What price an AAA-downgrade?

Posted by John McDermott on Jun 27 12:26.

The US has already received its verbal warning.

Now here’s its numerical one. From Monday’s FT:

Threat of $100bn hit if US top rating lost

Investors in the US government bond market could face losses of up to $100bn if the largest economy loses its triple A rating, according to a research arm of McGraw-Hill, the parent of Standard & Poor’s.

A ratings downgrade that results in higher bond yields and lower prices could also mean the US Treasury paying $2.3bn-$3.75bn a year more in interest on financing a $1,000bn annual budget deficit.

More red meat, then, for those pushing for a debt ceiling deal. But where do these figures come from? The FT article doesn’t describe the study’s methodology — how did it reach the $100bn figure?

Here’s the study’s authors, Michael Thompson, Robert Keiser and Steven Krull of S&P Valuation and Risk Strategies, in their own words:

We used the difference in current credit default swap (CDS) rates for the U.S. and the average CDS rates for other sovereign credits with potential new ratings–as implied by Standard & Poor’s Market Derived Signals (MDS)–in order to estimate how much rates on U.S. Treasuries might rise. Then, we combined our estimation with bond durations to compute the potential price drop. We used MDS since many of the sovereign credits have CDS rates that appear inconsistent with their credit ratings.

This could of course be a sign that some credit ratings are overly generous (there was silly us thinking, erm, price was the market-derived signal). But apparently not.

Instead, the authors took a bunch of CDS quotes for different sovereigns and obtained the difference between the US’s current CDS quote (52.5 basis points) and the “market derived signals” (MDS) quote for an average AA-rated (75.7bps) or A-rated (90bps) sovereign. These differences are then assumed to be proxies for additional interest costs and applied to the US’s outstanding 10-year and 30-year bonds:

Currently, the 10 year bond has a modified duration of approximately 8.5 years, and the 30 year bond, 16.8 years. All other things being equal, based on these durations, the 10 year bond’s price would drop by 2.0% and 3.2% for ratings reductions to ‘AA’ and ‘A’, respectively. Similarly, the 30 year bond’s price would drop 3.9% and 6.3%, respectively. Although the possible interest changes appear moderate, across the spectrum of U.S. Treasuries with maturities of two years and longer with over $4 trillion currently outstanding, the total loss to investors could easily range from $50 to $100 billion. Also, increased rates would result in higher borrowing costs for the Treasury in the future. For example, financing a $1 trillion deficit would add $2.32 to $3.75 billion per year in additional interest expense.

As the authors imply, their (“rough”) analysis could have been done on the back of a cigarette packet. And it all depends on these so-called ”market derived signals” ratings, which aren’t explained in the report or the FT article. Fortunately, there’s a guide to what this all means on the S&P website.

In the Market Derived Signals Model, the key parameters observed for a firm are its current five-year credit default swap spread, Standard & Poor’s long-term issuer credit rating and CreditWatch/Outlook status, Global Industry Classification Standard (GICS® ) sector, CDS document type, and currency denomination. We compile this information on a large number of firms on a daily basis, and then at the end of each day, we estimate a linear model that regresses the observed log of the CDS spread on each of the other variables. For purposes of the model, the credit rating on the underlying obligation is assigned a numerical score that corresponds to the credit ratings scale, i.e. ‘AAA’ = 1, ‘AA+’ = 2, etc.

The methodology was designed for firms but has been applied to sovereigns, with a small tweak to adjust for the different durations of CDS spreads (some had five-year and some had 10-year spreads).

It’s certainly wise for the rating agencies to incorporate several variables when assessing the impact of downgrades on interest rate costs. But it’s also interesting that it’s having to use this methodology because — as S&P admits — market prices are diverging from what its normal rating would expect.

Full report in the usual place.

Related link:
Threat of $100bn hit if US top rating lost – FT

This entry was posted by John McDermott on Monday, June 27th, 2011 at 12:26 and is filed under Capital markets. Tagged with , , , , , , , .

Post a comment or leave a trackback: Trackback URL.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: