Streetwise Professor

Podcast: Play in new window | Download

Subscribe: RSS | More

There are rumors flying around–all adamantly denied–that JP Morgan Chase will acquire tottering Bank of America.  In my opinion, that’s crazy talk.  It’s crazy because JPM has to be smarter than to buy a toxic loan dumping ground and lawsuit magnet.  (There’s a joke going around: “Will the last person to sue Bank of America turn out the lights?”)

The only way it would not be crazy in the sense of it might actually happen is that another shotgun marriage is being arranged, with Uncle Sam at the trigger.  I can see that.  This is a tried and true regulatory game.  It happened in 2008: Bear (with JPM), Merrill (BAC), Lehman (with Barclays, ultimately aborted).  It happened during the S&L Crisis.  Regulators like this because offering financial support to dodgy banks through the front door is politically radioactive, and that’s especially true now.   Mashing together a (relatively) healthy bank and a very unhealthy one kicks the problem into the future, and miracles–including resurrection–can happen in the future, dontcha know.  Moreover, regulators can reduce the risk of leaving fingerprints if they do it this way.  If it turns out bad, it looks like the dealmakers made the mistake (witness BAC’s Ken Lewis).

But it would still be crazy in the sense of economic irrationality.  The first order problem that needs to be addressed right now is rightsizing banks, not creating fewer, bigger ones.   Moreover, the analogy that I made with respect to Europe–amputation vs. gangrene–works here too.  Grafting the definitely gangrenous Bank of America on top of the still healthy (in comparison to its peers around the world) JP Morgan would extend the rot far more directly than would be possible along the normal inter-bank connections.  It would be better to ringfence, isolate, quarantine Bank of America and deal with the problem straight up (like the Swedes did in the early-90s, or the Resolution Trust Corporation did with S&Ls in the same time frame).  But that would be politically parlous, which means that the sensible thing might not happen.

I believe–and hope–that this rumor is only that, the product of heated imaginations dealing with possible resolutions of BAC.  If it isn’t, if there is there there, be afraid.  Be very afraid.  Because it will mean that things are very bad indeed, and that the political and regulatory will does not exist to deal with the problem in a rational, hardheaded way.

I was incorrect to blame the leak of confidential CFTC trade information on the agency itself.  I jumped to a conclusion, and that was wrong.

The malfeasor has been revealed.  It is US Senator Bernie Sanders (Socialist–really–VT):

Oil trading data that exposed the extensive positions speculators held in the run-up to record high prices in 2008 were intentionally leaked by a U.S. senator, sparking broader concern about industry confidentiality as Congress moves on Wall Street reform.

Senator Bernie Sanders, a staunch critic of oil speculators, leaked the information to a major newspaper in a move that has unsettled both regulators and Wall Street alike.

In a June 16 e-mail reviewed by Reuters, a senior policy adviser to Sanders discusses how his office received private data with the names and positions of traders and forwarded it exclusively to a Wall Street Journal reporter.

The e-mail, which also attaches two files with the data, was sent to Public Citizen’s Tyson Slocum asking him to review it and speak with the newspaper about his observations.

In a statement from Sanders provided to Reuters, Sanders said he felt the data needed to be publicly aired.

“The CFTC has kept this information hidden from the American public for nearly three years,” he said. “This is an outrage. The American people have a right to know exactly who caused gas prices to skyrocket in 2008 and who is causing them to spike today.”

No, Senator, the outrage is that you violated confidentiality to engage in political grandstanding.  Moreover, you provide no evidence to support your assertion that the firms and individuals whose identity you revealed “caused gas prices to skyrocket.”  None.

There you have it, ladies and gentlemen.  The Senator from Vermont, providing additional proof, as if any was needed of Mark Twain’s dictum that “America is a nation without a distinct criminal class…with the possible exception of Congress.”

Today’s WSJ has an article that lists the positions of several market participants in the oil futures market on 30 June, 2008, shortly before prices hit their all-time high  This information was leaked by somebody in the CFTC.  A CFTC spokesman “said the agency couldn’t comment because the list contains confidential information.”

You don’t say.  [I cleaned that up.]  Apparently somebody in CFTC didn’t get the memo.

This is egregious.  Truly egregious.  This is highly confidential information.  Its disclosure was clearly intended to advance some political agenda.

Let me hazard a guess what is driving that agenda: next month the CFTC is likely to move on its energy position limits proposal.  The disclosures about the positions held by financial players will no doubt help shape the battlefield for that fight.

I cannot emphasize enough what an abuse this is.

And it will hinder the CFTC going forward.  The agency often relies on voluntary cooperation by market participants to supply data that helps it understand what is going on in the market, or to help develop policies.  Is there any doubt that companies receiving such requests will now think twice about cooperating?

And don’t you feel so much better that under Frank-n-Dodd that the CFTC will have access to far more data from far more markets?

I just shake my head.   And speaking of heads, somebody’s should roll over this.  Maybe several somebody’s heads.

