1. The description is not accurate, but the market is “unmoored from the fundamentals.” The market is unmoored from the fundamentals because there’s no really good way to price credit default risks. It should fluctuate inverse with unsecured debt, but it doesn’t.

The remainder of the description is inaccurate, however. A CDS buyer may be able to show a profit if the value of the CDS goes up, but this is not because they get cash out of margin payments.

And the description also assumes that CDS traders are able to coordinate their activities. Why did they pick Greece anyway? More likely, the Greek CDS price rose because the market accurately perceived increased risks of a Greek default, but the short-term price fluctuations might time larger and more wild than they otherwise would have been because of large-volume technical trading.

2. You are correct, the argument against requiring an insurable interest is that not requiring one brings liquidity into the market, making CDS’ tradeable instruments rather than insurance.

3. To close the loop, the perceived relationship between (a) requiring an insurable interest, and (b) “fundamentals” pricing, is inaccurate. Insurers often price based on the return they expect to receive on the “float,” expecting to break-even, or close to it, on the insurance policies themselves. Insurance prices are set on a market. As an example, consider the workers’ compensation insurance market in the mid to late 90s, in which insurers turned out to have been writing policies at far below cost for years, there were a wave of bankruptcies, and rates skyrocketed. These kinds of cycles in insurance are not uncommon. See also, auto insurance, life insurance, etc.

Another misconception. Many parties both write CDS protection and buy CDS protection on the same bond (making money on the spread), which increases the notional value of CDS contracts without increasing the outstanding value of the underlying bonds.

Moreover, it stacks a “deep” market on top of a “shallow one” (at least in the case where the underlying bond is a small-issue, e.g. a muni).

Right, though my use of “dwarfs” was meant to indicate that if there are *many* multiples in CDS of the underlying assets, the market is not limited to parties buying insurance/protection for their investments (which is fine, and basically uncontroversial), but is also speculative and highly volatile (everyone that works with CDS and related derivatives knows this to be true). I also understand that without netting the total exposure written and purchased by the same counterparty with respect to the same risk, you don’t learn a lot from notional amounts in a vacuum. However, an examination of the contracts written by AIG’s problem unit, Lehman, and others does point to a large and problematic speculative market, with insufficent transparency and regulation (i.e. none) regarding concentration of counterparty risk and sensible margin requirements or exchange clearance.

Thales: In my experience this is not true, and logically it can’t be true in cases where the amount outstanding of a bond is dwarfed by the notional amount of the CDS that are traded on such bond.

Another misconception. Many parties both write CDS protection and buy CDS protection on the same bond (making money on the spread), which increases the notional value of CDS contracts without increasing the outstanding value of the underlying bonds.

Indeed, exempting CDs and similar “insurance” from the gambling laws may have been a huge mistake.

“2. The author also overstates the value of an insurable interest requirement. In his example the hedge funds buying protection don’t have an insurable interest, but most parties buying CDS protection do have an insurable interest. For example hedge funds will often buy the bond and the protection to arbitrage the difference between the two cash flows.”

In my experience this is not true, and logically it can’t be true in cases where the amount outstanding of a bond is dwarfed by the notional amount of the CDS that are traded on such bond. Many institutions, for regulatory or other reasons, have an incentive to take a “synthetic” position on a bond, i.e. one where the underlying security is neither purchased nor sold. It is for this reason that so many CDS contracts have been written, and (for better or for worse) that feature of the market would disappear if an insurable interest were required.

Here are my guesses:

1. Yes. The key issue is the time horizons for traders without insurable interest vs. traders with. Why should a rational counterparty’s time horizon not shrink to match the time horizon of the average trader (who, without an insurable interest) has time horizons to short to internalize any substantial portion of the ex ante risk of default?

2. Yes and no. The grand dream of insurance derivatives was a market in which any kind of risk can be sliced up and sold off to the cheapest risk bearers. The requirement of an insurable interest does not, by itself, defeat that dream in theory. But the fact that insurable interest holders need insurance tells you something about their willingness to accept unconventional risk relative to other entities.

On my view, the combination of no requirement of insurable interest with the proliferation of unregulated derivatives led to instruments too complex to be valued accurately by even the largest financial institutions. Once you had a situation where Bear, Lehman, Citibank, and half the rest of the institutional investors in the world were buying and selling CDO^2s without worrying about the fact that a back of the envelope calculation showed that an order of magnitude more money was flowing into subprime than could realistically be serviced by such borrowers the problem was obvious. But stopping the train before it got that far down the tracks would have required a far more conservative institutional culture than has prevailed on wall street for some time.

