Sunday, 27 June 2010

The Goldman Case – 4: CDSs and Synthetic CDOs

The CDO at the center of Goldman case is a synthetic CDO.
According to the New York Times, a paper that has won many awards for educational reporting – that would be reporting that educates the public:
During the later stages of the boom, banks began offering so-called synthetic CDOs. Instead of combining bonds, these combine credit default swaps written against specific bonds or pools of bonds. Credit default swaps were developed as a kind of insurance on financial instruments, albeit in an unregulated form. Essentially, one party swaps the risk of holding debt with another by paying a fee to that swapholder in return for a promise that a certain amount of money would be paid in case of default.
According to Wikipedia:
In technical terms, the synthetic CDO is a form of collateralized debt obligation (CDO) in which the underlying credit exposures are taken on using a credit default swap rather than by having a vehicle buy assets such as bonds.

According to Investopedia, a synthetic CDS is:
A collateralized debt obligation that invests in credit default swaps. This investment can lead to large returns for holders of the CDO; however, the nature of credit default swaps may leave the holders liable for more than their initial investments, should there be significant changes in the credit default swaps.
These descriptions are not, per se, inaccurate. They are useless. You do not understand anything about CDOs from them because their language is clinical, the way a torturer would describe the details of torture without communicating anything about the horrors of the act. Clinical language is pernicious because it masquerades as the detached language of the “expert”, while decontextualizing the event it is supposed to describe. So we get empty words and the impression of knowledge being imparted, while nothing of the sort happens. Note how every definition above has a reference to credit default swaps without saying what they are or how they relate to the synthetic CDOs. Obscurum per obscuris.

Acquiring knowledge is an active process. It demands the conscious participation of the learner, which is another way of saying that knowledge has to be arrived at; it cannot be given. Let us return then to our investor group and see if, starting from where we left, we could arrive at comprehension of synthetic CDOs. Recall that we had “shot” the housing market by pressuring lenders and underwriters to indiscriminately create mortgages so we would bundle them into the CDOs and make a bundle in the process. The process wrecked the underwriting standards and pushed housing prices ever higher. With chickens about to come home to roost, the time has come to reverse gears. Now is the time to make money from the crashing housing market and defaulting mortgages. But, how?

Every trader knows the tools of the trade when things start going south: short selling and buying put options. Short selling is selling a security you do not own. The idea is to buy it back later, after it has fallen in price, for a profit. It is the time-honored buy low/sell high rule executed in the reverse order. While selling something you do not own is fraud in the standard commerce, in the securities market, it is de rigueur. The reason has to do with the peculiarities of the concept of security. From Vol. 2:
As a condition of opening an account in a brokerage firm, customers must sign a form granting the firm the right to lend the securities in their accounts to short sellers. When a security is “lent,” short sellers must compensate the original owners of the securities – the customers – in all respects such as dividends and splits save one: they do not and cannot grant the voting power to the customers because voting power is attached to shares and not the accounts. If the shares leave, so does the voting power. The condition for purchasing ownership in a corporation is agreeing to give up the legal rights of that ownership!

This fact may strike many as an absurd manipulation by the brokerage houses, but it is in fact a consequence of the very nature of a security. A security is evidence of ownership of notional capital. It confers on its owner not the right of ownership of the physical production apparatus but the right to appropriate a pro rata share of profits only. Since the short seller compensates the security’s owner for the profit portion, the owner has no more “rights” left.
But to be shorted, a security must be widely held, otherwise we cannot borrow it from another holder for the delivery. Individual mortgages are held by a single party. It would be impractical – indeed, impossible – to short them. How could we buy them back if the sole holder did not want to sell?

The other bear strategy, buying put options, is likewise not an option. There are no put options on mortgages, not the least because options pricing is based on a process that depends on short selling.

All this, however, is beside the point. It is missing the point, really, as these methods would not suit our purpose even if they were available.

Short selling and buying puts, you see, are capital market strategies in the sense that they are derived from, and exist, on the assumption of capital markets being a going concern. They assume uninterrupted trading. They are bearish strategies, of course, but they are the logical flip side of the bullish strategies which, combined, make up and define trading and capital markets.