I was quoted in this WSJ/Dow Jones article about why sovereign CDS are not being cleared.  I think that legal risk has something to do with it.  But upon further consideration, I think the problem with sovereign CDS is far deeper than that.

The problem is not exclusively with clearing vs. bank intermediated-bilateral.  The problem is with wrong way risk which makes it very dangerous to have a bank as the counterparty selling protection in a sovereign CDS.  Clearing sovereign CDS through a bank-dominated CCP means that banks are the ultimate counterparty, and if the protection seller defaults there is a possibility that banks will be on the hook to pay out on the defaulted contracts precisely when they are in financial distress.

The recent crisis in Europe demonstrates that with current monetary arrangements, the default of any sovereign name of any size can trigger a financial crisis that threatens banks.  Note that bank CDS spreads tend to blow out whenever one of the PIIGS looks shaky. I don’t have the data at hand, but I would wager there is a very high correlation between Greek, or Portugese, or Spanish, or Italian CDS spreads and bank CDS spreads.  Which means that there is a substantially elevated risk that a bank or bank-backstopped counterparty (i.e., a CCP) won’t be there if a sovereign defaults.

There are structural reasons for correlation.  Banks tend to hold large amounts of sov debt in their portfolios.  A surge in sovereign default risk tends to lead to funding problems for banks–especially those with large sovereign exposures.  Thus, sovereign debt crises are associated with substantially elevated risk of bank failure.

This is actually a firmer basis for governments and regulators to be skeptical about–and perhaps even hostile to– sovereign CDS than their usual kill-the-messenger bleats about evil speculators.  Sovereign CDS are potentially beset with wrong way risk, and they pose a real systemic threat to banks if the banks sell protection: in contrast, banks buying protection makes good sense as a hedge if the counterparty is likely to be able to perform in a sovereign debt crisis.  The risk is right-way when banks are protection buyers.

The issue with clearing sovereign CDS is that clearing makes banks contingent protection sellers, and the contingency kicks in right when banks are least able to afford it.  But even a bank running a balanced book would still pose problems because if a counterparty from whom it bought protection defaults, it would be obligated to dip into its capital to pay off the parties to whom it had sold protection, and again precisely when it is least able to afford it.

This suggests that there are acute systematic risks inherent in sovereign CDS under both bilateral and cleared approaches.  Since we have seen that the potential default of any European sovereign name, even those of middling size, can precipitate financial panic, sovereign CDS are arguably the most problematic derivatives out there from a counterparty risk/systemic risk perspective.  They are likely beneficial as long as banks are not major protection sellers under any contingency.  That’s impossible to achieve with clearing, and likely very difficult to achieve in a bilateral market.

What this suggests to me is that (a) it is unwise in the extreme to clear sovereign CDS, and (b) banks should be subject to a hefty capital charge on their gross sales of sovereign protection.  Gross, not net, because the gross drives their contingent exposure.  Low capital charges on a positive net position in protection purchases would be defensible as that risk is right way and net protection purchases gives banks an incentive to lay off sovereign debt risk on those who can bear it at a lower cost than they can.

Looking at the issue of sovereign debt risk more broadly, this analysis again raises serious alarms about the wisdom of mandating or incentivizing banks to hold large sovereign debt positions, where the incentivizing comes in the form of liquidity requirements that must be met by sovereign debt holdings and favorable capital charges, and the mandating comes in the form of margin and clearing requirements.  But the financial repression pressures driven by the pressing need of governments to get somebody to hold their bonds is driving us in that direction.

Which suggests to me that the systemic risk of the future–and the systemic risk from hell–is the prospect for chronic fiscal crises around the world.

And on that cheery note . . . back to work.

Michael Giberson at Knowledge Problem has an excellent post that illustrates perfectly three fundamental problems with the metastasizing regulatory state.   EPA rules on air toxins and mercury are going to force–or at least accelerate–the closure of large numbers of coal-fired power plants especially in the East and Midwest.

Problem One: making uncoordinated policy decisions that have myriad impacts in tightly connected systems.  The EPA decision focuses on one thing: air quality.  But its chosen remedy has the potential to cause major disruptions in areas outside of EPA jurisdiction, considerations that EPA did not take sufficiently into consideration.  Specifically, the loss of such large quantities of baseload generation threatens electricity supply, and electricity reliability.  The stability of the grid depends crucially on the spatial configuration of load and generation; those have to be balanced tightly in order to ensure reliability.  Shutting down just one plant can upset that delicate balance in a way that greatly increases the system’s vulnerability.  So in its zeal to do good, EPA has created a huge potential for bad.

Problems Two and Three have their roots in EPA’s response to Problem One.  EPA says: “Don’t worry!  We’ll use our discretion to try to mitigate these bad effects!  We might give waivers!”

