Book Review: The Big Short
A few weeks ago at a little league baseball game in Lincoln, Nebraska, I was in a light-hearted discussion with two brothers-in-law about what a CDO was, how a credit-default swap worked, what trade exactly Michael Burry made, and how it all factored in to the 2008 crash. We were only doing our best to state facts but we did it from memory as confidently as we thought we could pull off. We had all seen the movie, but not within the last few years. I spent much of the car ride home with Google and Claude, doing my best to convince myself I had, in substance, been correct. When I got home, I bought The Big Short on Audible — the 2025 edition, narrated by Michael Lewis himself — because that’s what you do when you’ve talked a big game you can’t quite back up.
A few weeks ago at a little league baseball game in Lincoln, Nebraska, I was in a light-hearted discussion with two brothers-in-law about what a CDO was, how a credit-default swap worked, what trade exactly Michael Burry made, and how it all factored in to the 2008 crash. We were doing our best to state facts, working entirely from memory but sounding far more confident than we’d earned.
The Big Short is a story about a major failure of the market to efficiently aggregate information through pricing, facilitated by multi-dimensional bad incentives. That failure also came about, partly, because of a regulatory regime which ostensibly expanded access to homeownership in America while actually, maybe accidentally maybe not, giving birth to residential mortgage backed securities vampires. And of course, The Big Short is a story about a group of misfits that saw what others couldn’t, or wouldn’t, see. Who doesn’t love an underdog story?
Not coincidentally, The Big Short isn’t really a standalone–the first book in the series was Michael Lewis’ first book, a genre-defining Wall Street expose published in 1989 called Liar’s Poker. In that first book, Michael Lewis tells the story of how the mortgage bond market was created at Salomon Brothers in the 80s, and how Wall Street firms learned to package debt into instruments opaque enough to hide huge fees. The Big Short is where those instruments finally detonate. It is not the just conclusion however. The traders that lost billions of dollars didn’t lose their own dollars, after all. Shareholders and taxpayers covered the buck, before, during, and after the 2008 crisis. The doubtful question I’m sure echoes in every reader’s mind is, “Did we learn?”
To make sure I did, I’m writing up this Jack-brain-native explanation of the 2008 crisis, based on extensive paraphrased notes I took while listening to The Big Short. It is meant to be accessible to the pre-reading Jack Pond, who had little concept of what happened in 2007-08 (despite having seen the movie) and without the vocabulary to really follow along. I’ll mostly follow the ordering and the beats of the book, which illustrates the story through very interesting characters.
The good old RMBS
Organizations (corporations, charities, governments, any kind of institution) can raise money for their projects in two primary ways: Selling equity and selling debt. Both have pros and cons. When a company sells a share of itself, and doesn’t end up making very much, they’re not in the hole with their investors. If they make a ton of money, however, they end up needing to fork over quite a bit to investors–corresponding to however much equity they sold. If they sell debt, then they are obligated to pay a fixed amount back to whoever owns the debt, whether they make a lot of money or not. Some organizations, like governments, can’t sell equity. They raise money only by selling debt. A package of debt is often called a bond, and is a very simple kind of contract: “If you (the investor) give us (the organization) give us money now, we will give it back to you plus a little bit in a fixed amount of time.” The little bit is the interest on the loan. Take a US bond as an example. You give the government some money (you buy the bond), agree on an interest rate, and in five or ten years they give it back and then some, and you have yourself a nice return. Corporations are one organization that can sell equity and debt. If they are a public company, their equity is traded as stocks on the stock market. They also may issue debt in the form of corporate bonds.
If you buy debt, you are effectively in the business of making loans, even though it might feel a bit indirect. You will want to buy up a bunch of debt when interest rates are high, when borrowers are willing to agree to give you back a bunch more than you lent them. Then, when interest rates are low, you can be smug that you are earning high interest on loans you already made (back when interest rates were high!) while others are lending money to borrowers who won’t return much more than they received.
Mortgage debt was packaged into bonds as well, by taking a big pile of mortgage loans (each of which might be worth $200k-$800k, for example) and selling investors a piece of it. When monthly mortgage payments came in for that big pile of mortgages, they were split up and divided among investors according to how much of the pile each had bought. There was a problem though, one that kept big investors from buying mortgage bonds. Good, low-risk bonds got pre-paid; mortgage owners would wait until interest rates drop, refinance, and pay the loan back early, giving credit owners a bunch of cash when interest rates are low, when they least wanted it. Since this delivers an unpredictable and bad deal to investors, it wasn’t a great vehicle for credit investment.
To solve this issue, the packagers created a clever innovation and instituted a predictable trade off. Think of a high rise building in a flood prone neighborhood. The renters on the ground floor pay cheaper rent because they flood first. The second floor renters pay slightly higher rent but have less risk of flooding, up to the highest floor, which pay the highest rent but have the lowest risk of floodwater. The floor you rent is a bet about the size of the liability of a flood. For mortgage bonds, these are called tranches. The first tranch gets higher interest rates but their debt obligations take prepayments first, exiting them from the investment first. Their cash flow is high, but could get cut off if even a small percentage of mortgage owners pre-pay. The highest tranches have the lowest risk of prepayments but get the lowest interest rates. Investors got to pick their level of predictability and mortgage bonds got funded.
This trade took off in the 80s, partly because the existing buyers of mortgage debt, the savings and loans industry, was distressed. The predictable, structured1 mortgage bond made it a palatable class of assets for Wall Street firms who eagerly got involved. There is a lot of money in mortgage bonds–hundreds of millions of Americans put a lot of their money toward a mortgage. Because of the relative complexity and opacity of mortgage bonds, canny Wall Street firms were able to buy the debt relatively cheaply from loan originators, the companies that actually extended the loan to homebuyers, package the debt into structured bonds, and then sell to institutional investors like hedge funds or insurance companies at relatively high prices. They pocketed the difference and made a ton of money. They were the middle-men between the first and last buyers of the mortgage debt.
Rotten Seeds: The 1997-1998 mini subprime mortgage scare
Corporate bonds, government bonds, and, initially, mortgage bonds were all generally considered low-risk assets. This depended, of course, on which corporation, which government, and which mortgage at issue. Mortgage bonds started by packaging the low-risk stuff. These were home loans made to borrowers who had good FICO scores, demonstrable ability to pay, a good history of paying back loans, etc. For a few reasons, in the 90s and early 2000s, the appetite for mortgage bonds continue to grow. For one, countries like China and oil exporters such as Saudi Arabia had a lot of cash that they didn’t want to sit around, so they parked trillions of dollars in US Bonds. The demand for US government bonds drove the price up and the potential yield down. Many investors (insurance companies, money markets) had institutional requirements to both invest very safely but also to make at least a modest yield. That was increasingly hard to do with US bonds… but mortgage bonds were great for it. There just weren’t quite enough low-risk or “prime” bonds. The banks, and loan originators, soon turned their gaze to higher-risk or “subprime” mortgages. These were loans to people with lower-than-average credit scores, or with lower incomes, or with shorter credit histories. These borrowers were often poor or immigrants, and didn’t have as much experience in the US finance scene. Because of their bad credit, they could easily get stuck in a vicious cycle of only being able to secure loans with high interest rates that were difficult to repay, hurting their credit further.
