WEBVTT 00:00:00.000 --> 00:00:05.759 Welcome to the debate. Today, we are opening 00:00:05.759 --> 00:00:08.759 the hood on the financial engine that powered 00:00:08.759 --> 00:00:12.339 the global economy for a decade before spectacularly 00:00:12.339 --> 00:00:15.300 blowing it apart. We're talking about the collateralized 00:00:15.300 --> 00:00:20.359 debt obligation or, you know, the CDO. For most 00:00:20.359 --> 00:00:23.760 people, that acronym is a SCAR, a remnant of 00:00:23.760 --> 00:00:26.320 the 2008 financial crisis that just signifies 00:00:26.320 --> 00:00:29.320 greed and recklessness. But if we strip away 00:00:29.320 --> 00:00:31.539 the Hollywood narrative, the hindsight bias, 00:00:31.859 --> 00:00:34.619 we find something else. A piece of financial 00:00:34.619 --> 00:00:37.219 architecture designed to solve a really fundamental 00:00:37.219 --> 00:00:40.140 problem of capitalism. How do you move risk to 00:00:40.140 --> 00:00:42.159 the people best equipped to handle it? And I 00:00:42.159 --> 00:00:44.719 would argue that what we actually find is a mechanism 00:00:44.719 --> 00:00:47.659 designed to hide risk from the people least equipped 00:00:47.659 --> 00:00:51.000 to understand it. You call it architecture. I 00:00:51.000 --> 00:00:53.320 think the popular consensus and, frankly, the 00:00:53.320 --> 00:00:55.719 historical record suggest it was more like a 00:00:55.719 --> 00:00:58.340 magic trick, a way to take, you know, radioactive 00:00:58.340 --> 00:01:01.380 financial waste, wrap it in a complex legal structure 00:01:01.380 --> 00:01:04.180 and somehow sell it as gold. So the central question 00:01:04.180 --> 00:01:07.540 is whether the CDO was misused or whether an 00:01:07.540 --> 00:01:09.519 instrument that relies on that level of opacity 00:01:09.519 --> 00:01:12.980 should exist at all. Exactly. OK, so. I'll be 00:01:12.980 --> 00:01:16.180 representing the view that the CDO was a triumph 00:01:16.180 --> 00:01:19.140 of financial engineering, a tool for efficiency 00:01:19.140 --> 00:01:23.099 that was unfortunately compromised by bad inputs. 00:01:23.480 --> 00:01:25.939 And I'll take the position that the CDO is inherently 00:01:25.939 --> 00:01:29.260 flawed, a weapon of mass destruction, to borrow 00:01:29.260 --> 00:01:31.599 a phrase from Warren Buffett, that incentivized 00:01:31.599 --> 00:01:34.280 predatory behavior and made the entire system 00:01:34.280 --> 00:01:36.900 incredibly fragile. All right, let's get into 00:01:36.900 --> 00:01:39.799 it then. To understand the CDO, we have to go 00:01:39.799 --> 00:01:42.939 back. way before the housing bubble. We have 00:01:42.939 --> 00:01:46.540 to go back to the 1980s. The first private bank 00:01:46.540 --> 00:01:49.620 CDO wasn't about subprime mortgages at all. It 00:01:49.620 --> 00:01:52.920 was issued in 1987 by Drexel Burnham Lambert. 00:01:53.040 --> 00:01:56.540 And it was all about corporate debt. The problem 00:01:56.540 --> 00:01:59.040 it was trying to solve was liquidity. Banks had 00:01:59.040 --> 00:02:01.620 all these loans on their books, loans to companies, 00:02:01.900 --> 00:02:05.260 aircraft leases, emerging market bonds, and they 00:02:05.260 --> 00:02:07.870 were tying up their capital. Right. So they wanted 00:02:07.870 --> 00:02:10.889 to clear the decks. Exactly. They couldn't lend 00:02:10.889 --> 00:02:13.270 more money because their vaults were just full 00:02:13.270 --> 00:02:17.930 of these IOUs. The CDO allowed them to bundle 00:02:17.930 --> 00:02:21.909 these diverse loans into what's called a special 00:02:21.909 --> 00:02:24.069 purpose entity. Just think of it as a separate 00:02:24.069 --> 00:02:26.909 company created just for this pile of assets. 00:02:27.569 --> 00:02:30.590 This entity would then issue bonds to investors. 00:02:30.909 --> 00:02:34.830 And suddenly, the bank has fresh cash to lend 00:02:34.830 --> 00:02:37.759 to new businesses. and investors have a new way 00:02:37.759 --> 00:02:41.219 to earn interest. In its pure form, this is, 00:02:41.240 --> 00:02:43.340 you know, the lifeblood of a healthy economy. 00:02:43.400 --> 00:02:47.099 It keeps capital moving. Moving is a very generous 00:02:47.099 --> 00:02:49.919 term. I think laundering might be a bit more 00:02:49.919 --> 00:02:52.379 accurate. While I'll grant you that the early 00:02:52.379 --> 00:02:55.460 days of corporate debt CDOs were relatively stable, 00:02:55.759 --> 00:02:58.240 you're glossing over the core mechanism that 00:02:58.240 --> 00:03:00.900 made them so dangerous. It wasn't just about 00:03:00.900 --> 00:03:04.780 selling off loans. It was about tranching. This 00:03:04.780 --> 00:03:07.590 is the alchemy. You take a pile of risky loans, 00:03:07.770 --> 00:03:10.949 you cut them into slices, and then somehow you 00:03:10.949 --> 00:03:13.349 tell a pension fund that the top slice is as 00:03:13.349 --> 00:03:16.550 safe as a U .S. Treasury bond. That creates a 00:03:16.550 --> 00:03:18.729 false sense of security that just permeates the 00:03:18.729 --> 00:03:22.050 entire system. Well, it's not false security 00:03:22.050 --> 00:03:26.210 if the math holds up. And let's clarify tranching 00:03:26.210 --> 00:03:28.250 for everyone listening, because this is really 00:03:28.250 --> 00:03:30.789 the technical heart of the matter. Imagine a 00:03:30.789 --> 00:03:33.330 waterfall. The money coming in from all those 00:03:33.330 --> 00:03:35.949 loans, people paying their debts. That's the 00:03:35.949 --> 00:03:38.750 water at the top. It flows down into a series 00:03:38.750 --> 00:03:42.449 of cups. The first cup to get filled is the senior 00:03:42.449 --> 00:03:45.750 trench. This is the safest spot. If there is 00:03:45.750 --> 00:03:48.610 any water flowing at all, the senior investors 00:03:48.610 --> 00:03:51.610 get paid first. And they get that coveted AAA 00:03:51.610 --> 00:03:54.370 rating. Correct. Because they take the least 00:03:54.370 --> 00:03:57.050 risk, they get the lowest interest rate. Then 00:03:57.050 --> 00:03:59.509 the water spills over to the mezzanine cup, the 00:03:59.509 --> 00:04:02.620 middle layer. Higher risk, higher return. Finally, 00:04:02.740 --> 00:04:05.180 whatever is left over splashes into the equity 00:04:05.180 --> 00:04:07.919 cup at the very bottom. These investors, they 00:04:07.919 --> 00:04:09.939 take the first hit if loans start to default. 00:04:10.159 --> 00:04:12.840 If the water stops flowing, their cup stays empty. 00:04:13.039 --> 00:04:15.340 They take the losses so the senior investors 00:04:15.340 --> 00:04:17.920 don't have to. It's a brilliant way to segment 00:04:17.920 --> 00:04:20.500 risk. A conservative pension fund buys the senior 00:04:20.500 --> 00:04:22.860 cup. A high -rolling hedge fund buys the equity 00:04:22.860 --> 00:04:25.839 cup. Everyone gets what they want. That is the 00:04:25.839 --> 00:04:28.399 textbook definition, and it does sound elegant. 00:04:28.639 --> 00:04:31.000 But let's look at the reality of that mezzanine 00:04:31.000 --> 00:04:33.740 cup. This is where the whole scheme started to 00:04:33.740 --> 00:04:37.000 rot. By the mid -2000s, banks were churning out 00:04:37.000 --> 00:04:39.300 mortgage -backed securities. Everyone wanted 00:04:39.300 --> 00:04:43.480 the AAA slice. But, well, nobody wanted the mezzanine 00:04:43.480 --> 00:04:46.600 slice. the BB -rated bonds. They were too risky 00:04:46.600 --> 00:04:49.000 for the grandmothers and pension funds, but they 00:04:49.000 --> 00:04:51.500 didn't pay enough yield for the gamblers. So 