The difference between S&P and Moody’s

By Felix Salmon
August 9, 2011
not Moody's, which downgraded the US. I

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Amidst all the downgrade talk, one crucial point has been largely missing: there’s a very good reason why it was S&P, and not Moody’s, which downgraded the US. It’s this: the two companies don’t measure the same thing with their credit ratings.

An S&P ratings seeks to measure only the probability of default. Nothing else matters — not the time that the issuer is likely to remain in default, not the expected way in which the default will be resolved. Most importantly, S&P simply doesn’t care what the recovery value is — the amount of money that investors end up with after the issuer has defaulted.

Moody’s, by contrast, is interested not in default probability per se, but rather expected losses. Default probability is part of the total expected loss — but then you have to also take into account what’s likely to happen if and when a default occurs.

The difference, as it applies to the US sovereign credit rating, is enormous. No one doubts America’s ability to pay its debts, and if the US should ever find itself in a position where it’s forced by law to default on a bond payment, that default is certain to be only temporary. Bondholders would get all of their money, in full, within a couple of weeks, and probably within a few days.

Contrast that with, say, some incomprehensibly complex constant proportion debt obligation, which makes all of its payments by dint of clever leverage games, and which, if it ever does default, does so with utter finality, and will never pay out a single cent again.

Let’s say, then, that Tim bought Treasuries in 2006, with their triple-A rating, while Chris bought triple-A CPDOs. They both had the same rating, but Treasuries were safer, and therefore had a lower yield. Tim was paying a premium for the liquidity associated with Treasuries: he knew that there would always be a willing buyer for them, even in the event of a default. And it’s conceivable that there was a tiny premium too for the fact that the recovery value on Treasury bonds is likely to be very close to 100%: if there ever is a default, investors will ultimately get back everything they’re owed. Chris, by contrast, knew that in the event that his CPDO defaulted, he’d get no money back at all.

All of those are very good reasons for Tim to pay more for his bonds than Chris paid for his. But all of them are explicitly ignored by S&P. S&P doesn’t put itself forward as some kind of investment-advice company: it takes no position on which bonds are good buys and which ones should be sold. All it does is try to rate credits on the basis of how likely they are to default.

Moody’s, by contrast, appreciates that bonds are investment instruments, and tries to build into its ratings the likelihood that investors will end up getting all their money back at the end of the day, rather than simply measuring how likely it is that there might be a default.

Here’s David Levey, for instance, the former managing director of sovereign ratings at Moody’s:

US Treasury bills and bonds, along with government-guaranteed bonds and highly-rated corporates, will for the foreseeable future remain the assets of choice for global investors seeking a “safe haven”, due to the unparalleled institutional strength, depth and liquidity of the market. Although there are several advanced Aaa-rated OECD countries with lower debt ratios and better fiscal outlooks than the US, their markets are generally too small to play that role. Since ratings are intended to function as a market signal, it makes little sense to implicitly suggest to investors seeking “risk-free” reserve assets that they reallocate their portfolios toward these relatively illiquid markets.

This is a very Moody’s thing to say, and is quite different from how the people at S&P think. S&P bends over backwards to try to say that it is not sending a market signal, and that a downgrade is not the same as a “sell” rating. Moody’s, by contrast, is a bit more realistic and appreciates that people use its ratings in the context of deciding which bonds to buy and which bonds to sell.

If Moody’s were to downgrade the US, then, that would indeed be an implicit suggestion that investors rotate out of Treasury bonds and into safer credits like, um, France and the UK. Which is a pretty silly idea. But S&P isn’t Moody’s, and so I think that Levey is wrong to say that S&P is making that suggestion.

Similarly, Nate Silver has a long post on “why S&P’s ratings are substandard and porous” which starts with the point of view of “an investor looking for guidance on which country’s debt was the safest to invest in.” That’s something you (purportedly) get from Moody’s; it’s not what you’re getting from S&P.

Silver also has a big problem with the fact that S&P ratings are more correlated with the Corruption Perceptions Index than they are with things like GDP growth or inflation, or debt. That fact, he says, “suggests that S&P is making a lot of judgment calls about countries.” Which, well, yes. Sovereign defaults are always political, rather than economic: if you looked only at macroeconomic ratios, then Ecuador should be investment grade, as would just about any other country which has recently defaulted and wiped out most of its debt. A sovereign credit rating is therefore primarily a function of a country’s willingness to pay, rather than its ability to pay.

