Streetwise Professor

Last week it looked like the Obama administration had decided to be sensible on at least one energy issue-the export of LNG.  It approved a new license (for Freeport LNG) for export to countries with which the US has no free trade agreement.  But the WSJ reports that new Energy Secretary Ernest Moniz is thinking of putting on the brakes again.  Because we need more studies.  No.  Seriously:

Mr. Moniz showed caution about the existing studies. Speaking to reporters after a speech to an energy-efficiency conference here, he said he was “committed to doing a review of what’s out there in terms of impact analyses” before approving more applications to export U.S. natural gas. Critics have said last year’s study didn’t rely on the best data available.

“Right now we have no plans of commissioning new studies, but everything is on the table until I have done my analysis,” Mr. Moniz told reporters after his first public remarks as energy secretary

Sorry.  We don’t need no steenkin’ studies.  The whole idea smacks of central planning.  The presumption should be that if firms are willing to risk their private capital, the benefits exceed the costs.  Any analysis should be restricted to potential externalities.  Real externalities.  Not pecuniary ones.

But these “impact studies” are all about pecuniary externalities.  Namely, they focus on the effects of exports on prices of natural gas, and the effects of natural gas prices on consuming industries (like petrochemicals).  But these price effects are not true externalities that lead to inefficient allocation of resources.   Indeed, restricting exports because of these effects would cause a misallocation of resources.

Pecuniary effects do have distributive effects.  In the case of LNG exports, they affect the distribution of rents between gas producers in the US and foreign consumers on the one hand, and domestic gas consumers on the other.

And that’s what the need to get an export permit does: it permits the government to affect the distribution of rents.  That, in turn, gives rise to rent seeking.  And corruption.

In this context John Cochrane mentions the Indian “permit raj”: there you need to get a permit for everything.  This gives those with the authority to grant permits incredible power.  Power they use to enrich themselves and secure political support.

That is exactly what can go on here.  Those hoping to get a permit have an incentive to exert influence, through lobbying, campaign contributions, and supporting public campaigns on issues favored by the administration.  They also realize that they face substantial risks if they oppose the administration on other issues: “Nice little LNG terminal proposal you have here.  Would be a tragedy if something happened to it.”

The government has no business being in this business, beyond perhaps-perhaps-addressing real externality issues.  But even there, other mechanisms (e.g., liability for pollution) may be preferable to a permitting process.  (Look at how Russia used environmental regulations to drive Shell out of Sakhalin II: any power to permit can be used to expropriate of hold up the party seeking the permit.)

In the US, energy, and particularly the international trade of energy, is particularly raj-like: Keystone II is another example.  This destroys value in myriad ways.  Beneficial investments are delayed, or not made at all, either because the government stops them directly, or the risks and costs of getting approval undermine the economics.  Real resources are used to influence policy.  Since energy investments involve big dollars, the losses can be big too.

People often lament the lack of an American energy policy.  I disagree.  We do have an energy policy, and the Energy Raj is a big part of that policy.  A better policy, by far, would be no policy at all.  Would that the DOE adopt the motto: “Don’t just do something! Stand there!”

I’m not holding my breath, though.  The benefits of the raj to the rajahs are far too great.

Harry D’Angelo and Rene Stulz have an interesting paper about bank leverage.  It makes a simple point.  Banks’ liabilities-notably, deposits-are someone else’s asset. That asset provides a benefit, namely liquidity, that depositors can’t realize by trading in capital markets, due to some friction.  This intermediation is what makes banks special.  As a result, depositors are willing to hold bank liabilities in exchange for a lower return than bank assets. Banks therefore have an incentive to create these liabilities-that is, to leverage up-to the maximum extent possible, in order to capture as much of this liquidity premium as possible.  This leveraging is a good thing, because it supplies an asset that is highly valued by those who incur high costs to access capital markets directly.

In the D’Angelo-Stulz model, bank equity is the capitalized value of this below market borrowing rate/liquidity premium.  If the liquidity premium is modest, the  supply of liquid claims results in a high debt to assets ratio.  Again, this high leverage is a good thing, because it is the consequence of the efficient supply of intermediation.  Banks access the capital market effectively engage in a form of asset transformation that generates value: those who cannot access the capital markets, or at least not as efficiently, cannot engage in this asset transformation, and they are willing to pay banks who can.  Through diversification and risk management, banks can engage in asset transformation that is valued by depositors.

This paper is most useful as a corrective to the rather annoying Admati et al “Bankers’ New Clothes” arguments that banks are excessively leveraged and should therefore be subjected to far more rigorous capital requirements, e.g., 25 percent equity.   Admati et al rely heavily on Modigliani-Miller type arguments, and D’Angelo-Stulz show that high leverage can be efficient when there is a single deviation from MM assumptions.  Bank debt (at least some forms of it-more on this below) provides a valuable device in the presence of a friction, and therefore issuance of this debt (in lieu of equity) is socially beneficial.

That said, there are some issues with the paper.

First, the formal model is very rudimentary.  It posits that those without access to the capital market are willing to pay an exogenously specified premium for bank liabilities that offer liquidity, i.e., guaranteed purchasing power.  But willingness to pay does not determine the equilibrium price of liquidity.  In a competitive banking market, the equilibrium price of liquidity is determined by the cost of producing it as well.  This is not modeled in the paper.  The marginal cost of engaging in intermediation/asset transformation must be upward sloping in order for banks to earn producer surplus (which, when capitalized, would be the value of bank equity).

