

(It's not the phrasing that's the problem, btw. His point wasn't lost, it's just nonsensical).

[Rationalitate]

...and the suburbs that enact it need to open up their zoning.

This is precisely the type of land use policy that O'Toole and his fellow travelers ignore. Ask a normal person what zoning is, and they'll say that it's what suburbanite use to make sure nobody builds an apartment building or townhouses near them. Ask O'Toole, though, and he'll tell you that zoning is when Portland liberals decide that the state ought to mandate high density. He highlights a trend that's minimally relevant compared to the bigger problem - mandatory low density zoning.

[Jesse Walker]

Joe: It was sloppily phrased because it treated a particular kind of land-use regulation as land-use regulation in toto. But I don't think the basic point is nonsensical. The Glaeser research he links to suggests that those regs make the price cycle more volatile, and the Krugman column he links to makes a reasonable case that it's easier to get a housing bubble started when regulation restricts supply.

[Mike Laursen]

My understanding is that most of the sub-prime lending was in areas where it was relatively cheap to build. Places like Las Vegas or the exurbs of Los Angeles. Not in areas like the San Francisco Bay Area that tend to have high prices because of land use regulations.

In which case, O'Toole is missing the distinction that there were really two parallel housing price bubbles going on. One related to sub-prime lending, which is related to the current financial crisis. And one, still ongoing, where prices are rising in desirable areas that have lots of well-paying jobs.

[lupan sansei]

"The problem is that when markets freeze up because of fear and panic, there are a lot of securities that don't trade at all. And then the "mark to market" adjustment is not based on any actual trade but on models, guesses and presumptions."

This is absolutely not true. What you have just described is mark to model. mark to model is how many of these MBS's were valued in the first place because there was a lack of a market. As you say, they were based on assumptions, variables and inputs which were subjective to the seller. Once an MBS is sold on the market, the true market price of that MBS is determined and all holders of similarly risk profiled MBS's must mark those holdings to the market determined price. By definition, mark to model is "the act of assigning a value to a position held in a financial instrument based on the current market price for that instrument or similar instruments".

"Not all the debt instruments out there affected by this trading freeze are totally worthless. Some are backed by pools of mortgages that are not in foreclosure or default. Long term assets are being artifically drastically marked down as if they are nearly worthless."

That is true, that is why, at it's heart, IMHO, this is a credibility problem, not a credit/liquidity crisis. Those are just symptoms thereof. The rating process and grades have lost all credibility as the credit rating agencies incorrectly/incompetently/negligently/corruptly(take your pick) bundled higher risk debt instruments with lower risk instruments and then rated them much higher than they should have. Thus, nobody knows if the "lower risk" debt that others (as well as themselves) own are truly lower risked investments and are unwilling to loan money without this transparency. If correctly rated and bundled, then mark to market would not affect these lower risk instruments.

"A morgage backed security that is still generating incoming cash flow from principal and interest payments is not analgous to the depreciated value of an old car or computer.
It may be worth less than the purchase cost but it is not "near worthless"."

Mark to model only assigns the last market determined price for that asset. If it is "near worthless", then that is the market determined price. An MBS which was bought on the premise that it was going to pay 8% a year until maturity which is now only producing 5% due to foreclosures and delinquencies should be devalued. Why should it not be? And please don't tell me that they devalue long term assets for what they would be worth...it assigns current market values, as determined by the market, on long term assets which might be worth their projected value. It replaces a hypothetical with an actual.

[joe]

Mike Laursen,

That second one wouldn't be a bubble. If the demand really is there, it's not going to pop.

[Mike Laursen]

Good point.