I have a hazy memory of those exciting days in 2008 and 2009 when the world was going to be remade, shiny and new, with all of the risk gone from the financial system. The way we were going to get rid of all the risk, as I recall, was with the Volcker Rule and transparency around derivatives trading. That way, no shady prop trading of derivatives could blow up our financial system again, as long as you don't think too hard about the word "again."
Anyway, how'd it work out? Well, the New York Fed, who you'd think would have something to say about all of that, put up a note today about transparency in CDS trading. Their feelings about transparency in CDS trading can be summed up as "meh, we could take or leave transparency in CDS trading." Specifically:
Data on trading activity in the CDS market paint a mixed picture of the likely impact of trade reporting rules. The high levels of standardization of trading and contractual terms are apt to enhance the ability of market participants and policymakers to interpret the reported transactions. However, the low frequency of trading diminishes the potential price discovery benefits of real-time trade reporting.
So real-time trade reporting won't cause undue problems for dealers because most CDS is in fact a pretty standardized product that can be reported on a comparable basis. There's "an impressively high level of standardization of contract terms and market conventions" in single-name CDS, including things like standardized coupons, payment dates, and the fact that "47 percent of single-name transactions and 84 percent of index trades ... are in the five-year tenor."
But no one will care about that reporting, because there just isn't that much trading. This is based on a longer New York Fed paper from earlier this year, which we've mentioned before, but this graph is worthy of another look:
You could quibble with the data design (next graph I do will have three color-coded categories, "big numbers," "medium numbers," and "small numbers," and they will be interspersed randomly across the graph!) but let. it. go. Instead let's talk about the fact that for something like 1,200 of the 1,500-ish corporate CDS reference entities, the average trading frequency is less than once per day. If I want to buy CDS on FooCo and I see that it last traded a week ago at 300 and I go to my dealer and try to buy it at 300, he will tell me "yes, but CDX has widened 30 bps since then, and FooCo reported weak earnings, and look over there a dog just found a lot of counterfeit money*! Anyway, for you, because I like you, 450." And if I instead want to sell, he will say "yes, but basis is tighter and the sector is looking up and holy shit they ran that guy over with a car and a motorcycle! I can do 200." And I will be okay with that, because the reason he is a CDS dealer is because he's good at convincing me that he's got a great price in a nontransparent market.
But pricing transparency is not the only benefit of having standardized contracts that trade in a market that anyone can easily observe. Another one comes from Stephen Lubben's conspiratorial take on why Europe is so darn determined to avoid triggering CDS on Greek government bonds:
The really important question here about why the European Union cares so much about not setting off C.D.S. triggers in this exchange offer. The absurd lengths European leaders are going to in order to make this deal “voluntary” does raise a few eyebrows. And I have no really compelling explanations.
Still, would it be so hard to imagine that the E.U. wants to avoid setting off the C.D.S. because of aggregate exposure among European banks to Greek and other euro-zone sovereign debt? For example, what if the European banks have all been hedging their sovereign C.D.S. with each other – so that the German bank with the seemingly modest net exposure to sovereigns is really heavily exposed because the hedge is with a French bank?
So far so fair. Net notionals of Greek CDS are small, but gross notionals are $75 billion-ish, and while I'm in the sunny ISDA-can-do-no-wrong everything-is-collateralized-and-there's-tons-of-bilateral-netting permabull camp on net vs. gross, Lubben is not the only one who worries about what happens to people who bought Greek CDS from French banks. But he goes on into a bit of weirdness:
And let’s stop with the “Greek C.D.S. market is small” bit. Yes, the publicly acknowledged market is small. What about the bespoke market?
That link is to Lisa Pollack's heroic piece of detective work that has me, at least, convinced that the publicly acknowledged market is pretty much all there is. But even she doesn't know everything that could be known about Lubben's bogeyman:
What is a “credit derivative”? Does it, for example, have to be governed by an Isda Master Agreement to be defined as such? Where is the line being drawn, just so that we are completely clear? Will we one day find that there are instruments out there that look and smell like credit derivatives but are labelled as “guarantees” and hence not counted?