00:04:51.500 --> 00:04:54.220 Wall Street had billions of dollars of this BB 00:04:54.220 --> 00:04:57.680 -rated dross just sitting in warehouses. You're 00:04:57.680 --> 00:05:00.500 describing a market inefficiency. I'm describing 00:05:00.500 --> 00:05:04.220 a dumping ground. As the journalist Gretchen 00:05:04.220 --> 00:05:07.920 Morgenson put it, the CDO became the bin for 00:05:07.920 --> 00:05:11.230 the refuse of the mortgage boom. Wall Street 00:05:11.230 --> 00:05:13.629 realized they could take a pile of this unwanted 00:05:13.629 --> 00:05:18.149 BBB -rated mortgage debt, throw it into a CDO, 00:05:18.209 --> 00:05:21.569 and run it through that waterfall all over again. 00:05:21.870 --> 00:05:25.350 And here is the magic trick. The rating agencies 00:05:25.350 --> 00:05:29.050 Moody's S &P, they would look at this pile of 00:05:29.050 --> 00:05:32.009 junk and stamp the senior tranche of the new 00:05:32.009 --> 00:05:37.310 CDO as AAA. They turned lead into gold. Hold 00:05:37.310 --> 00:05:40.430 on. You're making it sound like they just arbitrarily 00:05:40.430 --> 00:05:43.490 stamped it. There was sophisticated math behind 00:05:43.490 --> 00:05:47.250 that decision. Specifically, the Gaussian copula 00:05:47.250 --> 00:05:50.329 models introduced by David X. Lee. The logic 00:05:50.329 --> 00:05:53.930 was based on correlation. Ah, yes. The formula 00:05:53.930 --> 00:05:56.769 that killed Wall Street. It was a tool for pricing 00:05:56.769 --> 00:06:00.389 correlation. The idea is simple enough. If you 00:06:00.389 --> 00:06:02.730 have a mortgage in Florida, a mortgage in Ohio, 00:06:03.050 --> 00:06:05.680 and a mortgage in California, What are the odds 00:06:05.680 --> 00:06:09.759 they all default at the exact same time? Historically, 00:06:09.779 --> 00:06:12.480 real estate is local. If Florida's market crashes, 00:06:12.639 --> 00:06:15.600 Ohio might be totally fine. So the model said 00:06:15.600 --> 00:06:18.420 that because the assets were diversified geographically, 00:06:18.500 --> 00:06:21.579 the probability of a total collapse, one that 00:06:21.579 --> 00:06:23.639 would drain the waterfall dry enough to hurt 00:06:23.639 --> 00:06:27.019 the senior investors, was statistically infinitesimal. 00:06:27.100 --> 00:06:30.579 And that is precisely where the intellect completely 00:06:30.579 --> 00:06:34.009 detached from reality. The model assumed house 00:06:34.009 --> 00:06:36.629 prices were like coin flips, you know, independent 00:06:36.629 --> 00:06:39.930 events. But they weren't. They were all connected 00:06:39.930 --> 00:06:42.810 by historically low interest rates and predatory 00:06:42.810 --> 00:06:45.529 lending standards. When the music stopped, it 00:06:45.529 --> 00:06:48.750 stopped everywhere at once. The correlation wasn't 00:06:48.750 --> 00:06:51.769 low. It was basically one. But the crucial point 00:06:51.769 --> 00:06:54.230 here isn't just that the math was wrong. It's 00:06:54.230 --> 00:06:56.649 that the incentives were designed to ignore that 00:06:56.649 --> 00:06:58.790 the math was wrong. You're referring to the rating 00:06:58.790 --> 00:07:02.389 agencies. I am. I mean, these agencies were operating 00:07:02.389 --> 00:07:05.949 with 50 % profit margins. And who paid them? 00:07:06.149 --> 00:07:10.029 The banks issuing the CDOs. So if Moody's said, 00:07:10.209 --> 00:07:13.629 hey, this pile of BBB subprime loans looks really 00:07:13.629 --> 00:07:16.850 risky, we can't give it a AAA rating, the bank 00:07:16.850 --> 00:07:19.550 would just walk down the street to S &P. It was 00:07:19.550 --> 00:07:22.389 a complete race to the bottom. They were paid 00:07:22.389 --> 00:07:25.029 to look the other way. You can't call that architecture. 