I wish to express my appreciation for all comments and suggestions made by readers of the first three parts; they were invaluable.
Part D: The Equity Risk Premium as Insurance Premium
It was shown previously that Credit Default Swaps are in fact insurance contracts and not bona fide swaps. Taking this insurance characteristic further, let's view the entire CDS market as a large insurance company (CDS Inc.) underwriting credit default insurance on the following entities:
- Global Governments
- Global Structured Finance
- Global Corporations
We can safely assume that central government credit represents only a small portion of the default insurance business. The largest government debt issuers in the world (US, UK, Germany, France, etc.) are themselves rated AAA and serve as risk-free benchmarks. The sole exception is Japan, which carries a AA- rating; it is highly unlikely, however, that hedgers and speculators are much interested in underwriting or purchasing CDS on Japanese government bonds (JGBs), given the extremely low nominal interest rates (e.g. the 10-year JGB is currently at 1.56%). The rest of the world's governments simply do not issue so much debt as to make a difference. In sum, I estimate that no more than 5% of CDS notional outstanding is against central government credits.
This leaves regional/municipal debt. In the US, by far the largest issuer of such bonds, this market is estimated at $1.7 trillion, spread among 50.000 separate entities that issue bonds (SIFMA). This is a market where credit insurance has been dominated by the monoline insurers (MBIA, AMBAC, FGIC, etc.), long before the advent of CDS. It is pretty much an end-user investor market, meaning that there isn't all that much speculation going on. With the exception of a few large issuers such as the States of NY, NJ, California and Florida the market is highly fragmented. This, in effect, limits the use of CDS to mostly hedging purposes and it is unlikely that notional CDS outstanding exceeds the face value of muni bonds, or approx. $2 trillion - if that much. This represents less than 5% of global CDS, bringing the total for all government-related credits to ~10%.
Next is structured finance, i.e. credit insurance written against CDOs, CLOs, etc. The total size of the asset-backed market is ~$2.6 trillion, with new bond issuance particularly heavy during the last 2-3 years. For credit insurance, this was also a hedging-related activity until rather recently, when speculation on the sub-prime MBS market created significant short selling interest, particularly against the ABX and CMBX indices. There is no way of knowing with any degree of certainty what multiple of face value is covered by CDS - but we do know that banks have been taking large, multi-billion losses on their portfolios of such paper, so it cannot be very large. Goldman is apparently an exception and so are some hedge funds who called this market right and benefited accordingly. I will go ahead and guesstimate the amount at ~$5 trillion, or another 10% of notional CDS.
This brings the total for the first two sectors to 20% of CDS notional, or $9.1 trillion, leaving 80%, or over $35 trillion, as CDSs written against global corporate names. The US Office of the Comptroller of the Currency in its latest survey found that US banks' positions in credit derivatives ($12 trillion) were broken down as follows: 73% investment grade - 27% sub-investment grade (chart below). It can safely be assumed that this split is pretty much the same across the entire CDS market.
Let's revert now to the original premise, that CDS Inc. is one very large insurance company writing all this coverage, 80% of which is insuring the credit, or business risk, of Global Corporations, Inc. We know that current credit spreads for investment grade corporations are approx. 80 bp (CDX IG) and for sub-investment grade approx. 450 bp (CDX HY). The weighted average for the market is thus 170 bp, or 1.70%. However, this is likely too high because we must subtract the far less risky government credits and GSEs included in the 73/27 split given by OCC. With all the estimating going on everywhere, I will reduce this rather arbitrarily to 150 bp, as representative of the entire "book" of corporate CDSs.
If you are still with me...bravo. There isn't much more left to go.
So, the annual GROSS insurance premiums changing hands for the entire corporate CDS market is 1.50% of $35 trillion, or $437.5 billion per year. Think of this as the annual premium paid on a contract to cover the risk of holding Global Corporations Inc. securities. The present value of such a contract for 5 years, discounted at 4.40% (the 30-year Treasury rate) is $2.3 trillion. If we view the CDS market as perpetual, then the present value rises to $10.5 trillion.
It is thus possible to argue that these sums represent the amount of risk that has been shifted from the global cash equity market onto the CDS market, thus making stocks appear less risky, or cheaper. Equivalently, we could say that total capitalization and P/Es for global markets are being understated by the equivalent amount. At the end of June, the total market cap of all major stock markets (i.e. stocks of corporations that are most likely to be covered by CDS) was approx. $47 trillion, therefore CDSs created an over-valuation of anywhere between 5% and 22% for stocks. While theoretically we could choose the 5% that represents the expiration of CDS insurance within 5 years, in practical terms it's more proper to view CDS as a perpetual market, adjusting instead for amounts outstanding.
The bottom line is that global shares may now be approx. 22% overvalued, due to the effects of the CDS market alone.
Again I will close by saying, please do take as many shots at this as you see fit. Comments can only help.
P.S. Correlation between S&P 500 and US credit spreads