Lecture Notes – December 10, 2012

A few bits and pieces today. First I will have additional office hours this week. I will plan on being in my office during my normal office hours and then I will also try to be there every day from around 12-2:30.

Second, please watch the Greenberg video before Wednesday. We will spend a bit of time comparing/contrasting what we covered in class vs. what is in the video on Wednesday.

Finally I have not had too many responses to my request for things to cover in our review sessions. Please let me know.

Mid-term grading

From talking to various people in the class I think there is some anxiety about where everyone stands going into the final. I spent some time going through the midterms this weekend and wanted to share with you my thoughts. First – everyone should realize that I am grading on a curve. If you recall the scores from the first midterm they were low but the curve is going to help a lot – recall I gave out indications of how I was going to divide the results.

The second point is that there are very few of you who consistently came well below average. If you scored in the bottom section on the midterm three times then I would say you are in a challenging position. What I intend to do is lower the weight of your worst mid-term by 20% and increase the weight of the best by 20%. This is going to help people that are on the cusp of going up a grade to go up that grade.

Finally I should point out that you have 60% of your grades. There’s an additional 40% to come. Looking at it if you have a C on the 60% then getting an A on the final will give you a

1.2+1.6 = 2.8 = B-.

Failing the final will date you down to a D+

If you have a D average then getting an A will give you

0.6+1.6 = 2.2 = C-

Having a B average then getting an A will give you

1.8+1.6=3.4 = B+

and so on.

If you are worried about where you stand in the class right now feel free to come see me and I am happy to go through your scores with you.

Credit Default Swap

What is a credit default swap? I like the definition in the COP report. Credit default swaps (CDSs) are privately negotiated bilateral contracts that obligate one party to pay another in the event that a third party cannot pay its obligations.

In essence, the purchaser of protection pays the issuer of protection a fee for the term of the contract and receives in return a promise that if certain specified events occur, the purchaser of protection will be made whole.

What does this really mean?

A basic example would be:

I would like to buy a Greek government bond. The bond is trading at par and yields 14%. I am concerned though that something might happen in the future and Greek government might default or restructure my bond so I want to buy protection from this event. I might be willing to pay 4% per year to someone in exchange for the right to exchange my bond for par in the next 5 years.

Before we get to the technical details we should look at some of the usage of credit default swaps:

General Usage

At this point liquidity in the CDS market for certain names may be better than that in the underlying corporate bond market. Thus it is more efficient to obtain an exposure to a reference entity through CDS than to buy the reference entity’s bonds.

  1. The liquidity of the CDS market compared to the corporate bond market makes it more efficient to obtain an exposure to a reference entity through CDS rather than through the purchase of the reference entity’s bonds.
  2. Conditions in the corporate bond market may make it difficult to sell a bond for which a manager is concerned about the credit of an issuer
  3. If a portfolio manager believes that an issuer may have credit issues in the future then CDS allow him to express this view. Alternatively the manager may short the bond but this is difficult in the corporate bond market
  4. A portfolio manager may seek a leveraged position in a corporate bond

Structure

Reference Entity and Obligations

A credit default swap starts with a reference entity and obligation(s). The entity is the issuer of the debt instrument to be protected. The obligation is the particular debt instrument for which credit protection is being sought. Generally there will be more than one obligation – for example senior unsecured IBM debt could be the reference obligations for a CDS contract.

Maturity

The typical maturity is 5 years but may be anywhere up to 20 years. The maturity dates are generally fixed to be March 20, June 20, September 20 and December 20 of any given year.

Premium payments

CDS are quoted as a spread – say 150 basis points but this is not how the premiums work. A protection buyer will pay either 100 basis points or 500 basis points throughout the life of the CDS.

The premium payments are made quarterly and will occur on March 20, June 20, September 20 and December 20.

Thus if you want to enter into a CDS contract you will pay accrued interest for the period between the last possible coupon date and today, as well as an upfront fee for the difference between the 100 basis point spread and the fair spread.

Let’s see what this means in practice. For example, let’s say that the fair spread for credit protection on a company is 100 basis points. If we are exactly half way through a coupon period you will pay accrued interest of 12.5 basis points to enter into the contract. Then you will make regular payments of 100 basis points.

On the other hand it could be that the fair spread is 50 basis points rather than 100 basis points. As the protection buyer you are paying too much for protection. Thus there will be an upfront payment equal to the present value of the additional 50 basis points that you will owe.

The reason for this standardization is to add liquidity to the secondary market. If I own a CDS on a particular name I will be paying 100 basis points or 500 basis points per annum. If I want to close out the position I can either try to cancel the trade with my counterparty or sell protection on the same name with someone else. I will be left with counterparty risk but no payments or market risk. Previously I might have sold protection on the same name but would have been left with an annuity stream representing the difference between the two spreads.

Credit Events

It is not just default that will trigger a credit event on a credit default swap. Credit events include:

•Bankruptcy

•Credit event upon merger

•Cross Acceleration

•Cross Default

•Downgrade

•Failure to pay

•Repudiation/moratorium

•Restructuring

Restructuring is the most complex of these events. From the perspective of the buyer of protection one wants as broad a definition of restructuring as possible. From the perspective of the seller one wants as narrow a definition as possible.

