The New Republic

There are two things wrong with securitization.  The first is that there is not the appropriate incentive to evaluate risk.  This is not likely to be mitigated by requiring originators to have "skin in the game" because, by any measure, the return to the originator is going to be enormous relative to its share of the risk.  This is precisely how the financial institutions got loaded up.

The second thing is that securitization allows financial institutions to evade leverage restrictions because the securitized loan portfolios have no capital requirement as financial institutions do.  Capital requirments have both a a per institution protective purpose and a systemic protective purpose.  By getting the stuff off their books, the institutions limit their own leverage -- the per institution purpose -- but aggravate systemic risk, as we have just learned.

The answer to both problems is simple:  Require securities portfolios to have the same equity capital percentage as banks, provided by bona fide third parties, not by the originators.  The ability to get equity investors to buy in will serve to assure quality and to prevent the systemic runaway problem.

Basically, it is STUPID to have leverage restrictions and then allow them easily to be evaded through securitization.  These portfolios are one-off banks with zero equity.  What do we think is going to happen?  Just what happens with banks with zero equity -- unreasonable risks.  Duhhh.

- roidubouloi

June 17, 2009 at 9:11pm

That's the sharpest comment I've seen on here, and one that will be missed without a deep understanfing of the securities industry.  Well played.

- jcooney

June 17, 2009 at 9:29pm

Roi nailed it.

- abrod

June 17, 2009 at 9:30pm

roid:

"There are two things wrong with securitization."

george:

Actually, the more pressing concern must revolve around comparing and contrasting what "securitization" is said to mean in places like this and how it will be "tinkered" with on Wall Street by the Gordon Gekkos.

And one way to calculate this is to go back to 1999 and see how it all unfolded in the sub-prime fiasco.

If there is one thing pundits fail to take into consideration it is the gap between the world of words they live in and the world in which the words are stuffed into the actual transactions on The Street.

There, money doesn't talk, it screams.

Roid is certainly a master of the microeconomic world, no doubt about it. But invariably he confuses the trees for the forest.

george

- iambiguous

June 17, 2009 at 10:48pm

Securitization is akin to reinsurance in the insurance sector - it allows the originator of the risk to spread it around among other parties. Risk spreading is generally good AS LONG AS risks aren't strongly correlated (e.g., fires/earthquakes/tidal waves damaging or destroying skyscrapers in NY and LA) and the parties taking on risks others originated aren't also originating their own similar risks.

Banks buying up each others' securitized mortgages is akin to reciprocal reinsurance in the insurance sector - this is common in continental Europe, but not in the UK or the US. There's no good reason for this other than having a larger pool of capital available to cover a larger pool of losses. Regulators may like it, and it may make sense for pooling very large but very rate potential claims/losses, but it makes no sense when there could be a lot of little losses - it's better for the economy as a whole for one insurer to fail spectacularly than for many or most to fail by just a bit.

The problem with CDOs is clearly underestimation of the correlation risk. If each new home mortgage default increases the likelihood of more defaults, it's difficult to estimate the potential down side, and therefore difficult to estimate appropriate risk premiums.

I'll admit I don't know the applicable risk based capital rules, but it seems likely they didn't contemplate CDOs and their attendant risks.

- hrlngrv

June 17, 2009 at 11:29pm

hringrv,

It is not just a question of correlations.  You could in theory diversify the risks within a sector completely perfectly if every investor were given a tiny sliver of every mortgage, but, if the entire sector goes bust, the diversification does not matter.  This is how the oil sector took down Continental Illinois National Bank & Trust Company of Chicago in the '80s (I like writing the whole name because I miss them.   At the operating level, they were far and away the smartest, most professional bankers I dealt with, and I dealt with all the majors.  They were undone by bad risk management policy at the top -- why does that suddenly sound so familiar?)

The problem is much worse, however, than mere sector risk because there is feedback between the asset and liability sides.  The availability of financing allows a bubble to develop and get way out of control.  When Glass-Steagall and margin requirements and the rest were imposed in the 30s in response to the Crash of '29, the people doing it were quite deliberately limiting the overall leverage in the system, not just protecting individual institutions.  In the Reaganite de-regulatory fever that swept over even the Clinton administration, the nation forgot.  (I didn't forget.)  When Alan Greenspan said that we could not have anticipated these events, this type of market failure, he was flat-out lying.  You only needed to know the economic history of the 20th Century to know that, freed of leverage restrictions via CDOs and derivatives, a bubble and a bust were inevitable.  That is what financial markets do if they are not controlled.  The have ALWAYS done this with regularity since they were invented.  

It was outrageous for Greenspan to say that this came as a surprise.  It was only a surprise to Greenspan, Rubin, Summers, et alia, because of their own arrogance and hubris.  Just like the geniuses at Long-Term Capital before them, they came to believe that they were so smart that they could get on the back of the leverage tiger and ride it successfully without leverage controls.  That is NEVER going to work for fundamental reasons having to do with the relationship between capitalization and demand that, frankly, these guys should understand but don't because it is heterodox economics and they are nothing but intellectual conformists.  They can only think the thoughts that everyone around them thinks.  

Without leverage controls for the financial system as a whole, imposed so that no instrument can escape, this will happen again, sooner rather than later.  But, of course, the financial institutions are already fighting to protect their phony profits -- the profits it looks like they are making until you factor in the losses they are building in and up.  They are so nuts, that I have not the slightest doubt that they are down in Washington at this moment pleading that they need to retain their "lucrative" businesses in securitization and derivatives and not lose them to regulation.  The proper response to that is, "If you think this business is lucrative, you are insane and should be locked up in a loony bin, not in charge of an important financial institution."

As ever, I cannot quite figure out what combination of venality, greed, cowardice, stupidity, incompetence, and social subornation by the monied elite accounts for the behavior of Geithner, Bernanke, and Summers.  Greenspan was appointed to be the puppet of the monied class, so it was never hard to understand his behavior.  For the other three, I know that they are screw-ups;  I just cannot figure out quite why.

- roidubouloi

June 18, 2009 at 10:32am

roidubouloi,

I should have said POSITIVE correlation. Negative correlation is the nirvana of diversification.

When there's strong positive correlation, it means as the first few individual risks in the sector go, so more often than not the entire sector eventually goes. I'll admit I'm more familiar with earthquake and flood insurance modeling than bank solvency, but there are no effective diversification strategies if you can only insure buildings in the L.A. basin or homes in the (flood) plains of Iowa.

Mortgage CDO diversification would require taking no more than small portions of mortgage portfolios from many different banks (so presumably many different levels of conservatism in assessing borrowers loan worthiness) in many different regions (to avoid localized economic slumps like in the Rust Belt) in many different countries (so presumably different levels zealousness of government regulatory oversight), and even then if the global financial system goes to Hell, you're screwed.

We made a big mistake repealing Glass-Stesgall. Commercial banking needs to become a boring, low ROE business again. Insurance companies insuring financial risks need to be subject to federal solvency regulation for that portion of their book. Even better, they should be subject to government regulation IN EVERY COUNTRY where they operate (that ought to provide some disincentive to globalization, which, from the hindsight perspective, seems to be adverse to diversification). Investment banks and hedge funds must be limited in how large they may become.

Clearly total exposure to loss (no matter how 'unlikely') needs to become a larger factor in solvency regulation for all financial businesses, not just for commercial banks and insurers.

- hrlngrv

June 19, 2009 at 5:28pm