As Richard Epstein pointed out, Government by Waiver has at least two serious defects.  First, it is an invitation to rent seeking, lobbying, and politicking–all just different flavors of corruption.  Resources are wasted in competing for favors.  And favors are allocated to those who have a comparative advantage in politicking, not those who have a comparative advantage in producing or consuming efficiently.

Second, it creates substantial uncertainty.  Investments in generation and transmission (which can be both a complement to or a substitute for generation) have very long lives.  The return on any investment can depend decisively on regulatory decisions.  Yeah, EPA can say it is striving “to provide a rational basis for utility planning,” but anyone out of diapers knows how the game is really played.  The outcome of the regulatory process is very difficult to predict because it depends on the uncertain outcome of the rent seeking process, fluke events (e.g., one utility is the subject of a damaging revelation, perhaps untrue, which makes it a political pariah), and other unfathomable possibilities.  “Regulatory discretion” is a recipe for radical uncertainty.  Anyone contemplating making a huge investment that can turn into a long-lived white elephant at a regulator’s whim will often decide to wait until that uncertainty is resolved.  Thus, this policy uncertainty tends to retard investment, leading to higher prices and lower reliability.

Markets rely primarily on the price system to coordinate the actions of millions of dispersed decision makers acting on little pieces of information.  Prices tell people about the consequences of their actions, give them an incentive to act efficiently and take these consequences into account even if they could not possibly articulate what those consequences are.

There is no analogous mechanism to coordinate the actions of regulators.  Regulators routinely take actions that are perfectly sensible within their narrow, legally circumscribed spheres, but which impose huge costs elsewhere.  There is a feedback mechanism of sorts, but the ways that regulators respond to the feedback creates its own problems.  In this instance, EPA did something in pursuit of its legislative mandate which will have severe adverse consequences outside of it.  The affected industry howled, providing feedback, so EPA is extemporizing a response.  But that extemporization brings on its own bad effects, in the form of corruption and investment-killing uncertainty.

Like Reagan said: The nine most terrifying words in the English language: “‘I’m from the government and I’m here to help.”

High frequency trading firms came under heavy fire in the aftermath of the May 2010 Flash Crash because they largely withdrew from the market shortly before the Crash.  Indeed, the lack of liquidity (which is nowadays largely provided by HFT firms) was likely a necessary condition for the Crash to occur.

In light of that, and of last week’s market craziness, it is therefore quite interesting to read this:

The stock market’s fastest electronic firms boosted trading threefold during the rout that erased $2.2 trillion from U.S. equity values, stepping up strategies that profit from volatility, according to one of their biggest brokers.

The increase from Aug. 1 to Aug. 10 over their 2011 average surpassed the 80 percent rise in U.S. equity volume, showing that high-frequency traders made up more of the market during the plunge, Gary Wedbush, executive vice president and head of capital markets at Wedbush Securities, said in a telephone interview. Wedbush is the largest broker supplying bids and offers on the Nasdaq Stock Market, according to exchange data.

“We’re seeing a tremendous amount of high-frequency trading,” said Wedbush, whose company is one of the biggest execution and clearing brokers catering to high-speed firms. “Their business is a trading business, and volatility creates far more opportunities. Some of their algorithms and automated systems are trading two, three or five times as many shares as they would have in a more normalized volatility environment.”

It is therefore incorrect to say that HFT firms are Sunshine Liquidity Suppliers who head for the hills when things get volatile.  Sometimes HFT reduce the amount of liquidity they supply.  Sometimes they increase it.  And volatility alone is not enough to spook HFT firms.  Indeed–and this shouldn’t be surprising, really–volatility can be a huge profit opportunity for them.  Note, moreover, that their profit-spurred entry almost certainly reduces volatility below what it would be otherwise and improves efficiency.

One should hope, therefore, that regulators and others who blame HFT for every market outcome they don’t like will take this more recent experience into account in order to develop a more mature understanding of HFT, and use that understanding to develop sensible policies that encourage beneficial HFT.  Policies that don’t throw the baby out with the bath.

A study of what caused the differences between the first two weeks of August and the Flash Crash could be quite illuminating.  Market making, whether old school on the floor of an exchange or new school on computer at high speed, is vulnerable to toxic order flow–that is, order flow driven by private information.  Any market maker who wants to survive has to have the ability to determine when order flow is toxic and when it is not.  Old school traders could intuit the likelihood that the guy who wanted to buy from or sell to them was trying to pick them off because he had better information.  HFT uses algorithms to try to sniff out such opportunistic order flow.

It is pretty clear that on May 6, 2010 HFT algos were flashing warning signs that the order flow had turned toxic,  and they cut back as a result.  It is also clear that despite all the volatility in these days of Downgrade and Debt Crises, the algos are not picking up evidence of an increased intensity of privately-informed trading.  As a result, the HFT firms are sufficiently confident that the risk of being picked off is low that they can profitably supply liquidity.

One interpretation of this is that HFT firms are primarily responding to market conditions, rather than creating them.  One implication of this is that regulators should try to develop tools that would allow them to pick up the same warning signals that HFT algos do in order to anticipate potential market disruptions a la 6 May 2010.  This would permit the implementation of precautionary policies that could reduce the likelihood and intensity of Flash Crashes.  This would be a far better policy than micromanaging HFT trading through things like order time limits or commitments to make markets even when that is unprofitable.  Micromanagement is likely to be counterproductive as it raises the costs of supplying liquidity, and inducing exit of market making capital–which would make prices more volatile during conditions like those that have prevailed this month.

Exchanges are in the business of facilitating the discovery of prices.  The whole rationale for exchanges is that discovering more accurate prices encourages the efficient allocation of scarce resources.  That’s the value that they produce for society.

Which is why it peculiar, to say the least, that exchanges consistently don’t use prices to allocate a scarce resource at the heart of their businesses: communications capacity.  The recent volume surge caused by the intense volatility over the past week has taxed this capacity heavily, as this FT article shows.  Four major exchanges faced technical problems due to an avalanche of message traffic, most notably Borsa Milano.

To the extent that exchanges react to message traffic surges, they do so through non-price allocation methods, such as “throttling.”

As I have been discussing since the very first SWP post, exchange bandwidth is a scarce resource.  There can be intense spikes in bandwidth usage.  Some market participants exploit the fact that this capacity is not priced to engage in trading strategies that may be parasitical, predatory, and destablizing.  Even non-opportunistic use of exchange bandwidth can cause serious technical problems, including problems so severe that exchanges must shut down.   There is every indication that there are periodic tragedies of the commons on exchanges.

Pricing capacity, including peak-load pricing, is the obvious way to eliminate tragedies of the commons and abusive uses of system capacity.  You’d think that idea would come naturally to those in the business of pricing everything from stocks to bonds to soybeans.  But despite recurrent, and arguably increasingly serious, problems, exchanges uniformly shrink from this solution.

It was a puzzle to me five plus years ago.  It puzzles me still.

Betting the Business has a recent post that is a paean to central clearing.  I have some major problems with it.

The problems start with the title: “Central clearing lowers end-user costs.”  That short, unqualified, declaration begs major questions.  Most importantly: If clearing is more efficient, why did cleared markets lose market share to the putatively less efficient alternative, bilateral uncleared OTC trades?

I’m not suggesting that such inefficient outcomes are possible, but I take it as a rebuttable presumption that competitive processes tend to result in the displacement of the inefficient by the efficient.  So I’d like to see a convincing story at least, backed by some anecdotes at least, that would represent a plausible rebuttal.  That’s a low bar: it would be preferable, of course, to have a rigorous model supported by empirical evidence.  But I’ve yet to see anyone clear even that low bar: indeed, very few have tried.  To be honest, BTB don’t in this piece.

And by “convincing,” I don’t mean stories that rely on bankers’ Svengali-like hold on their customers, or customers suffering from some variant of battered spouse syndrome.  These stories are facially implausible, and indeed, counterfactual during the period during which OTC derivatives markets grew dramatically, the 1980s and 1990s.  This was a period in which banks were suffering “disintermediation”: they were losing business and customers to capital markets that customers found more efficient sources of capital.  So, the Svengali thing wasn’t working so good when it came to traditional loan business.  Ironically, many of the big loan customers that banks lost simultaneously began trading OTC derivatives with them.  Also ironically, one of the major reasons that the CDS market began was that banks were losing their traditional loan business.  JP Morgan in particular made a strategic decision to develop CDS business aggressively because it was facing more intense competition from capital markets in making traditional loans.  Which all means that customers–especially big ones–have the ability to adopt more efficient ways of financing and managing risk.

My other big problem with the BTB piece relates to netting:

Central clearing, if managed appropriately, helps to reduce the volume of credit risk by cancelling out many offsetting positions, disappearing some credit risk and leaving us with a smaller, residual credit risk from derivatives trading. One of the problems with OTC markets is that brokers wanting to closely guard their clients don’t have an incentive to cancel out offsetting exposures to counterparties. Critics of the clearing system keep overlooking the fact that risk can in fact disappear. The derivatives market system, depending upon how it is structured and operated, can have more or less total credit risk.

The notion that multilateral netting necessarily reduces risk is incorrect regardless of whether one focuses on credit risks in derivatives alone, or takes the proper perspective of looking at credit risk in toto.

With respect to credit risk in derivatives taken in isolation, clearing some products and not clearing others–which is inevitable given the unsuitability of some products for clearing, and indeed, the refusal of CCPs to clear them–may exploit some scale economies from multilateral netting in a particular product, but can destroy cross-product scope economies.  For instance, clearing a credit index but not all the names in the index destroys the scope economies that a bank can realize by holding positions with customers in both the index and the components.  Or if an interest rate swap is cleared but an interest rate swaption is not, netting economies across these products are lost.  These are simple examples, but they could be multiplied ad infinitum.