The repeal of Glass-Steagall and the Volcker Rule are a good start on reform. But ultimately the aggressive types that get into managing wall street banks have to be given a compelling game to play that doesn’t put the entire world’s capital at risk. That should be another goal for reforming. Simply taking the marbles away won’t avoid another catastrophe.

I don’t see a problem inherently in this sort of risk-reallocation, and market participants who are purely speculating add liquidity, which has a positive value as well.

Interestingly, in this case it actually can decrease liquidity and increase volatility because the underlying security market that is being ‘insured’ is not deep enough. That is, if the value of the CDS’s on a particular debt are far in excess of that debt itself then a small movement of capital on the debt-market makes a much larger splash in the CDS market.

In particularly perverse cases, a party that holds a decent fraction of a shallow bond pool and a pile of CDS’ on that debt can start playing the market’s anxiety. When bad news come out they can pile it on by selling the bonds (taking a small loss) but pushing their CDSs way up. The same happens in reverse on days when the market dynamic looks good for the debtholder.

This only works, of course, if the bond market is not deep enough to iron out these wrinkles. IIRC, the total market value of all Greek sovereign debt CDSs outweighs the value of the actual debt by worse than 6:1 at this point.

I should add that I’m all for CDSs and CDOs under appropriate circumstances.

LN: And if your reply is that insurance regulators are better at policing reserves — do I need to remind you that AIG was an insurance company?AIG Financial Products was not regulated in the same way that AIG’s more mainstream insurance operations were. I believe that the holding company is not even an insurance company, so the overall institution was not regulated as an insurance company (although its insurance subsidiaries were in fact regulated like regular insurers).

That’s not quite true (and any way, how competent are insurance regulators if all you have to do to skirt them is set up a subsidiary?), but if AIG isn’t a good example for you then look at the monolines.

I don’t see a problem inherently in this sort of risk-reallocation, and market participants who are purely speculating add liquidity, which has a positive value as well. I do think, however, that counter-party risk isn’t sufficiently priced and disclosure isn’t particularly good on CDSs and derivatives, generally. There is definitely systemic risk created when the notional value of derivative trades exceeds the total capitalization of market participants. But that suggests the risk here isn’t on Greece (whose bonds are being speculated on), but rather on the institutions that are facilitating the speculation.

I’m not a trader but as I understand it, a CDS does not necessarily need to be linked to an actual tradable security which makes it much more useful as a way to hedge market risk. For instance, you will be pretty unlikely to find a CDS on a municipal bond issued by, for instance, Sunnyvale, CA. But you probably can find a CDS that is linked to a larger pool of California munis with about the same risk characteristics.

Requiring an insurable interest would be very difficult to enforce in a situation like this. If there is no trading asset that tracks the pool of munis the CDS issuer is tracking, what is an acceptable substitute?

In the case of Greece, the concern of bond holders is that the ability of the government to tax (and not to spend) is greatly imperiled.

AIG Financial Products was not regulated in the same way that AIG’s more mainstream insurance operations were. I believe that the holding company is not even an insurance company, so the overall institution was not regulated as an insurance company (although its insurance subsidiaries were in fact regulated like regular insurers).

Laura S.: Do CDS markets create harm? I don’t think so. Look, the mechanism of supposed harm is that A buys insurance on B. Then pushes B over. The trouble is that this is insurance only on default. The only reason B can be pushed over is that he’s already insolvent. i.e., debt is being issued to pay for existing debt. That’s a crisis. So I don’t think its far to say that B was pushed. B has already toppled in all but name only.

By that definition though, any financial company with less than 100% capital ratio (i.e. all of them) is permanently insolvent, so you’d consider them to have always “already toppled”. In fact almost all large companies are insolvent in this sense: their liquid assets can’t cover their outstanding debt, so they rely on a combination of it not being called in (i.e. banks hoping there is no bank run), and rolling over bonds in the capital markets when they come due (issuing debt to pay for debt). There are very few large companies that are able to cover all their outstanding debt with cash on hand. (A few cash-rich companies like Microsoft and Apple are exceptions.)

