Information Processing

Just another WordPress.com weblog

Archive for the ‘bubbles’ Category

Simple question, complex answer

with 4 comments

A former physicist (but non-financier) writes:

I have a question, and who better to ask than you. …something isn’t adding up.

I keep hearing that mortgage defaults are what is bringing down many financial institutions, and the the default rate in some particularly bad mortgage pools is up to 50%. Because housing prices are down about 20%, financial institutions can still regain 80% of the value of those loans, no? Actually the real value is probably a bit better, as most loans will be partially paid off. At any rate, doesn’t this imply that even in the worst loan pools, there is only a total 10% loss. And most financial institutions will have some higher quality loan pools also, and stocks, etc. So the total effect is going to be smaller than 10%, unless financial institutions were all constructing some sort of horrible options based high risk bets on housing prices, but I doubt this would happen except in some risky hedge funds.

Is this reasoning correct? If so, I don’t understand how our system can be so fragile that a few percent drop would bring everyone down. Of course everyone holding a share of these funds will have a small portfolio dip, but this happens every few years anyway.

Your calculations are reasonably correct (see comments for more detail). So why the crisis?

1) Leverage. Many I-banks had 30:1 ratios, so a small movement in value of a subcomponent of their portfolio could wipe them out. It’s like a guy who puts 10% down on his house, who can lose everything if the price goes down by 10%. Of course this only matters if he is forced to sell, or, in the bank case, if shareholders and counterparties start losing confidence. This is happening simultaneously in financial markets due to the second factor…

2) Complexity. No one knows who is holding what, who has sold insurance (credit default swaps) to other parties and is on the hook, etc. So trust is gone and credit markets are paralyzed — no muni bond issuance, no short term loans to businesses, no car loans, etc.

The efficient functioning of our economy is built on trust — I have to trust that the grocer will give me food in exchange for a dollar bill, that I can get my money out of the bank, that my employer will pay me at the end of the month, that its customers will pay it, etc.

We are nearing a dangerous point. Confidence, once destroyed, is very hard to rebuild.

Relative to the size of our economy, the amount of money involved is not that great. If we had perfect information we could solve the whole problem with about $1 trillion. (About the cost of the Iraq war; not bad for a bubble that involved housing — our most valuable asset.)

Written by infoproc

October 2, 2008 at 4:18 pm

Ask the expert

with 4 comments

Several panicked multimillionaires and financiers asked me recently if they should liquidate all their investments and go to cash.

My answer? “Beats me” :-/

Standard questions:

1) are you going to be able to time the bottom? what’s your investment timescale?

2) what kind of cash? Treasuries? Swiss Francs? Renminbi?

Generally I’m not a big believer in timing the market. On the other hand, I can think of numerous plausible scenarios for the next couple of years in which (some kind of) cash is by far the best asset. I can’t think of very many in which it’s not 😦

Written by infoproc

September 23, 2008 at 4:27 am

Academic trends in pictures

with 7 comments

From GNXP, some beautiful graphs which depict the rise and fall of certain academic fads. My wife is a professor in the humanities (she did her graduate work at Berkeley in comp lit during the height of “theory”), and she got a big kick out of these!

Judith Butler, your 15 minutes are over 🙂 (Bad academic writing awards; see below figures for sample.)

Certain higher dimensional theories of fundamental physics might be next 😉

I searched the archives of JSTOR, which houses a cornucopia of academic journals, for certain keywords that appear in the full text of an article or review (since sometimes the big ideas appear in books rather than journals). This provides an estimate of how popular the idea is — not only the true believers, but their opponents too, will use the term. Once no one believes it anymore, then the adherents, opponents, and neutral spectators will have less occasion to use the term. I excluded data from 2003 onward because most JSTOR journals don’t deposit their articles in JSTOR until 3 to 5 years after the original publication. Still, most of the declines are visible even as of 2002.

No, this is not a joke — at least as far as I know.

Professor Butler’s first-prize sentence appears in “Further Reflections on the Conversations of Our Time,” an article in the scholarly journal Diacritics (1997):

The move from a structuralist account in which capital is understood to structure social relations in relatively homologous ways to a view of hegemony in which power relations are subject to repetition, convergence, and rearticulation brought the question of temporality into the thinking of structure, and marked a shift from a form of Althusserian theory that takes structural totalities as theoretical objects to one in which the insights into the contingent possibility of structure inaugurate a renewed conception of hegemony as bound up with the contingent sites and strategies of the rearticulation of power.

