A Dialogue Concerning Two Rating Systems

The following is a hypothetical conversation between Constance “Connie” de Boudinville, the dashing and attractive spokesperson for the US credit rating industry, and Axel Raskolnikov, an eastern European banker on a fact-finding mission in the United States aimed at improving the practice of banking in his native land by emulating the American model. The setting is October 2009 in sunny Miami, Florida, on the last day of the annual ASF securitization conference at which Ms. de Boudinville had just delivered the closing address, a sort of Platonic apologia for the rating agencies.
N.B. This post is dedicated to my former student and current colleague Max Rumyantsev

Axel: Madame de Boudinville may I have a few words with you?

Connie: Well, you could call me Baroness, but just call me Connie! I’m feeling good today; and after all, this is America!

Axel: Well, Connie then, I noted that, in your address, you failed to define what a credit rating really meant. Please excuse my ignorance, for I am new to the term and to the world of finance. Could you define what is meant by a credit rating in words a mortal like me can understand?

Connie: Actually, it’s quite simple. May I call you Axel?

Axel: ???????! Of course, I mean.

Connie: A credit rating is a measure of the probability with which a corporation or legal entity will fulfill its financial obligations, assuming it has any.

Axel: What would happen if a company had no debt at all and wanted a credit rating?

Connie: Well, if it paid us a fee, it would get a rating of course, but here and now, I can’t tell you what that would be.

Axel: But wait, if a company has zero debt, the probability with which it will fulfill its obligations is 100% by definition! In that case, its credit rating has to be the best there is, right.

Connie: I can see you’re new at this game Axel! That’s not exactly how the system works, but let’s just say you’re right for now.

Axel: So, let’s assume I owe you 100 Rubles, Dollars I mean, and the rating agency estimates that the probability of my paying it back is 90%, my rating will be 0.9 right.

Connie: Yes, that would be nice Axel but in fact, we are much less precise than that. Instead of using numbers everyone can understand, we use letters of the alphabet, like A, B and C.

Axel: You mean like in high school.

Connie: That’s right Axel, just like in high school!

Axel: So, I imagine that the best rating is ‘A’ then.

Connie: You’re coming along wonderfully well Axel. In fact, the best rating is not A, but triple-A; the second best is double-A, and so on.

Axel: But if the ratings are probabilities of repayment, why don’t you just use the numbers themselves? Wouldn’t it be much easier to understand than letters?

Connie: I know Axel, it’s weird, but most people in America prefer the letter system; it’s much easier to remember than numbers. There are only fifteen or so letter-ratings and most investors have no trouble with that number.

Axel: But wait Connie, if I owe you 100 Dollars on Friday and I can only pay you on Monday, that’s not the same as if I paid you on Friday, is it not?

Connie: Obviously not, because you might need that money on Friday to pay someone else on Friday. Money today is not the same as money tomorrow. No argument there Axel!

Axel: So, how can there be only one kind of rating then?

Connie: I never said that Axel. In fact, there are many kinds of ratings, and each of them means something slightly different. For instance, take where you work, a bank. Banks have financial strength ratings, short-term ratings and more, and one of them addresses the Friday issue you just mentioned.

Axel: Amazing! How does one measure financial strength then?

Connie: We use ratios of various financial variables, more or less the same ones people use to tell the government about their own accounts, and then decide what that means in terms of a rating.

Axel: OK, got it! Speaking of banks, how do you account for corruption inside the bank? In my country, the absence of corruption is a major sign of “strength”, if not financial, at least moral.

Connie: Axel, in America corruption is exceedingly rare, and when it happens, we usually find out too late anyway. Besides, we can only go by what other people tell us. It’s their job to tell us the truth about the company. We are not policemen Axel, only judges.