TCO: He’s basically right. It’s just a bunch of derivatives, a bunch of bookies. Now that may be ok…or it may be wrong. But it’s what it is. That’s why the bailouts were so moronic. who CARES if some derivative crapshooters lose money? Bailing out hedge funds and “banks” (and they were essentially trading companies, not FDIC concerns) was very silly. Much smarter to let it all unravel and let people take their haircuts rather than general taxpayers paying for New York financiers.

I think the problem with just letting the gamblers lose is that anyone who had money in money market accounts would have lost money, not just the gamblers. These banks and other financial entities were using off-balance sheet transactions, i.e. the shadow banking system, to borrow in supposedly safe, short-term money market funds in order to fund their unregulated speculation. I believe when all the moms and pops who had thought their money was safe in their money market accounts awoke to a huge haircut due to rampant speculation by the financial community using these “safe” funds, off their books and outside the regulated traditional banking system, there would have been hell to pay. And well deserved hell to pay.

Think everyone taking their money out of money market funds all at once when they realized they weren’t safe at all. The entire economy would have shut down and pitchforks out. In fact, I believe that is what started the crash, people starting to remove their money from money market funds when they realized they weren’t safe.

4. Later in the article, the author claims that sellers of CDS protection are not regulated to ensure they have sufficient reserves. This is a common misconception. One would not buy CDS protection from a seller without thoroughly inspecting their books, and being satisfied with their capital on hand. Sure, and insurance regulator isn’t policing their reserves, but the buyer of the protection is. And if your reply is that insurance regulators are better at policing reserves – do I need to remind you that AIG was an insurance company?

There is so much more I can write about why this is a flawed article, but I’ll leave this be.

1. The biggest flaw is the line, “Eventually the money flow will be reversed, when a bail-out is announced.” This assumes that the participants in the market are expecting a bailout, which is ridiculous. See e.g. Lehman.

2. The author also overstates the value of an insurable interest requirement. In his example the hedge funds buying protection don’t have an insurable interest, but most parties buying CDS protection do have an insurable interest. For example hedge funds will often buy the bond and the protection to arbitrage the difference between the two cash flows.

3. To blame wall street for Greece is laughable. International economists have known about the Greek problem for at least a few decades. Greece’s economy has fundamental problems that the EU has complained about for a long long time.

Oren__: Germans with their life savings in low-interest nationalized banks probably prefer 2.5%. The Greeks would love 6% to inflate away their debt. How you can have predictability in that environment is beyond me. 

Actually, with a predictable level of inflation at any level, no one’s debt is inflated away. The current ECB inflation target is 2%. For the reasons explained in Krugman’s blog post, that could easily be increased to 4 or 5%. There’s not much of a practical difference between agreeing on a target of 2%, and agreeing on a target of 4%.

That wouldn’t give any creditors a break, but it would mean that you could have nominal wage development ranging from 2% to 6% averaging at 4%, meaning that technically nobody has to take a pay cut, and the average still ends up at 4%. If the target is 2%, it is much more difficult to get there without anyone taking a cut. (That’s the Akerlof et al argument.)

Also, there’s this: There have been some discussions about whether a negative nominal interest rate is possible, and IIRC even some experiments, but effectively the minimum stays at zero. If your ordinary interest rate is a few points higher, there is more room for interest cuts.

The absolute level of inflation is much less interesting than the question of whether it is predictable. So I agree with Krugman that an inflation target of, say, 5% would be better.

Well predictability is fantastic but getting Europe to agree on a number would be worse than herding cats. Germans with their life savings in low-interest nationalized banks probably prefer 2.5%. The Greeks would love 6% to inflate away their debt. How you can have predictability in that environment is beyond me.

In practical terms, one of the problems of writing contracts in real terms is that it causes more inflation, since it causes more prices to be adjusted more frequently upward. That is why, IIRC, many countries used to ban contracts written in real terms in 1920s and 1930s. (i.e. Such contracts were unenforceable in court.)

Only in the presence of either huge economic growth or insane monetary expansion.

I can see it being an issue if they are unbalanced — merchants have to pay their employee in real dollars but their suppliers in nominal dollars, for instance. The answer there would be to hedge it out, of course. If real dollar contracts were, in fact, universal then prices would steadily go down since efficiency/productivity/technology are steadily increasing.

