The Debt Rating Problem

by: Carl Dincesen

It is bigger than you think. The world's financial system depends on accurate risk ratings whether assessed in-house by institutions or by commercial rating agencies. It might surprise you to know that in the 21st Century, the world's financial system does not have a uniform statistical standard measure of long-term credit risk.

The "standard" rating definitions used by the two largest rating agencies are a century old, and, not surprisingly, make no mention of any statistical measure of default risk.

Here they are for the three highest in a side-by-side comparison.

Click to enlarge

We measure and communicate future unknowns in terms of probabilities. Probability of default or range of probabilities is the measure of credit risk. Without it, raters cannot communicate the difference in risk or probability of default between BBB and AA, as an example.

The overriding question is how much capital must a lender have to support X amount of AA rated risk versus the same amount of BBB rated risk? As defined, today's ratings are virtually useless in providing a sound basis for setting risk based capital charges for lending institutions worldwide.

In 2010, I developed a rating scale for use in providing investor paid credit rating opinions on municipal bonds. I then realized that it was applicable to private as well as government debt. The only necessary distinction is how many years is your forecast valid. Expiration requires another review in order to keep the rating current i.e. valid.

That uniform standard measure of long-term credit risk looks like this in table format.

Click to enlarge

As a point of reference, recently the U.S. Federal Reserve Bank published a study on historical municipal bond defaults. A more than cursory review of the data shows a ten-year historical default rate on tax and government revenue secured bonds of 0.5%, rated and unrated bonds combined. Combined being appropriate because ratings are typically withdrawn prior to monetary default and so are unrated at the point of default.

In the rating table, a single A rating denotes a 1% maximum ten year probability of default and a minimum or unstressed rate of 0.5%, that also being the maximum default denoted by a double A rating. Once calibrated correctly for either government or private debt they are applicable as defined without regard to debt sector.

The scale tilt's a bit conservative all the better in favor of investors i.e. the buy side.

U.S. and overseas financial regulators must act to require all institutions and rating agencies use a uniform standard measure of credit risk like the one referenced above. Without that, the financial system continues to be significantly more at risk than it should be.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.