Not all of them where signed off as fighting fit. The assumption was that less than 10% would fail so you could break the risk out so someone who want's zero risk and is willing to only make 4% is safe as long as 80% of them live. The math would have been fine if each morgage was independent but because the economy as a whole tanked the default rate went way up and people who where not willing to take much risk still got a beating.
The math would have been fine if each morgage was independent but because the economy as a whole tanked the default rate went way up and people who where not willing to take much risk still got a beating.
Close, but I'm going to disagree with you on one detail. The subprime debacle began in the spring of 2007, and really began to spiral out of control on July 26 with various calamities in the CDO markets. By August 6-10 (the week of quant turmoil) it was starting to have an effect on equity markets-- which were strong at the time, and didn't start taking a real beating until Jan. 2008.
The economy was strong in 2007, with 4.9% (annualized) growth in Q3. Subprime didn't blow up because the economy tanked, because the mess began during a great economy, and turned it into a bad one. Of course, the tanking of the economy has made the credit situation even worse and will lead to even more defaults.
The default rate goes up during a bad economy, and everyone knew that this risk was there, and that a cyclical recession would occur sooner or later. What caused the rate to be "unexpectedly" high was the prevalence of speculation. In a bad economy with falling house prices, a resident homeowner is still going to make his best efforts to avoid the embarrassment and hassle of foreclosure, even if he has negative equity. Negative equity is irrelevant if he's going to live there for 10 years. On the other hand, a greedy and unscrupulous speculator has no reason not to default.
To clarify people that where taking the high risk stuff got hit early when the economy was doing "well" but for the most part the "AAA" stuff could withstand a fair amount of foreclosures with little impact. The crap nobody wanted was really high risk so banks kept it which is one of the reasons why Lemon Brothers failed. Granted there was also a fair amount of crap rated like gold, but late 2007 to now we have been in a recession which kept feeding off of more and more people getting bitten as an ever increasing number of mortgages failed while nobody wanted to buy the property's.
Let's say a bank loans 300k on a house that was worth 200k but got inflated to 300k in a bubble fine somebody takes a 100k hit when it fails but they are only down 33%. However, if nobody is going to buy it for six months and by then it's only worth 150k and the spiral continues.
I also think the economy was fubar as early as 2005 with the fed using low interest rates to make everyone pretend that the economy was fine this was really helped by CDO's and people taking out risky mortgages. Then there was a little bump in interest rates which finally set this ball rolling but it was obviously going to happen at some point. Basically, when any little thing is going to upset the house of cards there is already a problem even if it looks ok.
PS: When the economy tanks it's a lot harder to prop up stupid companies so we are seeing a lot of things that where on their death beds for a while all fail but in the long term this is a good thing.
Explanation fail. This works perfectly well so long as you know how diseased the cows are, and the odds of any one cow dying are independent from the next. It's called math.
The actual problem with CDOs is that the cows were all more diseased than the historical track record suggested (because after they started putting together CDOs, a flawed incentive structure created a new motive to buy sicker cows) and that the probability of one cow dying wasn't independent of the next cow dying.
They didn't know how bad the mortgages really were, and the failure of one mortgage is not independent from the failure of the next. We're seeing this now in neighborhoods were foreclosures are driving housing prices down and making people want to walk away from underwater mortgages. This is starting to happen in California.
"The actual problem with CDOs is that the cows were all more diseased than the historical track record suggested"
I think you are splitting hairs here. The point remains, all the cows were sick, so Adams' analogy is appropriate.
The fact that everyone just ASSUMED that the cows were OK, because they usually weren't sick in the past, instead of just going out and looking at the cows, really doesn't help to make this whole situation less worthy of ridicule. Everyone was saying things like you are saying to convince themselves "no, no, it's more complicated than that, see we have all this math, you don't understand" when the simple truth was that all the cows were sick and just looking at the cows instead of doing the fancy math would have made more sense.
Sometimes it is actually better to oversimplify things and just get the big picture and not get lost in all the details. This is Adams' underlying point, and I think it is valid in this case.
As I understand it[1], there were people that made their living by taking the BBB tranches based on sick cows and re-securitizing them (dividing them into yet more tranches, some of which would be upgraded again). The more you did this, the closer you could get to having all of the original subprime loans be AAA rated.