And so the same structuring was used when sub prime mortgage bonds entered the market, but with different risks. The rate of delinquency and default was higher for the loans in these pools, so the lowest tranch took the loss first (rather than the risk of prepayment). The second tranch wasn’t at risk of losing any of their investment until the first tranch’s investment was completely wiped out by defaults, and so on up the chain. That’s a fundamentally different kind of risk. However, it was judged that the highest tranches of these pools were also very unlikely to lose money, and were rated as very low risk
Steve Eisman was a bit of a maverick analyst at a financial firm called Oppenheimer & Co. in the 90s. He was a sell-side equity analyst, meaning that he produced analysis and research that other firms bought in order to know what investments to make. His specialty was subprime loan companies–he would root out the bad ones and identify the good ones so other investors could buy/short their stock.
A subprime loan company was in the business of lending money to subprime borrowers, generally meaning they were at higher risk of not paying back the loan. This risk was mitigated by a higher than normal interest rate. The high interest rates on subprime loans that do get paid, in aggregate, should edge out the amount that never gets paid, plus some for a return to the lender. The mortgage pool is nice because it is the aggregate. The interest needs to be high enough to match the risk though, or else money is lost on average and overall.
Steve Eisman understood the social value of subprime mortgage lenders. By connecting homebuyers with Wall Street investors, they made home loans more competitive and efficient, and truly lowered home prices for the everyday subprime buyer. However he was also skeptical of many specific companies, and sensed the vulnerability of home buyers to predatory practices. For instance, loan originators sold almost all the mortgage debt they originated firms, usually investment banks, to package the debt into bonds. Since they offloaded most of the risk, they were incentivized to extend loans that they knew wouldn’t be repaid, because they could sell the rotten loan to someone else who would lose the money, and hurt the borrower in the long run too.
Around 1997, Eisman smelled something fishy. The mortgage bonds had long been fairly opaque assets–little was published about the actual mortgages they were made of. However, when Moody’s, one of the ratings agencies, finally published a ratings report that gave access to the kinds of loans in a pool of loans, Eisman assigned a newly hired accountant to dig through it and report what he found. That accountant was named Vincent (Vinnie) Daniels. He found high rates of delinquency in the pool of loans the mortgage bond referenced, as well as lots of weird euphemistic accounting. For instance, “manufactured housing” instead of trailer homes (which depreciate like cars rather than rise in value like homes usually do) of “involuntary prepaid” instead of defaults. He diagnosed an odd disconnect, where interest rates were far from meeting the risk of the loans. The small fraction of loans that subprime loan companies kept on their books instead of selling to investors were counted as assets with the (very bad) assumption that they would be paid in full and not early. They could count this as a profit, inflating their earnings before anything was realized. This worked as a kind of Ponzi scheme, where companies inflated their earnings before any investments were actually realized, and they could raise more capital and make more subprime loans without anyone realizing they were bleeding badly.
Armed with this evidence, Steve wrote a scathing report and uncovered a lot of the poor practices of subprime loan companies, much to everybody’s ire. He was right. A year later Russia defaulted and a big hedge fund (LTCM) went under. In the ensuing panic, many firms and investors tightened their books and became much more conservative. Subprime loan companies lost a lot of capital flow and pretty much all went bankrupt. Most people blamed their bad accounting, which counted unrealized investments as profit. Vinnie saw a deeper problem of refusing to properly account for risk. He saw the whole snafu as a suspense of normal practices in order to cheaply (and lazily) solve a social problem: Getting cheap loans to low income individuals. He considered it a way to make lower-income individuals feel wealthy even while wages were stagnant.
Afterwards, Steve Eisman left to join a hedge fund called Chilton as a trader, but was soon demoted to analyst again. Subprime loan companies had effectively gone extinct… except for a consumer loan2 company called Household Finance. They had begun to grow a ton in the subprime mortgage space. Even after the dot com crash, when the economy was not in a position to be making these kinds of high-risk loans, their subprime sector ballooned. Steve discovered they were selling a fifteen year second mortgage3 as if it were a 30 year mortgage.
That tricky move went like this: Household finance would take the fifteen year payment plan, spread out the same dollar amount over 30 years, calculate the interest rate in this hypothetical scenario, and then tell the borrower that effective interest rate. Which was a baloney nonsense interest rate around 7%. In actuality, they were paying off the loan in 15 years, with something like a 12.5% interest rate.
Steve Eisman got his hands on a sales document and was amazed at the brazen deception–thought it was fraud, and found other people (many were borrowers from Household Finance) who wanted to sue or press charges against Household Finance, including the Washington state attorney (Household had made a lot of these loans in WA) who had been blocked by a judge. Household Finance ended up settling a massive multi-state legal action for $484 million in 2002. Attorneys general from all 50 states were involved! At the time it was the largest consumer protection lawsuit in history.
Steve Eisman mostly acted as a whistleblower in the case, but was disillusioned by the ultimate lack of consequences. Months later Household Finance was bought by the huge British firm HSBC for about $15.5 billion. The CEO personally made $100 million. Much of the value of Household Finance was their large portfolio of subprime mortgage debt. This incident filled in much of Steve Eisman’s cynical worldview, especially with consumer finance4, which he felt existed only to rip off poor people.
Cracks
Despite the mini-crisis, the appetite for subprime loans only grew. In the mid 90s, $30 billion in subprime loans was a good year. In 2000, there were $130 billion in subprime loans, with around $55 billion ending up in mortgage bonds. A year later there was around $600 billion in subprime loans with ~$500 billion in mortgage bonds. The market was growing incredibly quickly, even though interest rates were climbing. Part of this market was because of a federal administration that prioritized increased homeownership accessibility, and the market for subprime loans gave low-income families cheaper capital.
By this time, Steve Eisman had gotten sick of analysis and had started his own hedge fund, technically within Morgan Stanley, but completely autonomous, called Front Point. He figured something fishy was going on, and wondered who could be buying all these bonds.
At the same time, in the early 2000s, the originate-and-sell model caught on for subprime loan companies. They had learned the wrong lesson from the mid-90s failure mode, where subprime loan companies were sunk by the fraction of subprime loans they kept on their books. Starting in the early 2000s, they would originate subprime loans and get them all off their books. They would no longer count any profit from the interest of these subprime loans, they would only pocket the spread between origin price and sale price, which factored in the expected value of the fixed-income asset and supposedly the baked-in risk. The big banks such as Lehman Brothers, Morgan Stanley, and Goldman Sachs all got into this business through originate-and-sell shells which obscured who was running the business. It was nontrivial to link these companies to the entities that were actually running them.
