Explanations of the various bizarrities of the financial world
So, monolines. What are they, and why are
people worried about them? Why do they have such a strange name? Monolines take their name not from the disease that they
are trying to spread but rather from the fact that they do only one line of business. That business is to insure bonds. Now
it is not unusual to hear suggestions that bonds themselves do not lead a particularly dangerous life. In the old days they
were pieces of paper which spent most of their lives in musty safes. Now most don’t even take these sorts of risks living
a solely virtual life in hyperspace, merely an electronic transfer. However monolines are not insuring the bonds against damage
but rather their cashflows. In truth they aren’t looking to help bonds at all, they are looking to help the people who
buy bonds (bondholders). They guarantee the bond’s interest payments and principal payments.
So if you buy a bond from some dodgy corporate and it goes bust, if a monoline has insured the bond they will pay the
interest and principal on the bond on behalf of the now defunct corporate. This process is known as wrapping a bond because
the bond is encased in the soft warm loving embrace of a monoline that ensures it will pay.
Monolines grew up to services states and cities in the US. They would guarantee the bonds issued
by these organisations which helped them pay for essential public services and wildly luxurious expense accounts for politicians.
They have also been heavily involved in the mortgaged backed securities market wrapping some of these bonds. This as you can
imagine is not considered to be an activity that contributes greatly to their financial strength at this point. It does get
worse though. They have also been involved in CDO of sub prime ABS, the most evil of all financial creations with the exception
of poverty and death (the later is not entirely as financial creation).
Now it is essential for monolines that they retain a good credit rating as this ensures that there insurance is worth
something. If they had low credit ratings no one would pay them to insure bonds, and if by some freak of nature they did,
no one would buy them. Most monolines have AAA ratings. One, ACA, has only an A rating but this is probably just because it
is part owned by Bear Stearns. If a downgrade were to happen to a monoline this could be disastrous. Rating agencies are reviewing
the monolines in light of their activities in the sub prime markets, but they are being very careful about it.
The monolines themselves are sanguine. The biggest, MBIA, says it has no problems and is instead concentrating on coming
up with the kind of corporate slogans that make you think they want a job at Hallmark. Their main slogan ‘We never mistake
borders for barriers’ does perhaps lead you to believe that they might mistake barriers (sub prime) for opportunities.
The other ‘Sink or Swim? That depends on how well your bond insurer knows the water’ is crying out to be reused
as an epitaph.
In our first articles in finance explained we look at some of the mechanisms
used in credit structuring (the stuff that got us into this mess, well somewhat anyway), beginning with tranching.
Tranching is one of the fundamental wizardries (some might say alchemy) that makes up credit structuring. It is a mechanism
that allows one to produce some AAA bonds from a group of BBB bonds. In this sense it can turn trash into gold whilst preserving
some of the benefits of trash. What are the benefits of trash? Well in the bond world they usually come in the form of high
yields. Tranching is fundamental to CDOs. If I were a credit structurer I would refer to it as part of the technology of CDOs.
This is a strange reference, it is almost as if credit structurers are trying to turn themselves into IT geeks, as if they
consider them higher in the social hierarchy.
The fundamental idea of tranching is allocation of losses. Suppose one has a corporate bond portfolio which is owned
by a group of investors who have each staked an amount of money to buy the bonds. The question now is how are you going to
allocate the revenue coming from the bonds (which comes in the form of coupons). You could do it evenly, giving everybody
the same, a somewhat communistic approach (though truthfully to get to a communistic allocation I suppose you would allocate
on their need for bond coupon a subject Marx seldom commented on, but anyway this is capitalism). Then you would, in this
extremely even society, allocate any losses on the bond portfolio (coming from defaults on the bonds) evenly across all the
beneficiaries of the portfolio. Now society doesn’t work that way; its uneven ( I could be going a bit far with this
parallel but somehow I feel sociologists have missed an opportunity by not seeing what bond structuring reveals about people’s
perception of social hierarchies). So inspired I could construct a different way of distributing coupons and losses.
