Streetwise Professor

In a weird counterpoint to the LME nickel story, another big clearing-related story that is causing a lot of consternation in derivatives circles is FTX exchange’s proposal to move to a direct clearing model that would dispense with FCMs as intermediaries. Instead of having an FCM interposed between a customer and the clearinghouse, the customer interfaces directly with the FTX Derivatives Clearing Organization (DCO).

What is crucial here is how this is supposed to work: FTX will utilize near real time mark-to-market and variation margin payments. Moreover, the exchange will automate the liquidation of undermargined positions, again basically in real time.

The mechanics are described here.

FTX describes this as being the next big thing in the derivatives markets, and a way of addressing systemic risks. Basically the pitch is simple: “real time margining allows us to operate a pure no credit/loser pays system.”

FTX touts this as a feature, but as the nickel experience demonstrates (and other previous episodes demonstrate) it is not. Margining generally can be destabilizing, especially during stressed market conditions, and the model FTX is advancing exacerbates the destabilizing potential of margining.

The mechanical means of addressing margin shortfalls on a real time frequency increases the tight coupling on the exchange, and is tailor made to create destabilizing positive feedback loops: prices move a lot leading to margin shortfalls in real time that trigger real time trades that accentuate the price movement. It is like seeding the market with huge numbers of stop orders, which are inherently destabilizing. Further, they can create incentives to manipulate. Anyone who can get some idea of where the stops are can “gun the stops” and trigger big price moves.

This instability potential can be exacerbated by the ability of traders to hold collateral in the form of the “underlying” (i.e., crypto, at present). Well, the collateral value can fluctuate, and that can contribute to margin shortfalls which again trigger stops.

Market participants can mitigate getting stopped out by substantially over-margining, i.e., holding a lot of excess margin in their FTX account. But this is a cash inefficient way of trading.

It’s not clear to me whether FTX will pay interest on collateral. It seems not. Hmmm. Implementing a model that incentivizes holding a lot of extra cash at FTX and not paying interest. Cynic that I am, that seems to be a great way to bet on higher interest rates! Maybe that’s FTX’s real game here.

I would also note that the “no leverage” story here reflects a decidedly non-systemic view (something that I pointed out years ago in my critiques of clearing mandates). Yes, real time margining plus holding of substantial excess margin reduces to a small level the amount of leverage extended by the CCP/DCO. But that is different than reducing the amount of margin in the system as a whole. People who have borrowing capacity and optimal total leverage targets can fund their deposits at FTX with leverage from other sources. They can offset the leverage they normally obtain from FCMs by taking more leverage from other sources.

In sum, FTX is arguing that its mechanism of direct clearing and real time margining creates a far more effective “no credit” clearing system than the existing FCM-intermediated structure. That’s likely true. But as I’ve banged on about for years, that’s not necessarily a good thing. The features that FTX touts as advantages have very serious downsides–especially in stressed market conditions where they tend to accelerate price moves rather than dampen them.

Insofar as this being a threat to the existing intermediated system, which many in the industry appear to fear, I am skeptical. In particular, the cash inefficiency of this mechanism will make it unattractive to many market participants. Not to be Panglossian, but the existing intermediated system evolved as it did for good economic reasons. It trades off credit risk and liquidity risk. It does so in a somewhat discriminating way because it takes into account the creditworthiness of market participants (something that FTX brags is unnecessary in its system). FTX is something of a corner solution that the market has not adopted despite the opportunity to do so. As a result, I don’t think that corner solution will have widespread appeal going forward.

If you use the official LME nickel and copper prices from Monday, before the exchange stopped trading of nickel, you can determine that the value of the metal in a US nickel coin is worth a dime. As the shutdown lingers, one wonders whether the LME is too.

The broad contours of the story are understood. A large Chinese nickel firm (Tsingshan Holdings, largest in the world) was short large amounts of LME nickel, allegedly as a hedge. But the quantity involved seems very outsized as a hedge, representing something like two years of output. And if the position was concentrated in nearby prompt dates (e.g., 3 months) it involved considerable curve risk.