As we talked about in the beginning, the fixed income world (bonds) was much much bigger than the equity world (stocks) for structural reasons. Everyone can issue debt, but only corporations issue equity. There were also some incentives to get into the fixed income world, though. Debt can be marked as much safer, which means it can be bought by large conservative institutional investors like insurance companies and money markets. Fixed income departments became the backbone of a lot of major banks. It had become the mainstay of the financial system in just two decades.
Let’s zoom out for a moment. The market for credit was, and is, ostensibly a marketplace that allows for people who have cash to lend to people who want cash in order to build value. Creditors can pool together to provide a lot of cash to people who are building a lot of value, like companies or houses. This is the kind of price-discovery and capital allocation that capitalist markets are meant to be excellent for. However, in an ecosystem of pro-housing-accessibility regulation and underdeveloped risk evaluation for some of these new securities, fixed-income became predatory. We already saw this with Household Finance, where the originators and middle-men could make a lot of money buy lending money to borrowers who were unable to pay back the loan, and then selling that debt (and risk) to someone else. All the while, cash is being allocated to places where value is not being created. This is the kind of market mismatch that only a few could see.
One of the few was named Michael Burry, who ran a $550 million hedge fund called Scion Capital out of San Jose. He ran it differently than most managers. He had a long lock up period after initial investment, which made it hard for investors to take money out of the fund. On the other hand, he didn’t take the traditional 2% carry fee.5 He wanted to align incentives and make the fund good for investing, rather than good for customer service. And good for investing it was. He had more flexibility than managers who had to maintain liquidity for investors who wanted out, and was free to make trades that looked bad up front (and which made investors nervous) but paid off big. He was an incredible stock picker, and did the reading. A famous and early on example is when he learned about a company which was going through a lawsuit, but which still had good underlying fundamentals. He learned that most employees were Chinese nationals and couldn’t leave the company easily. So he bought cheap, quickly became a majority shareholder, and advised the company on changes to make. After weathering the lawsuit, a steady revenue flow won out and the stock shot up, delivering a huge return to Scion Capital.
Michael Burry was a value investor, somewhat after the tradition of Warren Buffet. He looked for under-valued companies, and bought and stuck around. Likewise, he looked for over-valued companies to short. He originally trained to be a doctor, and maintained an obsessive value-investing blog throughout his residency, which gained him notoriety and eventually, when he opened his hedge fund, early investors.
He realized early–around 2002–that the subprime loan market was rotten, and said so in his letters to investors for years. In 2005, he crunched the numbers about teaser interest rates.
Teaser interest rates were short-term beginner interest rates meant to entice borrowers to take on larger loans than they could pay back. Usually for the first two years of the loan, borrowers would only pay interest on the loan, without paying down the principal. At the end of two year, the interest rate would suddenly rise as the borrower was expected to actually start paying down the principal. A standard playbook for borrowers who couldn’t afford these payments was to take a second mortgage against the increased equity of the home, assuming its value had appreciated, in order make the original payments. Perhaps the second mortgage had a teaser rate as well…
Michael Burry recognized this for the dumpster fire it was. He realized that housing prices didn’t need to go down to crash the market–they only had to not go up. Teaser rates and second mortgages both had the effect of lessening the borrower’s skin in the game, so they would lose less if they walked away. Burry also saw subprime mortgages take up more and more space in bonds without the changing the ratings of those bonds. He knew that current default rates were at 4%, and that it would only take a 7-8% default rate to completely wipe out BBB tranches of the structured bonds.
So he did what any confident investor who foresaw the crashing and burning of a massive market–he invented a way to bet against mortgage bonds. He met with several banks to create a standardized new derivative for mortgage bonds called a credit default swap. Credit default swaps are confusing because they aren’t really a swap.
The canonical swap is for fixed vs. floating interest rates. Say that Alice owns debt that pays a steady fixed interest rate, and that Bob owns debt that pays a floating interest rate, meaning the amount depends on federal interest rates or some other variable in flux. Alice could sell Bob a swap that entitles each to the other’s interest revenue, effectively trading incomes. A credit default swap is really closer to an insurance policy. Credit default swaps for corporate bonds had been around for a while, but were considerably more straightforward. The buyer of a credit default swap that referred to a corporate bond paid regular payments to the seller of the swap (like an insurance premium). If the issuer of the corporate bond (the corporation) declared bankruptcy or otherwise defaulted, then the value of the bond went to zero and the swap seller would pay the lost value of the bond to the swap buyer. The less likely the reference bond is to fail, per the risk model of the seller, the cheaper the regular payments.
However, mortgage bonds aren’t all or nothing the way corporate bonds are. They aren’t binary. A mortgage bond refers to a pool of mortgages, some of which might default, most of which probably (*massive air quotes*) won’t. The bond’s value is on a spectrum from 0% to 100%. The credit default swap Burry helped the banks design for him factored this in. They designed it to be “Pay As You Go”, meaning that the seller of the swap paid the buyer of the swap the writedowns on the bond. If on a Tuesday 5% of the mortgages in the bond fail, the swap seller pays that 5% of the bond’s value to the swap buyer. If on the next Thursday an additional 10% fail, the swap seller pays that 10% of the bond’s value to the buyer. And so on. The most the seller could pay out is 100% of the bond’s value; the least they could pay out is 0%. At either end it resembles a corporate bond referencing credit default swap, but in between it paid out however much the bond had fallen.
Burry very carefully picked specifically the very worst, absolute garbage BBB tranches of mortgage bonds, and bought credit default swaps on $1 billion worth of them. That means his small regular payment (2.5% annually, split over four quarters) insured $1 billion worth of bonds, the most that could pay out if all the bonds went to zero. He was ever nervous that the banks would wonder why on earth he wanted to insure so many bonds, and catch on to what he was up to. They didn’t. On the flip side, his investors were furious that he was leaking 3-4% of his holdings every year, or that he had strayed into fixed income. After years of prophesying doom for the mortgage bond market, he had put his money where his mouth was.
Burry’s trades helped invent a whole new dimension to the bond market. One story about these CDSs is that they were insurance policies. Burry paid a premium and got a big payout in the event of a “rare” occurence. The bank (or whoever sold the swap) received payments but also paid out the big payout. In Lewis’ words, this was more like “buying fire insurance on a dry slum where fire breaks out often,” a slum that Burry didn’t own! Another, perhaps more accurate, story is that a CDS is a pretty pure derivative with a short side and a long side. A third, more clever, story is that a CDS is a kind of synthetic mortgage bond. The seller receives a fixed income of payments, and loses the worth of defaulted mortgages, without ever having to originate new mortgage loans! This freed up big banks from the tedious work of actually originating mortgage loans and allowed them to sell ever bigger bets on the already existing mortgage loans as pure gambles.