Silver goes on to complain that credit ratings are a lagging indicator: upgrades and downgrades tend to lag the market, rather than anticipate it. Again, this is a complaint only if you think of the ratings agencies as being some kind of guide to help people beat the market. But they’re not. Sovereign upgrades and downgrades are big, important things, and the ratings agencies take their time over them — they’d much rather err on the side of caution and act too late than jump onto some wave of excitement and then regret doing so a few weeks later. That kind of activity they’re happy to leave to markets.

This is also the reason why S&P doesn’t much go in for multi-notch downgrades. Silver is right when he says this means that a country which has been downgraded to AA is a worse bet than a country that has been upgraded to AA: the former is much more likely to get another downgrade than it is an upgrade, while the latter is on an upgrade path and is more likely to get another upgrade than a downgrade. So they’re not exactly the same.

But the ratings agencies are very good at emphasizing that two countries with the same credit rating are far from identical in other respects. Again, S&P — and even Moody’s — would never say that investors should be agnostic when it comes to choosing between countries with the same credit rating. They’re just being cautious when it comes to their ratings moves, going slowly rather than quickly because that way they won’t make major multi-notch mistakes and they’re giving countries the opportunity to stop the deterioration in their ratings. The markets love to give an immediate verdict on creditworthiness: if you want that kind of thing, just look at bond prices or CDS spreads. Credit ratings are something different, which is a good thing.

Silver’s main thesis seems to be that the markets are a better guide to the markets than the credit rating agencies are. Which is true as far as it goes, but misses what it is that the ratings agencies in general, and S&P in particular, actually do. They’re a datapoint, not a financial advisor: ratings are more of a constant, in contrast to bond prices, which are highly variable. If you want a guide to bond prices, look at bond prices. If you want a guide to default probabilities, however, then the ratings agencies are still a good place to start.

More From Felix Salmon
Post Felix
The Piketty pessimist
The most expensive lottery ticket in the world
The problems of HFT, Joe Stiglitz edition
Private equity math, Nuveen edition
Five explanations for Greece’s bond yield
Comments
19 comments so far

True as far as it goes, but Silver found ridiculously simple models (debt-to-GDP ratio) that were better predictors of medium-range default than S&P.
So if S&P only predicts default probabilities and simple, back-of-the envelope calculations are better at calculating default than S&P, you have to wonder what value S&P brings to the process.

Posted by RZ0 | Report as abusive

Poor Washington Post information management:
1.Posted at 01:34 PM ET, 08/08/2011 Standard & Poor’s inconsistencies By Ezra Klein

[Standard Poor’s downgrade of the U.S. economy is beginning to look like a kamikaze mission. The credit-rating agency delivered its payload, but at the cost of destroying its own credibility.]

2.Posted at 11:55 AM ET, 08/09/2011 A little perspective on S&P’s downgrade By Robert Samuelson

[Next, let’s try to understand S&P’s $2 trillion “error” that’s received so much publicity. What did S&P do wrong? Well, the Budget Control Act — the legislation with the budget deal that raised the debt ceiling — specified that one category of spending (“domestic discretionary spending”) would grow at the rate of inflation over the next decade. By contrast, S&P’s analysts assumed incorrectly that this spending would grow at the rate of the economy (gross domestic product). GDP almost always grows faster than inflation, and the difference — when compounded over a decade — came to $2 trillion.

That’s a lot of money, even in Washington. This is a serious mistake, and once the Treasury pointed it out, S&P might have paused to see if its rating conclusion warranted rethinking. S&P didn’t pause, but it does have a rejoinder. Despite the revised spending, prospective budget deficits — the annual gaps between spending and tax revenues — remain so large that government’s debt burden keeps rising. It goes from 74 percent of GDP in 2011 to 85 percent of GDP in 2021. The basic trends, S&P argues, don’t change. (Under the original assumption, the debt-to-GDP ratio in 2021 would have been 93 percent.)

Is Klein on S&P’s “error” no longer operative, then? Does Samuelson’s explanation make sense on its own terms? Who knows.

Posted by ClaytonBurns | Report as abusive

Moody’s is also wholly owned by Warren Buffett, a longtime cheerleader of Obama. There’s your difference.

Whereas Buffett for years railed about US spending problems and the twin deficits, he completely did a 180 when Obama got into office and now pronounces America AAAA.