Suppose I select 1 person, let’s call him Bert, and say he will receive all of the first losses of the portfolio
up to a certain limit, say ten percent. Now no matter where in the portfolio the losses come from they will go straight to
Bert. In society the person who’s taking all the risk would probably also be paid the least, but this isn’t the
real world this is credit structuring. It gives to people in proportion to the risk they take on. Clearly Bert is taking the
most risk. Therefore he gets allocated more of the coupons than the others. Now once the first ten percent of the portfolio
is eaten up by losses the losses have passed through the barrier of Bert’s loss allocation (and eaten his entire investment),
and the next person starts to get losses allocated to him. Now obviously if the portfolio is a good one there is every possibility
that the losses will never exceed 10% and all the rest of the people will be fine. But if they do exceed 10% the next person
in line starts to get hit by the losses. He gets the second most amount of coupon in the allocation. And so you can work up
the ladder of people.
Now say the
bonds are pretty poor, paying high yields but likely to go into default. Bert’s allocation will be very dodgy, he’s
likely to get hit with a lot of losses. Even the next person up the chain is likely to get hit. But even with dodgy bonds
the third and fourth are unlikely to get hit. The fifth, sixth and higher will have very little, they will have the equivalent
of a AAA bond made from very poor bonds. Of course Bert and his fellow low level guys will have the equivalent of very poor
quality bonds. The wizardry of the structurer is that given all the ratings of the bonds and the effective rating of the position
that the investors find themselves in you can create much higher yields that with similar quality bonds. So the fifth person
will receive a much higher yield than a normal AAA bond. The reasons for this is correlation. Correlation is the food the
magic of credit structuring feeds off. Next week we will look at how it works.
is the food of the magic of credit structuring. Tranching (see below) allows credit structurers to ‘create’ extra
coupon from bonds. This is the magic. What drives the magic is correlation. Well more precisely the rating agencies image
Correlation is broadly
a measure of the relationship between two (or more) things. So if two events are highly correlated they are likely to happen
together. If we say two bonds have high default correlation they are more likely to default at the same time than two bonds
with low default correlation. In practise it is quite a tricky idea, for example what does it mean for two bonds to default
at the same time? Is it the same minute, day, century? But we will follow the rating agencies and ignore these complexities.
Now consider two potential investment ideas; investment number one is to invest in just one bond with $100, the second is
to invest $100 equally in 100 bonds. Which is the better investment? Well of course it depends upon the bonds, but suppose
we have a way of rating them (what a great idea eh?), and they are all pretty much the same. In 1 if the bond goes bust you
lost all you’re investment (ignoring for the moment any recovery on the bond). But that is very unlikely. If one of
the 100 in number 2 goes bust then you will lose only a small part of your investment, but clearly there is more chance of
that happening, because there are 100 of them. Now lets split investment number 2 into a 2 further choices, call them A and
B. In A all the bonds are high default correlation. In B they have low correlation. In A you have a higher chance of a big
loss, but a low chance of a small loss, whilst in B it is the reverse. Now as explained in tranching from the 100 bonds the
CDO securitisation process (and at a lower level the same thing was used in asset backed securities, though in this case mortgages
were the underlying rather than bonds) can produce a series of new bonds who, when compared to normal corporate bonds with
the same rating (from the credit rating agencies) produced higher spreads. And this was because the rating agencies believed
that a portfolio of bonds, well diversified with low default correlation, was a better investment than a single bond investment.
And at the end of the day,
that is what produced the demand for CDOs and in fact mortgage backed securities, the idea that a well diversified portfolio
was better than a single investment. The rating agencies built complex models to measure diversity (at least that’s
what they say, they could just be picking a number out of a hat ,except for the fact that this would produce a more realistic
estimate) and to rate the bonds produced by securitising bonds and mortgages.
Never overestimate the benefit of unadulterated truth
Regulators, they're better than nothing...well almost