Axel: The truth, ah yes, that ingenious concoction of desirability of appearance. The truth is that the probability of payment can be any number between zero and one. Therefore, there could be thousands of credit ratings, not just fifteen. If you use letters to express ratings, some companies will have different payment probabilities and yet have the same letter rating, and unscrupulous people might take advantage of that misinformation to make money. I think they call it “arbitrage” or something, but I wouldn’t pretend to teach you French Connie!

Connie: Axel, you wanted to know how it works, please leave the truth alone!

Axel: I’m sorry Connie; I just get carried away. I won’t do it again, promise! How do we know credit ratings actually work Connie?

Connie: Finally! It’s quite simple Axel: we have bond defaults studies!

Axel: ??? ???? What’s that Connie?

Connie: Well, that’s the proof that credit ratings really do work. Over the last 100 years or so, we have accumulated data on the default of corporations in the United States, so we now know what the ratings mean in terms of default frequency.

Axel: Default frequency? Please explain Connie.

Connie: OK, here we go Axel. Instead of expressing the rating as a payment probability, we use the non-payment, or default probability, because the numbers are much more manageable that way. You see, default probabilities are very small anyway, so it’s easier to quote them than the corresponding payment probabilities that would all be essentially equal to one. No one would see any difference between all these ratings, and we can’t have that, now can we Axel?

Axel: What’s a default Connie?

Connie: Basically, a default is a bankruptcy, although we also include “mini” bankruptcies and special situations, and call them defaults too.

Axel: You mean like what Mother Russia did in 1998?

Connie (laughing): That was a scandal Axel, not a default. OK, it was a default.

Axel: So a hundred years ago, rating agencies did not know what ratings really meant since they did not have any default data. What did people do then?

Connie: I don’t know Axel, but somehow the Republic survived!

Axel: So then, rating agencies only have ratings for the United States right. You said that the default data were for US companies only.

Connie: Actually Axel, we have ratings for most large countries and most large companies in the world.

Axel: But how do you know that the default probabilities associated with those ratings are the same as those for the US companies of the same rating? Life in Europe is quite different from life in the US Connie; you know that.

Connie: It does not matter Axel, because we are the worldwide arbiters of credit. As long as we say it’s OK, it’s OK Axel. See what I mean. One day, we will know what these European ratings really mean, but in the mean time (please excuse the pun) there can only be one referee in any game. Can you imagine a soccer match with two referees?

Axel: No, I really can’t. However, I can imagine a biased referee in the game Connie. Many European companies are now competing with US companies and if the default frequencies of two companies are in fact the same it wouldn’t be fair to give the US company a higher rating than the European one, now would it?

Connie: No it would not, but we are fair, Axel, that’s why we’re in charge.

Axel: This is great Connie! May I trouble you some more?

Connie: Of course Axel, the rating business is an open book.

Axel: I also noted that you spoke of senior secure and senior insecure ratings. What are these things Connie? Where I come from, everything is insecure.

Connie (laughing): I said “unsecured”, not “insecure” Axel. This means that corporations can have securities with different priorities of access to cash should they go bankrupt, and we feel that investors who own the securities with the higher priority will lose less of their investment than those who own those with a lower priority. We assign a lower rating to the second type of securities to reflect that belief, which belief, I might add, is also verified in practice.

Axel: Loss? You never said anything about loss before. I thought credit ratings were about default frequencies, not losses. Even if I lose less than the next guy, the company still defaulted, right Connie?

Connie: But how then are we going to tell investors that they will lose more money if they invest in unsecured than in secured bonds? We have to make a difference, Axel dear.

Axel: Certainly, Connie dear. However, the place to introduce the concept of loss is where it makes sense, i.e. not in the definition of default frequency but rather in your estimate of recovery. You should simply state that unsecured investors would experience higher loss upon default than secured investors, but that the ratings are the same since the default frequencies are obviously the same. Right now, people who price loans using the current system are clearly overpricing since they are double-counting the reduction in recovery.

Connie: Touché Axel. I’ll have to bring this up at the next steering committee meeting. Tell me Axel, would you be available to come out and tell our management that they have been wrong for the last hundred years?

