Funny story.

Three years ago I started working for an investment bank that raised money for firms that sold CDOs - one of the financial derivatives that caused the grand market shit storm in whose splatter we are still living. As a fledgling young analyst, I experienced a feeling of disquiet because I couldn't understand how CDOs created value. I felt dumb. I was especially embarrassed because I had a degree in Economics and did well in school, so I assumed I should be able to grasp the idea behind financial products at least at a high level. But my doubts were soon drowned out by a torrent of work. I blamed my ignorance on my inexperience and moved on.

After all, "they" wouldn't build a trillion-dollar market based on bullshit, would they?

For those not current on financial jargon, a Collateralized Debt Obligation (CDO) is made by taking a $100 million worth of mortgages, throwing them in a box, and then selling pieces of the box for a total of about $105 million. ~ $1 million gets paid to the investment bank who sells the pieces and the rest of the profit gets kept by the firm who owned the mortgages. Not a bad gig.

In fact, CDOs weren't always composed of mortgages. You could put other financial assets in the box, too. You could even buy pieces of other CDOs, throw them in a new CDO, and sell the pieces for more money. These are called "CDOs-squared". I-shit-you-not.

By now you might understand why I was puzzled by how this activity constituted a valuable economic enterprise. Theoretically, CDOs created value because owning half of two mortgages is safer than owning 100% of one mortgage. And that's true. If one mortgager defaults, you still have 50% of a mortgage that's sending you monthly payments. But that line of reasoning is a lot more true if you're the kind of person who buys one mortgage than if you're the kind of person that buys 10,000 mortgages. If you have a large budget, then you can already diversify pretty well on your own. CDOs are a convenient source of diversification, but is the convenience enough to justify paying some yahoo middleman a 5% cut?

As it turns out, the answer is "no". Part of that 5% value creation was generated from convenient diversification, but part of it came from the fact that once you throw all the mortgages in the box it's harder to figure out how risky it is to own them. So "value" was created by temporarily hiding the down-side behind a chain of opaque securitizations. This later blew up in our collective faces.

When the economy started to collapse partially from the collective weight of toxic unpriceable CDO slices, I remembered my earlier confusion and laughed. Maybe I wasn't so dumb after all.

Share this

Part of the problem was also

Part of the problem was also the real-estate bubble traveling upstream. A mortgage has two aspects, payment and property value. Investing in someone else's mortgage is a dual risk, but it's easy to mislead yourself into thinking that these risks are independent and offset each other (in reality they can be linked, as we've seen). Which is what, I'm certain, a lot of investors in these repackaged mortgages believed. Even if you accept the fact that people who currently have insufficient income on the books to pay their mortgage through its conclusion are likely to default (a pretty straightforward calculation) you may still rest comfortably in the notion that the result of such a calamity for the individual will still see right by you, the distant investor, due to the foreclosure of their property, still good for the remainder of the value of the loan (or more if you're lucky!).

This is, of course, a perfect example of greed overriding sound rationality and business sense, classic prerequisites for a speculative bubble. So we had a bubble on top of a bubble, and when it collapsed all hell broke loose.


Hey guys, each of these posts provide a really clear, helpful analysis of these issues. This is a true public service.

there's a word for that

"So 'value' was created by temporarily hiding the down-side behind a chain of opaque securitizations."

The word for that is fraud.

When you have customers for a demand deposit scheme who put $500,000 into your bank, you are agreeing to keep those funds "on deposit" so they can have them back "on demand." Thus the term demand deposit.

When your bank then lends out, because the government licenses such lending, $3 million in auto, home, and commercial loans, guess what that is? That's fraud. Your bank never had $2.5 million of those dollars to lend out, under any conditions. That part of the $3M was created out of nothing. Your bank shouldn't have lent out the other $0.5M because that was on deposit and available on demand.

Saying that the market allows for fractional reserve banking because most of the time the loans are paid back and most of the time the customers don't all want their deposits back is bullshit. It is of the same order of bullshit as the collateralized debt obligations boxed up and packaged as sausage, but actually just sheep guts filled with shit.

Bankers don't work for a living. They fuck over everyone else for a living. Like in Amsterdam, you can make a living at it, but you need a license from the government. Unlike in Amsterdam, where the health service comes around and checks for venereal diseases, nobody gets to look at what the banks are doing.

Now, it doesn't take a government agency to check for venereal diseases. I'd prefer a "clean prostitutes aren't skanky" seal of approval from CPAS, the clean prostitutes aren't skanky testing team.

Maybe we should develop a "this financial institution offers sound money" seal of approval. It would be soooo easy. After all, we won't ever find a bank or financial institution that actually offers sound money in the traditional finance sector. And how long would it take to evaluate the open source software behind Loom and Trubanc?

Actually the problem arose

Actually the problem arose because they did not expect the housing markets across the country to correlate so closely in a fall because "it had never happened before" (BS - great depression, but since then i guess). But yeah, diversification broke down because of high countrywide housing market correlation going south.

I maintain that the

I maintain that the underlying cause of the bust was the original price bubble. But it was made worse by the excesses of the financial markets, which both fueled and fed off the price bubble.

Ah the beauty of hindsight

I'm puzzled you worked with CDO's but you don't understand what a tranche is?

The purpose of CDO is not just to make a basket of security but to create tranches, the most senior tranche gets paid first, not matter what, so even if many mortgage default, it still gets its coupon. Then if the senior tranche can be paid the other tranches are paid in order etc.

This approach allows funds who are legally obligated to invest in AAA security to access a larger pool of fixed income securities. It changes the risk structure to cater to a market where some investors are not very risk averse and some are very risk averse.

The correlation between mortgages was vastly underestimated which caused the CDO's to not be as safe as expected.

I think what made the mispricing catastrophic is the following:

You're a banker, you come up with a product with a superior yield that seems safe. What do you do? You borrow short term and invest in that product. In a free market, most of the short term liquidity is pumped into these products and the short term interest rate rise accordingly. The "cost" of the mispricing is more expensive loans for other ventures. If the short term interest rate is manipulated to stay artificially low, the arbitrage can only stop when the yield of these assets becomes lower, which means in the case of mortgages that so many houses are built that people are no longer willing to enter mortgage agreements. It's a much longer process and the result is a bubble.