The Big Short: Michael Lewis – A Book Review

The Big Short is a very insightful book viewing, from a different angle, the financial structure collapsed due to the bubble in the mortgage markets. Michael Lewis does a wonderful job at tackling the subject from the point of view of market participants that believed that the housing bubble was just not sustainable. The book introduces us to various characters that are not very famous and until now I personally hadn’t heard about: Michael Burry of Scion Capital, Jamie Mai and Charlie Ledley of Cornwall Capital, Steve Eisman of Frontpoint (probably the best known of the group), and participants from Wall Street such as Greg Lippman from Deutsche Bank, Howie Hubler from Morgan Stanley, and a group of ex-Drexel traders inside AIG. The story of each of these players alone is worth picking up the book.

Of course in addition to reading how some of these market participants concluded that the housing market and specifically the way that Wall Street packaged the mortgages was going to blow up, you will learn and come to understand how these products worked, why they where created, who was feeding the pipeline, and how the inverse pyramid structure of subprime and CDOs resulted in bringing down the financial system not just in the USA but across various countries.

Here is a brief review of what was going on, and how the scheme worked; from initiating a plain vanilla mortgage to create Mortgage Bank Securities (MBS), followed by Asset Back Security (ABS), to Collateralized Debt Obligations (CDO) and then finally the Credit Default Swaps (CDS) and synthetics. In order for this to have worked for so long, especially the miss-pricing of risk, a belief was held that because of low historical correlation, home prices could not fall across the US at the same time and with the same magnitude; this is essence was the main cause of the bubble – that and greed of course. This belief set into motion a mechanism that would not stop as long as some investor (yield chasers) believed that premise.

The bubble grew as Investment Banks (IB) started to package pools of mortgages into MBS that received AAA ratings from the rating agencies. As the housing boom grew, lower quality mortgages started to be issued, to feed the IBs products: sub-prime, interest-only, alt-a (no doc), etc normally the market would have re-adjusted risk and higher prices would have slowed down the process, but these lower quality mortgages started to get packages into MBS and then as MBSs got sliced into different tranches became part of ABSs that continued to receive ratings from AAA to BBB despite the deteriorating quality of the underlying mortgages. The models used by the rating agencies where assuming low correlations between regions and types of mortgages, and they all assumed that these bonds would be fine as long as default rates stayed at historical levels (around 5%) – both assumptions turned out to be very wrong. At first everybody was on the same page since everybody was making money; home owners were happy as prices rose, IB made fees packaging these bonds, investors received slightly higher yields then currently available in Treasuries, and rating agencies were making a nice fee rating all the MBS, ABS, and CDOs. By the time the rating agencies changed the model it was too late since the amount of paper outstanding was enormous.

As BBB ABS became more difficult to sell by the IBs, a division of AIG got involved to provide insurance against default on these bonds. AIG is after all an insurance company, so its job is to find ways to collect premiums by pricing risk correctly, in this case the risk of default. Back in 2005 the premiums where more or less 20bps on AAA securities, 50bps on A rated securities and 200bps on BBB securities. In other words, a CDS owner would pay a premium of say $200,000 per year to in theory to “hedge” a portfolio of $10mm BBB MBS/ABS/or CDO depending on the structure. AIG collected a premium, which they did for a couple of years producing nice profits, but took on the risk that the bonds would default resulting in a loss of the face value of the insurance, in this case $10mm. The interesting part of these models was that they assumed a 6% default level and very low correlations; basically everybody was using the same flawed model.

Another interesting situation was that once these BBB ABSs were once again sliced up into different tranches that somehow became AAA CDOs ($10bn BBB ABS somehow became part of AAA CDO). The CDS insurance on these structures soon found their way to the short sellers and not to the holders of the securities that should have been the natural buyers of this insurance. Once AIG stopped supplying CDS insurance in 2007 (they had about $80bn in exposure by then), and less mortgages and MBS or ABS where around to create CDOs, the long buyers of mortgages securities (the yield chasers) started to buy synthetics, since they still wanted to be net buyers of any product with a higher yield – which required a natural short seller on the other side of the trade. The buyers of these products, mostly fixed income institutional money managers were not as sophisticated as their titles indicated, for the most part they trusted blindly the rating agencies. And short sellers where more than willing to comply by taking the other side – towards the very end of the musical chair game the IB where the ones taking the other side of these synthetics.