Axel: Now, now, you wouldn’t want me to subvert the American way of life Connie! Communism is really over, you know!

Connie: Don’t worry Axel! Americans truly respect honesty, especially in foreigners.

Axel: So how does it work Connie? I need a rating, so I go see a rating agency, pay them a fee and get a rating. Is that it?

Connie: That’s about it. See how simple it is!

Axel: What if I don’t like it? Can I pay more and get a better rating?

Connie: Axel, this is your second warning! One more and you’re on watch! Of course not! If you could do that, everybody would do it. Then, credit ratings would lose their meaning and the country would be in utter chaos.

Axel: That’s amazing Connie! In my country, everything has a price. I can see a situation where someone would borrow money from the government to look good for the rating contest and then give the money back afterwards. This way, they would get the rating they want.

Connie: That would not work Axel, because as soon as they gave that money back to the government, we would lower their rating.

Axel: But wait a minute Connie; are you saying credit ratings can change?

Connie: But of course Axel; they can change at any point in time.

Axel: Do they?

Connie: Actually, they don’t because most of the time, corporations don’t borrow money from the state, or from anybody else for that matter, just to get a credit rating. If they did, that would be fraud, and that’s illegal. Most of the time, their financial condition is like our method, an open book. Credit ratings are more or less the same all the time. Of course, every now and then, something awful happens and we have to lower the rating. We call it downgrading. Or else, something very good happens and we can raise it; we call that upgrading. Credit rating is quite a sophisticated business Axel, one where angels fear to tread.

Axel: But things happen all the time in this world Connie, so ratings should really change quite often. Why don’t they?

Connie: Well that’s just it Axel! Our clients: investors, corporations and bankers that is, really don’t like it if we downgrade too quickly, or upgrade too quickly for that matter. They say it creates uncertainty because companies borrow money at rates largely determined by their credit rating. Can you imagine if ratings could change every day? CFO’s would never get any sleep! Unless things are truly bad out there, we leave it alone. Besides, if we had to monitor all the companies we rate our analysts wouldn’t get any sleep either! This way, everybody is happy.

Axel: Yes, but if you downgrade only if things get really “bad”, as you say, everyone will be taken by surprise, except of course those that already knew it. How can you say that a default probability has suddenly increased tenfold overnight? Is it not likely that all that bad “stuff” was there all along for everyone to see? If you slowly reduced the rating as things unfolded over time, would this not be a better way to tell investors about it. In fact, would this not avoid having to do the very thing you seek to avoid, i.e. a catastrophic downgrade?

Connie: Perhaps so Axel, but you have to pick you poison. Either a credit rating is a long-term view of the company, or it is a current view of the same company. It can’t be both Axel. As long as everybody knows what we do, it’s really their fault if they misjudge the rating! Moreover, as I explained to you ten minutes ago, credit ratings really do work. On average, they really express the default probability of obligors. It’s not our fault when some investor out there owns the one corporation that actually defaults.

Axel: That’s strange, because in my country, people don’t care if someone else loses money, only if they lose money. Wouldn’t it make more sense to tell people that they are about to lose money rather than telling them that they are just fine on average? To say the least, it seems to be misleading. Could you not provide timely ratings instead of long-term ratings? No one really cares what happens on average, if they are dead now!

Connie: Axel, this is your last warning! Don’t you remember what the greatest writer of all time, Voltaire, once said: the more things change, the more they stay the same! Well, that’s our policy in credit ratings Axel. That’s all I have to say.

Axel: I do remember Voltaire, but the greatest writer of all time was Dostoievsky, not Voltaire! Joking aside, you mentioned something about structured finance ratings in your talk. What is structured finance anyway?

Connie: Glad we changed the topic Axel! Structured finance is a way for people who can’t get money to get it anyway. You can imagine how popular that might be!

Axel: And how does that happen? We’d really like to use it in Russia if possible.