We have an entirely different focus. We do not want a mere big win, a 10, 20, or 30 point gain from the price decline of some securities. Those are for boys -- and birds. We are men. Our aim is accordingly high. Why settle for mortgages declining when mortgages defaulting is in the cards? Since default is the extreme case of decline, our profits should be accordingly – and unprecedentedly – bigger. Our byword is implosion. We are shooting for a clean sweep that will come about in consequence of the collapse of mortgages.

Our strategy, then – and, by extension, the “products” designed to execute it – cannot be of capital markets. Implosion of securities negates securities trading.

Very little by way of contemplation is needed to realize that the only tool that fits the bill is a bet. We need a bet in which, if mortgages default, we would win big. But we need an intermediary, a financial powerhouse with contacts, which could find a counterpart to the bet and design and execute the plan. Would this intermediary let us pick the mortgages as well? After all, if we are betting big time, we need a fighting chance. It's only fair. Remember the $700,000 mortgage given to that Mexican strawberry picker who had made all of $14,000 in the prior year? How about something marginally better?

So, we take 5 “good” mortgages to the intermediary – Goldman, for example – and explain our plans. Here they are. We saw them in Part II:

1: $200,000 [$1,150]
2: $240,000 [$1,250]
3: $250,000 [$1,300]
4: $300,000 [$1,700]
5: $260,000 [$1,600]

The total value of the mortgages is $1,250,000; their combined monthly payment, $7,000.

The first step is creating a CDO in 3 tranches. Here is the CDO in three tranches: super senior (SS), mezzanine (MZ) and first loss (FL). The total principal and the monthly payment [in brackets] of each tranche is also given. We are familiar with this one as well from Part II:

SS: $700,000 [$3,600]
MZ: $500,000 [$3,000]
FL: $50,000 [$400]

Recall that the annual yield on the super senior (SS) tranche is 6.17%:

SS: 3,600 x 12 / 700,000 = 6.17%

With the raw material in hand, the intermediary calls upon its financial engineers, structure finance specialists, quants, wizards, rocket scientists and gurus to satisfy our demand, i.e., deliver to us a bet that would have the potential of a clean sweep. The focus, for the reason that will become clear in the sales pitch, is on the SS tranche only.

This is how the bet’s design process progresses:

What is the “other side” of the bet, the opposite of mortgages defaulting?

Well, it is mortgages not defaulting.

Good. How do we create the payoff? Who wins what if mortgages default – or don't?

Let us begin with the other side. What does the other side win if the mortgages in the SS tranche do not default?

The answer is, what the tranche pays, as long as there is no default. That is, $3,600 each month or $43,200 a year.

What would our client – here they are talking about us, the investor group – win, if SS mortgages default? Keep in mind that the client wants to win BIG.

What is the biggest number “pertaining” to the SS tranche? Yes, the absolute largest number “related” to the SS tranche, even if the question does not make sense.

The answer is the total principal of all mortgages in the tranche, equal to $700,000.

That is it then. The bet is set. Here is how it works.

One side gets $3,600 a month. The other side – we, the investor group – undertakes to pay it. Now, suppose mortgage No. 3, with the principal amount of $250,000 defaults. Nothing would happen in theory – and practice – because that does not affect the SS tranche. We would still have to pay $3,600 a month to our counterpart.

Then mortgage No. 4, with the principal amount of $300,000 defaults. Still, nothing happens to the SS tranche – in theory. We still would have to pay $3,600 a month to our counterpart.

Now, however, $550,000 worth of mortgages out of the total $1,250,000 have defaulted. The cushion that was protecting the $700,000 SS tranche is gone. One more default, and it will hit the SS tranche.

Suppose now mortgage No. 1 defaults. In that case: i) our monthly payment to the counterpart is reduced by $1,150, which is the monthly payment of mortgage No. 1; and ii) our counterpart has to pay us $200,000, which is the principal value of mortgage No. 1!

It gets better.

