Monday, October 06, 2008

A primer on the financial crisis

I saw last night that 60 Minutes tried to explain the great finance crisis that's hitting the world (but, and let's put the blame where it belongs, it started in the U.S.). They didn't do a great job of it. They used sound bites to explain what caused the problem and why it was bad without explaining what really happened and how. Not that this is surprising for television reporting.

I did some digging and found the following information, culled from several sources. Special thanks to a user named Tavella on for posting an explanation, from which this post is culled.

* * *

Banks like giving out mortgages. This shouldn't come as a surprise to anyone. They are a safe vehicle for providing credit. Traditionally the failure rate is around 5%, peaking at around 15%. When mortgages do fail, they are attached to collateral: actual property, for which there is always very good demand. The value of property generally trends upwards. There are occasional bubbles and bursts that affect home owners and investors in the short term, but that shouldn’t be enough to sink a bank.

So, what happened?

With the rise of China as a source of the world's cheap crap, and the increase in oil revenues, and several other things, the 1990s saw the rise of a lot of investment money. The investors wanted a safe investment vehicle. Mortgages were a safe investment vehicle! So, the banks started selling mortgages. After all, people usually paid their mortgages (and if they hit hard times, they'd usually pay their mortgage before any other debt), and if they didn't the bad debt was tied to collateral.

The way banks sold mortgages was through something called a Collateralized Debt Obligation (CDO). This is one of those things you've heard about called a "derivative".

CDOs work based on the default rate of mortgages. The banks took a whole bunch of mortgages and lumped them all together. The banks knew how much money they would make from those mortgages, both as monthly payments and as interest. However, between and 5% and 15% would fail, based on historical data. 5% were likely to fail, and there was some risk that up to 15% would fail. So, the banks took this lump of mortgages and split it into what they called tranches. They took 85% of the mortgages and put them into the A tranche. Of the remaining 10% they put that in the B tranche. The leftovers were put in the C Tranche.

These tranches were then sold to investors (usually large institutions). The A tranches got the highest AAA ratings. The B tranches got BBB rating. No one touched the C tranches, so the banks kept that for themselves and paid themselves nice interest rates. The risk was spread across the tranches. The difference, as an investment, between the tranches was based on how they were paid out. When a mortgage is paid out, investors get a return based on the tranche they bought. The C tranches got the highest percentage return, then the B, then the A. However, if some of the mortgages defaulted, what money was recovered went to the A tranche first, then the B tranche, then the C. The A tranche, therefore, was considered safer, though it generated a lower rate of interest.

These CDOs were so popular that investors started asking for more and more of them. Unfortunately for the banks, only so many people could afford to buy homes. Too bad, as these things were generating a lot of income for the banks.

They were also generating a lot of income for the investors, because housing prices kept increasing during this period (with the occasional bubble, but mostly moving upward) which increased the dollar value of mortgages, and the payout to investors. This put even greater pressure on the banks to generate more of these securities.

This is where things start to go wrong.

You can’t just “create” more mortgages. You only get more mortgages if you find ways to lend more people money to buy property. The banks started to get "creative" in the types of mortgages they produced, in an effort to get more people to take out mortgages.

The most infamous of these creative mortgages is the sub-prime mortgage. You got a mortgage at a really low interest rate for a set amount of time. At the end of that time, the rate would go up, sometimes drastically. You could then take out a loan on the equity that you'd built up or sell the house for a profit before the mortgage came up for renewal. Or, you sucked it up and paid the higher interest rate.

Now, I want to mention something about Canadian mortgages. In Canada, most mortgages are variable rate. I was quite surprised to see that American banks would offer fixed rate mortgages that never changed for the lifetime of the mortgage. In Canada, you typically take out a variable rate mortgage. Every 3 to 5 years your rate changes. This is common, but it’s not a “sub prime” mortgage. The initial amount of the mortgage is at a competitive rate. The rate doesn’t go sky high after the initial term in order to make the bank more money. The bank is happily making money right from the get go, while the homeowner has a really good idea of what they got themselves into. (Not always. A lot of people got hit with high interest rates in the late 70s. I remember my parents being happy about renewing their mortgage at 14% for a couple of years back when some people were hitting 18% and higher.)

The sub-prime mortgage wasn't the only type of creative (perhaps “shady” is more accurate) mortgage. There were some where your payment wasn't even enough to cover the interest, or banks wouldn’t look too closely at someone’s credit history. The point was to sign more mortgages to fuel the demand for CDOs. It didn't hurt that both the Democrats and the Republicans saw it as a good thing for more people to buy homes.

The wheels didn’t fall off the cart with CDOs and shady mortgages, but they certainly became very wobbly with the introduction of Credit Default Swap, or CDS. 60 Minutes concentrated on CDSs last night.

A Credit Default Swap is a derivative in the form of insurance. You have an investment and you think it might not completely pay out. You’re going to lose money. Another company comes along and says, “Yep, you’re right, I think it’s going to fail. Tell you what, if it doesn’t pay out I’ll pay you the difference. In the meantime, you give me monthly payments.” The first company buys a CDS. They make payments to the other company, the seller. If the security falls through, the buyer either hands over the collateral to the seller for the value of the security, or the seller pays the buyer the difference. In the meantime, the buyer makes payments to the seller.