It is to be noted that he was financial advisor to Governor Schwartzenegger during the 2000s. California presently has one of the shabbiest fiscs of all states.

Posted by DanHess | Report as abusive

You have pretty much sanitized S&P’s role, Felix.

msnbc.com: Was S&P downgrade an act of revenge? …

[It's hard to view the monumental ratings downgrade in context without understanding the long-running feud between the government and ratings agencies. In April, Sen. Carl Levin, D-Mich., issued a scathing 650-page report contending that malfeasance at ratings bureaus like Standard & Poor’s was as much to blame for the housing bubble as any bank, and included a series of smoking gun e-mails that suggested that the firms knew they were profiting from unethical behavior. A little-known section of the Dodd-Frank financial reform bill also hits the rating agencies with new limits destined to undercut their lucrative business; the Securities and Exchange Commission is discussing right now just how to implement the new rules. The public comment period on new rules ended Monday.]

S&P is like Kaplan, a parasite on the system. It is striking just how much parasitism there is in American education and in wider circles. Apparently, precise measures of the opportunity costs of such practices are not in place.

Posted by ClaytonBurns | Report as abusive

Ratings are intentionally lagging indicators, and they are also path-dependent, as Moody’s research has demonstrated. Market-implied ratings are much more volatile. Ratings try to dampen the noise. Ratings are authoritatative and methodology-driven. They have very limited alpha.

Posted by nixonfan | Report as abusive

[Sorry, I inadvertently misposted this an earlier entry in Felix's blog. It properly belongs here.]

It’s clear Felix Salmon is getting desperate on his earlier lame defense of the indefensible S&P downgrade.

On the “econometric” side, Salmon now treats us to an explanation of how S&P does what it does in sovereign debt ratings which amounts to one of the best – though surely unintended – arguments for why S&P’s methodology is so one dimensional as to be, practically speaking, useless when it comes to rating sovereign debt. Why even bother when the methodology is so terribly flawed from an investor standpoint, especially since there can be undeserved consequences for a reckless performance, such as the one we have just seen.

On the political side, Salmon is either being disingenuous, or simply not a very perceptive reader of what Nate Silver wrote. A side-by-side comparison of what Silver wrote and what Salmon represents him as having written unfortunately strongly suggests the former.

Salmon writes:

“Silver also has a big problem with the fact that S&P ratings are more correlated with the Corruption Perceptions Index than they are with things like GDP growth or inflation, or debt. That fact, he says, “suggests that S&P is making a lot of judgment calls about countries.” Which, well, yes. Sovereign defaults are always political, rather than economic: if you looked only at macroeconomic ratios, then Ecuador should be investment grade, as would just about any other country which has recently defaulted and wiped out most of its debt. A sovereign credit rating is therefore primarily a function of a country’s willingness to pay, rather than its ability to pay.”

Silver wrote:

http://fivethirtyeight.blogs.nytimes.com  /2011/08/08/why-s-p-s-ratings-are-subst andard-and-porous/#more-14347

“What factors is S.&P. looking at when it rates sovereign debt? A country’s debt-to-G.D.P. ratio? Its inflation rate? The size of its annual deficits?

“S.&P. does look at each of these factors. But it also places very heavy emphasis on subjective views about a country’s political environment. In fact, these political factors are at least as important as economic variables in determining their ratings.

“For instance, the S.&P. ratings have an extremely strong relationship with a measure of political risk known as the Corruption Perceptions Index, which is published annually by Transparency International. These ratings have been the subject of much criticism because they are highly subjective, relying on a composite of surveys conducted among “experts” at international organizations who may have spent little time in most of the countries and who may instead base their judgments on cultural stereotypes.

“I don’t know whether or not S.&P. looks at these ratings. But the fact that the two sets of ratings are so closely related is troublesome. It suggests that S.&P. is making a lot of judgment calls about countries they have no particular knowledge about. Keep in mind that even when it comes to the United States, S.&P. made a $2 trillion error that reflects their lack of understanding of the way that bills are scored by the Congressional Budget Office. Are we to expect that they add value based on their perceptions of the political climate in Kazakhstan, or Cyprus, or Uganda?”

Is there anyone reading this blog who thinks that Salmon has not distorted what Silver said?

Notice in particular the deliberate – there is no other word for it — butchering of Silver’s sentence:

“It suggests that S.&P. is making a lot of judgment calls about countries they have no particular knowledge about.”