If mortgage No. 5 also defaults, the payment to our counterpart will be reduced by a further $1,600. Furthermore, the counterpart has to pay us an additional $260,000, which is the full principal of mortgage No. 5

Finally, if the last remaining piece, mortgage No. 2, also defaults, we would owe nothing in monthly payment to the other side. But the other side has to pay us $240,000, the mortgage’s principal amount.

If these scenarios were to happen, we would pay $3,600 a month for a few months, or a few years, until the mortgages began defaulting. In return, we would receive $700,000. That is the clean sweep, the big play. If you multiply the amounts in this example by about 1,500, they would approximately correspond to the transaction in the center of the Goldman case.

Note what happens to the mortgages when they are made the subject of a bet: they lose their characteristics as securities. To bet on their demise or survival, we no more need to own the legal title to them than a bettor in the race track needs to own the legal title to the horse he is betting on. This follows from the fact that our “strategy” is a mere bet and hence, outside the realm of finance. The sole role of the mortgages is now being a reference point – reference securities, they are called – for the determination of the winner and loser of the bet. In this way, our original, cash CDO becomes a synthetic CDO, where the mortgages as securities have been turned into mere virtual indices.

How is a synthetic CDO to be priced? If legal ownership existed, the question would be a simple matter of bond pricing. But there is no ownership, only the privilege to receive the monthly payment of mortgages, as if one owned them. So the question becomes, how much is that privilege worth?

The raison raison d'être of synthetic CDOs answers that question. The purpose of a synthetic CDO is generating a potential windfall for the seller, should the mortgages default. So, in return for receiving $3,600 a month, the buyer of a synthetic CDO tranche undertakes to pay to the seller the full principal amount of any bond in the tranche that defaults. This latter is a credit default swap: paying the full principal of a bond in case it defaults.

Synthetic CDOs, CDSs and their relations to one another are now clear. Return to the top and read the definitions. They will make sense.

All that is left is the small matter of selling this bet to the prospective buyers of the synthetic CDOs, say, ACA Management and IKB Deutsche Industriebank.

Here is the pitch:

Interest rates are at historically low levels. Five year Treasury rates are just over 4.17%, so if you invest $1 million, you will get $4,170 a year. How would you like to better that by 2 percentage points, to 6.17%, with the same credit quality?

Yes, really! Of course, in the fixed income area where the managers fight for two-hundredth of a percentage point,
2 percentage points does seem beyond the realm of possible, especially without sacrificing the credit quality and in fact improving it.

How do we do that? Well, we wouldn’t be a financial powerhouse if we couldn’t.(Smiles all around). But we offer this only to our best clients; not everyone can get into these deals, you know. Correct, exclusive deal, just like Madoff's.

But permit us to say that our deal is in fact better than it sounds. This will become clear if we get into the details.

How do you get $4,170 a year in Treasuries? You have to first invest $1 million. You will not earn interest unless you own the securities.

In the deal we are proposing, you invest nothing. Zilch. Zero. But receive $3,600 a month,
as if you had invested $700,000 in these securities.

The catch? There is no catch. You pay for these securities, so to speak, by selling a credit default swap to the party that pays you the monthly $3,600.

What is a credit default swap? That's an excellent question. The best way to describe it is to say that it is an insurance policy. Basically, if the mortgage defaults, you have to pay the mortgage principal to the other party. As per the SEC requirements, we must warn you that this is risky. There is a potential that you will lose some money – or even a sum equal to the principal of mortgages – blah, blah, blah; you know how it works. As if life had no risk. As if there were any guarantees.

But, look! The securities are the super senior tranche of a synthetic CDO. The tranche is rated AAA by both Moody's and Standard & Poor's. The probability of default of a AAA-rared security, based on the historical default tables, is almost zero. They are like the U.S. Treasuries in terms of credit worthiness. Between us, though, they are probably safer, what with this runaway U.S. expenditure and the beating the dollar is taking in the international arena.