(Simplified example:) Let’s say you have a mortgage security for $1 million. You think you’re only going to get $900,000 out of it. You pay the seller, oh, $20,000 a year for five years in monthly payments. If the security falls through within 5 years, the seller of the CDS takes the foreclosed property off your hands and pays you the $1 million. Or, you keep the property and the seller pays you the difference between the property and the value you can get for it. If the mortgage security doesn't fall through, the seller of the CDS makes money in the form of monthly payments.

The big banks and investment houses had their own CDSs, so sometimes they would just trade one CDS for another, betting that the investment they were giving the CDS against wouldn’t collapse. CDSs were sold as investment opportunities. Banks and corporations sold them to manage risk. If the risk was lower than anticipated, the seller of the CDS made money on the “premiums”. If the risk was higher than anticipated, the seller had to cover the loss, but presumably the banks and investment houses did some math to estimate how much money they were really risking. This is the big risk with these items: the seller of the CDS is now on the hook for bad debts. They’re going to owe money if the CDS falls through.

Now that the banks had a way of mitigating against loss, they started making securities based on even riskier mortgages. And they started playing around with how they created CDOs. This is where the wheels start to fall off.

Lets look at those CDOs. The A tranches sold well, because the risk was pretty low. The B tranches, with the riskier mortgages, not so much. Investment analysts needed a way to sell more of the B tranches. They decided that not all of those B tranches were really going to fail. They figured, oh, 20% of the B tranches would be bad, but the rest were good.

So, they bundled a bunch of B tranches together and created a new security out of them, called a second stage derivative. They took 80 percent of this bundle and called it an A tranche. The credit rating agencies nodded their heads and declared they were nice and safe as A tranches. Now investment companies were buying a whole lot of B tranches thinking that they were as secure as A tranches.

Wouldn’t you know it, the banking houses built CDSs out of these second stage derivatives, pretty much like they did with the first stage. Now, you and I can see that these second stage derivatives are riskier than first stage. And, we can see that there are already CDSs out there for the first stage derivatives, and now they’ve created them on the second stage, too.

But it gets better (or worse, if you will). Remember that with a CDS, the seller gets a premium each month. Well, that’s income, so the banks turned around and created a CDO out of that, too, which in turn was used to anchor a CDS. Then the investors got really inventive and came up with even weirder — and riskier — derivatives.

Now you had more and more securities built around existing securities. If a security did well, as it would while housing prices increased, you had a magnified effect of greater profits.

The housing boom and easy-to-get mortgages brought a lot of people into the market who shouldn’t have been there. It also brought about things like “flipping” (where someone bought a house cheap, spent money to fix it up, and then sold it at a profit; it seems like the TV channel TLC is based around this whole flipping thing). This fueled an increase in housing prices, making the bubble bigger. People took out loans on the equity on their house and used it for consumer spending. Other people bought property out of panic. More and more mortgage were produced, with more and more securities based on them, and with the securities themselves leveraged so the banks could invest even more heavily.

This worked well while housing prices increased.

Then the bubble burst.

A glut of new homes hit the market, slowing the rate of increase in home prices as homes took longer to sell. Subprime mortgages came due, and suddenly people were shocked at the increased interest rate. Homes weren’t worth quite as much as the owners thought. Refinancing became an issue. Some people got into a negative equity situation (the mortgage was for more than the house was worth), and had to make up the difference when it was time to renew the mortgage. Instead, people walked away from the home.

Banks foreclosed, but they couldn’t easily get rid of the properties they had acquired; the bubble was busily bursting.

All those CDSs were supposed to ease this sort of risk. Except the securities were too highly leveraged. There were too many securities based on the same crappy mortgages. The investment banks couldn’t afford to pay out the A tranches, let alone the B and C. They didn’t have enough money, and couldn’t liquidate enough securities to get it, because those securities had dropped so much in value.

That’s when the banks and investment houses started to collapse. I over simplified a lot of this, but that's essentially what happened. I'm not sure anyone really understands everything that happened in great detail. If they did, they were a mental oddity: intelligent enough to figure out these derivatives but not intelligent enough to figure out this house of cards would crash, and crash hard.

Peculiar political ad

There's a peculiar political ad floating around on Louisiana TV. I just noticed it this weekend, partly because of Alana's surgery, and partly because I mostly avoid the local channels.

The ad is an attack ad by Republican Louisiana Treasurer John Kennedy against incumbent Democratic U.S. senator Mary Landrieu (Kennedy is running against Landrieu). The ad accuses Landrieu of voting 81% of the time the same way as Barack Obama.

What's odd about this ad is that, somehow, the Republicans think this revelation is going to come as a surprise. Really? Two Democrats voted the same way? Say it ain't so, Joe... I mean, John!

Well, actually, yeah, it was a surprise. I mean, she only voted the same as Obama 81% of the time? What about the other almost 20%? Which one of them is less liberal, Obama or Landrieu. I'm actually semi serious about this; I'm really curious about the bills where they didn't vote the same.

Anyway, I really can't imagine why the Republicans are wasting money on this ad. Surely they could have come up with a better ad than that, one that doesn't verify what most voters already knew, that Landrieu was in the same party as *gasp* Obama.