What the short sellers found out after researching the structure and adjusting their assumptions was that these models were pricing risk very wrong. The rating agencies didn’t differentiate very much between high quality prime mortgages and low subprime mortgages, as long as the mix of the mortgages inside a MBS or ABS had low correlations, ie coming from different regions, but every month the quality of these mortgages was dropping. The shorts sellers recognized that more of these structures carried floating rate mortgages that would reset in 2years, and that somebody with 2 or 3 properties was not going to be able to pay back any of them (in 1996 65% of mortgages had fixed rates, by 2005 75% were floating rates). That simple though process was what kept the short seller involved in the trade for what they thought would be a two or three year horizon. They were willing to pay the premiums on the CDS for two years, feeling that the payout when default occurred was worth the cost. Short sellers were expecting three things to occur: 1) floating rate become higher fixed rates (producing defaults), 2) correlations to be much higher than 30% once defaults started to occur, and 3) they did the math and found out that ONLY 7%-8% of the mortgages in these ABS or CDO had to default in order for the whole structure to go to zero! That was the main reason they had so much confidence in their positions – they felt like the risk reward was completely in their favor; pay a premium for two years and collect a huge payoff, some of the bets where basically risk $1mm to make $80mm, and they felt like the market was mispricing that risk. I for one remember mentioning many times that even if some subprime mortgages defaulted they only made up less than 8% of total mortgages outstanding, clearly not enough to bring a whole system down. Well like many normal people we didn’t know about was the inverse pyramid that was created on top of these low quality mortgages. In 2000 less than 50% of subprime mortgages where packaged into other bonds, by 2007 80% of them where packaged and repackaged and even became the benchmark for synthetic securities (imaginary portfolios of mortgage bonds – which required an entity to be long and another to be short).

If this explanation was a bit too short, then definitely pick up the book, Michael Lewis does a great job at explaining the structure and the thought process behind all the structures. Also of interest was the unwinding of the shorts, since now the CDS counterparties where AIG, Bear Stearns, Deutsche Bank (DB), Morgan Stanley (MS), Goldman Sachs, HSBC, etc all of Wall Street, which if they went under would not pay out on the CDS insurance policies. The Big Short covers some unique stories of a DB and MS traders that where in the center of the markets. One found himself short to “take one for the team” but soon convinced himself and others that short was the right trade; the other trader, in MS, started short and had to hedge his position because of the costs associated with paying the premiums on BBB CDS, just to take the other side of AAA CDS thinking he had a spread hedge that would still pay off more on the short side – he was wrong and turned a trade that would have paid $2bn+ into a $9bn loss.

It’s hard to pass negative judgment on the short sellers, at least the original ones; maybe when the IB started to take the other side and went out of their way to pile in the worst bonds into synthetics you can make a good case of unethical practices. The rating agencies seen to be clearly at fault, if only for not understanding the products they where rating, and not understanding the enormous size of the markets – mistakes on small abstracts structures is one thing, but this market was huge, more care should have been taken to make sure stress tests made sense; also the role of the regulators, not existing really, and the Government agencies such as Fannie Mae and Freddie Mac, all made the issue worst. However another book has to come out on the rating agencies, I am looking forward to that one.

Michael Lewis starts the book with this quote of Leo Tolstoy in 1897:

The most difficult subject can be explained to the most slow-witted man if he has not formed any idea of them already; but the simplest thing cannot be made clear to the most intelligent man if he is firmly persuaded that he knows already, without a shadow of a doubt, what is laid before him.

So true…