Oren: Finally, the quote about inflation is quite interesting. I suppose this is because wages are (foolishly) written in terms of nominal currency and not real currency. I’ve long been of the opinion that almost all contracts/statutes should be assumed to talk about real currency unless explicitly nominal (case in point, the 7A to the US Const). I digress….

The absolute level of inflation is much less interesting than the question of whether it is predictable. So I agree with Krugman that an inflation target of, say, 5% would be better.

Writing contracts in real terms is problematic because that raises the question of the most appropriate measure of inflation. Krugman uses the GDP deflator, which is my favourite as well, but that measure, like the Paasche and Laspeyres CPIs, and their progeny, has its flaws, not to mention that there is often no principled reason to prefer one over the others.

In practical terms, one of the problems of writing contracts in real terms is that it causes more inflation, since it causes more prices to be adjusted more frequently upward. That is why, IIRC, many countries used to ban contracts written in real terms in 1920s and 1930s. (i.e. Such contracts were unenforceable in court.)

Wall Street was complicit in this dishonesty, and for that I have mixed feelings. One should not put a stumbling block in front of the blind (as it were) but one should also not object to a sovereign government arranging weird financial deals either.

Finally, the quote about inflation is quite interesting. I suppose this is because wages are (foolishly) written in terms of nominal currency and not real currency. I’ve long been of the opinion that almost all contracts/statutes should be assumed to talk about real currency unless explicitly nominal (case in point, the 7A to the US Const). I digress….

Oren: They can cook the books in the short term but eventually will have to either pay up or default.

That’s too simple. What Greece can’t do, that it otherwise could, is compensate for its loss of competitiveness by devaluating the currency in order to boost exports and stifle imports, thus making it easier to “pay up”. Since they can’t do that, they’ll either have to go through an internal devaluation, or they’ll have to get the rest of the Eurozone to catch up with them, inflation-wise.

Oren: The plain truth is that the government of Greece needs to be put into protectorship of some sort because they are incapable of honestly writing their own budget.

They have been, to the extent possible. (In case you’re not familiar with diplomat speak, the relevant language is in the 4th paragraph, which essentially establishes a system of monthly Commission assessments.) The problem is that they can’t be put into protectorship by the people who are really their bosses, because the Greek people fully support this approach.

Every government cooks it’s books. It seems the Greek problem though is that they can’t manipulate their way out of their public overspending via the usual methods becuase [sic] they are trapped in the EU.

Trapped? Are you kidding me?

Every *other* government in the world must borrow at the prevailing market rate for their bonds. They can cook the books in the short term but eventually will have to either pay up or default. The latter, of course, will wreck their currency and any future lending prospects. See, e.g. Russia’s default in the late 90s. Greece is blessed with the fact that Germany is shackled to its corpse.

For insanely wealthy countries like Japan, they can coast for a long time in gradual decline. Japan is also still wildly productive as well, which gives the government the tax revenue to spend on the downside.

Actually, the original post isn’t really about the Greeks cooking the books, but about Wall Street cooking them. Greece cooks its own books as follows (from Afoe, today):

Well, I think Wall Street marketed particular financial arrangements to Greece that were designed to make the deficit look smaller. For instance, taking one-time payments for future regular proceeds (airport landings, lotto).

My first instinct was to say the Greek government is (metaphorically) mature enough to enter into voluntary financial arrangements however they see fit. On reflection, this is plainly incorrect — they act like teenagers who will do anything for immediate gratification and push off liabilities into the future. The plain truth is that the government of Greece needs to be put into protectorship of some sort because they are incapable of honestly writing their own budget.

SenatorX: Every government cooks it’s books. It seems the Greek problem though is that they can’t manipulate their way out of their public overspending via the usual methods becuase they are trapped in the EU.

Actually, the original post isn’t really about the Greeks cooking the books, but about Wall Street cooking them. Greece cooks its own books as follows (from Afoe, today):

In a fascinating article in today’s New York Times, journalists Louise Story, Landon Thomas and Nelson Schwartz begin to recount the mirky story of just how the major US investment banks have been able to earn considerable sums of money effectively helping European governments to disguise their growing mountain of public debt.
(…)
So the question naturally arises, just how much in debt are our governments, really? As the NYT team point out, Eurostat has long been grappling with this matter, and as far back as 2002 they found themselves forced to change their accounting rules, in order to try to enforce the disclosure of many off-balance sheet entities that had previously escaped detection by the EU, since up to that point the transactions involved had been classified as asset “sales”, often of public buildings and the like. Following advice paid for from the best of investment banks many European governments simply responded to the rule change by reformulating their suspect deals as loans rather than outright sales. As we say in Spain “hecha la ley, hecha la trampa” (or in English, when you close one loophole you open another). According to the NYT authors:

“As recently as 2008, Eurostat…. reported that “in a number of instances, the observed securitization operations seem to have been purportedly designed to achieve a given accounting result, irrespective of the economic merit of the operation.””