Indeed, these cross-product netting opportunities are one very plausible explanation as to why clearing lost ground to bilateral markets dominated by dealers.  Dealers trading multiple products can exploit cross-product netting economies and thereby offer their customers lower costs.

What’s more, when we consider clearing-as-it-will-be-in-this-messy-world, as opposed to the Platonic ideal of clearing, even within a single product clearing mandates may reduce netting economies.  Jurisdictionalism is likely to result in the creation of multiple CCPs for a given product type.  For instance, it is likely that regulators in the US and Europe insist that CCPs for particular products be established in their jurisdictions, and that financial institutions subject to their jurisdiction will use these “home” CCPs.  That creates operational headaches and reduces the netting benefits of clearing.

A proper, systemic approach also demonstrates that the disappearance of risk, like a magician’s trick, is often an illusion that works because the audience’s eyes are fixed in one place, and therefore they miss where the real action is.  As I pointed out several years ago, taking derivatives positions and the capital structures of market participants as given, netting redistributes risk from derivatives counterparties to other holders of liabilities on these derivatives market participants.  Yes, multilateral netting (and bilateral netting supported by collateral, rights of offset, etc., for OTC deals under bankruptcy laws) reduces the credit risk in derivatives trades, but the risk doesn’t disappear down the magician-CCP’s hat: it is reallocated to those who have lent money to the firms that hold positions in the netted derivatives.  Netting puts derivatives counterparties closer to the front of the line of creditors in bankruptcy.

Now the issue of the proper priority of derivatives is a sticky one, and I don’t claim to have the answer.  Giving derivatives counterparties priority tends to reduce their incentive to run, but it can increase the incentive of other creditors to run.  And in equilibrium, capital structures will depend on priority rules.  So it may be the case that clearing reduces the risk of runs on derivatives dealers by their derivatives counterparties, but increases the risk of runs by their repo counterparties, or the buyers of their corporate paper. It’s unclear how that cuts.  (No pun intended.)

Though I don’t have the answer as to the optimal priority rules, I do realize that netting does affect priority–and that this means that statements about netting “making risk disappear” are not generally true, and are typically wrong, because these statements are based on only a portion of the total risks.  The risk no more disappears due to a change of priorities achieved through netting than the rabbit really disappears into the magician’s hat.

It should be noted, moreover, that this issue of netting and priority has long been understood in other contexts, notably in payment systems.  Work in the 1990s and early 2000s (by Charlie Kahn, among others) emphasizes that netting systems and real time gross settlement have different affect the priority of payment system participants relative to other creditors of those systems.

There are some other problems with the BTB analysis, but I’ll just mention them in passing.  They say:

Margins are paid by those who use the system. In the OTC markets, banks either correctly price risk, or trade with clients by granting these subsidies from the taxpayer, or from wealth transfers by their bond holders. Clearing and margining is a potent dissuader of free riding on the system by negligent brokers, and also helps to mitigate agency conflicts between bonus motivated bankers and the bank’s stakeholders, which, we have learned, includes the taxpayers/citizens of each country.

I can’t say that I fully track that paragraph, but I will say that I am struck by the free riding analysis.  It suggests a sort of transubstantiation, where bankers are agency-conflicted, or perversely incentivized due to too big to fail subsidies when acting individually, but somehow become less conflicted and better incentivized when they act collectively through a CCP (and most CCPs will be bank dominated).  It’s not obvious how that works.  Indeed, collective action problems and multi-task/multi-principal agency problems usually result in weaker, not stronger incentives.

In this regard, it should also be noted that the homogenization of credit risk in a CCP creates problems.  CCPs make all trades by all members fungible, even if those members bring differing amounts of credit risk to the system.  This tends to reallocate trading activity away from the more creditworthy to the less creditworthy.

I also consider this unrealistic:

And to protect the public, it is necessary for regulators to see and be able to manage the true volume of credit supporting derivatives trading. Clearing is part of making that possible.

Given the fungibility of credit, you will never unambiguously know how much credit is used to support derivatives trades, even in a cleared system.

I have other issues (I can just imagine you exclaiming: “I’ll say!”), but I’ll leave it at that.  Suffice it to say that there are major holes in this clearing paean, and that it certainly does not provide a firm basis for mandating the adoption of central clearing.

Zero Hedge provides a reminder of the approaching automatic triggering of an increased haircut on Spanish and Italian government bonds in RepoClear.  Earlier this year LCH indicated that it would increase margins on Irish bonds when the spread hit 450 bp, and Italian and Spanish debt is reaching that level.

This is the kind of action that is individually rational for a clearer to take, but which can be destabilizing, pace my earlier posts on the hike in silver margins earlier in the year.

The haircut spiral plays a central role in a recent paper by BOE Financial Stability Executive Director Andrew Haldane and two co-authors.  The paper is nicely done, and the use of numerical methods to study the highly non-linear dynamics of complicated financial networks is the right way to go.