Presumably, equity is one of the means by which banks are able to engage in asset transformation that provides reliable liquidity to those holding bank liabilities.  In essence, equity is a means of bonding contractual performance (a point I learned from reading Yoram Barzel years ago).  In the banking context, equity provides a cushion that ensures that depositors will be able to realize the face value of their claims at will-which is the essence of liquidity.  Thus, the reliability of the liquidity banks supply, and hence the premium that depositors are willing to pay, depends on the amount of equity (as well as on the asset side of bank balance sheets).  The paper does not address this interaction, taking equity value as a pure residual value driven by an exogenous liquidity premium that does not depend on bank equity.

Second, although deposits that provide liquidity to investors without ready access to the capital market are one part of bank leverage, banks, and especially systemically important ones, borrow in other ways, and the resulting liabilities do not have the same money-like characteristics as deposits.   The paper does not explain the entire capital structure of banks.  Indeed, it predicts that banks should be financed almost exclusively by the issuance of money-like liabilities.  They aren’t, so the paper doesn’t explain why banks issue debt that does not provide liquidity benefits instead of equity.  (Perhaps it could be argued that banks provide liquidity indirectly, e.g., by issuing corporate paper that is purchased by money market funds which provide money-like claims to investors.  But this still doesn’t explain the issuance of longer term debt.)

Third, this analysis relates primarily to commercial banks that issue deposits.  What about investment banks?  They are highly leveraged, and it’s not so clear that their liabilities offer the kinds of liquidity benefits that drive the results in the paper.  (Theories that argue that banks are too highly leveraged because of deposit insurance subsidies or access to central bank liquidity (at subsidized rates) don’t apply to investment banks either, because those didn’t have access to deposit insurance of central bank liquidity support.)

Fourth, the paper discusses systemic risk, but this discussion is a little glib.  Debt that provides valuable liquidity services under normal circumstances is fragile, and susceptible to runs.  When runs occur, money-like bank liabilities do not provide liquidity.  Presumably, this affects the liquidity premium, which means that it is unclear that banks issue too much debt.  But the paper can’t really address this question because the value of liquidity is specified exogenously.

All this said, the paper is still valuable because it makes an important point.  Existing theories of banks posit that banks are leveraged because debt addresses some sort of agency or information problem.  These are essentially supply-side factors. The D’Angelo-Stulz theory identifies a demand-side driver of bank leverage.

Bank leverage is a big issue now, with Basel III and Brown-Vitter.  Good policy regarding leverage (and hence, capital requirements) requires an understanding of its costs and benefits. Calling attention to the demand for bank liabilities, and the benefits they provide, is an important contribution to such an understanding.

This story is very bizarre, and I can’t figure out what is going on, exactly.  Or put differently, what has put a bee in the CFTC’s bonnet about CME Clearport’s Exchange of Futures for Swaps (EFS) facility after all these years.

The Commodity Futures Trading Commission has issued a “special call” asking Wall Street banks and other traders to provide documents that would prove recent derivatives transactions known as “exchanges of futures for swaps” were legal. Lawyers at the CFTC enforcement division are also scrutinising the trades for possible violations.

. . . .

The new inquiry centres on whether large traders and market-makers used unregulated over-the-counter swaps markets to trade what were in fact futures, strictly regulated contracts that are economically identical to swaps.

Trading futures off an exchange is illegal, and regulators are concerned that traders may have used these deals, known as EFSs, to agree prices that did not reflect the market.

“They’ve made information requests to everybody that’s ever traded an EFS. They’re saying, ‘prove to us that the swap was legitimate’,” said a recipient of a CFTC document request

The only thing that makes sense is that the CFTC believes that market participants engaged in EFS transactions without having a legally binding swap agreement in place first, meaning that the parties would have engaged in futures trades off-exchange.  Or something.

Note that even if the parties had entered a swap, it may have been in effect a very short period of time-just as long as it took to execute the deal and submit it to Clearport for clearing.

I also find it curious that the article mentions that the CFTC is looking only at deals done post-Frankendodd, even though deals have been done this way since the 2002 time frame, if memory serves.  One explanation is that CFTC believes the alleged conduct was permissible under CFMA, but not under DFA.  Another guess on my part.

If there is a violation here, it seems to be a highly technical one.  The end result is pretty much the same if they did or they didn’t execute a binding swap first: each party has futures positions obtained at a privately negotiated price.

CFTC has a lot on its plate already: is this really a priority? Really?

Moreover, the party that usually screams the loudest about off-exchange trading of futures is the futures exchange.  But Clearport is a CME system, and EFS is a CME procedure, and it seems that CME is totally fine with this.  Actually, if the following quote relates to the EFS issue (and it’s not clear that that’s the case from the article), CME is actually hacked at this:

Terry Duffy, CME executive chairman, said in a letter to CFTC last week: “In our view, nothing is served by piling on duplicative reporting mandates.”

For certain, though, CME was perfectly satisfied with the way market participants were doing EFS deals.

So who is the victim here?