The whole point of that sentence was at the backend, where Nate says: “about countries they have no particular knowledge about.” Salmon leaves off the “they have no particular knowledge about” part and immediately segues into “Which, well, yes. Sovereign defaults are always political, rather than economic,” as if poor Nate had just said something so anodyne and obvious that any intelligent person would wonder why he said it in the first place.

Give it up Felix.

S&P shot itself in the foot on the econometric side, as Krugman et al pointed out,* and Salmon agreed, though he tried to trivialize this egregious error … and then take refuge in the validity – in his mind – of S&P’s political analysis.

And when Silver provided a powerful argument for why we should discount S&P’s ability in the area of political analysis as well, Salmon’s response was to grossly under- and outright misrepresent what Silver said on this point.

Very Republican of you, Felix.

Why not just admit “Hey, I sometimes make mistakes,” and move on?

*BTW, there was a similar pattern of misrepresentation in Salmon’s treatment of Krugman in one of his previous posts on this subject.

http://blogs.reuters.com/felix-salmon/20 11/08/06/the-credibility-and-integrity-o f-sp%E2%80%99s-ratings-action/

Salmon juxtaposed a quote from Krugman in which Krugman was excoriating S&P’s past econometric sloppiness – and current $2 trillion mistake — as disqualifying it from doing its downgrade with a quote from Tyler Cowen supposedly preemptively answering methodological criticisms such as Krugnam’s as being those of a “commentator … trying to muddy the broader issues.” Salmon – and, of course, Cowen – are free to disagree with Paul Krugman. But who except one of the autodidact trolls that frequent Krugman’s blog would ever accuse Krugman of “muddy”[ing], let alone not bringing into the discussion quite explicitly, the “broader issues”? In particular, I cannot imagine Tyler Cowen saying that Krugman was one of the commentators to whom his quote applied. Though that is exactly what Salmon disingenuously suggested by his juxtaposition.

Posted by billyblog | Report as abusive

Just goes to show that S&P don’t know what the hell they are talking about. The probability of a US default is basically zero for at least another 10-20 years (after which the probability is still absurdly low).

Posted by JasonDick | Report as abusive

If you are certain the US will not default, ever, then why did the Tea Party Congressmen take the action they did by taking things down to the wire over raising the debt ceiling? One has to assume that they were willing to default because they said so, and by saying so it meant S&P had to downgrade.

OR, it was a hollow threat, in which case why did they use it? The damage it has caused is real and has wiped trillions off World stock markets. Just for a bit of attention.

Posted by FifthDecade | Report as abusive

I don’t care what S&P says. Regulators say that they are a nationally recognized statistical ratings agency. If said regulators can explain why banks’ capital requirements should depend on the probability of default only and can explain why required capital should be changed with caution because it is better to have an undercapitalized bank than demand over capitalisation, then their decision to so recognize S&P might be defensible.

But they can’t so it isn’t. S&P is part of the club. Therefore regulations are written such that their pointless ratings are valuable. I can see no justification for letting Moody’s do one thing and S&P something completely different. It seems to me clear that Moody’s approach is the one with some regulatory relevance. Is there any reason to allow S&P to rate as it does and treat its rating as useful for Basel ? I can see the point of avoiding a duopoly, but I can’t see any reason to give the same legal force to ratings which are not the same at all.

I really wish that I were confident that “And it’s conceivable that there was a tiny premium too for the fact that the recovery value on Treasury bonds is likely to be very close to 100%” is a joke. Yes indeed, if most money is handled by traders with horizons such that eventual recovery is irrelevant (not in their annual bonus and won’t keep them from being fired) then are huge difference in expectable recovery ratios would make a tiny difference in prices. It sure looks that way.

This means that a patient investor who cares about hold to maturity or recovery losses can make excess returns by buying lower rated debt. I believe that this is true of corporate bonds in the USA (by rating and vs Treasuries and using Moody’s rating).

I think that many investors overlook the detail that recovery ratios are not always exactly zero (so it is considered a financial joke when a bank in Korea wrote CDSs on its own debt).

Investors are more short sighted than you imagine possible even taking into account that they are more short sighted than you imagine possible (note I am open to the possibility that actual traders are perfectly rational given their contracts).

Posted by robertwaldmann | Report as abusive

Gordon Brown does not mention S&P, unless indirectly. By its precipitate and poorly-thought out action, S&P has cornered itself. It is obsolete, an artifact of anachronistic thinking. Traditionally, America has paid the gentleman’s taxes: absurd college practices, parasitic habits in business, and an inability to orient to the future.