I am not persuaded that the paper focuses on the real sources of financial crises, however.  The reason is directly related to my post earlier in the week on clearing and systematic risk.  Haldane et al study how an idiosyncratic shock to the bank-specific haircut of a particular bank can propagate through a complex, densely connected network.  In some cases, such a shock can lead to a systemic crisis resulting from a cascade of banks deciding to hoard liquidity.  Not surprisingly, they find that the likelihood and intensity of such a contagion depends on the connectedness of the network, the skewness of the sizes of firms in the network, and the size and connectedness of the bank hit by the shock.

But I remained unconvinced as an empirical matter that major financial crises, at least in recent years, have developed in this fashion.  For this story to work, a single bank–and in the Haldane et al results, that bank would likely have to be a very big, highly connected one–is hit by an idiosyncratic shock that results in it having to pay higher haircuts on its secured lending.  In the real world, this could happen as the result of a rogue trader risk or some other major operational failing.

Perhaps the canonical example of such an episode would be the Kervial debacle at SocGen.  But even though that occurred when the markets were already stressed (in January, 2008), it didn’t lead to contagion.  I would be interested to hear of other examples, but I can’t think of any that have led to the cascade of failures like that that plays the central role in the Haldane et al analysis.

Instead, the major financial crises, most notably that of 2007-2008, the one that may be impending,  1998, etc., have been marked by an adverse shock hitting virtually all major financial institutions due to the similarity of their balance sheets.  European banks today are on the brink because they are all–all–heavily exposed to Eurozone government debt, and in particular to PIIGS debt.  This is a system-wide shock–a systematic shock, not a diversifiable, idiosyncratic one that is magnified through the web of interconnections among banks.  Such interconnections, margin spirals, liquidity hoarding, and so is a real possibility and could exacerbate the effects of the shock, but the point is that what is going on now, and what went on in 2007-2008, and what went on in 1998 is very different than the mechanism studied in the Haldane et al paper.

Perhaps many of the recommendations that Haldane and his coauthors make based on their analysis of the propagation of an idiosyncratic shock carry over to the case of a common shock hitting virtually all banks.  For instance, bigger liquidity cushions might be stabilizing in this situation as they are in the Haldane et al setup.

But maybe not.  Especially when one considers that for a variety of reasons–financial repression among them–sovereign debt is, and likely will remain a major component of any liquidity buffer.

That doesn’t work so well when the collapse of the creditworthiness of the governments issuing the debt is the root of the crisis, as is the case today.

I recommend the Haldane paper, but I think that it,  and many other papers that are based on tracing out the implications of idiosyncratic shocks to individual banks in a dense network, aim well wide of the mark.  I would like to see empirical work–hell, a case study, or even an anecdote–showing this mechanism working in practice.  If it is instead the case, as I am highly confident that it is, that financial crises are typically the result of an economy-wide, non-idiosyncratic, shock, the mechanism will be different and the policy implications will almost certainly be different.

A conjecture: in such a world, the greatest crisis accelerant is the similarity of balance sheets across major financial institutions.  To some degree, this cannot be avoided.  But there is no doubt in my mind that uniform regulations–and the Basel standards and other coordinated risk-based capital requirements–encourage even greater similarities in balance sheets.  This is a real danger, and if my conjecture is correct, many of the efforts currently underway to harmonize capital and liquidity requirements could make crises more rather than less likely . Basel Faulty and all that

The self-interested role of governments in cramming down more of their burgeoning debt will only exacerbate this problem, as regulators have an incentive to tilt capital and liquidity requirements in favor of holding government debt.  As dodgy as that debt is these days, it’s hard to imagine a better way to create the next crisis.

Central counterparties work on the principle of mutualization.  CCP default funds are a mechanism for mutualizing–sharing–counterparty loss risk, where the CCP-insurer has some protection in the form of a “deductible” that consists of the defaulter’s margin plus default fund contribution.

Insurance works when risks are diversifiable.  So, to the extent that the risks of default exhibit relatively little correlation across clearing members (CMs), mutualization through a CCP can improve the allocation of risk: through diversification, the variability of the average default loss across all CCP members is smaller than the variability of any CM’s default loss profile.

It should be noted that these diversification effects work in OTC markets too: market participants who deal with multiple counterparties diversify exposure to counterparty losses.  Moreover, in OTC markets, it is often possible to hedge counterparty risk through the CDS market.  [Which raises a question: how do you hedge CCP credit risk?  It is highly unlikely that CDS will be traded on CCPs.  Perhaps using a portfolio of CDS on CMs?  But who clears those CDS?  Sounds like the wrong way risk from hell.]

CCPs have been sold as a means of reducing systemic risk.  The basic story is that in bilateral OTC markets, the failure of one entity can cause the failure of other entities in falling-domino-like fashion.  In this telling, a CCP reduces interconnectivity of the financial market, and thereby reduces the likelihood of that the failure of one firm–even a big one–will cause the failure of others.  Put differently,  in this view default/counterparty risk is insurable because it is diversifiable, and therefore mutualization can improve the allocation of risk.