It’s actually ironic that EFS was the CME’s way of implementing clearing for energy and metals.  And the CFTC is rah-rah about clearing.  But it is looking askance at the CME’s way of implementing clearing.  I guess it’s a case of that was then, this is now.

CFTC also effectively killed EFS as a mechanism for facilitating OTC clearing by determining that a swap, no longer how short its existence before conversion into futures, counted towards a firm’s annual $8 billion (to be reduced to $3 billion) de minimus swap volume for the purpose of determining whether it is a swap dealer.  As a result, market participants are moving to block futures trades rather than EFS to clear energy transactions: block futures trades that are negotiated away from the central market, just as the allegedly phantom swaps were. So this is last year’s war.  If traders weren’t doing papering officially a swap deal, they were effectively engaging in block futures trades, which is what they are doing now.  If it’s OK now, other than the technical violation, was it so horrible then that it requires a full blown investigation?

So what’s up?  A burdensome, intrusive “Special Call” to investigate a possible technical violation that the exchange that would be hurt by a violation doesn’t seem to care about, and which is of little relevance going forward.

Wow.  That seems like a totally reasonable use of scare resources-resources Gensler claims he doesn’t have nearly enough of.

Javier Blas of the FT just posted an article about a report “written by a leading academic on commodity markets” on whether commodity trading firms (like Cargill or Vitol) are sources of systemic risk.

What, haven’t heard about that report?  Well, that’s the main point of the story.  The report was spiked by the GFMA, a banking trade association, which commissioned it.  According to Javier:

However, the report was never completed and remained in a “draft” status, after its conclusions went against the interest of the lobby group, three people familiar with the matter said.

So yes, perhaps you’ve guessed by now that the academic in question is me (though you have to read 2/3s of the way through the story to get to my name).  And yes, that’s pretty much what I understood to have happened, though I was never told that in so many words.  It’s nice to have it confirmed by “three people familiar with the matter”, even though it was blindingly obvious to me at the time. *

I call them like I see them.  GFMA didn’t like that.  I wouldn’t change the call, so they sat on the report.  So it goes.

I think GFMA handled this badly even from the perspective of its own interests, though I guess I am not really surprised: this is the way organizations like this tend to behave.  I am sure this has given the report more visibility than it ever would have achieved otherwise, and makes GFMA look bad in the bargain, at least in my (probably biased) opinion.  The regulatory body they were trying to influence-the FSB-was briefed on the findings, and had a draft of the report, so deep-sixing the report only signaled to the FSB that GFMA didn’t like the results, which it probably knew anyways.  Spiking the report also serves to validate the independence of the findings-and of the finder of the findings.  That’s definitely an upside for me.

Working on the report helped me learn a good deal more about the global commodity trading firms, so that’s also a good thing.  I look forward to learning and writing more in the future.

*For the record, the copy of the report “seen by the Financial Times” didn’t come from me.

There are reports that the CFTC will vote on the SEF rule next week.  The rule had been in limbo for months due to Gensler’s insistence that the rule require those requesting a quote solicit them from five potential counterparties.  Gensler has apparently relented because he could not get the new Democratic commissioner, Mark Wetjen, to join with Chilton and Gensler to vote out the 5 RFQ rule.

The compromise will require users to solicit two quotes for the next two years, and then three thereafter.

Whatever.

On the 1 year anniversary of the DFA, I named the SEF mandate as The Worst of Frankendodd. I haven’t changed my mind on that, though the competition is fierce.  And the RFQ requirement is the Worst of the Worst.  It is defended as a way of  improving competition.

This is at best paternalistic.  It presumes that those who want to enter into swaps don’t know their own interests.  Perhaps Gensler thinks that the buy side suffers from some sort of Stockholm Syndrome after years of captivity to the dealer banks.

In reality, buy side firms-most of whom are extremely experienced and sophisticated-are making trade-offs between competition and information leakage.  They are trying to minimize cost of execution, and have the information and incentive to do that.  Note too that they are required to do this for every trade, regardless of instrument, size, and other factors that may influence the trade-off.  But nope, one size fits all. They should be allowed to make that trade-off themselves, without any guidance from Gary.

RFQ5?  How about RFQ0?

Here’s an analogy.  How would you like it if the government told you how many stores you had to visit before making a purchase?  You know, to make sure that you get the best price.  Call it the CS5 rule.  You have to comparison shop at five stores before making a purchase.  On everything.   Of course, when deciding on whether to shop at one store or five, you trade-off the potential savings (which will depend on the value of the purchase, the good you are shopping for, and other factors) from shopping around more, against the cost (which will vary with the value of your time, how hurried you are, your income, the price of gas, where you live, etc.)  But none of that matters under the CS5 rule.  Want to buy a quart of milk?  Shop at five stores.  For your own good.

Yeah.  It’s that bad.  CS2 would be bad, but not that bad.

Once the SEF rule goes into effect it will be interesting to see how the structure of the industry involves.  There will be a land rush of new SEFs.  I predict there will be a shakeout, and there may well be only a single dominant SEF for each major instrument.  The SEF rule does not, as I understand it, require a SEF to send an order to another SEF offering a better quote.  Which means that the network effects of liquidity will tend to cause trading activity to “tip” to a single SEF for products big enough to support order book trading.

But the whole SEF landscape will also be shaped by the margin rules, the Bloomberg suit over those rules, block trading rules, and on and on.  The rule is not the beginning of the end, it is barely the end of the beginning.

