How Collateralized Debt Obligations Work

By now, you've heard of the subprime mortgage lending problem.

Back in the 1980s, banks held mortgages themselves, instead of selling them to other investors. When interest rates shot up to 10% or more, a bunch of small banks suddenly found themselves insolvent. They owned mortgages that were only yielding 6% while they were borrowing at 10%-15% or more. Of course, only small banks were forced into bankruptcy. As usual, large banks had the capital base to withstand the crisis and survive until they were bailed out by a Federal Reserve interest rate cut.

Rather than fixing the fundamental structural flaw in debt based money, a different solution was proposed. Banks should sell mortgages immediately after issuing them. The problem was not the unsoundness of debt-based fiat money after the abandonment of the gold standard and skyrocketing interest rates (artificially raised by the Federal Reserve). The problem was that banks were subject to interest rate risk. Banks were borrowing at the short-term rate from depositors or the Federal Reserve. Banks were lending at the long-term rate. When the Federal Reserve is forced to jack up interest rates to prevent hyperinflation, this long term debt at a lower interest rate loses much of its value. This, combined with use of leverage, makes banks insolvent.

The solution is that, instead of keeping mortgages on their balance sheet, they are packaged into bonds and sold. However, mortgages are an inherently risky investment. There are three risks associated with a mortgage. The first risk is that interest rates will rise, making the long-term loan an unattractive investment. The second risk is that the debtor will default on his loan. The third risk is that interest rates will fall. When interest rates fall, all outstanding loans are worth more. However, if the debtor still has sound finances, he will refinance his mortgage at a lower rate. There's nothing that can be done about the first risk, but the other two can be alleviated with the invention of the "collateralized debt obligation" (CDO).

To reduce the risk, a bunch of mortgages are packaged into a single bond and sold. Typically, hundreds of thousands of mortgages are packaged into a single group of bonds. This means that there is zero scrutiny of the creditworthiness of individual debtors. However, the risk is not spread equally. The risk is divided into "tranches" A, B, and C. In the examples I read about, there were 3 tranches, but it could be any number.

The "A" tranche is the highest quality. The "B" is in the middle, and the "C" is the lowest. When defaults occur, the "C" tranche loses first, then "B", and finally "A". Similarly, if debtors choose to refinance, the "C" tranche takes the hit first, then "B", and finally "A".

Even when the economy is great, there are going to be a certain number of people who refinance because they move, or a certain number of defaults.

The "A" tranche gets a great credit rating. It can be purchased by insurance companies, who are required by law to invest in high-quality bonds. The "B" tranche is speculative. The "C" tranche is junk.

These mortgage bonds yield around 6% for the top tranche, 8%-10% for the middle, and 15% or more for the bottom tranche. Hedge funds will borrow at the Fed Funds Rate and buy these bonds. Their usual leverage trick enables them to earn huge profits.

That's another important point. Most of the hedge funds who buy these bonds are *NOT* taking a simple long position. They are taking a *LEVERAGED* long position in these bonds, with leverage ratios of 10x or more. When the price of the bonds crashed, they were unable to meet their margin calls. *EVERYONE* who bought these bonds was using leverage. When the hedge funds were forced to sell due to margin calls, there were no buyers. In fact, all the hedge funds were afraid to sell at all, lest all the other hedge funds be forced to "mark down to market". During the crash, if these bonds were fully marked down to their current market value, a huge number of banks and hedge funds would have been forced into bankruptcy.

The problem was extensive use of leverage *COMBINED* with defaults *COMBINED* with a sharp drop in market price. There were literally no buyers, because all of the potential buyers had already maximized their leverage.

The size of each tranche and the risk are calculated using complicated mathematical models. Under the assumption is that the defaults are uncorrelated, the "A" tranche has earned its great credit rating. Even though each individual mortgage may have a junk rating, the tranche structure shifts almost all the risk to the "C" tranche. According to the mathematical model, the "A" tranche has practically no risk.

Do you see the fallacy yet?

SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE
SPOILER SPACE

The fallacy is the assumption that defaults are uncorrelated. Due to the Compound Interest Paradox, economic busts are inevitable. The weakest class of debtors is forced into bankruptcy during each bust phase. This time, the weakest class of debtors were people holding subprime mortgages.

The "A" tranche got an investment grade rating UNDER THE FALSE ASSUMPTION THAT DEFAULTS ARE UNCORRELATED. However, defaults *ARE* correlated. It's not because the debtors got stupid or lazy. A certain number of debt defaults ARE GUARANTEED BY THE RULES OF THE MONETARY SYSTEM. During *EACH* bust phase of the economic cycle, it is GUARANTEED that the weakest class of debtors is forced into bankruptcy.

Even though fancy and sophisticated models gave the "A" tranche of the CDO its investment grade rating, this was nonsense. The problem is that the "defaults are uncorrelated" assumption was WRONG.

With CDOs, banks no longer have any responsibility if a mortgage they issue goes bad. They are immediately selling the mortgage and liability to someone else. A bank is now merely a broker who collects a fee for issuing the mortgage and selling it. However, that is *NOT* the underlying problem. The problem is that the CDOs were structured under the false assumption that debtor defaults were uncorrelated. This, combined with extensive use of leverage, caused the crisis.

Banks also issued "mortgage insurance", guaranteeing the investment grade rating of the "A" tranche. If defaults were uncorrelated, this insurance would have been worth something. However, when a bunch of defaults occurred all at the same time, then the companies that issued mortgage insurance wound up going bankrupt.

The important point is that the CDOs were structured under the false assumption that defaults were uncorrelated. Due to the fundamental structural flaw in debt-based money, defaults ARE correlated. A debt crisis like this happens every 5-10 years. It is no accident. The rules of the monetary system *GUARANTEE* that such a crisis will periodically occur.

Naturally, people who buy CDOs are going to be a lot more cautious for awhile. The spread between the Fed Funds Rate and the rate charged on mortgages will increase to compensate. As one bubble bursts, another is being inflated. There *WILL* be another similar crisis in 3-5 years. You don't know where it will be until it occurs. From the point of view of the average person, by the time news of a popping bubble occurs, it's already too late to sell. Only insiders get to profit immensely from each boom/bust cycle.