The Russian invasion juiced the price of nickel, not surprising given Russia’s outsized presence in that market. That triggered a margin call (allegedly $1 billion) that the firm couldn’t meet–or chose not to. That led its brokers to try to liquidate its position in frenzied buying on Monday evening. This short covering drove the price from the close of around $48,000 to over $100,000.

That’s where things got really sick. The LME shut the nickel market. It was supposed to reopen today, but that’s been kicked down the road. But the LME didn’t stop there. It decided that these prices did not “[reflect] the the underlying physical market,” and canceled the trades. Tore them up. Poof! Gone!

Now in a Back to the Future moment echoing the 1985 Tin Crisis, the LME is trying to get the longs and shorts to set off their positions. “Can’t we all just get along?” Well likely not, because it obviously requires agreeing on a price. Which is obviously devilish hard, if not impossible given how much money changes hands with every change in price. (In my 1995 JLE paper on exchange self-regulation, I argued that exchanges historically did not want to intervene in this fashion even during obvious manipulations because of the rent seeking battles this would trigger.)

So the LME remains closed.

Some observations.

First, told ya. Seriously, in my role as Clearing Cassandra during the Frankendodd era, I said (a) clearing was not a panacea that would prevent defaults, and (b) the clearing mechanism was least reliable precisely during periods of major market stress, and that the rigid margining mechanism is what would threaten its ability to operate. That’s exactly what happened here.

Second, clearing is supposed to operate under a “loser pays/no credit” model. That’s really something of a misconception, because even though the clearinghouse does not extend credit, intermediaries (brokers/FCMs) routinely do to allow their clients to meet margin calls. But here we evidently have a situation in which the brokers (or Tsingshan’s banks) were unwilling or unable to do so, which led to the failure of the loser to pay.

Third, by closing the market, the LME is effectively extending credit (“you can pay me later”), and giving Tsingshan (and perhaps other shorts) some time to stump up some additional loans. Apparently JPM and the Chinese Construction Bank have agreed in principle to do so, but a deal has been hung up over what collateral Tsingshan will provide. So the market remains closed.

For its part, Tsingshan and its boss Xiang “Big Shot” Guangda are hanging tough. The company wants to maintain its short position. Arguably it has a strong bargaining position. To modify the old joke, if you owe the clearinghouse $1 million and can’t pay, you have a problem: if you owe the clearinghouse billions and can’t pay, the clearinghouse has a problem.

The closure of the market and the cancelation of the trades suggests that the LME has a very big problem. The exact amounts owed are unknown, but demanding all amounts owed now could well throw many brokers into default, and the kinds of numbers being discussed are as large or larger than the LME’s default fund of $1.2 billion (as of 3Q21 numbers which were the latest I could find).

So it is not implausible that a failure to intervene would have resulted in the insolvency of LME Clear.

The LME has taken a huge reputational hit. But it had to know it would when it acted as it did, implying that the alternative would have been even worse. The plausible worst alternative would have been a collapse of the clearinghouse and the exchange. Hence my quip about whether the exchange that trades nickel is worth a dime.

Among the reputational problems is the widespread belief that the Chinese-owned exchange intervened to bail out Chinese brokerage firms and a Chinese client. To be honest, this is hard to differentiate from intervening to save itself: the failure of the brokerages are exactly what would have brought the exchange into jeopardy.

I would say that one reason Xiang is hanging tough is that the CCP has his back. Not CCP as in central counterparty, but CCP as in Chinese Communist Party. That would give Tsingshan huge leverage in negotiations with banks, and the LME.

So the LME is playing extend and pretend, in the hope that it can either strongarm market participants into closing out positions, or prices return to a level that reduce shorts’ losses and therefore the amounts of variation margin they need to pay.