The FSB is contemplating designating commodity trading firms like Cargill and Glencore as systemically important.  No, not that FSB: The Financial Stability Board, a group of global regulators established in the wake of the financial crisis.

The reason for this is, apparently, that these firms are engaged in shadow banking.  The linked Reuters story mentions two types of activity.

First, commodity trading firms extend trade credit and working capital to their customers.  Second,  some use securitization to raise funding.

The FSB is rightly concerned about shadow banking, but the kinds of activities that commodity trading firms engage in is quite different from some of the activities implicated in the crisis.

Some forms of securitization contributed significantly to the crisis.  Most of the most dangerous forms involved significant maturity transformation, and direct ties to big banks.  SIVs that funded portfolios of mortgage securities with short-maturity corporate paper, that were backed by liquidity puts provided by sponsoring banks are the prime example of this.  When investors began to doubt the value of the underlying assets, there were runs on the SIVs: they could not rollover their paper, and sponsoring banks ended up taking the now toxic assets back onto their balance sheets.  This was a big problem because the banks were highly leveraged; their main business is to provide credit; and they are essential components of the payment system.

The securitizations that commodity trading firms have engaged in, like the Trafigura program mentioned in the Reuters piece, are (a) far more limited, and (b) very different.   In particular, these structures do not involve the kind of maturity mismatch that was the Achilles heel of the SIVs.  Indeed, if anything, to the extent there is maturity transformation it is the opposite of the type that proved problematic in the crisis.  The underlying assets are very short dated (such as receivables) and/or very liquid (like inventories of aluminum in LME warehouses).  The assets typically have shorter maturities than the liabilities issued to fund them.  Indeed, one of the challenges of these structures is to replenish the assets as they mature.  Traditional SIVs had to rollover their liabilities: commodity trade securitizations have to replenish their assets.  The former is far more problematic than the latter because the run risk is far greater.

Moreover, historically the default rates on the trade receivables that are securitized are very, very small.  The extent data are for all trade receivables, not just commodity trade receivables, but the credit losses are trivial.

I also find little reason to be unduly concerned over the  the granting of trade credit and extension of working capital funding through prepayments and other arrangements.  Consider prepays, where a trading firm may provide funding to a refiner, say.  The commodity trader may fund the input (crude oil), and in exchange receive refined products.   The refined product is essentially collateral for the financing provided to buy the input.  Moreover, the value of this collateral can be hedged in most cases, limiting the credit exposure of the commodity firm extending the credit.   This is often quite different than extending credit to fund illiquid, hard to value, long maturity assets that cannot be hedged.

The commodity trading firm has a comparative advantage in marketing the output.  Moreover, due to its knowledge of the industry and the particular firms involved, it likely has information that makes it the most efficient creditor.  It knows more about the market and the particular borrowers than most banks.

The foregoing suggests that the credit risk and maturity transformation risk involved in commodity trading firm shadow banking activities is far less than that involved in the kinds of shadow banking that proved highly problematic during the crisis.  But credit losses are conceivable.  What would happen if commodity trading firms suffered big credit losses?

It’s hard to see how this could seriously threaten the stability of the financial system or the global economy.  Commodity trading firms aren’t that big: if Glencore is too big to fail, so is Kraft, or a similarly sized company.  Their assets are dwarfed by those of the big SIFI banks: they are about as big as many middling banks you’ve probably never heard of.  Moreover, the commodity trading firms are far less leveraged than big banks.  Furthermore, their capital structures are far less fragile, because they involve little maturity transformation: their short term liabilities are largely matched against short term assets, such as hedged commodity inventories.  The firms provide financial intermediation, but unlike banks, that’s not their primary function, so they are not as vital to the credit supply process as banks.  They are not essential elements in the payments infrastructure.

Commodity firms provide valuable logistical services, but the assets, human and physical, that they utilize are readily redeployed.  If one company goes bust, its assets can be redeployed so that the trade in commodities can continue.

Some of the shadow banking activities that commodity firms engage in actually take risk out of the banking system.   A major reason for the development of securitization in commodity markets was that big banks-especially French banks facing difficulties in securing dollar funding-cut their funding to commodity firms.  So the firms turned to securitization and thereby transferred the risk to the capital markets.  And unlike the case with SIVs sponsored by banks, there isn’t a backdoor by which this risk can make its way back to bank balance sheets.

I would also note that commodity trading firms do not benefit from deposit insurance, and so don’t pose the same moral hazard concerns as banks that do.

So I find few parallels between the kinds of shadow banking that proved so dangerous during the crisis, and the kinds of shadow banking that commodity firms engage in.

Commodity firms are first and foremost logistics specialists that engage in financing transactions to facilitate that business. A couple of other historical experiences suggest that such logistical intermediaries are not systemically important.

In 2002, the entire merchant energy sector in the US imploded.  These firms-companies like Enron, Dynegy, Mirant, and Williams-were primarily logistical intermediaries that provided financing and risk management services to their clients, just as global commodity trading firms do.  Not just one of these firms imploded.  The whole industry did: the stock prices of the firms in this business fell by about 90 percent between April and July of 2002.  But gas and power continued to flow. The impact on the US economy was barely measurable.