Connie: It all starts with a special purpose company or an SPC as they say. Then, you sell your assets to the SPC and <span style="text-decoration: underline;">it</span> borrows money, not you. This way, the people that give you the money don’t really care if you go bankrupt, since they have your assets locked up in the SPC. They get comfortable with the situation and let you have their money.

Axel: So, all I need is an SPC and assets for sale?

Connie: That’s right Axel! However, make sure you really sell those assets, and not just pretend to sell them. Otherwise, it’s not structured finance but corporate finance.

Axel: That should be easy. It must be obvious when you are selling something, is it not?

Connie: In fact, it’s not Axel; and that’s why we’re on top of things in that department. Darling, don’t you now see how useful rating agencies are?

Axel: I have so much to learn Connie. What about structured ratings then?

Connie: Structured ratings are different from corporate ratings because structured securities are specifically constructed to have a default frequency of zero, so we can’t give them normal ratings now can we.

Axel: If they all have the same default frequency, i.e. zero, what do these ratings really mean then?

Connie: Remember the loss discussion from a few minutes ago? That’s the key. Structured ratings are loss estimates, not default estimates, and we express that loss in basis points, which is banker-talk for percentage Axel. Unfortunately, some of us in the industry are still not convinced and insist on using default frequency as the measure of structured ratings. I’ll have to talk to them about that some day (sigh).

Axel: But how can this be, isn’t everybody looking at the same deals?

Connie: Yes Axel, but that’s politics. Please don’t tell me you don’t know about that!

Axel: Politics is one thing, money quite another. One of these people is clearly wrong. Can’t the other guys talk to him and work out a uniform definition of structured ratings?

Connie: Actually, it’s more or less already happened. As far as I know, everybody now looks at losses.

Axel: Thank God Connie! You scared me for a split second! Come to think of it though, it wouldn’t be so bad after all to have two incompatible definitions. We Russians have lived with a split-brain for centuries and we are doing just fine, right?

Connie: Yeah, right!

Axel: Well, if structured ratings don’t mean the same as corporate ratings, there is no reason why they wouldn’t change all the time right, or is Voltaire making magic here too?

Connie: Yes, I’m afraid that most illustrious of world scholars is right on again! Structured ratings are different from corporate ratings in their definition, but they are still supposed to remain the same forever. Just like for corporate ratings, if we find out something nasty is going down, we can downgrade the bonds. In fact, even if we only believe something is going to happen, we can take preemptive action, just like the US army Axel!

Axel: But how do you find out what people are up too if you don’t monitor the deals?

Connie: We do monitor the deals Axel! However, we are not responsible for other people’s crimes, just our own. If we know what’s going on, we adjust the rating for sure. Otherwise, our policy is the same as the army’s: don’t ask, don’t tell.

Axel: But if structured ratings are loss estimates, isn’t it true that one knows more about losses four or five years after closing than one year after, and so the rating should change to reflect that increased knowledge, should it not? After all, deals can also get better, not just worse. How does it go again? The sun sets, but the sun also rises!

Connie: I see you read a lot of books. That’s quite unusual for an investment banker. Axel really! Even if we did know more about a deal, and I’m not saying we do, what’s the difference? After all, investors are still in as good a shape as they were four years ago. There’s nothing new in that. Who cares if the rating is really better than before? You can’t lose less than zero right. If the issuer has too much capital in the transaction, big deal! We are an investor-service Axel, not an issuer-service.

Axel: You do have a way with words Connie! If structured ratings are already different in one respect, why can’t they be different in another? Why can’t structured ratings be dynamic instead of static?

Connie: Because we are already on record as saying they don’t change. Don’t you hate it when people keep changing their mind? I surely do. This way, everything is just fine. Of course, every now and then we make a mistake. To err is human Axel, and to forgive divine, remember that!