It is impossible to pay these taxes any more. The US Senate Banking Committee should take the lead in investigating S&P to help put an end to its miserable existence, and focus on how to initiate new international standards.

Opinions Washington Post
Global crisis calls for G-20 growth pact
By Gordon Brown, Tuesday, August 9, 4:55 PM

[...] [And if the problem is truly a global matter, are not many national conversations by definition incomplete and perhaps even self-defeating?

The world’s interdependence has once again been demonstrated in recent weeks. A loose statement anywhere can sink stock markets everywhere. [...]

Now, following a loss of confidence in the safest asset — sovereign debt — can the world come together to stop a new round of asset fire sales as liquidity in markets dries up? Either we cooperate globally, reforming the plumbing of our financial system, or the way banks and quasi-banks lend to each other will render monetary action ineffective, intensifying rather than resolving the crisis.]

Posted by ClaytonBurns | Report as abusive

OP-ED CONTRIBUTOR NEW YORK TIMES The Revenge of the Rating Agencies By JEFFREY MANNS Published: August 9, 2011

[...] [Meanwhile, the rating agencies have begun a guerrilla campaign of behind-the-scenes lobbying to weaken the commission’s efforts to carry out other parts of Dodd-Frank.

The S.& P. downgrade has elevated this simmering standoff to an overt clash. Politicians will be tempted to wave a white flag by granting the agencies a pass from tough regulation in exchange for the agencies’ not downgrading federal debt further. While that approach may give the United States breathing room in the short run, the government should not give in to such extortion.]

The subtext in S&P’s machinations is unsavory. Not only would many reporters rather not consider the subtext, they do not even want to bother reading the text. I have never seen so many blind byways in story coverage. Putting the microscope to the economic crisis as a text is productive. It might even work for CJR.

Posted by ClaytonBurns | Report as abusive

Why is everyone shredding S&P?. The key point here is that USA probability of default went up; therefore, they got demoted. Whether, default will occur or not is IRRELAVENT here! The default is highly unlikely and I believe would never occur, but the probability went up therefore they have the right to downgrade. How much more debt can a nation take….

Posted by Variation | Report as abusive

Strange that I do not think a country should rely on double-digit trillion dollar federal deficits. It reminds me that I should switch of calling astronomical numbers ‘economical’ numbers instead.

It does cost money to service the interest on the debt – and all projections are that the debt increases. They differ how much. What if the projections are incorrect and the debt increases larger than projected? Relying on some of the politicians to make the numbers meet is speculative business.

What is the off chance that the debt increases from 15 T to 20 T, in say four years? Everything happens and you want to have (some) cushion of flexibility. See LTCM, it did not work out as well.

Posted by jsmith6423 | Report as abusive

I’m surprised nobody is giving you flack on the following point, Felix: when securitizations went south, Moody’s was quick to say, clearly and loudly, that their securitization ratings were not intended to be comparable to plain bond ratings because they reflected only default probability. They abandoned their methodology just when it might have been useful. Inter alia, one can infer that this fact was not widely appreciated ex-ante. And a cynical mind might suppose that Moody’s junked their standard methodology because it was expedient in winning business.

Posted by Greycap | Report as abusive

Basically S & P or Moody’s or Fitch’s, or anybody else for that matter, should be able to issue ratings on anything they want to as long as it’s non-sovereign debt. If anybody wants to pay attention to them on non-gov stuff, that’s their business. But CRA’s should not be allowed to issue ratings on government securities just because they are CRA’s, maybe especially because they are CRA’s. Rating’s on all sovereign government securities should adhere to a strict set of guidelines established and universally knowable in advance. It doesn’t matter if they are an SEC recognized CRA or not. Right now all the CRA’s out there are flying in surreal space as far as their credibility is concerned, they blew it so badly in the run-up to the last finacial fiasco. Why the hell does the SEC get to “nationally recognize” CRA’s anyway? The SEC’s performance up to the meltdown is one of the most horrific examples regulatory oversight in all of history. Why doesn’t S & P rate their own performance and complicity in all this? We’ll see who is really full of BS then ..

Posted by Woltmann | Report as abusive

Interesting bit. A couple of points….