Sure, why not. There is real reason to worry that someone is writing a thing that is sort of a creepy credit insurance contract but is called something else. ("European Banks Get ‘False Deleveraging’ in Seller-Financed Deals." Hmm. Not exactly what we're looking for, but hmm.)
But if, like Lubben, you're just looking for a scary thing that you're not sure exists yet, there's no reason to believe that that thing should be triggered by an ISDA determination of a credit event. If I'm writing a bespoke contract to offload credit risk on bonds that I hold, that contract says "this pays off if my bonds lose money" - I don't care what ISDA decides. Similarly, if I'm writing a bespoke contract to take credit risk on bonds that someone else holds, it probably says something like "you can't do dumb shit to reduce the value of your bonds, like hand them in to Greece at 50 cents on the dollar." But in either case I'm not looking to an ISDA determinations committee. The ISDA DC is for people who want to avoid bespokery. If you're paying for bespoke, you want it to fit well.
Much of the history of finance is:
(1) identify a risk ("hey, we want to make loans, but we don't like keeping all the credit risk!")
(2) find a way to separate and hedge that risk via a bespoke product ("what if we get an insurance company to write us a contract on the loans?")
(3) standardize that product and roll it out to everyone ("what if everyone wrote each other insurance contracts payable on a standardized list of default events on the cheapest-to-deliver pari passu debt instruments of every issuer, with an arbitrary 5-year tenor and fixed coupon payments?")
Standardizing is good for dealers, because it lets them scale up volume; it's good for customers, because volume drives down price; and it's good for regulators, because they can easily understand what's on your balance sheet. It has some downsides, like the fact that not everyone has the exact same risks that are addressed by the standardized product. But once it's been accomplished, that's kind of it for the asset class. When the standardized product has enough scale, the benefits of doing a tailored trade don't outweigh either the transaction costs or the funny looks you'll get from regulators for doing something abnormal. Publicly traded shares of corporations with limited liability are awesome, so that's pretty much how companies sell equity. (Pretty much.)
CDS keeps being in headlines in part because it is right on the cusp of that breakthrough into standardization: it's all pretty much ISDAed, half of it is 5-year, most of it has fixed coupons, you could probably trade report it without too much bother. On the other hand, some of it doesn't have those standardized terms, and even much of what is standardized still trades by appointment. If you want to make a $100mm equity investment in Jefferies, and aren't Warren Buffett, then of course you go buy the shares that trade on NYSE just like everyone else. If you want to buy $100mm of CDS on Jefferies, you could go buy a standard ISDA contract (and you probably would), but it'd take some work because they don't trade all that much. So when you finally got someone on the phone willing to sell it to you, you'd be a bit more tempted to ask for something a bit bespoke: maybe customize the maturity to your bond holdings, say, or have the insurance trigger on an Egan-Jones downgrade.
And that's why we have a world where 99% of credit derivatives are reported to DTCC, and no one believes that. And where there's a standard mechanism for determining credit events for all reference entities, which is actually based on pretty clear contractual language, and no one is quite sure that it won't be manipulated to screw them. CDS, has many of the problems of a standardized, regular-way market: banks have bought a one-size-fits-all thing that may not protect them from some of their bespoke risks. But because it's still in the infancy of realtime reporting and complete standardization, it's easier to believe that it hides all of the scary features of an off-the-run, fully customized market where no one on the outside can ever know what risks might be hiding within.
How Might Increased Transparency Affect the CDS Market? [FRBNY/Liberty Street Economics]
Why Not Pull the Trigger on Greek C.D.S.? [DealBook]
* A tragedy in my life is that I can't link this, but seriously, look on page C2 of today's Journal, it's adorable. He's so happy! Because he found all that money!