00:07:25.389 --> 00:07:28.920 That's systemic corruption. OK, I will concede 00:07:28.920 --> 00:07:31.939 that the rating agencies were overwhelmed and 00:07:31.939 --> 00:07:34.199 the conflict of interest is a valid critique. 00:07:34.519 --> 00:07:38.240 However, to say it was purely corruption, I think 00:07:38.240 --> 00:07:40.240 that misses the demand side of the equation. 00:07:40.699 --> 00:07:43.860 We were in the middle of a global savings glut. 00:07:44.100 --> 00:07:47.399 You had trillions of dollars from Asia, from 00:07:47.399 --> 00:07:50.180 oil producing nations, all looking for a home. 00:07:50.319 --> 00:07:53.120 They couldn't just buy U .S. treasuries because 00:07:53.120 --> 00:07:56.120 the yields were at historic lows, 1 percent or 00:07:56.120 --> 00:07:59.149 less. These investors were desperate for anything 00:07:59.149 --> 00:08:02.189 that looked safe but paid a decent return. So 00:08:02.189 --> 00:08:04.670 the solution was to manufacture fake safety? 00:08:04.930 --> 00:08:08.129 It was an attempt to engineer a product that 00:08:08.129 --> 00:08:11.930 met a global need. The CDO bridged the gap between 00:08:11.930 --> 00:08:15.550 global savings and American borrowers. It lowered 00:08:15.550 --> 00:08:18.610 the cost of borrowing for everyone. If you bought 00:08:18.610 --> 00:08:22.329 a house in 2004, you got a low rate partly because 00:08:22.329 --> 00:08:25.660 this machinery was working. It efficiently allocated 00:08:25.660 --> 00:08:28.120 capital from where it was abundant to where it 00:08:28.120 --> 00:08:31.180 was needed. But it allocated capital to people 00:08:31.180 --> 00:08:34.279 who couldn't pay it back. I mean, this is the 00:08:34.279 --> 00:08:38.139 most damning part of the CDO legacy, the originate 00:08:38.139 --> 00:08:41.159 -to -distribute model. In the old days, if a 00:08:41.159 --> 00:08:43.899 local bank lent you money, they kept that loan 00:08:43.899 --> 00:08:47.340 on their books. If you defaulted, they lost money. 00:08:47.480 --> 00:08:49.679 So they checked your income. They checked your 00:08:49.679 --> 00:08:52.799 credit. They cared. The originate -to -hold model. 00:08:53.080 --> 00:08:56.070 Sure. But with the CDO machine hungry for inputs, 00:08:56.330 --> 00:08:59.149 the local lenders suddenly didn't care. They 00:08:59.149 --> 00:09:01.529 originated the loan and sold it to Wall Street 00:09:01.529 --> 00:09:04.429 five minutes later. The risk wasn't theirs anymore. 00:09:04.649 --> 00:09:07.850 It belonged to some pension fund in Norway. This 00:09:07.850 --> 00:09:10.470 broke the fundamental feedback loop of lending. 00:09:10.669 --> 00:09:14.289 We saw the rise of ninja loans, no income, no 00:09:14.289 --> 00:09:17.669 job or assets. We saw that famous case documented 00:09:17.669 --> 00:09:20.870 by Michael Lewis, the strawberry picker in Bakersfield, 00:09:20.970 --> 00:09:23.659 California. I know the one you mean. Income of 00:09:23.659 --> 00:09:29.779 $14 ,000 lent $724 ,000. Exactly. A man who didn't 00:09:29.779 --> 00:09:33.039 speak English, earning poverty wages, given three 00:09:33.039 --> 00:09:36.399 quarters of a million dollars. Why? Not because 00:09:36.399 --> 00:09:38.759 the bank thought he was a good credit risk, but 00:09:38.759 --> 00:09:41.600 because the CDO machine needed that loan. It 00:09:41.600 --> 00:09:43.840 needed the raw material to package into bonds 00:09:43.840 --> 00:09:46.639 to sell to the global glut. The strawberry picker 00:09:46.639 --> 00:09:48.940 was just a cog in a machine that had totally 00:09:48.940 --> 00:09:52.960 lost its moral and financial compass. Okay, that 00:09:52.960 --> 00:09:55.340 is a striking example, and it was certainly a 00:09:55.340 --> 00:09:57.559 failure of underwriting, but I want to push back 00:09:57.559 --> 00:10:00.100 on the idea that this invalidates the CDO as 00:10:00.100 --> 00:10:03.240 a concept. You're conflating the fuel with the 00:10:03.240 --> 00:10:05.860 engine. If you build a Ferrari engine, a high 00:10:05.860 --> 00:10:07.899 -performance piece of machinery, and you pour 00:10:07.899 --> 00:10:10.620 sugar into the gas tank, the engine will explode. 