This is worth restating, as it is the inflection point of absolute absurdity. Burry’s credit default swap created another security, another contract, that looked almost exactly like a mortgage bond. Burry, the buyer, owed regular payments to the seller, who only lost money if mortgages defaulted. The main difference was the order. For a mortgage bond, the money for all the underlying houses was offered up front, and if a house defaults, investors lose their up-front investment. With a credit default swap, no cash for the houses changes hands up-front–but if a borrower defaults, the seller pays the buyer the worth of the house. This symmetry, between mortgage bonds and credit default swaps that reference mortgage bonds, was a huge deal. Up to this point, if banks wanted to participate in lucrative trades including mortgage bonds, they (or the loan originator) would have to wade through the tedium of making new loans to new home-buyers, and then package them up into structured securities. The new CDS allowed them to essentially buy and sell new bonds on already existing mortgages. It multiplied the amount of money riding on the mortgages referenced by the bond that the credit default swap referenced. You can see why traders’ heads were spinning.
As an unfortunate side note, Burry was so confident about this position that he tried to raise money to start a hedge fund to exclusively buy up CDSs. He was never a very personable salesman, and quickly failed. His pitch of a market cataclysm was just too unbelievable.
It’s *mumble mumble* all the way down
Around this point, a couple guys at a quant desk at Goldman Sachs had the incredibly (I mean this genuinely, though ruefully) bright idea to take garbage BBB mortgage tranches and repackage them into other CDOs.
A CDO is a collateralized debt obligation, and is the same idea as a mortgage bond, but one level higher. A mortgage pool is a pile of perhaps thousands of individual home mortgages, and a chunk of that is a mortgage bond. A CDO is a pile of perhaps a hundred mortgage bonds, diversified across several mortgage pools, and a chunk of that is called a note. Each aggregation is meant to have a diversifying effect. Buying a bond with slices of a thousand mortgages is more likely to hold its value than any one mortgage; and buying a note with slices of a hundred mortgage bonds is likely to hold its value more than any one mortgage bond.
Recall that a BBB tranch of a mortgage bond is among the very first to go–it is effectively the first or second floor of the skyscraper in the middle of a flood plain. Goldman could take a big pile of BBB mortgage bonds, sometimes classified as mezzanine mortgage bonds, and package them into yet another structured security, and carve it into tranches again. They could pretend they had diversified and get the most senior notes rated AAA. They hadn’t really diversified, of course. They had taken the lowest floor from buildings all on the same flood plain, to use Lewis’ metaphor, and bundled them together, then pretended some were higher than others.
It worked. Rating agencies rated 80% of these as AAA securities. Incredibly, Goldman Sachs could take the remaining 20%, bundle them together with some other remaining 20%, and get 80% of those rated as AAA securities! Rating agencies didn’t know what they were rating and sometimes used Goldman Sach’s own models to evaluate risk. Goldman didn’t even have to originate all these mortgage loans to get billions of dollars worth of BBB mortgage bonds because they could use Burry’s credit default swaps instead. CDOs consisting of credit default swaps that referenced mortgage bonds rather than actual mortgage bonds were called synthetic CDOs, but functioned much the same.
Buyers of the senior tranches of these CDOs thought they were getting smart, safe yields. Nobody understood that even though their investments were rated AAA, they would get wiped out at the same time as the chumps who owned the same BBB mortgage bonds that the AAA CDO notes referenced. There was enormous demand for this kind of “safe” yield, and Wall Street firms effectively colluded with credit rating agencies to provide a counterfeit, fueled by investors who bought credit default swaps like Michael Burry.
AIG FP was a British firm that was a big seller of corporate CDSs, and had made a large profit essentially insuring against lots and lots of corporate risk. In the early 2000s, they got fooled into thinking that the mortgage-bond referencing CDSs were more of the same. Goldman Sachs essentially had unlimited supply of CDSs because AIG FP would always sell them more. They just needed to find demand… which is why they became eager to sell to Burry and likeminded investors. They could sell to Burry at 250 basis points (2.5% of the insured amount annually), and buy from AIG FP at 12 basis points, pocketing 2% of these huge trades and transferring all risk to AIG FP. They would also roll together AAA mortgage bonds with CDSs from AIG FP and sell them as though they were riskless,6 though in reality buyers risked AIG’s insolvency.
Deutsche Bank got involved in the market soon after Goldman, and essentially forced their bond director Greg Lippmann to be $1 billion short subprime mortgages in order to create the sell-side securities, to sell. At first Lippmann was not happy, but he asked Eugene Shu to crunch the numbers on how different home appreciation regimes would affect the subprime bonds’ values… and was happy to be short after that. He became an opinionated market maker, and created a presentation to urge investors to buy the CDSs that he owned. He would collect fat fees and possibly be the only buyer if these investors wanted to sell their positions, and most got spooked. He was losing lots of money and his position became worth even less as prices for subprime CDOs boomed, because demand for subprime CDOs was bolstered so much by AIG.
Within AIG, a bond trader named Gene Park disagreed with AIG’s policy of always saying yes to selling CDSs. When offered the CDS director role, he refused. Just about everyone at AIG assumed that selling insurance for AAA tranches of CDOs was free money, but he strongly disagreed, and told his boss Cassano what he thought. Initially furious, Cassano agreed to a tour around Wall Street to ask some tough questions, and he came around to the same conclusion–that the whole market was so much smoke. AIG stopped selling, but actually held the positions they had already sold. Lippmann thought that would bring a market to a halt, and the value of subprime CDOs and the underlying bonds would drop. He was wrong. Wall Street continued to find new sellers, new nooks and crannies to stuff the risk from mortgage loans and associated side bets.
Lippmann was the one who got Front Point involved, headed by Steve Eisman. Vinnie was working with him again, along with Front Point’s head trader Danny Moses. Initially they had very little trust for Lippmann, and that never really changed. They were well primed to understand and accept his position, since they had been cynical of the subprime mortgage market for years. But they couldn’t figure out his angle. They decided to buy when rating agencies announced they would be changing the risk assessment model they used to evaluate subprime mortgage referencing securities, and firms started to crank out all the bonds they could while they could still get them rated highly with the old bad models.
After buying the short position, Vinnie and Danny headed to a conference for the subprime mortgage loan market in Orlando, just to sniff around. They were blown away. They managed to organize a meeting with a few employees from rating agencies, and a woman from Moody’s explained some of the rating process to them. She told them that identically terrible bonds got rated differently, and that her boss wouldn’t allow her to downgrade 75% of the bonds she wanted to. They couldn’t believe it.
Lippmann really was patient zero for the shorting virus. Everybody heard about it from him or from somebody that he told about it. Perhaps most notable was an investor named John Paulson. He heard about the short position from someone who heard about it from Lippmann, and was able to do what Burry couldn’t pull off. He started a hedge fund specifically for buying CDSs and sold it as a way to hedge investors’ exposure to subprime loans. He was short billions of dollars worth of subprime loans, and later said he had never seen a market where he could transact billions of dollars and not move the price of the market. He started a little cautiously but saw there was no point, and was incredulous that it was so much easier to buy CDSs than to short bonds.