Axel: I agree with you Connie. Consistency in spite of overwhelming evidence is truly a virtue. What I now see so clearly is that the problems of two people like us don’t amount to a hill of swaps in this crazy world! If they change that rating, they’ll regret it, maybe not today, maybe not tomorrow, but some day! This way, they’ll always have plausible deniability (intense gaze)!

Connie: Axel, this could be the start of a beautiful friendship (longing look)!

Axel: But wait Connie, if rating agencies believe structured ratings are dynamic, then what’s the problem telling investors the truth?

Connie: It’s not about the truth Axel, it’s about deal flow. At any rate, before one can tell the truth, one has to know it!

Axel: Amazing! It’s starting to look more and more like Russia. My boss is going to love this!

Connie: You’re quite a fast learner dear. Make sure you let me have your resume when we’re done.

Axel: Thanks Connie! By the way, I couldn’t help reading about what those horrible people at Enron did! Apparently, they used structured finance to get money they shouldn’t have. However, you said that’s what structured finance is all about Connie. So what did they really do wrong?

Connie: You put your finger on the problem Axel! What they did was not structured finance at all, but something that was made to look like structured finance. Investors were all fooled because structured finance is really a good thing. These wretched individuals used the good name of structured finance to get money from unsuspecting investors. That’s what went wrong!

Axel: Isn’t there something that could have been done to see what was going on, like following the rules of structured finance for instance?

Connie: That’s easy to say Axel! Have you ever tried to find out what is going on inside a small company, let alone a big one? We can do nothing about fraud Axel! We trust that what accountants are telling us is true. If it’s not, we are off the hook anyway, and we simply move on. Of course, there’s a bad apple in every barrel Axel, but we’re still the only game in town. In fact, we are protected by the first amendment to the US constitution because we are only offering our opinion, nothing more. Like ABC News or something! We have no liability for that opinion. If we mess up, we just say buyers beware and that’s it. It’s a very solid business plan.

Axel: It’s getting better all the time it seems. Coming back to Enron, were the rating agencies not watching these bad people in action Connie? If it was not structured finance, and the rule makers of structured finance, i.e. the rating agencies, were reviewing the deals, then how could they make it look like it was?

Connie: But Axel, structured finance is complicated nowadays. We are not Gods, although you might think we are. In many cases, our people are simply overwhelmed and just can’t cope with these complex structures.

Axel: What do you mean by a structure Connie?

Connie: A structure can be many things but essentially, it’s a set of rules that determine who gets the cash that comes into the deal. Much like a financial “auto pilot” you might say. It works like Aeroflot, except that these structures don’t crash all the time. Come on Axel, I was just kidding!

Axel: That’s great! Because Russians know a lot about rules, in fact, we practically invented the rule business. You know, like chess and all! Wouldn’t it be easier to cope with all these rules by putting them inside a cash flow model and letting the chips fall where they might. That way, you wouldn’t really have to worry about how complicated the interactions between the rules might be, for they would simply come out of the wash naturally. All your analysts would have to do is keep a stable of models for the various types of deals associated with the assets, and voilà! Isn’t this a better way to go?

Connie: Perhaps it would be in the end Axel, but as surely as I’m standing here today, it would also destroy the romance of the old ways, when a seasoned analyst could simply eyeball a deal and guesstimate the required enhancement at a glance. Everybody’s just talking about the good old days, but at the rating agencies, we are truly living them. Don’t you feel like our society has lost too much of its charm with all that financial wizardry? Enron is just another example of why we should move backward, not forward.

Axel: Yes, technology does dehumanize Connie, but in my opinion, losing money is much worse. On the other hand, I freely admit that romance is hard to give up.

Connie: Axel dear, don’t you think we’ve talked shop long enough? Why don’t you buy me a drink, and then maybe you can show me your e-bit, da?

Axel: Ahhh Constance ma chérie! Détente can be beautiful. ?????!
Enjoyed this very much. Not sure I understand the pun at the end. I mean, I know what EBITDA is, but seems like shes going for a double meaning? What is the double meaning? It hints at being something, shall we say "dirty" :D