00:10:10.840 --> 00:10:12.759 That doesn't mean the internal combustion engine 00:10:12.759 --> 00:10:14.820 is a flawed concept. It just means you used the 00:10:14.820 --> 00:10:17.480 wrong fuel. But the engine was designed to run 00:10:17.480 --> 00:10:21.340 on sugar. You can't separate them. The structure 00:10:21.340 --> 00:10:25.039 of the CDO, specifically the opacity, it encouraged 00:10:25.039 --> 00:10:29.360 the bad fuel. By 2006, investors weren't even 00:10:29.360 --> 00:10:31.919 looking at the loans. They couldn't. A single 00:10:31.919 --> 00:10:35.019 CDO might contain slices of a hundred other mortgage 00:10:35.019 --> 00:10:37.559 bonds, which in turn contain thousands of individual 00:10:37.559 --> 00:10:40.120 loans. It was impossible to do due diligence. 00:10:40.399 --> 00:10:43.740 They were buying the rating, not the asset. The 00:10:43.740 --> 00:10:46.720 complexity was a feature, not a bug. It allowed 00:10:46.720 --> 00:10:48.759 the banks to hide the fact that the fuel was 00:10:48.759 --> 00:10:52.250 toxic. Transparency was an issue, absolutely. 00:10:52.629 --> 00:10:55.629 But markets evolve. We saw the market try to 00:10:55.629 --> 00:10:57.710 become more efficient with the introduction of 00:10:57.710 --> 00:11:00.350 synthetic CDOs. Now, this is where I think the 00:11:00.350 --> 00:11:03.230 genius of the financial engineering is most misunderstood. 00:11:03.710 --> 00:11:07.330 People hear synthetic and they think fake. But 00:11:07.330 --> 00:11:10.309 in finance, synthetic just means simulated. I 00:11:10.309 --> 00:11:12.570 think fake was a lot closer to the mark, personally. 00:11:12.809 --> 00:11:16.590 Well, hear me out. A normal CDO takes months 00:11:16.590 --> 00:11:20.429 to build. You have to go out and buy actual bonds, 00:11:20.710 --> 00:11:24.769 warehouse them and structure the deal. A synthetic 00:11:24.769 --> 00:11:29.429 CDO uses credit default swaps or CDS. It doesn't 00:11:29.429 --> 00:11:33.029 buy the bond. It sells protection on the bond. 00:11:33.169 --> 00:11:37.009 It mimics the performance without all the friction 00:11:37.009 --> 00:11:39.889 of ownership. It allowed for price discovery 00:11:39.889 --> 00:11:43.110 to happen at the speed of light. It allowed investors 00:11:43.110 --> 00:11:46.350 to express a view on the market. without needing 00:11:46.350 --> 00:11:49.490 the physical assets to change hands. Expressive 00:11:49.490 --> 00:11:53.210 you? You mean gamble? This is where the CDO went 00:11:53.210 --> 00:11:56.129 from being a tool for funding mortgages to a 00:11:56.129 --> 00:12:00.169 massive, unregulated betting parlor. With a cash 00:12:00.169 --> 00:12:03.230 CDO, you are limited by reality. There are only 00:12:03.230 --> 00:12:06.350 so many subprime mortgages in existence. Once 00:12:06.350 --> 00:12:08.850 you package them all, you have to stop. But with 00:12:08.850 --> 00:12:12.259 a synthetic CDO? You can reference the same pool 00:12:12.259 --> 00:12:16.220 of bad loans over and over and over again. You 00:12:16.220 --> 00:12:18.940 can take positions on them, yes. That is how 00:12:18.940 --> 00:12:22.480 markets find equilibrium. No, that is how markets 00:12:22.480 --> 00:12:26.480 build a doomsday machine. Look, if I insure my 00:12:26.480 --> 00:12:29.820 house and it burns down, I get paid once. But 00:12:29.820 --> 00:12:32.139 in the synthetic market, all of my neighbors 00:12:32.139 --> 00:12:35.080 can buy insurance on my house burning down. Suddenly, 00:12:35.259 --> 00:12:37.639 you have a thousand people hoping my house catches 00:12:37.639 --> 00:12:41.700 fire. You had deals like the Abacus 2007 AC1 00:12:41.700 --> 00:12:44.879 transaction by Goldman Sachs or the Norma CDO. 00:12:45.200 --> 00:12:47.919 These weren't built to fund housing. They were 00:12:47.919 --> 00:12:50.460 built to fail. Now, that's a harsh characterization. 