Don’t assume malice if you can assume incompetency
The engine of the market was the group of rating agencies. Their general incompetency allowed subprime mortgage securities to be marked safe, and to trade at the prices they did. Wall Street traders came from the best schools and were wily and clever. Rating agents… were mostly people who failed to become Wall Street traders. One Morgan Stanley trader said that the asset-backed people were “brain-dead.” There were too many loopholes for subprime loans and traders found and exploited them all.
Take FICO scores as an example. For a pool of loans, the rating agency only looked at the average, which needed to be about 625 to pass muster. Already FICO doesn’t look at income, which is a shortcoming on its own. The fact that you could have half 550 and half 675 and still get rated highly meant traders could buy trashier, cheaper loan pools and sell it for just as much. A person with a credit score of 550 is almost sure to default on a mortgage loan, but traders were asking originators for these loans. It’s pretty easy to find a person who can default on a mortgage. But how to get all these loans to borrowers with 675? One strategy was thin-file credit scores. People with no history of credit could have pretty high credit scores, and the rating agencies never checked.
One loan Lewis highlights was a mortgage loan made to an immigrant worker strawberry picker who bought a house worth $724,000. The migrant worker is not the problem here, the problem was a system of incentives that made banks want to give loans to low-income individuals who couldn’t pay.
Rating agencies preferred low teaser rate ARMs7 to fixed rate loans, because they assumed that each loan would be paid in full, and low teaser rate ARMs paid more interest. They didn’t mind no-doc loans (no income verification), or liar loans (also no stated income loans). Every time a trader found a loophole like this, they would covertly let the originators know what they were looking for and get as much of it as they could, before the rest of Wall Street figured it out. These loopholes allowed them to stuff ever worse loans into highly rated tranches and sell to big investors.
What happens in Vegas…
Charlie Ledley and Jamie Mai are two more important names on our eccentric cast. They were drop-out type investors, young, post gap years, who started a money management firm called Cornwall Capital with $110,000 in a Schwab account. Their investing insight was two-fold: 1) Most institutional arbitrageurs limited their sights to just a few markets, and 2) option contracts were sometimes very poorly mispriced. They made tons of money on event-driven asymmetric bets, mostly with long-term call option contracts that were cheap but allowed them to buy stock for very cheap if the stock shot up.
Charlie and Jamie noticed that call-options were priced assuming that price changes followed a normal distribution whose spread depended on past volatility. One of their first big bets was on Capital 1 following a regulation scare. They figured that the stock would either stay low or get lower if it turned out Capital 1 had been dishonest, or it would shoot back up to pre-scare prices if it turned out they had been honest. Since call options price models considered any large change unlikely, Charlie and Jamie could buy call options for very cheap… even though there was some event (an announced resolution to the regulation question) that would either drive the price up a ton, or not at all. This is the event-based part in event-driven asymmetric bets. They spent a little on call options, which capped their downside on these bets at a small loss, and they could make five, ten, twenty times the cost of call options if the stock shot up. They trawled worldwide markets for this kind of deal, even correctly predicting a coup d’etat in Thailand and betting on it8.
Cornwall Capital was on the lookout for long-shot bets, events that were probably 10-to-1 but had associated derivatives that were priced as if the event was 100-to-1. That attitude set them up to stumble across the CDS for subprime loans trade with very open minds.
Michael Burry had pushed for the standardization of a contract for CDS that referenced specific mortgage bonds. Those had become the standard way to short the housing market, and almost all shorts bought CDSs on BBB or BBB- bonds. Cornwall Capital noticed, however, that while those CDS were indeed crazy mispriced, they weren’t the most mispriced. These same BBB and BBB- mortgage bonds had been laundered into AAA tranches of CDOs. The market lied and cheated to get these subprime-loan-referencing securities to get this rating, but then they believed the rating when they priced CDSs that referenced them. Cornwall realized that the real question when shorting a CDO was “Why should I pay 2% for insurance on this BBB tranch of a CDO when I can pay 0.5% for insurance on the AA tranche, which is a pool of the same kinds of bonds?” By buying 4x cheaper insurance, they could make 4x the profits. The market for CDSs for AA tranches of CDOs referencing RMBSs didn’t really exist (nobody had bought them yet), and Rich Rizzo, a bond trader at Deutsche Bank who worked for Lippmann, didn’t think they would be able to sell when they wanted to.
Cornwall wanted to do due diligence on this trade, and check what was even in the CDOs–what underlying mortgages the CDO notes referenced. As far as they could figure out, they were among the first to do so. Each CDO had a pool of 100 bonds, each of which referenced thousands of mortgages. They had to peel back the layers. They asked the rating agencies for more information, but they didn’t know either. They hadn’t gotten that far yet! $400 billion worth of CDOs had been created in just a few years, and they had never been properly vetted.
Charlie and Jamie eventually found a website that listed some of the info they wanted, at least aggregate stats, and they started trying to buy credit default swaps for CDOs with one of two characteristics: 1) CDOs that contained the highest percentage of bonds backed entirely by recent subprime mortgage loans (which, due to the practices of the day, were more likely to be terrible), and 2) CDOs that contained the highest percentage of other CDOs. After all, CDOs tended to be incestuous! When traders had a hard time selling some particularly nasty tranches of one CDO, they would package it up into another CDO to get better ratings on the same underlying subprime mortgages. Most traders laughed at them when they asked to buy these swaps, but eventually sold to them.
Hilariously, Charlie and Jamie didn’t even realize that they had bought CDSs for tranches of synthetic CDOs, which held the opposite side of CDS contracts for the bonds that Burry and Eisman were shorting! Although they felt like they were flying blind, they ended up making a short side bet on several of the same bonds handpicked by Burry and Eisman to fail first. However since they were shorting AAA tranches, Jamie and Charlie were paying a discounted rate for their short position. However, they couldn’t shake the feeling that they were maybe morons, like everybody told them they were. It was only weeks to go before the crisis, and they headed to a big asset-backed securities conference in Las Vegas, where Eisman and the other Front Point guys were headed as well.
Investors that had bought the short side of credit default swaps from Lippmann were getting nervous, and he took a golden opportunity in Vegas to reassure them. At a paid dinner for investors, Lippmann sat Eisman next to Wing Chau at a Teppanyaki restaurant to introduce him to a guy on the other side of the bet–and hopefully make him feel better about his short position. It worked. Wing Chau was a CDO manager. He selected Wall Street firms to supply him with subprime bonds that were packaged into CDOs so he could vet them, and was also supposed to actively replace bonds that were about to go bad with better ones. He did no such thing. The selection process of the firms was meant to be competitive, but more often the firms would find a good front man to just buy all their rotten bonds on behalf of investors, who often could only invest in AAA securities, such as pension funds or insurance companies. Wing Chau would just rubber stamp everything and pass on all the risk from the firms to his hapless investors, who thought their money had been parked in safe, reasonable yield investments.
