The Toxic CDOs

It's amazing to me that the big issuers of mortgages at the top -- Merrill Lynch, Lehman Brothers and Bear Stearns, as well as New Century Financial and NovaStar -- didn't know more about what they were packaging and that they were so reliant on the agencies and insurance for protection. Nor did the insurers, like AIG (AIG) , know what they were insuring.

It is amazing because all of these entities relied on the ratings agencies, which in real life they never believed to be doing a thorough job anyway, and on their own mortgage originators, whom they didn't trust either.

No one really believed in this stuff. All of these loan packages that were put together? The smart people knew that the default rates could be huge. But every one of them had the same model. They all figured they could get away with it, that they could foist it on clients. The clients were often more than happy to trade it.

For one simple reason: They just used it for arbitrage.

I am thinking about this great arbitrage trade because of the fallen trader at Deutsche Bank, who, as the Journal says this morning, "left behind $1.8 billion hole."

You know how to leave behind a $1.8 billion hole? You buy one security and you short another and profit from a small difference. It works only if you borrow money, 20, 30 times over, and profit each month from the difference. A consistent profit. Just as consistent if you bought a CDO with Treasuries and held it. And again, borrowed 30 times your capital to profit from it.

The reason these mortgages are so intractable and can't be pulled apart is that they were only put together for this arbitrage. All of the banks, all of the hedge funds, all of the bizarre clients who bought this junk were relying on a model of unemployment -- as long as unemployment stayed low, people don't default. The model of course was wrong, and that's what wiped out the arbitrage, because the high-interest CDOs suddenly had principal risk, and the low-interest loans that they took against their collateral from the brokerages didn't.

This is all so pertinent because the CDOs weren't packaged with an eye toward maintaining the mortgages underneath, they were packaged strictly so the brokers could make profitable loans to clients to buy the CDOs to do the arbitrage.

Now 80% of the mortgages issued during 2005 to 2007 are in those CDOs, and given that most were issued by brokers who knew nothing about issuance and by scam artists who are out of business, there's simply very little worth to them.

And Merrill knew it. Lehman knew it. Bear knew it. They knew it because all they had to do was read the papers to know where their mortgages were from. They knew it because they knew the loan-to-value of the mortgages within in the CDOs. They knew it because in many cases they had funded hedge funds -- literally started hedge funds with their own borrowed capital -- which then bought this stuff from them on a 30-to-1 basis.

The credit guys knew it, but they were only making about $100,000 and they are easily dismissed. They knew that the hedge funds had to put up more margin to make this thinly capitalized trade, so therefore the bosses knew it too, but they looked the other way.

The only reason Goldman was the only firm that didn't get hit hard -- because they packaged mortgages -- was that they listened to the credit department because it was filled with partners who made a huge amount of money. That was always the Goldman way: to reward people who did a great job who saved the firm every day, not just those who drove revenues.

So Goldman wasn't caught with much inventory, and what they were caught with they were hedged against, and not with AIG for the most part. The stories that said that Goldman demanded that AIG be saved and beseeched Hank Paulson were wrong. They had some counterparty risk in this stuff, but for the most part they just ran a hedged position, and that's what saved them. (They didn't hedge the private-equity bond exposure, which is where their pain came in.)

The CDO arbitrage was largely unwound by the end of 2008, which is why there is some hope in the system. But as we see from Deutsche Bank, the arbitrage came in many forms and with many bonds, and that's not done. It is a huge user of everyone's capital, and it isn't talked about much.

It should be made aware that it was one of the intractable aspects of this era, because the CDOs were never meant to be anything but pieces of paper to leverage against Treasuries. There is nothing to look through them, and there is no master trust staff to try to fix them. There's just a bunch of guys in a room who send out foreclosure notices and don't make any money anymore.

Until all of these servicers are dealt with the way Goldman Sachs is dealing with things, by slashing the principal, the $2 trillion in issued mortgages will simply keep poisoning us (I am not even talking about whole loans, which are only 20% of the problem).

The reason housing should be fixated upon is because of the servicer problem. They keep pumping out foreclosures as fast as we draw down home inventory. Now that unemployment has spiked -- and the number this morning with revisions is much worse than people are making it out to be -- they will do it faster than there is demand unless Bernanke buys mortgage bonds and we get that $15,000 tax credit to buy just old homes. The new provision includes old and new homes, which is really bad because the homebuilders will be able to stay in business because of that clause. Another example of bad policy condoned by Obama.

The servicers are the problem. They are a legacy of the arbitrage.

They aren't being dealt with other than in the Goldman Litton servicing arm. They are choking us.

All because some hedge funds and brokers figured out a consistent way to make money month after month that was immediately inconsistent when the first homeowners started defaulting and the credit departments issued the margin calls that made the Merrills and Lehmans and Bears take back the CDOs that ultimately sunk the first two ... and now threaten to sink Bank of America, too.