00:12:50.980 --> 00:12:53.419 Hedge funds like Magnetar, who were involved 00:12:53.419 --> 00:12:55.539 in these deals, they were shorting the market. 00:12:55.740 --> 00:12:59.039 Short selling is a valid, vital function. It 00:12:59.039 --> 00:13:02.000 exposes overvalued assets. They believe the housing 00:13:02.000 --> 00:13:04.500 market was a bubble and they use synthetic CDOs 00:13:04.500 --> 00:13:06.740 to bet against it. If the market hadn't been 00:13:06.740 --> 00:13:08.620 a bubble, they would have lost a lot of money. 00:13:08.740 --> 00:13:11.399 But they helped create the very things they were 00:13:11.399 --> 00:13:15.080 betting against. Magnetar helped select the worst, 00:13:15.100 --> 00:13:19.259 most toxic assets to put inside the CDO, specifically 00:13:19.259 --> 00:13:22.740 so they could then bet it would explode. It's 00:13:22.740 --> 00:13:24.799 like a mechanic cutting the brake lines on a 00:13:24.799 --> 00:13:27.440 bus and then buying life insurance on all the 00:13:27.440 --> 00:13:31.179 passengers. It's not price discovery. It's predatory. 00:13:31.700 --> 00:13:34.740 And the scale of it. Because these were synthetic, 00:13:34.940 --> 00:13:37.200 the notional value of the derivatives completely 00:13:37.200 --> 00:13:40.580 dwarfed the actual value of the real estate. 00:13:40.860 --> 00:13:43.980 That's why the crash was so devastating. It wasn't 00:13:43.980 --> 00:13:46.419 just that people lost their homes. It was that 00:13:46.419 --> 00:13:49.779 the bets on the homes were 50 times larger than 00:13:49.779 --> 00:13:53.240 the homes themselves. I see the emotional weight 00:13:53.240 --> 00:13:55.879 of that argument, but we have to look at the 00:13:55.879 --> 00:13:59.379 aftermath with a cool head. The crash was severe. 00:13:59.600 --> 00:14:03.159 We saw Bear Stearns collapse. We saw Lehman Brothers 00:14:03.159 --> 00:14:08.460 vanish. We saw $542 billion in write -downs. 00:14:08.460 --> 00:14:11.480 It was an incredibly painful correction. But 00:14:11.480 --> 00:14:14.539 here is the crucial data point that gets ignored. 00:14:14.899 --> 00:14:18.659 Did the CDO die? The market for mortgage CDOs 00:14:18.659 --> 00:14:21.340 certainly did, for good reason. But the structure 00:14:21.340 --> 00:14:25.580 survived. Look at CLOs, collateralized loan obligations. 00:14:26.000 --> 00:14:28.759 These are the exact same structural waterfall. 00:14:29.360 --> 00:14:31.379 but they're backed by corporate loans instead 00:14:31.379 --> 00:14:34.539 of mortgages. That market is thriving today. 00:14:34.740 --> 00:14:38.039 It performed remarkably well, even during the 00:14:38.039 --> 00:14:41.179 depths of the 2008 crisis, compared to the mortgage 00:14:41.179 --> 00:14:44.100 stuff. This proves my core thesis. The waterfall 00:14:44.100 --> 00:14:47.559 works. The trenching works. The special purpose 00:14:47.559 --> 00:14:51.659 entity works. The failure in 2008 was a specific 00:14:51.659 --> 00:14:55.480 failure of U .S. housing assets in a black swan 00:14:55.480 --> 00:14:58.210 event. where safe assets suddenly acted like 00:14:58.210 --> 00:15:01.629 volatile ones. Calling it a black swan is a total 00:15:01.629 --> 00:15:04.230 cop -out. A black swan is something unpredictable. 