Written by infoproc

September 23, 2008 at 4:04 am

Treasury to socialize entire mortgage loss

with 3 comments

Complexity has won! More precisely, fear of systemic failure due to overcomplexity has trumped fear of moral hazard.

Is the overcomplexity due to insufficient regulation?

We’re going to put the whole mess on the US and taxpayer balance sheet. Paulson estimated hundreds of billions, but it could easily be a trillion. It all depends on home prices.

WSJ has a nice blow by blow account of the last week of crisis, focusing on key players like Paulson, Thain, Fuld, etc. (See also here.)

WSJ: …In a private meeting with lawmakers, according to a person present, one asked what would happen if the bill failed.

“If it doesn’t pass, then heaven help us all,” responded Mr. Paulson, according to several people familiar with the matter.

Paulson statement

Written by infoproc

September 19, 2008 at 5:35 pm

Orders of magnitude and timescales

with 13 comments

As a physicist I can’t help making some comments about orders of magnitude and timescales 🙂

If home prices return to normal (historical) levels, total mortgage debt losses will be about $1 trillion. This is a staggering sum, but won’t destroy our economy. After all, our misadventure in Iraq will end up costing us about the same amount. [Insert anti-Bush diatribe here.] If necessary, we could socialize the whole loss like we’ve done with Iraq — put it on the nation’s and taxpayers’ balance sheet.

The problem is that the credit bubble losses are concentrated in financial firms, which are getting hit with a huge shock as their portfolios approach the day of mark to market reckoning. This shock is going to have to be worked through the system over a relatively short timescale if we are to avoid systemic paralysis, or worse. Once a particular entity becomes insolvent, the entire web of counterparty transactions between it and the rest of Wall St. is in jeopardy.

Dealing with that relational web is the real challenge — can we recognize the losses without impairing the functioning of our financial and banking system?

Written by infoproc

September 16, 2008 at 2:56 am

A trillion in the balance

leave a comment »

Via Calculated Risk. If house prices return to historical norms, mortgage credit losses will be roughly $1 trillion.

From Jan Hatzius, Goldman Sachs chief US economist, presented at the Brookings conference, Beyond Leveraged Losses: The Balance Sheet Effects of the Home Price Downturn. Here is a short excerpt on estimating mortgage credit losses (note that Goldman is now forecasting prices to decline another 10%):

If nominal home prices remain at their 2008 Q2 level until mid-2009, before reverting to a +3% annualized trend, our model implies that mortgage credit losses realized in the 2007-2012 period will total $473 billion. If nominal home prices fall another 10% through the middle of 2009, the model projects losses of $636 billion. Finally, if prices drop another 20%, predicted losses increase to $868 billion. Moreover, the table suggests that losses peak in the third quarter of 2008 if home prices are flat going forward; in the fourth quarter of 2008 if prices drop another 10%; and in the second quarter of 2009 if prices drop another 20%.

Written by infoproc

September 12, 2008 at 9:53 pm

Credit crisis: half way through?

leave a comment »

Case-Shiller is down about 20% since the peak in real terms, and we have another 20% to go before the metrics are back in normal range. This would give back all the housing gains since the late 90s. Commercial real estate and consumer debt crises have yet to materialize but are likely to follow.

10 million Americans are underwater on their mortgages. The UK is in even worse shape, with a 70% drop in mortgage approvals.

Bear is gone.

Paulson and Treasury looked closely at the GSEs and decided it would be better to move now than to pay more later. Was there a trigger event? Were there scared Chinese bankers on the phone with Paulson? (Who do you think are the biggest holders of GSE paper?)

The Korea Development Bank looked closely at Lehman and took a pass. Lehman’s stock is now in free fall.

Merrill financed 75% of its recent sale of $30B in mortgage securities at 22 cents on the dollar. The counterparty is a straw man, so Merrill will have to take the securities back if the price falls even further. That means Merrill could still go the way of Bear and Lehman eventually.

No mark to market, no resolution in sight.

Socializing losses after privatizing gains in the fat years.

Of related interest: great podcast interview with Robert Shiller. Act now, because Bloomberg has a tendency to take these down after a while.

Written by infoproc

September 11, 2008 at 12:55 pm