- If they aren’t roughly equivalent, then why do they usually match? S&P and Moody’s have similar looking ratings scales that, in my experience at least, most people treat as roughly equivalent. For example, the highest S&P rating (AAA) is considered more or less equivalent to the highest Moody’s rating (Aaa). This makes sense given that S&P and Moody’s most often give the same ratings to debt that they both rate. Split ratings (situations in which one agency provides a lower or higher rating than the other agency) do occur, but the vast majority of the time the ratings are the same. Given that split ratings are the exception, one might reasonably assume the ratings are roughly equivalent.

If Felix is going to explain the split rating on US debt obligations by asserting that Moody’s ratings include recovery rates and not just default rates, then how does he explain the fact that in most cases the ratings aren’t split?

- Recovery rates have more to do with general economic conditions than credit rating, per se. Both agencies publish annual default studies. For S&P it is the Annual Global Corporate Default Study and for Moody’s it is the Default and Recovery Rates of Corporate Bond Issuers. They used to be free but I think they’re charging for them these days. These studies present the average cumulative default rates by rating over the last 20 years or so. Moody’s also has a small discussion on recovery rates. Moody’s is very careful to state that recovery rates are much more highly correlated with general economic indicators than with the credit ratings themselves. Therefore, I perceive it is a common practice to assume a relatively similar (or even the same) recovery rate across all entities regardless of rating. So if you are Aa1, I assume a recovery rate of .45 and if you’re Baa1 I might assume a recovery rate of maybe .40. I assume a slightly higher recovery rate for the more highly rated entity, but not much. I should adjust those recovery rate assumptions in my model up and down based on general economic conditions. If that’s true, then the impact of embedding recovery rates in the Moody’s ratings isn’t likely to have a material impact on the ratings. Simple loss given default = Credit Exposure * default probability * (1- recovery rate). A 5% difference in your recovery rate just isn’t going to make a huge difference on a per dollar basis.

- Ratings are, in part, relative measures, not absolute measures. If Moody’s considers expected recovery rate in its credit ratings and recovery rate is highly correlated with general economic conditions, shouldn’t Moody’s ratings be more conservative than S&P’s given the weak state of the global economy? No, because the ratings are relative measures, not perfectly absolute measures. Moodys’ and S&P I’m sure have some standard guidelines about the characteristics of obligations at different ratings levels, but those guidelines shift over time. So it’s not a perfectly hard standard, to some extant it’s a relative representation of who’s the strongest and weakest in a given sector.

- Ratings are highly subjective. The ratings process is highly subjective. All kinds of guesswork and assumptions go into them that may or may not be true. Even though I personally think they do a good job on average, I have certainly seen them make some terrible assumptions in the energy sector, in particular with energy trading shops. My hunch is that the subjective nature of the rating process is likely to explain a much large portion of the variance in the credit ratings from S&P and Moody’s than the consideration of recovery rates.

Cee-Jay

Posted by Cee-Jay | Report as abusive

You are mistaken on what an S&P sovereign rating means. It is not a measurement of their probability of default. The academic literature handily refutes this for all of the major credit agencies. S&P’s revised “Sovereign Government Ratings Methodology and Assumptions (6/30/2011) makes it clear that their ratings “pertain to a sovereign’s ability and willingness to service financial obligations to nonofficial, in other words commercial, creditors.”. S&P now makes no claim to being able to determine the probability of default, just their assessment of how well and how willing the country is to meet its obligations. Worse still, the revised criteria actually downgrades a very strong predictor of sovereign default – the type of currency regime followed.

Posted by RGreg | Report as abusive

None of the rating agencies have any idea what the difference is between a AAA rating and a AA+. If you take a few minutes to look at the default statistics they publish it is clear that, no matter how long a time period you look at, there is no empirical difference. It doesn’t matter whether the rating is relative or absolute, or is assessing default or loss.

S+P’s decision to downgrade the US is no more than a hunch that it is somehow riskier than it was in the past. But you don’t need the rating agencies to tell you that.

Posted by BoringCanadian | Report as abusive

Why do you think if the US defaults that the holder of 30 year treasuries are going to 100% of their principal back? On what basis are you making this claim? You think the holders of this debt have call over what assets is exactly?

Chris on the other hand has call on the assets whose cashflow is his debt is securitised against.

Also how liquid do you think the 30 year off the run treasuries are? Remember S&P downgraded US long term debt not t-bills.

Posted by Danny_Black | Report as abusive
Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/