But is the domino-model really a good description of systemic risk?  Is the system really threatened by the isolated, idiosyncratic failure of one firm, even a big one?

Here’s an alternative story, in which a big, system-wide shock hitting many institutions simultaneously is the source of systemic crisis.  Some examples: Black Monday in 1987, the Asian and Russian crises in 1998, and the subprime meltdown of 2007-2008.

In such episodes, there are large price moves across asset classes.  These price moves often blow through margins–and by a lot.  During the subprime crisis, for instance, even well-before the Lehman episode, some prices moved by huge amounts.  In August, 2007–more than a year before Fannie-Freddie, Lehman, and AIG–David Viniar of Goldman lamented that there had been several 25 standard deviation moves in multiple markets.  In a “normal” (Gaussian) world, 25 sd moves don’t occur.  Period.  People don’t set margins based on the possibility of 25 sigma moves.  So when such moves occur–and they occur in the real, non-normal world–margins get blown through.  Meaning that if someone defaults, they’re well into the default fund, and the protection of the “deductible” is gone.

Moreover, such earth-shaking episodes are typically associated with adverse shocks to bank (CM) balance sheets.  What’s more, such events raise serious doubts about the solvency of banks/CMs: adverse selection problems can run rampant, causing markets to freeze.  This tends to cause funding and liquidity to dry up.  Due to actual insolvency, and illiquidity driven by widespread fears of potential insolvency, those facing big margin calls may face substantial difficulty in meeting them.   Thus, they are more likely to default.  This is what happened to AIG and Lehman.

Since the shock hits all financial institutions simultaneously, and since the shock is big, this risk of default is not idiosyncratic; it is highly correlated across firms.  Given such a shock, if one firm defaults, others are at elevated risk of default too.

If defaults that breach margins occur, the surviving CMs who contribute to the default fund must absorb the loss, and take a hit to their capitals.  But this hit occurs exactly when capital is very valuable/costly for them.  That is, hits to the default fund are likely to come at the absolute worst times–during adverse moves in the overall market.

It is possible that a CCP’s default exposure in such a circumstance could crater the clearinghouse.  But even if the CCP survives, its surviving members suffer a loss (via to their exposure to the default fund) when they can afford it least.

This has some interesting implications.  In particular, it means that the systematic risk of default fund capital is very high, which in turn means that default fund capital is extremely expensive.  The default fund is like a deep out of the money short put on the market.  Such puts are essentially bets on systematic risk, and hence are very expensive.

The expense of such capital means that CMs are likely to want to minimize the amount they post.  They have every incentive to try to construct risk measures that make CCPs look safe, with very low risk of default, in order to reduce their exposure to the default fund.

Think that this can’t be done?  Think again: that was essentially the entire game in CDOs.  As Coval-Jurek-Stafford show in an AER paper, “super senior” CDOs were claims that had low probabilities of default, but that defaults would occur precisely during periods of crisis–analogous to the foregoing characterization of the risk profile of default funds.  C-J-S present empirical evidence that these CDOs–which were AAA rated–were grossly underpriced, meaning that those who were buying these securities didn’t fully appreciate how heavily they loaded on systematic risk because analysis–and ratings–were framed almost exclusively in terms of unconditional default probabilities, and the rating agencies and many investors and regulators ignored the fact that defaults would occur when the value of capital was very high.  I can see the same thing happening with CCPs–who will reflect the interests of their members–focusing on default probabilities to justify their capitalization to regulators.

Another interesting implication is for what instruments should be cleared.  Instruments that have big losses during big systematic shocks–those that load heavily on systematic risk–are exceptionally dangerous for CCPs.  This means that clearing things like senior or supersenior tranches (or derivatives on them) is exceptionally dangerous.

This has a corollary: anybody that uses AIG to justify clearing mandates doesn’t know what the hell they’re talking about.  AIG’s positions in derivatives on supersenior tranches were, as C-J-S show, systematic risk time bombs.  Clearing those things would dramatically increase the systematic risk of default funds, because they would expose default funds to losses precisely during systemic crises.

Some people know this.  Jon Gregory, for instance.  When discussing “market coverage of a CCP”, he states:

Finally, CDO tranches (especially supersenior) would be expected to have serious correlation [i.e., systematic risk] impact since in a scenario where there are losses on these tranches it is likely that many CCP counterparties [and CMs] are in a distressed state.

The above analysis is to some degree subjective but one general point–probably beyond argument–is that the most important trades products to be traded through CCPs are arguably the most risky from the point of view of the stability of the CCP.  There has been much recent interest to trade all CDS index products and single-name products through CCPs. . . . The general point is that the products that market participants will most want and need to trade through a CCP will be the precise products that are most difficult to handle in this way (pp. 381-382).

In sum, the belief that CCPs can reduce systemic risk is predicated on a particular narrative of  the likely source of a future crisis.  This view is that the failure of a single institution, arising from some shock idiosyncratic to it (a bad trading decision, an operational risk such as a rogue trader) can lead to the failures of other institutions via a contagion mechanism.