I seriously wonder why anyone would trade on the open LME markets (e.g., copper) for reasons other than reducing positions–and therefore reducing their exposure to LME Clear. The creditworthiness of LME Clear is obvious very dodgy, and it is potentially insolvent.

Fourth, in an echo of the first point, this episode demonstrates that central clearing, with its rigid “no credit” margining system is hostage to market prices. This is usually presented as a virtue, but when markets go wild it is a vulnerability. Which is exactly why it is–and always was–vain to rely on clearing as a bulwark against systemic risk. It is most vulnerable precisely during periods of market stress.

All commodity markets are experiencing large price movements that are creating extraordinary variation margin flows, potential positive feedbacks, and the prospect for troubles at other clearers. Further, the broader economic fallout from the Ukraine war (which includes, for example, a large recession resulting from the commodity price shocks, or a Russian debt default) has the real potential to disrupt equity and bond markets. This would put further strains on the financial markets, and the clearing system in particular. Central Banks–notably the Fed–had to supply a lot of liquidity to address shocks during the Covid Panic of March 2020. Two years later, they may have to ride to the rescue again.

Former Treasury Secretary Timothy Geithner–better known as Timmy! to loooooongtime readers of this blog–is back, this time as Chair of the Group of 30 Working Group on Treasury Market Liquidity. The Working Group was tasked with addressing periodic seizures in the Treasury securities market, most notoriously during the onset of the Covid crisis in March 2020–something I wrote about here.

This is a tale of two reports: the diagnosis is spot on, the prescription pathetic.

The report recognizes that

the root cause of the increasing frequency of episodes of Treasury market dysfunction under stress is that the
aggregate amount of capital allocated to market-making by bank-affiliated dealers has not kept pace with the very rapid growth of marketable Treasury debt outstanding

In other words, supply of bank market making services has declined, and demand for market making services has gone up. What could go wrong, right?

Moreover, the report recognizes the supply side root cause of the root cause: post-Financial Crisis regulations, and in particular the Supplemental Leverage Ratio, or SLR:

Post-global financial crisis reforms have ensured that banks have adequate capital, even under stress, but certain provisions may be discouraging market-making in U.S. Treasury securities and Treasury repos, both in normal times and especially under stress. The most significant of those provisions is the Basel III leverage ratio, which in theUnited States is called the Supplementary Leverage Ratio (SLR) because all banks in the United States (not just internationally active banks) are subject to an additional “Tier 1”leverage ratio.

Obviously fiscal diarrhea has caused a flood of Treasury issuance that from time to time clogs the Treasury market plumbing, but that’s not something the plumber can fix. The plumber can put in bigger pipes, so of course the report recommends wholesale changes in the constraints on market making, the SLR in particular, right? Right?

Not really. Recommendation 6–SIX, mind you–is “think about doing something about SLR sometime”:

Banking regulators should review how market intermediation is treated in existing regulation, with a view to identifying provisions that could be modified to avoid disincentivizing market intermediation, without weakening overall resilience of the banking system. In particular, U.S. banking regulators should take steps to ensure that risk-insensitive leverage ratios function as backstops to risk-based capital requirements rather than constraints that bind frequently.

Wow. That’s sure a stirring call to action! Review with a view to. Like Scarlett O’Hara.

Rather than addressing either of what itself acknowledges are the two primary problems, the report recommends . . . wait for it . . . more central clearing of the Treasury market. Timothy Geithner, man with a hammer, looking for nails.

Clearing cash Treasuries will almost certainly have a trivial effect on market making capacity. The settlement cycle in Treasuries is already one day–something that is aspirational (don’t ask me why) in the stock market. That already limits significantly the counterparty credit risk in the market (and it’s not clear that counterparty credit risk is a serious impediment on market making, especially since it existed before the recent dislocations in the Treasury market, and therefore is unlikely to have been a major contributor to them).