Eisman realized over the course of dinner that the lowest BBB bonds were getting repackaged over and over into AAA securities (shares of CDOs), and when he bought CDSs to short those loans, those were also packaged into synthetic CDOs! Wing Chau told Eisman that he appreciated the volume and hoped that the market would continue to get worse, so more people would keep shorting the market and provide synthetic bond (CDS sell side) volume to keep his CDO management growing, because he personally had almost no exposure to bonds. He just got a percentage of the pot when money went in and when it went out. He was making tens of millions of dollars a year managing a $15 billion CDO fund, selling all the CDSs that AIG FP stopped picking up. He facilitated the transfer of housing market risk to the final buyer, the large institutional investors. And all along he was hoping for the same thing Eisman was–for the market to keep getting a little worse. He was the kind of financier keeping the price of bonds stable despite the market failing, since his appetite for bonds never slowed down. He provided the demand for loan originators to continue making terrible loans to borrowers who never should have gotten a loan and never should have bought that house. His incentives were beyond misaligned–he was actively hoping the market he invested in failed.
As a side note and callout to the movie, one of Danny’s friends really did meet a stripper with five home equity loans.
At the same conference, Cornwall searched to find people who could make an argument that they had bet wrong, but they couldn’t find anyone to do so. They arrived in Vegas thinking they were buying a cheap long-shot; Maybe it was actually 10-to-1 but it was priced as 200-to-1, and they liked that. The more they wandered around the conference and chatted with other attendees, the more they came to think of the crash as likely, better than even. They went to a panel discussion where John Devaney, a big time investor in subprime backed securities, went on a drunken rant about how worthless all the securities were, how rotten everything was. Everybody ignored it.
Front Point attended a talk by the CEO of Option 1, a loan originator. Option 1 originated loans and immediately sold to Wall Street firms who packaged mortgages into bonds. They weren’t meant to really hold any risk, but Option 1 had lost a chunk of money seven months previously. Turns out, there was a provision in the contract that allowed the banks to put a loan back to Option 1 if the borrower doesn’t make the first payment.
“Who takes out a mortgage and doesn’t make the first payment?” Danny Moses asked. “Who the f*** lends money to someone who doesn’t make the first payment?” Steve Eisman asked.
Thin ice
As 2007 continued, Front Point and Cornwall continued to rush to increase their short positions, but sellers dried up. The disconnect widened. Even as firms stopped selling to shorts, they continued creating and selling CDOs. Bond prices net fell, but analysts continued to preach the strength of the market, claiming risk was low even though they were unwilling to sell insurance.
Front Point shorted Moody’s and talked to the CEO, who said he was confident the ratings were correct—even though their surveillance team (which monitors loan performance) had no access to loan level data, because issuers wouldn’t give it to them. As part of their investigation into the ratings process and how firms gamed ratings, Front Point found more and more dumpster fires. For instance, ARM loans were rated higher than fixed rate loans because everyone assumed the higher payments would be made and that would increase the value of the bond for the bond investor. HSBC dumped their loans. Bear Stearns’ CDO fund dumped their bond holdings. Front Point and Cornwall kept wondering if the firms or the agencies knew something they didn’t. Despite asking around in wider and wider circles, nobody knew what they were talking about, or didn’t want to know.
Then Merrill Lynch announced losses from subprime mortgage backed securities, which was a big revelation to Front Point. The kind assumption was that these firms were dumping these trashy assets as quickly as they could, but apparently they were holding them on their balance sheet. Obliviously! And Eisman knew they were holding assets with more and more borrowed money. He realized subprime losses could mean more than just lost revenue, it could sink these ships. Front Point increased their short exposure by shorting stock of Bank of America, UBS, Citi Group, and Merrill Lynch.
Eisman even set up a conference call to broadcast his position and thesis, and 1000 people tuned in, but it was hard to tell if he was taken seriously by his listeners. That same day, investors in collapsed Bear Stearns hedge funds found out their 1.6 billion in AAA rated sub prime backed CDOs hadn’t just cratered, it was completely gone. At this point, there was still nobody who really knew what was in a CDO!
An illiquid market
Michael Burry had his own bizarre issues. He had rocky relationships with his investors, who were upset that he had bought a $1.9 billion short position in CDSs against a portfolio of $550 million. He knew he just had to wait, but he was never the best at customer relations. Additionally, he quickly found that even as the mortgages underlying the bonds he had insured defaulted at higher and higher rates, the banks–the supposed middlemen–were not changing the prices of the CDSs or the bonds. He asked them to find potential buyers, to get bids and do some price discovery. The banks responded that the swaps hadn’t appreciated at all, and that the price for CDSs was actually declining. He asked to buy at that price. The banks, of course, refused.
An important aspect of Burry’s CDS deals with these banks was that they had a mark-to-market collateral provisions, which are typical of derivative contracts. These provisions meant that when the value of the swap increased (if the bonds looked shaky), Burry’s position would be “in the money,” and the banks would post collateral–in bonds or cash–to Burry. On the flipside, if the swap position lost value (if the bonds looked steady), Burry would need to post collateral to the banks. This collateral is like a security deposit, to assure the counterparty that the money will be there. And if Burry’s portfolio dropped below a certain value, the banks could take that as a sign he may not be able to pay, and cancel the contract.
Burry realized the banks must have had some long exposure to these same bonds on their books, and that their balance sheet was talking, not potential buyers. They called the shots on what mattered for the worth of these swaps, not any kind of free market. His investors were furious with him and planned to withdraw anything they could at the end of a lockout period. Part of Burry’s deal with the banks was that they could cancel the contract if Burry’s portfolio dropped below a certain threshold, hurting his ability to pay.
To avoid this scenario, Burry declared that his CDS bets were in an illiquid, non-public market and exercised his right as manager to “side-pocket” the money necessary for that bet, and lock it down. Investors accused him of all kinds of crimes, but Burry held to his conviction against all sorts of pressure. He increasingly suspected intentional fraud in the banks. He suspected they were negligent in their duties as middlemen, because they were the other side of his deal, and refused to engage in a free market by artificially inflating the value of bonds and deflating the value of swaps. Gotham, a hedge fund that was an original partner in his firm, wanted their $100 million investment, and began litigation. Rumors began to swirl about Burry–that he had stolen the money or that he was in hiding.
In the first half of 2007, the value of the mortgages underlying the bonds diverged further and further from the prices of the bonds, and the prices of insurance on those bonds. Banks just chose to ignore this. In March 2007, Burry’s salesman at Citi Group sent, for the first time, serious analysis on a pool of Alt-A mortgages.9 It had only taken years.