Furthermore, the experience of the Fukishima earthquake and tsunami suggests that even if the failure of one or more major commodity firms did disrupt commodity logistics for a time, the implications of this for the global economy would likely be small.  The earthquake and tsunami created huge disruptions in supply chains, especially in automobiles and electronics.  Trade was severely disrupted.  But the impact on the global economy was almost immeasurably small.  Studies undertaken by several governments found that the Japanese disaster shaved a tenth of a point or two off global GDP growth.  The financial aftershocks were minimal, even in Japan.  If the world economy can survive such a literal seismic shock that seriously disrupted supply chains in high valued manufacturing, it can almost certainly survive the failure of a big commodity trader.  Or two. Or three.  Especially since the Fukishima disaster destroyed or disrupted physical assets in a way that the financial distress of a commodity firm with redeployable assets would not.

I’d be more persuaded by the FSB’s concerns if it would provide a description of the mechanism by which commodity trading firms can be the source of financial contagion, or the channel through which contagion can be communicated from the financial sector to the real economy.  As those who have read my writings on clearing can attest, I can have a pretty vivid imagination about how contagion can propagate, and despite giving this considerable thought, I haven’t found a plausible mechanism.

In some sense, it seems that the FSB is like Mark Twain’s cat that wouldn’t sit on a hot stove after it had been burned, but it wouldn’t sit on a cold one either.  Once burned by shadow banking of one kind (the kind tightly tied into the banking system), it seems afraid of any kind of shadow banking.

And I have a pretty good idea who is stoking those fears.  I know, as the result of personal experience, that major banks involved in commodity markets are chafing under the restrictions imposed on them, and resent the fact that commodity firms that are their competitors in certain activities are not subject to the same restrictions.  I know they have been importuning the FSB to identify commodity trading firms as systemically important, and to impose bank-like disclosure and capital requirements on them.  All the better to hamstring the competition.

How I know this, I can’t say.  But I’m just sayin’.  You can take it to the bank, as it were.

Too big to fail is a self-fulfilling phenomenon, in large part.  It would be far less of a problem if governments could credibly commit not to bail out certain firms.  It becomes much harder to make such credible commitments when those firms are identified as systemically important.  Therefore, regulators should be very reluctant to confer the “systemically important” label.  Very reluctant.   Objectively, there are few reasons to consider big commodity trading firms-even the biggest ones-systemically important. All the more reason to eschew conferring that designation on them.

In other words, sitting on a cold stove isn’t dangerous.  The FSB should be smarter than the cat in Puddin’ Head Wilson.

The most recent controversy over HFT stems from this WSJ story about the CME.  In a nutshell, computerized traders receive confirmations of their trades before information about those trades is disseminated to the market at large.  As in a few milliseconds before.  But in an electronic world, a few milliseconds can be decisive.

One example of a particularly informative trade is when an away-from-the-market limit order is executed.  This means that a market order of sufficient size to blow through the quote size at the inside market was submitted.  Given that orders and communicate information, and that the bigger the order, the more informative it is, knowing before anybody else that such an order has been executed can provide valuable information.

The implications of this depend on how the information is used.  A trader (or, more accurately, a bot) that gets this information can use it to take liquidity aggressively.  For instance, it can use information gleaned from a big, price-moving crude oil buy to submit an aggressive order in heating oil or RBOB, thereby picking off resting limit orders that cannot adjust to the new information.  Or, as the WSJ article suggests, the bot can use the information derived from the NYMEX CL trade to take liquidity from ICE Brent or NYMEX lookalike futures.

This kind of trading exacerbates information asymmetries, and all else equal, increases spreads, reduces depth, and increases trading costs for the uninformed.

But the “all else equal” part of the statement doesn’t necessarily hold.  This presumes that the amount of capital devoted to HFT is constant.  But that’s not true in the long run.  If these sorts of advantages generate profits, that will attract more capital into HFT.  Moreover, note that the strategy just outlined involves placing limit orders, and then reacting when those limit orders are executed.  Competition to get the information advantage will lead to more aggressive quotes, and quotes in bigger size.  In the long run equilibrium, this competition will dissipate the rents from the information advantage.

Therefore, if there is any reason to reduce this speed advantage (either by slowing down some traders or speeding up the dissemination of trade execution information to the market at large), it is to prevent the investment of excessive capital into HFT.  The effect on spreads and depth in equilibrium is ambiguous.

Moreover, there are other possible uses of the information advantage that are clearly socially beneficial.  An HFT market maker-who is likely making markets in a variety of contracts-can utilize the information to revise limit orders either in the market in which the execution occurred, or in other markets, especially those that are closely related (again, consider the CL/HO or CL/RB example).  Using the speed/information advantage in this way reduces the HFT market maker’s vulnerability to getting picked off, and makes it willing to supply liquidity more aggressively.  This tends to reduce trading costs, and does not lead to the rent seeking that in the long run equilibrium tends to result in an inefficiently large HFT presence.

We also need some perspective here.  I consider it beyond hilarious that the WSJ has a video embedded in the online version of the story that has many images from the floor.  (And these days, one of the floor’s main functions is to provide visuals for stories on trading-especially the trader’s-head-in-his-hands shot on days when the market falls a lot.  Pictures of servers aren’t nearly so dramatic.)