00:15:04.990 --> 00:15:07.950 This was entirely predictable if you just looked 00:15:07.950 --> 00:15:10.509 at the incentives. You mentioned efficiency earlier, 00:15:10.610 --> 00:15:12.730 but we have to talk about the cost of that efficiency. 00:15:13.149 --> 00:15:16.570 I mean the fees. In some of these deals, 40 -50 00:15:16.570 --> 00:15:19.929 % of the cash flow was stripped out to pay bankers 00:15:19.929 --> 00:15:22.350 and managers and lawyers before the investors 00:15:22.350 --> 00:15:25.620 ever saw a dime. Well, high fees reflect high 00:15:25.620 --> 00:15:28.639 demand for a product. High fees reflect a parasitic 00:15:28.639 --> 00:15:31.539 relationship. This wasn't about helping the economy. 00:15:31.720 --> 00:15:34.460 It was about fee extraction. And when you ask, 00:15:34.559 --> 00:15:38.240 did the CDO die? I say it just mutated. You mentioned 00:15:38.240 --> 00:15:41.700 CLOs, but we also saw the absurdity of CDOs squared. 00:15:42.019 --> 00:15:45.919 A CDO composed of tranches of other CDOs. Yes, 00:15:46.139 --> 00:15:49.710 that is complex. It's insanity. It's a derivative 00:15:49.710 --> 00:15:52.750 of a derivative. At that point, you are so far 00:15:52.750 --> 00:15:55.490 removed from the underlying asset, the actual 00:15:55.490 --> 00:15:58.590 business or homeowner, that risk assessment is 00:15:58.590 --> 00:16:00.990 literally impossible. It becomes a pure abstraction. 00:16:01.409 --> 00:16:04.649 You're trading math, not value. And when finance 00:16:04.649 --> 00:16:07.309 detaches from the real economy, it stops serving 00:16:07.309 --> 00:16:10.529 society and it starts consuming it. Let me summarize 00:16:10.529 --> 00:16:14.470 my position then. The CDO at its core is a mechanism 00:16:14.470 --> 00:16:17.799 for capital efficiency. It allows diverse assets 00:16:17.799 --> 00:16:20.639 to be pooled and risk to be distributed to those 00:16:20.639 --> 00:16:24.580 who want it. The failure of 2008 was a failure 00:16:24.580 --> 00:16:28.080 of inputs, specifically subprime mortgages, and 00:16:28.080 --> 00:16:30.220 a failure of oversight from the rating agencies. 00:16:30.500 --> 00:16:33.639 But the concept of securitization, it remains 00:16:33.639 --> 00:16:36.039 a vital part of the global financial plumbing. 00:16:36.320 --> 00:16:39.639 It's a brilliant design that just requires responsible 00:16:39.639 --> 00:16:43.570 operation. I'll summarize by saying that complexity 00:16:43.570 --> 00:16:47.669 is not a virtue, it is a risk factor. The CDO 00:16:47.669 --> 00:16:50.169 severed the tie between borrower and lender, 00:16:50.330 --> 00:16:53.909 it hid toxic risk under layers of math, and it 00:16:53.909 --> 00:16:56.830 turned the financial system into a casino where 00:16:56.830 --> 00:16:59.990 the house rigged the game. As Buffett said, they 00:16:59.990 --> 00:17:03.309 are financial weapons of mass destruction. Innovation 00:17:03.309 --> 00:17:06.910 is valuable, but not when it serves only to obfuscate 00:17:06.910 --> 00:17:10.960 reality. So a triumph of engineering versus a 00:17:10.960 --> 00:17:14.180 weapon of mass destruction. We may have to agree 00:17:14.180 --> 00:17:17.500 to disagree on that, but there's no denying its 00:17:17.500 --> 00:17:20.819 impact. On that, we can definitely agree. It 00:17:20.819 --> 00:17:23.819 changed the world. The markets have rebuilt and 00:17:23.819 --> 00:17:26.200 the structures remain. Perhaps we have learned 00:17:26.200 --> 00:17:28.579 to check the inputs more carefully this time. 00:17:28.799 --> 00:17:32.099 Thank you for listening to The Debate. We invite 00:17:32.099 --> 00:17:34.599 you to look at the architecture of risk in your 00:17:34.599 --> 00:17:37.819 own portfolio and decide, is it efficient? Or 00:17:37.819 --> 00:17:38.900 is it just obscure?