The report recognizes this: “Counterparty credit risks on trades in U.S. Treasury securities are not as large as those in other U.S. financial markets, because the contractual settlement cycle for U.S. Treasury securities is shorter (usually one day) and Treasury security prices generally are less volatile than other securities prices.” Geithner (and most of the rest of the policymaking establishment) were wrong about clearing being a panacea in the swap markets: it’s far less likely to make a material difference in the market for cash Treasuries.

The failure to learn over the past decade plus is clear (no pun intended!) from the report’s list of supposed benefits of clearing, which include

reduction of counterparty credit and liquidity risks through netting of counterparty exposures and application of margin requirements and other risk mitigants, the creation of additional market-making capacity at all dealers as a result of recognition of the reduction of exposures achieved though multilateral netting

As I wrote extensively in 2008 and the years following, netting does not reduce counterparty credit risk or exposures: it reallocates them. Moreover, as I’ve also been on about for more than a fifth of my adult life (and I’m not young!), “margin requirements” create their own problems. In particular, as the report notes, as is the case in most crises the March 2020 Treasury crisis sparked a liquidity crisis–liquidity not in terms of the depth of Treasury markets (though that was an issue) but liquidity in terms of a large increase in the demand for cash. Margin requirements would likely exacerbate that, although the incremental effect is hard to determine given that existing bilateral exposures may be margined (something the report does not discuss). As seen in the GameStop fiasco, a big increase in margins in part driven by the central counterparty (ironically the DTCC, the parent of the FICC which the report wants to be the clearinghouse for its expanded clearing of Treasuries) was a major cause of disruptions. For the report to ignore altogether this issue is inexcusable.

Relatedly, the report touches only briefly on the role of basis trades in the events of March 2020. As I showed in the article linked above, these were a major contributor to the dislocations. And why? Precisely because of margin calls on futures.

Thus, the report fails to analyze completely its main recommendation, and in fact its recommendation is based on not just an incomplete but a faulty understanding of the implications of clearing (notably its mistaken beliefs about the benefits of netting). That is, just like in the aftermath of 2008, supposed solutions to systemic risk are based on decidedly non-systemic analyses.

Instead, shrinking from the core issue, the report focuses on a peripheral issue, and does not analyze that properly. Clearing! Yeah, that’s the ticket! Good for whatever ails ya!

In sum, meet the new Timmy! Same as the old Timmy!

One of the few compensations I get from a Biden administration is that I have an opportunity to kick around Gary Gensler–“GiGi” to those in the know–again. Apparently feeling his way in his first few months as Chairman of the SEC, Gensler has been relatively quiet, but today he unburdened himself with deep thoughts about stock market structure. If you didn’t notice, “deep” was sarcasm. His opinions are actually trite and shallow, and betray a failure to ask penetrating questions. Plus ça change, plus c’est la même chose.

Not that he doesn’t have questions. About payment for order flow (“PFOF”) for instance:

Payment for order flow raises a number of important questions. Do broker-dealers have inherent conflicts of interest? If so, are customers getting best execution in the context of that conflict? Are broker-dealers incentivized to encourage customers to trade more frequently than is in those customers’ best interest?

But he misses the big question: why is payment for order flow such a big deal in the first place?

Relatedly, Gensler expresses concern about what traders do in the dark:

First, as evidenced in January, nearly half of the trading interest in the equity market either is in dark pools or is internalized by wholesalers. Dark pools and wholesalers are not reflected in the NBBO. Moreover, the NBBO is also only as good as the market itself. Thus, under the segmentation of the current market, nearly half of trading along with a significant portion of retail market orders happens away from the lit markets. I believe this may affect the width of the bid-ask spread.

Which begs the question: why is “nearly half of the trading interest in the equity market either is in dark pools or is internalized by wholesalers”?

Until you answer these big questions, studying the ancillary ones like his regarding PFOF an NBBO is a waste of time.