Why hilarious?  Well, the floor was the epitome of time and space advantages to a select few.  A select few who paid for the privilege.  I remember distinctly a trader telling me: “Why do I spend $500,000 on a seat? Because I get to see the price before anybody else.”

Exactly.  The floor was the meat version of colocation.  Or the carbon based life form version, if you like.  Those on the floor could see the execution prices, and bids and offers, and order flows, that those off the floor could not.  They profited accordingly.  Which is why the marginal guy on the floor-the least efficient trader-was willing to pay hundreds of thousands of dollars in some cases to get on the floor.

In 2002 or so I wrote a paper titled “Upstairs, Downstairs” (still a working paper) which showed that floor traders earned a rent as a result of their time and space advantage: upstairs traders could not supply liquidity as effectively as floor traders due to their information disadvantage, and this meant that floor traders faced limited competition in supplying liquidity.  Moreover, exchange limits on membership meant that entry could not dissipate these rents.  But by reducing the time disparities between liquidity suppliers advantage, electronic trading increased liquidity supply: upstairs traders were no longer operating under a time and space handicap.  Trading costs and rents decline. And that decline in rents is precisely why floor traders fought electronic trading so fiercely for years.

So yes, in today’s electronic markets some traders have a speed advantage.  But this disparity is nothing when compared to that which existed in the floor days.

Which is why I can’t really get all that spun up over the WSJ story, or most of the other stories about how unfair markets are.  Everything is relative.  No, the playing field isn’t perfectly level today, and along the lines of yesterday’s post, it may be in the interest of the CME to take measures to make it more level.  They say that they are.  But arguably the field is more  level than it has ever been.  It’s certainly far more level than in the heyday of the trading floors.  Don’t get nostalgic for the days when market makers were meat, not machines.  The table was tilted in their favor, bigtime.  Much more than today.

HSBC’s Gary Dunn said something at the ISDA AGM that I think is very important:

Dunn pointed out the risk characteristics of a CCP are very similar to that of collateralised debt obligations, the tranched credit products that were prevalent in the run-up to the 2008 financial crisis.

In the CCPs default waterfall, the initial margin payments from clients and default fund contributions from clearing members are comparable to the equity or first loss and mezzanine tranches of a CDO. In other words, these are the first sources of funds that get eaten into to cover any losses.

According to Dunn, the super senior tranche (which in the case of CDOs tended to attract Triple A ratings, but often subsequently sustained losses during the credit crisis) is the additional steps the clearing house might take when all other funds are exhausted, whether it is haircutting, asking for more capitalisation from clearing members of possibly even a government bailout.

“If you start modelling the risks of a CCP as a CDO, you realise the correlation risks in a CCP aren’t at the moment fully appreciated, very much like they weren’t when we had CDOs and CDO squareds,” said Dunn.

“The probability of a CCP failing is still relatively low, but there is a reasonable probability that people or banks lose money even if a CCP doesn’t fail,” he added.

I can hear you saying: “Where have I read that before?”  Or: “When have you said that before?”  In January, 2011, now that you ask

Let’s see how this relates to CCPs, and in particular to the default funds of CCPs that are the ultimate backstop of cleared contracts.  Default funds are analogous to protection written on supersenior tranches.  The collateral (margin) that firms must post to CCPs when they hold derivatives positions absorbs most of the losses due to movements in market prices.  Indeed, margins are usually set to absorb 95-99 percent of market moves.  Beyond margin, CCPs often have their own financial resources to cover the losses associated with the default of any member firm not covered by margin.  If the margin and CCP-resource elements of the CCP “waterfall” (note the similarity of terminology used to describe tranched structures and CCPs) are breached, then the members must absorb the remaining default losses, up to some pre-established commitment level.

This means that CCP members must cover defaults only in extreme circumstances, just like writers of protection on supersenior tranches must cover defaults only under extreme circumstances.  Indeed, the oft-touted features of CCPs, such as collateralization create the seniority/out-of-the-moneyness that gives rise to wrong way risk if the pre-requisite dependencies also exist.

So it seems that CCPs are potentially vulnerable to wrong way risk.  They are effectively financial structures of a type that can, given the right (or wrong, depending on your perspective) dependence, give rise to a serious wrong way risk problem.  Which raises the question: are the dangerous dependencies likely to be present?  That is, are the contributors to a CCP default fund likely to be in bad financial straits just when their contributions are needed to absorb default losses? Are those that are insuring default losses (through mutualization) likely to be in dodgy condition precisely when they are most likely to have to pay off on that insurance?

This is a big deal. Wrong way risk is particularly poisonous, and a source of systemic risk in systemically important institutions like big CCPs. Policymakers have been largely oblivious to this very important problem.  I’ve been on about it for over two years, but we see how that matters.  Maybe if people like Gary Dunn start raising the alarm this issue will get the attention it deserves.

But, of course, Anat Admati, etc., will dismiss this as self-interested scaremongering by banks, and will criticize me as being some sort of tool (which she has, by the way).

I understand perfectly that such self-interested scaremongering occurs.  But I also understand that sometimes there is a wolf.  This is one of those times.

There are all sorts of proposals out there to rein in HFT.  The Europeans in particular-and in particular, particular, the Germans-want to do so.

I’ve long argued that there is good HFT and bad HFT, and that trading platforms have an incentive and the information to adjust fee structures and trading protocols in order to mitigate the latter.  For example, some exchanges have cracked down on excessive cancellations or the entering and canceling of orders far away from the current inside market.