Why even safer than the Treasuries, you may ask? We are not exaggerating. Let us look at the facts. The mortgage default rates in the U.S. stand around 3%. The $700,000 SS tranche we are offering is part of $1,250,000 CDO. So, it is protected by a $550,000 cushion of mortgages, so to speak. That is, 44% of mortgages in the CDO must default before the SS tranche is affected. Now, do you believe that the historical mortgage default rates in the U.S. could suddenly jump to 44%? You decide.

Yes, we have the list of mortgages. You are welcome to perform your own due diligence. But we must warn you that it is not always easy to get background information about mortgages. Underwriters are not known for keeping the best of records. (Smiles all around.) Now, if you could sign here.
Game. Set. Match.

Everything said or written about the CDOs and CDSs comes from this pitch, which goes to show how the Wall Street controls the narrative of anything related to finance. That is why I started this series with a treatise of sorts on the importance of knowing and the need to penetrate the surface of the phenomena.

Note how a “synthetic CDO” is presented as a “you can have it all” product: getting the monthly payments without buying the securities that pay those payments. That is what is taught in the business schools and promoted in the financial press as one of the benefits of modern finance.

Note, also, how a CDS is presented as an insurance product. To the best of my knowledge, not a single person anywhere has challenged this spin. The most vocal critics of the CDSs start from the premise of a good idea of insurance gone bad by the misuse by bad people.

Bond insurance existed in the U.S. and many western countries for the longest of times. The businesses were known as monolines, mono because they only insured bonds. The business was low-margin but also extremely low risk. Ambac, MBIA and SCA were among the more well-known names that were sucked into the orbit of the CDSs and paid dearly for it.

A CDS is the “other side” of a bet that involves the default of a bond. It has absolutely nothing whatsoever to do with insurance. A CDS is designed to insure a bond in the same way that gallows is designed to support the weight of a man. The moment you speak of insurance, you are lost. You have understood nothing about them.

Look at this childish nonsense from Floyd Norris of the New York Times. I mention him because he is the most perceptive and competent among the financial reporters.
Credit default swaps are, in reality, insurance. The buyer of the insurance gets paid if the subject of swap cannot meet its obligation. The seller of the swap gets a continuing payment from the buyer until the insurance expires. Sort of like an insurance premium, you might say.

But the people who dreamed up credit-default swaps did not like insurance. It smacked of regulation and of reserves that insurance companies must set aside in case they were claims. So they called the new thing a swap.

In the antiregulatory atmosphere of the times, they got away with it. As Humpty would have understood, Wall Street was master. Because swaps were unregulated, calling insurance a swap meant those who traded in them could make whatever decision they wished.
A word is changed and the entire regulatory apparatus is fooled and, as a result, something bad is introduced to the financial system. Just like that.

The creation of CDSs as a bet is the logical extension of cash settlement that took the physical delivery of the underlying security out of financial transactions. From Vol. 2:

To make themselves more appealing to speculative capital, derivatives undergo changes which are intended to bring out and emphasize their betting aspects. One such change is cash settlement.

In cash settlement, the product on which the derivative is based does not change hands. Rather, the profit and loss of the trade is settled by exchange of money, calculated as the difference between the purchase and sale price. The commodity itself need not – actually, cannot – be delivered to satisfy the contract.

On the surface, this arrangement seems as a mere technicality, a change in rules for the sake of promoting “efficiency.” Otherwise, the P&L of the two sides remains unchanged … But the P&L is not the only thing that “matters” in trading derivatives … In cash settlement, the price of a commodity becomes a mere index that is used to determine the winner or loser of the transaction.

The implication of this change becomes apparent if we look at the flow of money. Previously, the seller delivered, and the buyer took delivery of, the underlying commodity [i.e,] always buyers paid, and sellers received, money. In cash settlement, not only the seller cannot deliver the commodity, he must also think of a possible payment obligation.

Cash settlement accommodates speculation.
The Goldman case is now clear, not based on the SEC allegations or the firm’s denial but “as God sees it”. The role of the firm, the role of us, the investor groups – that would be Paulson and his hedge fund – and the role of the buyers of the synthetic CDO – ACA and IKB, which lost a combined $1 billion – is understood.

We are in a position to judge.

I will return with the epilogue.