But, in my view, that isn’t an accurate characterization of the actual crises that have occurred over the years.  These have been associated with broad-based shocks that hit large numbers of institutions simultaneously.  Again: Black Monday, Asia/Russia, subprime.

There have been multiple examples of large, one-off collapses of highly interconnected and leveraged institutions that have not resulted in a cascade of failures that led to a systemic crisis.  To name several off the top of my head: MG (1992), Barings (1995), Amaranth (2006). For a while, it looked like the Hunt collapse in 1980 could have broader, systemic consequences, but that fear evaporated rather quickly.

The contrast between the effects of the collapse of Amaranth in 2006 and the crisis surrounding the problems of LTCM in 1998 is particularly instructive.  Amaranth was actually bigger than LTCM (though it should be noted that the markets were bigger too), and had large OTC positions.  The difference was that Amaranth’s collapse was not caused by a systematic shock, whereas LTCM’s was.  LTCM engaged in strategies that blew up because of major shocks arising in Asia and Russia.  It was thus on the verge of failure precisely when liquidity had dried up, and its (big bank) counterparties were in bad financial shape because of the crisis.

The Amaranth collapse was effectively a diversifiable risk, and potentially clearing improved the allocation of that risk (many of Amaranth’s positions were cleared).  But the bilateral markets could have likely handled it quite well too.  In contrast, LTCM’s risk was not diversifiable.  Clearing would have still passed the risk of its failure onto stressed financial institutions, and likely resulted in some similar emergency measure to mitigate the impact of its huge losses.  (I’ve long argued that the same would have been the case with AIG.)

Along these lines, FTAlphaville has a very interesting post describing some forthcoming research which shows that incorporating “common jumps” into models of index option prices go a long way towards explaining the volatility skew, and reconciling the skews of index components and the skew of the index:

According to a technical paper published on Risk.net by Alex Langnau, global head of analytics at Allianz Investment Management and Daniel Cangemi, head of FICC trading at EFG Financial Products, Wall Street’s next top (risk) model is actually a copula that attempts to explicitly link correlation skew to systemic risk so as to improve tail risk management of large portfolios.

The basic premise lies in the idea that the downside risk of a portfolio of assets is generally substantially higher than the downside risk of its components — because assets tend to correlate more strongly in times of crisis.

. . . .

As the authors conclude this could become extremely useful for trying to figure out the additional risk attached to asset portfolios:

We have presented a generalisation of the Merton jump option formula to multi-assets as well as diffusive skews and found that common jumps are an appropriate and intuitive way to define copulas that describe the systemic risk of a portfolio.

Financial institutions should start treating systemic risk as market data across asset classes and start to mark, risk-manage and price these effects in a systematic way.

This suggests that these common jumps are of first order importance.  And these common jumps aren’t diversifiable.  What I argue above is that the major risk that CCPs face arises from such system-wide, common shocks.  And in the event, mutualization is all but irrelevant.  Except, perhaps, to the extent that CMs attempt to limit their exposure to this risk via the default fund by undercapitalizing CCPs, thereby threatening the ability of these vital–and soon to be even more vital–pieces of market infrastructure to survive a systematic shock.  In which case, mutualization makes things worse because if some CCPs don’t survive, the knock-on effects from the original shock will be even more catastrophic.

In some respects, the issues I’ve discussed here reflect an older literature on bank runs.  One view of bank runs is that they are idiosyncratic in origin (i.e., a bad shock–perhaps even a sunspot–hits a single bank, leading to a run), but have systemic effects due the interconnectedness of the financial system.  In this view, widespread panics spread like a wildfire from a single, isolated spark.

An alternative view, associated most strongly with the work of Calomiris and Gorton, is that runs occur when depositors receive adverse information about the aggregate economy, which raises concerns about the viability of all banks.  In this view, panics involving runs at many financial institutions reflects the fact that all banks are exposed to a common risk.

The Calomiris-Gorton view, and the evidence they amassed to support it is persuasive.  In more recent times, major periods of crisis have been associated with a common shock hitting many financial institutions, rather than from the contagion of an idiosyncratic shock to one big institution laying low otherwise healthy institutions.  In these types of events, mutualization of risk isn’t of much help, and can actually make things worse.  Meaning that clearing mandates were sold largely on the basis of a misleading portrayal of, or misunderstanding of, the real dynamics of financial crises.

One takeaway from this is that there should be a very hard look at CCP capitalization, and at measuring the cost of default fund capital.  The exposure that default funds face to systematic risk affects the cost of capital–and hence the cost of clearing.  That cost should be taken into consideration when deciding the scope of clearing mandates–that is, when deciding what should be cleared.   For the analysis suggests that clearing can be very expensive, and that a major source of cost has been largely overlooked.  Forcing too much to be cleared can create costs with no corresponding benefit.