Today’s FT reports another example, and a rather dramatic one.  One of the biggest FX trading platforms, EBS, is jettisoning the venerable time priority (“first come-first served) system with continuous trading.  Instead, it will collect orders that arrive during a period lasting a few milliseconds, and then execute them in a batch.  Instead of a continuous market, it appears to be a high speed sequence of call markets.  This eliminates the advantage of getting your quote in a millisecond sooner than someone else.  Perhaps it will also deter forms of gaming, such as strategies that (allegedly) attempt to create and exploit latency.

Will it work?  Who knows?  But that’s the point.  Markets facilitate the process of discovery.  Market participants compete to find solutions to problems.  EBS has identified a problem, and are trying to fix it using a fairly innovative change to the matching process.  If they’re right that some kinds of HFT are detrimental to market performance (and thereby reduces the demand for to trade on the platform), and their replacement of continuous trading with high speed calls impedes this detrimental HFT without impeding the good kind, they’ll make money.  And others will likely imitate, or take the basic idea and try to improve on it.

Or maybe it won’t work as planned. In which case they can try something else, or lose business to a platform that devises a better protocol.  The point is that a trading platform identified a problem with HFT, and is doing something about it, on its own.

All of this is far superior to top-down, one-size-fits-all government mandated “solutions” that are driven by political economy considerations first, efficiency considerations second . . . or third . . . or Nth.  Trading platforms may not internalize all the costs and benefits of better trading rules and fee structures, but their information and incentives are far better than regulators or legislatures.  Between competition among HFT firms, and the efforts of trading platforms to optimize the demand for their services, it’s likely that HFT’s rough edges will be smoothed out.  And if trading platforms don’t adopt such measures, you have to doubt seriously whether there’s a problem in the first place.  Because if there is a problem, they’d be the ones in the best position to know, and the best position to fix it.

Pre-Crisis, there was very little academic writing on clearing.  Post-Crisis, with questions about the role of derivatives in creating systemic risk, and the mandating of clearing of derivatives as a means of mitigating this problem, this is changing.  This is a good thing, but unfortunately, this burgeoning academic literature is at risk of irrelevance, and worse, of being misleading, because the theoretical models of clearing are nothing like clearing as it actually is.  These models tend to focus on the mutualization of risk within CCPs.  That’s important, but as I’ll discuss in more detail below, mutualization is not the most important feature of CCPs.  Collateralization is.

I’ll just talk about a couple of papers in detail, Adam Zawadowski’s Entangled Financial Systems in the most recent RFS, and Clearing, Counterparty Risk, and Aggregate Risk by Biais, Heider, and Hoerova.

These papers have some important insights, and I don’t want to seem overly critical.  I just want to persuade scholars that the focus of these papers, and many others, is misdirected, and to suggest where they should direct their attention.

“Entangled Financial Systems” presents a model of the periodic collapse of the financial system through the channel of inter-bank derivatives exposures.  The paper is rough sledding, in part because it tackles a complicated issue, and in part because the exposition and especially the proofs are hardly paragons of clarity: I have my doubts about some of the proofs.  That said, the story is a plausible one.  Banks use fragile capital structures (short maturity debt to fund long-lived assets) to solve an agency problem.  They use derivatives to manage risk in order to protect non-pledgeable income.  Banks are at risk of blowing up due to an idiosyncratic, exogenous shock: perhaps an operational risk, like a rogue trader.  If a bank blows up, its counterparties don’t get paid in full on their derivatives.  If these counterparties don’t insure against this risk, they may fail, and so on, with the result being a daisy chain of failures.  Thus, one idiosyncratic failure can lead to the collapse of the entire system.

In the model, the original risk of a blowup is idiosyncratic and insurable.  But in equilibrium, banks don’t buy insurance against a counterparty failing because they don’t internalize the impact of their failure on their other counterparties.  Thus, there is a “market failure”: the system blows up periodically because it is privately efficient but socially inefficient not to buy counterparty insurance.

One crucial issue with the paper is that most things are, laudably, endogenized, but one crucial thing is not: each bank can trade with only two counterparties located adjacently on a circle.  The choice of counterparties is exogenously specified. This concentration of counterparty exposure is crucial in making the system vulnerable to collapse.

Zawadowski recognizes that greater diversification of exposures across counterparties reduces the fragility of the system.  Although the externality may induce insufficient diversification across counterparties (because the systemic benefits of this aren’t internalized), market participants in reality have a variety of reasons to spread their trades across many counterparties.  Meaning that real financial systems may be less fragile than in the model, with its exogenously imposed concentration of exposures.

I’m also skeptical that an idiosyncratic risk at a single institution can bring down the entire financial system.  Look at some of the rogue trader losses-Kerviel at SocGen, Adoboli at UBS.  These guys cost their banks billions-but even such huge losses didn’t lead to a financial system meltdown via any channel, including a derivatives channel.  Instead, the 2007-2008 Crisis was related to a systemic shock-a decline in US real estate prices-that hit multiple financial institutions and investors.  It’s hard to identify an actual episode where the channel analyzed in the model-an idiosyncratic shock at a single financial institution-led to a financial meltdown.  The idiosyncratic nature of the risk is important: that’s what makes the risk insurable.  The paper therefore has little to say about non-insurable risks.