For an example, say you had 100 subprime loans. You divide them up into tranches, and the top 86 become AAA, and the other 14 become something else. Then you combine this left-over 14 with the left-overs from other funds, and end up with, say, 100 of these lower rated securities. You then divide this set of 100 up into yet more tranches, 86 of which become AAA and the remaining 14 become something lower. Combine this 14 with left-overs from yet other funds, repeat many times, and eventually you can get nearly everything to be highly rated.
There are claims that some groups of sick cows had dozens of levels of securities built up upon them, allowing ridiculously unsafe stuff to eventually become AAA.
15 comments
[ 3.1 ms ] story [ 38.0 ms ] thread"You'll get the last one standing!"
Close, but I'm going to disagree with you on one detail. The subprime debacle began in the spring of 2007, and really began to spiral out of control on July 26 with various calamities in the CDO markets. By August 6-10 (the week of quant turmoil) it was starting to have an effect on equity markets-- which were strong at the time, and didn't start taking a real beating until Jan. 2008.
The economy was strong in 2007, with 4.9% (annualized) growth in Q3. Subprime didn't blow up because the economy tanked, because the mess began during a great economy, and turned it into a bad one. Of course, the tanking of the economy has made the credit situation even worse and will lead to even more defaults.
The default rate goes up during a bad economy, and everyone knew that this risk was there, and that a cyclical recession would occur sooner or later. What caused the rate to be "unexpectedly" high was the prevalence of speculation. In a bad economy with falling house prices, a resident homeowner is still going to make his best efforts to avoid the embarrassment and hassle of foreclosure, even if he has negative equity. Negative equity is irrelevant if he's going to live there for 10 years. On the other hand, a greedy and unscrupulous speculator has no reason not to default.
Let's say a bank loans 300k on a house that was worth 200k but got inflated to 300k in a bubble fine somebody takes a 100k hit when it fails but they are only down 33%. However, if nobody is going to buy it for six months and by then it's only worth 150k and the spiral continues.
I also think the economy was fubar as early as 2005 with the fed using low interest rates to make everyone pretend that the economy was fine this was really helped by CDO's and people taking out risky mortgages. Then there was a little bump in interest rates which finally set this ball rolling but it was obviously going to happen at some point. Basically, when any little thing is going to upset the house of cards there is already a problem even if it looks ok.
PS: When the economy tanks it's a lot harder to prop up stupid companies so we are seeing a lot of things that where on their death beds for a while all fail but in the long term this is a good thing.
The actual problem with CDOs is that the cows were all more diseased than the historical track record suggested (because after they started putting together CDOs, a flawed incentive structure created a new motive to buy sicker cows) and that the probability of one cow dying wasn't independent of the next cow dying.
Like if you put them in a pen together ...
Explanation fail maybe, but joke succeed!
That's a AAA tranche right there ;-)
I think you are splitting hairs here. The point remains, all the cows were sick, so Adams' analogy is appropriate.
The fact that everyone just ASSUMED that the cows were OK, because they usually weren't sick in the past, instead of just going out and looking at the cows, really doesn't help to make this whole situation less worthy of ridicule. Everyone was saying things like you are saying to convince themselves "no, no, it's more complicated than that, see we have all this math, you don't understand" when the simple truth was that all the cows were sick and just looking at the cows instead of doing the fancy math would have made more sense.
Sometimes it is actually better to oversimplify things and just get the big picture and not get lost in all the details. This is Adams' underlying point, and I think it is valid in this case.
As I understand it[1], there were people that made their living by taking the BBB tranches based on sick cows and re-securitizing them (dividing them into yet more tranches, some of which would be upgraded again). The more you did this, the closer you could get to having all of the original subprime loans be AAA rated.
For an example, say you had 100 subprime loans. You divide them up into tranches, and the top 86 become AAA, and the other 14 become something else. Then you combine this left-over 14 with the left-overs from other funds, and end up with, say, 100 of these lower rated securities. You then divide this set of 100 up into yet more tranches, 86 of which become AAA and the remaining 14 become something lower. Combine this 14 with left-overs from yet other funds, repeat many times, and eventually you can get nearly everything to be highly rated.
There are claims that some groups of sick cows had dozens of levels of securities built up upon them, allowing ridiculously unsafe stuff to eventually become AAA.
[1] Described here: http://www.portfolio.com/news-markets/national-news/portfoli...