There Had to Be a Short


wall streetI found a great article on on Goldman Sachs and CDOs. The article also explains what CDOs are and what synthetic CDOs are. The author gives a fair and balanced view on what Goldman Sachs actually did, and whether it was illegal or not. I loved the article so much I asked for permission to reprint the article in full on Value Walk. I am happy that the CEO Robert Huebscher granted me permission.

Below is the article, if you are really interested in the story with Goldman Sachs I suggest you read it, I rarely ask permission to reprint a full article unless I think it is very interesting and important. Enjoy!

Somebody needs to brief Senators before they get on TV and ask irate questions which demonstrate they have no idea what they are talking about. Expressing shock that someone was short on the trade in question shows you don’t understand the trade. Let me see if I can offer some clarity.

Corsair Capital Adds 17.5% In 2021, Notes “Change In Leadership” In Markets

According to a recent interview, Corsair Capital's founder Jay Petschek did not plan to be a hedge fund manager. After holding various roles on Wall Street, Petschek decided to launch the fund in January 1991, when his family and friends were asking him to buy equities on their behalf. He realized the best structure for Read More

Normally, you think of a Collateralized Debt Obligation (CDO) as a pool of mortgages. This pool is broken into anywhere from 6 to 15 tranches. The highest-rated tranches get their money back first, and the rating agencies made them AAA. While the lowest level would be called the equity portion and be first in line to lose, in theory it paid a very high yield. It was usually not rated. But the level just above that is BBB (just barely investment-grade), and that was typically about 4% of the total deal, but paid a much higher yield than the “safe” AAA portion.

Now, here is where it gets interesting. Investment banks would take the BBB portions of these Residential Mortgage-Backed Securities, which were not as easy to sell, and combine them in a CDO, which the rating agencies then rated using models based on data provided by the investment banks themselves. Since this combining of BBB tranches supposedly created diversification that the rating firms’ models indicated would drastically limit delinquencies and defaults, the AAA tranche of the CDO was jacked up to 75% of the total capital structure, with 12% rated AA. Only 4% was typically considered BBB. So pools of mortgages that probably should have been rated below BBB were miraculously turned into a CDO with 87% of its capital structure rated AAA and AA and only 4% rated BBB, with a chunk as equity. (I wrote about this in January of 2007, based on material from Gary Shilling and others, plus my own research, although I think I wrote about it in an earlier letter as well.)

Who would buy this stuff? Mostly institutions that were reaching for yield in what was, in 2007, a very low-yield world. Yield hogs. And institutions that trusted the rating agencies.

But the CDO in the Goldman case was not this type of CDO. It was hard to find enough BBB pieces to put together a CDO of the type described above, and the demand was high. Remember, everyone knew that housing could only go up. So, what’s an investment bank to do? They create a synthetic CDO. Follow this closely. The various investment banks – it was way more than just Goldman; rumors are it was up to 16 of them – would construct an artificial CDO fund based on the performance of BBB tranches in other deals.

Let me see if I can simplify this. It is as if I had a very negative view about a particular industry for which there was no future or index or liquid security. We could go to an investment bank and ask them to create a “hypothetical” index that would mirror the performance of this industry. I would be willing to short that index. But unless the bank wanted to be long that index, they would have to find a buyer who would take the long position. Presumably the buyer would have a different view than me.

Now, by definition there has to be a short for the long, and vice versa. This is a synthetic index. It exists only as a spreadsheet and performs in conjunction with the components it’s modeled upon.

wall streetNumerous hedge funds did not think the rating agencies knew what they were talking about when it came to the mortgage ratings. They also believed we were in a housing bubble. So they went to a number of investment banks and asked them to construct synthetic (derivative) CDOs that they could short. And there were buyers on the other side who wanted the yield, who trusted the agencies, and who believed that housing could only go up.

As to the Goldman deal, the buyers had to know there was someone short on the other side. By definition there was a short. Besides, they had a guarantee from ACA on the AAA portion (which of course went bad, as I wrote about later that year) – there was a guaranteed AAA yield a few points higher than with normal AAA debt. What could be better? Except of course that it was too good to be true. Learn a lesson, gentle reader. Don’t reach for yield.

The hedge funds that shorted the synthetic CDOs took real risk. They had to pay the interest on the underlying tranches to the investors who were long. And if the housing market continued to rise, and the bubble did not burst, they could easily lose a lot, if not all, of their money. No one knows when a bubble will burst. The markets can be irrational longer than you can remain solvent.

Let’s be very clear. This was purely gambling. No money was invested in mortgages or any productive enterprise. This was one group betting against another, and a LOT of these deals were done all over New York and London.

The SEC alleges that there was material lack of disclosure. I must admit that I would want to know that the person who was taking the short position had a hand in the creation of the pool of BBB paper I was buying. And if Fabrice Tourre told someone that Paulson was $200 million long when they were actually net short, that could be problematic. Now, if he just said that Paulson bought the equity portion of the synthetic CDO (there has to be one), that will be a different matter.

The prosecutor for the SEC is by all accounts a very solid and serious person who would not move this case forward if he did not think they would win. This is not one the SEC will want to lose. On the other hand, I hope that Goldman takes this to the Second Circuit Court of Appeals (the final decision maker in a long and arduous process), as there are some very interesting aspects to this case that I would like to see resolved, as an individual in the industry. On someone else’s legal bill.

I wonder why Goldman’s witnesses seemed ill-prepared. Did their lawyers tell them to keep it simple and not get into a spirited defense? My instinct says that a lot more will come out about this case. If it was just this one deal, then Goldman should pay the fine and walk away. Done all the time. I suspect there is more here. Or maybe it was just that they didn’t want to explain why they were doing a synthetic CDO. We’ll see when someone writes the book.

Reprinted with permission from

Updated on

No posts to display