Watch the Parking Meters

The big news on the blogosphere today concerns parking meters in Chicago.

Matt Yglesias is thrilled. Kevin Drum less so. Drum writes

A private company has agreed to give City Hall an upfront payment of almost $1.2 billion to run Chicago’s parking meter system for the next 75 years.

75 years seems a wee bit excessive to me, and will almost certainly bite Daley in the ass when Morgan Stanley, which put together the winning consortium, packages up the parking meter revenue, securitizes it, rolls it into an asset-backed CPMO (collateralized parking meter obligation), puts the super-senior tranche into an off-balance-sheet vehicle, hedges the rest via a CDS-backed synthetic CDO, and then resells the whole thing within 12 months to a sovereign wealth fund in Dubai for $5 billion.

From vehicles which aren’t on the curbe to vehicles which aren’t on the balance sheet in the bat of an eye.

My honest reaction is “Huh wha why would Morgan Stanley put a super-senior tranche ino an off balance-sheet vehicle?

I think I’ve been trying to learn about contemporary finance too fast. After the jump, I will use my very new knowledge of what the H a super senior tranche is to explain why I would guess that Morgan Stanley kept them on its balance sheet.

Felix Salmon explained what the super senior tranche is (as did ??? in comments here)

OK so the idea of tranches is that one can make highly (say AAA) rated instruments out of lowly rated instruments by pooling them and then selling claims of different seniority on the pool. A security which pays in full unless, say 15% of the underlying BB rated securities default will be AAA rated even if it pays zero if 16% of the underlying securities default. So will a security which pays in full unless 16% of the underlying securities default. What a waste. That means there are tranches worth more than the lowest value security that barely scrapes an AAA rating.

Those tranches are called “super senior” because they are senior to AAA. Sometimes they were informally called AAAA rated. This is nonsense. The lowest value “super senior” tranche isn’t even the senior tranche. The point is that, since there are a finite number of ratings, instruments with the same rating have different default probabilities. Also, due to the miracles of modern finance, the value of a fixed income estimate is not closely approximated by the default probability. That would be the case if the value of all instruments in default was drawn from the same distribution. As Lehman Brothers bond holders now know, the’re not your fathers defaulted bonds (they are now worth 10 cents on the dollar). In this example a tranche which combined the (pays in full if only 15% default and pays zero if 16% default) tranche and the pays if only 16% default tranche would have a default probability equal to the (pays in full if only 15% default and pays zero if 16% default) tranche, but it would be worth more compared to its face value. The fact that instruments with the same risk of default have different values is another reason that instruments of a range of value have the same ratings.

The innovation lead to the translation from AAA, which means at least as valuable as the least valuable AAA rated security to “AAA” which means the least valuable AAA rated security.

Now it is possible for tranching to be profitable even if the ratings agencies estimate default probabilities perfectly. The reason is that different firms face different regulations. For example, European banks have capital controls which depend on “risk adjusted” capital. To a European banker the ratings matter even if he or she doesn’t believe them. The capital controls have been binding, that is they would leverage more if they were allowed to. Thus they are eager for assets which are barely AAA. They are more risky than average AAA rated instruments and pay higher yields. However, for the Basel II regulations they count as if they were average AAA rated securities.

From 2004 until now (or at least until recently) Morgan Stanley did not face binding capital controls. Therefore they care about actual risk, not Basel rules weighted risk. This means they want the top ends of ratings intervals, including, I would guess the top end of AAA, on their balance sheets where they don’t waste valuable legal rights to increase leverage.