The market started to turn around June 2007, when ARM loans made in 2005 started to hit their rate reset, exactly when Burry said it would happen. Banks finally made moves to clear risk from their books, and began to mark Burry’s bets correctly (finally!)–after Burry suspected they had gotten in on the trade, possibly via early insider information from relations with issuers. In the latter half of 2007, Burry’s position began to pay off in a big way. The market was finally waking up and pricing things (both bonds and swaps) accordingly. Big banks began trying their best to disgorge themselves of bad mortgage backed securities.
The Rise and Fall of Howie Hubler
This section of the book was one of my favorites–the poetry and ridiculousness of it is staggering.
In 2003, Morgan Stanley was a pioneer in adapting financial technology used for corporate debt to package and structure home equity debt. This looked like innovations with bonds and CDOs, as we have seen. Howie Hubler was the bonds guy at Morgan Stanley. It was natural to also dream up a credit default swap trade for mortgage backed securities, because they existed for corporate debt backed securities. The mortgage desk at Morgan Stanley had to warehouse these loans, sometimes for months, in order to package them. They created CDSs as a way to offload the risk associated with these loans, and kept them arcane, opaque, and illiquid, so that nobody would be able to price them effectively.
By 2004, Hubler had grown skeptical of the market for subprime loans and was looking for a way to bet against subprime mortgage bonds. He teamed up with some other traders and quants and cooked up a CDS contract that would pay off in full if the least-likely to be repaid loans defaulted. This was a far cry from the more reasonable pay-as-you-go contract Burry would standardize about a year later. In Lewis’ words, it was “like buying flood insurance that paid out in full if a single drop of water so much as grazed the house.” This was to fix the issue of borrowers continuing to borrow to pay off initial mortgages. It would only take a 4% default rate for the bet to pay off, which could happen in good times! They found one client stupid enough to be willing to take the long side. By 2005, he had found enough clients to buy insurance on $2 billion worth of these CDSs, paying 2.5% a year. Apparently it especially appealed to German institutional investors. Around this time, in 2005, Burry actually got in the way by requesting a standardized, more liquid version, which made it more difficult for Hubler’s group to peddle their murkier version.
In 2005, Hubler’s desk at Morgan Stanley was generating ~20% of Morgan Stanley’s profits and was only growing. As we have noted before, the fixed-income sector was taking the finance world by storm. Hubler wanted to leave to spin off his own fund and keep more of his increases, but Morgan Stanley cut him a deal. He could start a proprietary fund inside of Morgan Stanley and spin it off later. One of Hubler’s conditions was that he kept his CDS position when he moved to the new fund.
Now observe. Howie Hubler and his sketchy group of mortgage bond goblins were among the first to create a CDS position to short the housing market. However, that position required him to pay the 2.5% payment annually, which made it hard for his team to hit their $2 billion a year goal… so he sold insurance on $16 billion of AAA rated mortgage backed CDOs to fund this $2 billion position on BBB rated RMBS backed bonds. Howie Hubler, the tricky trader who deceitfully bought some of the first credit default swaps on mortgage bonds sold credit default swaps on $16 billion of AAA rated mortgage bond backed CDOs to make payments for his $2 billion credit default swaps. If he saw further than others, it’s because he was on his tiptoes and only saw 3 inches further than others. Holding these positions at the same time amounted to a bet that the default rate nationwide would hit 4% but stay below 8%.
Incredibly, as the market started to fail, he managed to sell some of his long position off to UBS before the bets really cratered, but his group still ended up eating the largest proprietary loss in Wall Street history–$9 billion.
The crash
The end of the saga is a bit anticlimactic, if only because the reader saw it all coming from so far away. The tragedy isn’t any less because it’s expected, though.
By the middle of 2008, Wall Street banks were outrageously over-leveraged, from Bear Stearns at 40-to-1 all the way to Goldman Sachs at 25-to-1. When their AAA so-so-safe holdings all failed at the same time, they were on the hook for much more than they could pay. Steve Eisman was asked to give a “Bear’s report” talk at Bear Stearns, about why nobody should own its stock. During that talk, the Bear Stearns CEO released a statement about liquidity (claiming to have plenty) that ended up crashing the stock 45% that day. Hedge funds pulled funds, other banks refused to engage, overnight loans disappeared. It was a run.
When AIG FP stopped selling credit default swaps, many wondered if that stoppage of demand would slow down the market. It didn’t. A handful of firms did pick up where AIG FP left off… but a lot of the risk was just tucked away in the big Wall Street firms. Since it was rated AAA, practically riskless, they didn’t have to disclose their positions. They were far, far more over-leveraged than anybody realized. The opaque contracts, fee structures, and inflated ratings had allowed an appetite for risky bonds to grow until the entire market was gobbling up riskier and riskier garbage mortgage backed securities with no warning light and no way to brake.
JP Morgan ended up buying Bear Stearns for $10 a share. Lehman Brothers collapsed entirely. Other big banks declared that they were also going under, but the US government ended up guaranteeing assets, and giving loans and gifts to prop up big firms. This is the kind of policy called quantitative easing, pumping cash into a struggling economy to promote growth. For instance, the US gave a $180 billion loan to AIG FP.
This may have been the right policy. The US market certainly bounced back from the 2008 crash faster than other countries that didn’t ease quantitatively, like the UK. Perhaps it prevented an ongoing depression. Regardless, it reinforced the narrative of “socialized losses and privatized returns.” When Wall Street wins, it gets rich. When it loses, you get poor.
The first of the large Wall Street firms to go public was Salomon Brothers in 1981. This was a bit of a scandal at the time, a violation of the tacitly understood moral code of Wall Street. In a private partnership, the partners’ money is on the line, they have skin in the game. A risky bet risked wiping out a chunk of the partners’ personal net worth. A public firm, however, risks the public shareholders’ capital, not their own. The traders and executives making the bets still kept a chunk of the upside when they made good bets, but suffered none of the downside when they made bad ones. Shareholder ownership led to asymmetric risk for traders and incentives to engage with exotic risk—heads I win tails you lose. The story of Wall Street since the 1980s isn’t a story of excessive greed or occasional financial panics, it’s a story about a financial system that became disconnected from the public it represented, in a way.
Howie Hubler stayed a millionaire–he just had to change jobs. Same with Wing Chau. That is a markedly different story from the pension funds that invested with them (that got wiped out) or the insurance company funds that paid their bonuses (they got wiped out too).
The brilliant, condensed question Michael Lewis asks with this masterpiece of financial explainer / expose is:
“What are the odds that people will make smart decisions about money if they don’t need to make smart decisions–if they can get rich making dumb decisions? The incentives on Wall Street were all wrong; they’re still all wrong.”
Reflection
After finishing, I called my parents to ask about their reaction to the 2008 housing market meltdown, and the kinds of conversations they had at the time, the reasons to which people attributed the crisis. They got the shape of everything right–too much leverage, too many high-risk loans that incentivized walking away, the unsustainable “high risk high reward” regime. They didn’t seem to grasp that most people didn’t know it was high-risk, that rating agencies had also been befuddled, or that there was a widening drift between the finance industry and retail investors and borrowers.