Where does clearing come in?  In the paper, clearing is a form of counterparty insurance.  Mandating clearing internalizes the externality.

There are several problems here.  The first is that in the model, counterparty insurance is supplied by a continuum of investors who can diversify away the risk.  A CCP in theory can also diversify the idiosyncratic risk by mutualizing it.  But note that CCPs are voluntary cooperative arrangements among financial institutions.  If there is a gain from collective action-internalizing the externality through cooperation-why don’t financial institutions voluntarily cooperate to form CCPs?   (Though in the context of the model, sharing the risk among financial institutions is more costly than passing the risk to investors.  Still, there is a collective benefit from cooperation among the financial institutions.)  Such cooperation increases joint wealth: why don’t market participants cooperate to internalize the externality?  What stands in the way of consummating such mutually beneficial bargains?   Moreover, why does it happen in some markets, not in others?

But the mutualization issue generally is the bigger problem, and the root of my qualms about the developing literature on CCPs.  Most of the models, including the Biais et al paper, formalize CCPs as a mutual risk sharing/insurance mechanism.  (Hell, I’ve done that myself.)

But mutualization is only one of the functions of CCPs As We Know Them.  Indeed, I am increasingly leaning to the view that it is the most problematic of their functions.

CCPs operate on the “defaulter pays” principle.  That is, real world CCPs attempt to choose margins (collateral) and default fund contributions so that they almost always cover a defaulter’s losses, and that non-defaulters’ default fund contributions are seldom used to make good a defaulter’s losses.  That is, default funds are tapped-and risk mutualized-only in rare, and arguably extreme, situations.

Put differently, only tail risk is mutualized in real world CCPs, and the primary function of CCPs is to set margins so that default losses are NOT mutualized, except in extreme circumstances.  CCPs are like monoline insurance of supersenior positions well down on the default waterfall.

In the context of the _ paper, this is particularly problematic, as he shows that collateralization is an inefficient way to address the externality problem.  It ties up valuable resources that could be used to fund positive NPV projects.   This is just one problem with collateral: the “initial margin problem”, if you will.  There’s also the variation margin problem that I’ve written about over the years.

I consider the tail-risk mutualization aspect of CCPs highly problematic because of the wrong way risk problem.  Like super-senior tranches of a CDO, losses hit the default fund during systemic episodes when those exposed to the default fund (the members of the CCP) are under stress.

This all means that the academic literature, which is modeling CCPs as mutual insurers, has two big problems.

The first problem is one of positive economics.  Existing models are not able to predict (a) why clearinghouses form, and (b) why they are primarily mechanisms to net and collateralize exposures, and only mutualize extreme risks.  They do not predict why market participants sometimes cooperate to implement a “defaulter pays” model, and sometimes don’t-and why they have never implemented a fully mutualized insurance scheme.

We need models that help us understand why market participants sometimes cooperate to implement a defaulter pays mechanism, supplemented by mutualization of the extreme risks; why they sometimes don’t; and why they never fully mutualize.  That is, we need to understand why so few risks are mutualized, even when market participants choose to form CCPs.

With all due modesty, I think the answers will will be found in my original analysis from the 90s: that the usual bugbears of insurance-adverse selection and moral hazard-make it uneconomically costly to mutualize risk, and that these problems also make centralized/delegated setting of collateral levels more costly than bilateral arrangements for doing so, depending on the characteristics of the traded instruments and those trading them.

The second problem is one of normative economics and policy prescriptions derived from models.  Policy recommendations based on models of CCPs that are flatly contrary to the way CCPs really operate are highly misleading, and dangerous.  Eliding from a model that says “mutualization of risk is socially efficient but privately unprofitable” to prescribing a policy of mandating CCPs is deeply flawed, when in practice CCPs won’t mutualize risk as in the models, but will instead implement a defaulter pays model.  (NB: of late regulators are telling CCPs that their main source of concern is that CCPs will set margins too low.  That is, regulators want to make sure that CCPs really make defaulters pay.  So in practice, mandated CCPs will mutualize only extreme risks, and collateralize the rest.  Nothing like what’s in the models.)   It’s a sort of bait-and-switch.

Prudent normative policy recommendations need to be based on a model with good positive content. We need to understand much better why market participants eschew implementing the defaulter pay model before mandating it.  For that’s what clearing mandates do: they impose the implementation of defaulter pays, NOT the implementation of the kinds of mutualization in many formal models of clearing.   Given the costliness of collateral (IM-not to mention the destabilizing effects of VM), it is particularly misleading to advocate policy measures that will lead to increased collateralization based on models in which collateral plays no role whatsoever.  Again-a bait-and-switch.

In sum, I’m pleased that many talented scholars are turning their attention to clearing. Especially when they cite me But it will be a shame if these scholars go on a wild goose chase, and construct models of CCPs that are completely disjoint from real world CCPs.  This goes double when they make policy prescriptions based on their models.

We need to understand the costs and benefits of defaulter pays, with the mutualization of only extreme risks.  For that’s what CCPs are really about.  We need models that are based on an understanding of collective action issues-because non-mandated CCPs are institutions that arise from collective action.  Only when we understand these issues should we have much confidence about making policy recommendations.