(…)

Now, at the end of the day, you may ask “what is wrong with all of this”? Well quite simply, like Residential Mortgage Backed Securities these are instruments that work while they work, and cause a lot of additional headaches when they don’t. I can think of three reasons why debt aquired in this way in the past may now be problematic.

a) they assume a certain level of headline GDP growth to furnish revenue growth to the public agencies committed to making the payments. Following the crisis these previous levels of assumed growth are now unlikely to be realised.
b) they assume growing workforces and working age populations, but both these, as we know, are now likely to start declining in many European countries.
c) they assume unchanging dependency ratios between active and dependent populations, but these assumptions, as we also already know, are no longer valid, as our population pyramids steadily invert.

Given all this, a very real danger exists that what were previously considered as obscure securitisation instruments, so obscure that few politicians really understood their implications, and few citizens actually knew of their existence, can suddenly find themselves converted into little better than a glorified Ponzi scheme.
(…)

For that matter financial companies also cook their books and I have to assume technically most are insolvent. The repeal of mark to market accounting rules and a return to mark to myth last year wasn’t for no reason. We have Japanese style zombie banks right now.

But it was a trap.

The total notional value of the CDS contracts sold on this bond was many times greater than the total amount of the covered bond. So the CDS seller had enough money that it could, as a seemingly disinterested third party, pay off the bond (preventing the expected default event) and walk away with a tidy profit.

The lack of an insurable interest requirement led to the CDS buyers getting ripped off.

EandJsFilmCrew: I think we see a much more objective measure of Greece’s public finances in CDS spreads than we’ve ever seen in the Greek gov’ts accounting statements.

True, obviously. But then, what’s wrong with simply looking at the bond spreads? Which, BTW, is by far the most commonly observed variable these days: the Greek-Bund bond spread. Like here, on A Fistful of Euros, or here in Martin Wolf’s FT column. CDS markets are just too all-over-the-map to be useful these days. (If ever.)

Laura S.: CDS market is well known among all participants to be at risk of manipulation. Mostly, this is because they are too small.

Who’s “they”? Surely not the market, which runs into the tens of trillions. (At least it used to.)

All that said, the notion in this particular case that “when we look behind CDS prices, we don’t see an objective measure of the public finances of Greece” strikes me as hilarious. Here we have a gov’t that for decades has falsified it’s books in exactly the same ways that AIG did, i.e. generating near term revenue by engaging in hugely risky long term derivative contracts, yet we are supposed to believe that the problem is unscrupulous market participants preying on the isolated members of the herd. What a joke. Greece would have long since been unable to roll its debt were it not for game playing like this. The roll of derivatives here has been to enable them to be irresponsible liars for longer than the otherwise would have been. I think we see a much more objective measure of Greece’s public finances in CDS spreads than we’ve ever seen in the Greek gov’ts accounting statements.

Suppose the total U.S. Mortgage market, approximated by what Fannie & Freddie cover, is about $5-$7 trillion. Then, even if 20% of those mortgages were to become completely worthless, the loss would be $1-$1.4 trillion.

But, because AIG and others could sell CDS to anyone, and because buying CDS was a lot like shorting the mortgage backed security market, lots of folks who thought there was a housing bubble wanted to go short, so AIG and others faced potential losses substantially exceeding $5 trillion.

Not that being able to see & measure short interest in an asset class is a bad thing, per se, but the absence of any insurable interest requirement, plus poor regulation, plus people underestimating the counterparty credit risk of even large insurers, plus failure to see the importance of short interest led to disaster.

I buy material from A and I like buying my material from A. B owes A money. If B goes down I risk not being able to buy my material from A, so I estimate the cause of that risk and ensure it.

But ‘not an insurable interest’ is meaningless, and leads to everything you see above.