My dad was concerned about the Trump administration’s loosening of banking regulations, allowing them to leverage more. Despite the looser regulation and my dad’s worries, however, the banks are re-tranching, claiming ever more senior positions and passing risk off to hedge funds and pod shops, who are filling the riskier roles investment banks used to.
I think the deeper lesson isn’t really about banks or bonds specifically. It’s about something more structural in how human institutions work at all.
In the masterwork Sapiens, Yuval Noah Harari’s central claim is that humans conquered the world through an unmatched ability to coordinate at scale, and that cooperation runs almost entirely on shared fictions — money, corporations, nations, credit ratings. An AAA rating is exactly this kind of fiction: a shared story that lets a pension fund in Ohio trust a bond built from mortgages issued to a strawberry picker in California they’ll never meet. That’s an incredible technology–millions of strangers trading and building with millions of strangers! It’s also, obviously, a terrifying liability. The story only works as long as enough people believe it, and once it’s load-bearing for trillions of dollars, almost nobody downstream has the incentive to be the one who says the story is wrong, least of all the insiders who helped write the story. David Epstein calls this cognitive entrenchment, when experts get so fluent in a system’s existing rules that they lose the ability to question the rules themselves and lose flexibility under even small perturbations of those rules.
Bond traders knew everything about tranching and didn’t worry about whether a strawberry picker could actually make a $700k mortgage payment. Rating agency analysts stayed in their lane and never asked whether the models’ inputs made sense. It took outsiders — a doctor running a hedge fund on the side, two guys managing money out of a Schwab account in their garage — to ask a dumber, better question. That’s not a coincidence. It’s close to the mechanism.
Malcolm Gladwell argues in Talking to Strangers that we’re wired to default to truth. We assume the people and institutions around us are basically acting in good faith, because doubting everyone constantly is exhausting and makes ordinary life impossible. That’s not a bug — I don’t re-derive the FDIC’s solvency requirements every time I deposit a paycheck. But it means the system stays systematically vulnerable to anyone, or any incentive structure, willing to exploit that default trust. Wing Chau and the rating agencies didn’t need to lie outright, and they never did. They just needed everyone downstream to keep defaulting to truth for one more quarter.
This is the same shape as the Bernie Madoff story, which happened in the same years and barely gets mentioned alongside the housing crash even though it’s the same disease. The man who caught Madoff, Harry Markopolos, wasn’t a regulator or a fraud investigator — he was an obsessive, borderline paranoid quant who ran the numbers on Madoff’s reported returns, found them mathematically impossible, and spent nearly a decade trying to get the SEC to care. Nobody listened, for the same reason nobody listened to Burry: the story was working, everyone downstream was getting paid, and there was no institutional appetite to be the one who called it a lie. When I first watched The Big Short film, one of my friends joked that “Since 2000, Burry has called 11 out of the last two crashes.”
Markopolos, relatedly, kept a gun in his home office in case the SEC raided his house to cover up their own incompetence. He would check under his car for bombs and carried a smaller gun everywhere. His skepticism was ludicrous, expensive, and often misplaced. There’s a real cost–personal and social–to defaulting to suspicion instead of trust. Paradoxically, the extreme skeptics play an essential role for keeping systems trust-able.
Today, Michael Burry is still shorting, but he’s shorting NVIDIA, Tesla, even Caterpillar, who makes power generators that data centers use. He doesn’t trust these AI/computer-hardware companies’ forecasts about their profits or about how long their products will last and be useful, and doesn’t put stock in their stretched valuations. I’m not sure if he is right about these things, or if he has the timing right. Skeptics and outsiders, like Michael Burry, are often wrong. But sometimes they are so so right.
Footnotes
“Structured” refers to the set-up I described: A pool of underlying assets (mortgage loans, in this case) is sliced up into different tranches, each with its own risk, return, and priority of payment.↩︎
Consumer debt is a large umbrella which covers pretty much all retail debt–credit cards, car loans, student loans, home loans, and the innovative burrito loan all fall into this category. A consumer loan company is in the business of loaning to individuals rather than to other businesses.↩︎
When a homebuyer takes out a mortgage to buy a home, they pay for part of the home up front (a down payment) and then take out a mortgage for the rest. After a few years, they have paid down the loan a bit and own some equity. Additionally, the value of the home has likely increased. Both of these contribute to the home-buyer’s increased equity in the house. A second mortgage is a loan taken out against this extra equity, and is junior to the first mortgage. That means that in the event of default, the home is sold and the proceeds cover the first mortgage entirely, first, and then any extra goes towards paying back the second mortgage. Thus, second mortgages are considered higher risk, not only because they have a junior claim, but also because if the home-buyer doesn’t have any equity in the home… there is less stopping them from just walking away from it.↩︎
The consumer finance industry can deliver lots of convenience because it ruins some peoples lives. Take credit cards, for example–Chase can offer me a 10,000 point ($1,000 value) sign-up deal on top of 2% back on all spending because there are some card users who are paying a backbreaking 19-27% interest on their credit card loans. One of my brothers-in-law from the baseball game calls this the “irresponsibility subsidy,” since it effectively funnels money from the financially disadvantaged/illiterate to the financially literate. What can you do. I took the sign-up deal.↩︎
The traditional set up for a hedge-fund manager is to take a 2% carry fee, which is a flat 2% of all assets under management, and a 20% cut of the increase on investment. Burry thought the 2% carry fee rewarded him even if he didn’t increase investment, which he thought was unfair and incentivized him to accumulate more investors rather than make good on investments he already had.↩︎
Since an investor could buy the mortgage bond and insurance at the bond at the same time, some small yield was in theory guaranteed: If the bond doesn’t fail, they get the interest yield, and if it does fail, they get the insurance payout.↩︎
Adjustable Rate Mortgages were similar to teaser rate mortgages in that they started with a temporary pretty low interest rate, which typically lasted five years, and then floated after that to follow market rates plus a set margin. The failure mode was similar to teaser rates: Either a borrower couldn’t make the payments after the rate reset, or they took out a second mortgage in order to pay before struggling to make both combined payments and walking away.↩︎
Funnily enough, the bet on the unrest in Thailand didn’t go as planned. They had bet that the change in power would trigger a panic sell off of Thai currency, but the exchange of power was swift and smooth, and business carried on as usual. The Thai baht didn’t budge. Jamie Mai quipped, “We predicted a coup, and we lost money”↩︎
Alt-A mortgages weren’t loans to the highest rated borrowers, but they weren’t to the worst either. Alt-A borrowers, ostensibly, had FICO credit scores above 680, and subprime borrowers had FICO credit scores below 680. However, Alt-A loans were poorly documented and by 2007 they were really just subprime loans. For instance, a loan could be classified as Alt-A even if the borrower hadn’t disclosed their income. In practice, Alt-A loans made in the US between 2004 and 2008 totaling $1.2 trillion were as likely to default as subprime loans, totaling $1.8 trillion.↩︎