Settlement

On a default event the protection buyer pays the accrued premium up until the default date and delivers a number of bonds with face amount of the notional of the CDS.

The protection seller pays the protection buyer the notional amount of the CDS.

Note, there are often many more CDS trading compared to the amount of bonds. Thus upon default there can be a challenge for the protection buyer to acquire a bond in order to deliver it. Thus in a very active name one can have significant price increases in the reference obligations as naked protection buyers scramble to purchase the bonds.

NOW THERE ARE A COUPLE OF SLIDES ON MECHANINCS

Size of market

In December 2011 there was an estimated 25.9tr gross notional of CDS written. This translates into a 2.7tr net notional. It is an OTC market so is difficult to measure precisely.

The market is monitored by the International Swaps and Derivatives Association. This is a trade organization that strives to make the derivatives marketplace safer and more efficient. As they say…

Bad decisions about credit have been expressed through many different instruments, but it is important to note, it is the decisions NOT the instruments that led to losses.

Some data

The next couple of slides show a bit about the largest CDS notionals. From a gross notional perspective it is dominated by CDS on sovereigns. On a net notional basis a couple of different names are added in – residential MBS and GE.

Securitization

To talk about AIG we need to go back to my favorite securitization picture. I found a good picture on the web so I don’t have to draw it on the board. Basically you have lots of mortgages, they are pooled together and securitized. There are two things to discuss here. The first question is who buys the tranches of the securitization. There are lots of businesses that want to hold AAA rated paper – low risk money market funds, insurance companies, securities lending businesses looking for yield. The real problem is who buys the lower rated paper. The reality is that especially for RMBS containing sub-prime – no one wants it.

The second thing to talk about is how much risk is there in the AAA rated paper. From a bank capital perspective a bank has to hold more capital against AAA asset backed paper than one does against sovereign debt. There’s a solution to this that we will get to shortly.

What do you do with the non-AAA tranches

You securitize them!!!

Multi-sector CDO

A multi-sector CDO is a transaction that securitizes a variety of structured finance collateral. Incuding:

•asset-backed securities (e.g. securitizations of auto receivables, credit cards, etc.),

•commercial mortgage- backed securities,

•CDOs and

•various types of residential mortgage-backed securities including prime and subprime RMBS.

As the markets went through 2000-2006 more and more subprime RMBS was securitized in multi-sector CDOs.

Intuitively these structures should be more risky than the underlying RMBS but that’s the trick of securitization. You can simply choose a higher level at which your AAA tranche starts and you should be fine.

Underpinning the valuation of a multi-sector CDO though is a correlation assumption.

What does your correlation assumption do to the value of your tranches?

Let’s first think about the AAA tranche. If the correlation between the underlying assets is higher then it reduces the value of the AAA tranche because this increases the probability that more assets become impaired together and thus the AAA tranche could suffer losses.

On the other hand it increases the value of the equity tranche. This is because the more correlation there is the higher the probability that both higher losses than expected will occur but also there is a higher probability that lower than expected losses will occur. Thus the equity tranche has a greater chance of a payout.

AIG Credit Default Swaps

AIG was fairly unique in that it almost exclusively sold protection. It sold protection on super-senior, high grade and mezzanine tranches of muti-sector CDOs. Another player in this business was MBIA which also got into trouble during the financial crisis.

The idea for purchasers of the protection was that they could hold less capital against AAA pieces of multi-sector CDOs if they purchased insurance against them. One of the challenges in buying insurance is making sure that you purchase insurance from a company that is going to be able to pay. AIG was a AAA rated company – seemed like one of the safest.

The idea for AIG was that these things never had issues anyway so they were getting free premium for transactions for which

•It is hard for us, without being flippant, to even see a scenario within any kind of realm of reason that would see us losing $1 in any of those transactions

Think about how hard it is to fight against this as a company CEO. You have various divisions of the bank telling you about their performance and AIGFP tells you that they are earning risk-free premiums of X per year.

There were two problems.

Collateral – even though there had not been defaults the assets underlying the multi-sector CDOs were becoming impaired and thus the value of the multi-sector CDOs were being written down. This in turn increased the value of the CDS contracts. This in turn required AIG to post additional collateral.

Maiden Lane III

So AIG are in trouble. They have written CDS against multi-sector CDOs for which they have to post enormous amounts of collateral. The Fed are bailing them out but the collateral posting keep increasing. They chase the counterparties of the insurance who all tell them – these assets are close to default and we purchased the insurance for a reason. We need the contracts in order to ensure the viability of our business. In the end Maiden Lane III was set up to purchase the multi-sector CDOs from AIG’s counterparties so that the credit default swaps could be cancelled.

The Fed invested $30bn into these securities. The last of the securities were sold in August 2012 and the Fed made a profit of $6.6bn or 22%.

Greenberg Video