The economics are actually very straightforward. In competitive markets, customers who impose different costs on suppliers will pay different prices. This is “price discrimination” of a sort, but not price discrimination based on an exploitation of market power and differences in customer demand elasticities: it is price differentiation based on differences is cost.

Retail order flow is cheaper to intermediate than institutional order flow. Some institutional order flow is cheaper to intermediate than other such flows. Competitive pressures will find ways to ensure flows that are cheaper to intermediate pay lower prices. PFOF, dark pools, etc., are all means of segmenting order flow based on cost.

Trying to restrict cost-based price differences by banning or restricting certain practices will lead clever intermediaries to find other ways to differentiate based on cost. This has always been so, since time immemorial.

In essence, Gensler and many other critics of US market structure want to impose uniform pricing that doesn’t reflect cost differences. This would be, in essence, a massive scheme of cross subsidies. Ironically, the retail traders for whom Gensler exhibits such touching concern would actually be the losers here.

Cross subsidy schemes are inherently unstable. There are tremendous competitive pressures to circumvent them. As the history of virtually every regulated sector (e.g., transportation, communications) has demonstrated for decades, and even centuries.

From a positive political economy perspective, the appeal of such cross subsidy schemes to regulators is great. As Sam Peltzman pointed out in his amazing 1976 JLE piece “Toward a More General Theory of Regulation,” regulators systematically attempt to suppress cost-based price differences in order to redistribute rents to gain political support. The main impetus for deregulation is innovation that exploits gains from trade from circumventing cross subsidy schemes–deregulation in banking (Regulation Q) and telecoms are great examples of this.

So who would the beneficiaries of this cross-subsidization scheme be? Two major SEC constituencies–exchanges, and large institutional traders.

In other words, all this chin pulling about PFOF and dark markets is politics as usual. Furthermore, it is politics as usual in the cynical sense that the supposed beneficiaries of regulatory concern (retail traders) are the ones who will be shtupped.

Gensler also expressed dismay at the concentration in the PFOF market: yeah, he’s looking at you, Kenneth. Getting the frequency?

Although Gensler’s systemic risk concern might have some justification, he still fails to ask the foundational question: why is it concentrated? He doesn’t ask, so he doesn’t answer, instead saying: “Market concentration can deter healthy competition and limit innovation.”

Well, concentration can also be the result of healthy competition and innovation (h/t the great Harold Demsetz). Until we understand the existing concentration we can’t understand whether it’s a bug or feature, and hence what the appropriate policy response is.

Gensler implicitly analogizes say Citadel to Facebook or Google, which harvest customer data and can exploit network effects which drives concentration. The analogy seems very strained here. Retail order flow is cheap to service because it is uninformed. Citadel (or other purchasers of order flow) isn’t learning something about consumers that it can use to target ads at them or the like. The main thing it is learning is what sources of order flow are uninformed, and which are informed–so it can avoid paying to service the latter.

Again, before plunging ahead, it’s best to understand what are the potential agglomeration economies of servicing order flow.

Gensler returns to one of his favorite subjects–clearing–at the end of his talk. He advocates reducing settlement time from T+2: “I believe shortening the standard settlement cycle could reduce costs and risks in our markets.”

This is a conventional–and superficial–view that suggests that when it comes to clearing, Gensler is like the Bourbons: he’s learned nothing, and forgotten nothing.

As I wrote at the peak of the GameStop frenzy (which may repeat with AMC or some other meme stock), shortening the settlement cycle involves serious trade-offs. Moreover, it is by no means clear that it would reduce costs or reduce risks. The main impact would be to shift costs, and transform risks in ways that are not necessarily beneficial. Again, shortening the settlement cycle involves a substitution of liquidity risk for credit risk–just as central clearing does generally, a point which Gensler was clueless about in 2010 and is evidently equally clueless about a decade later.

So GiGi hasn’t really changed. He is sill offering nostrums based on superficial diagnoses. He fails to ask the most fundamental questions–the Chesterton’s Fence questions. That is, understand why things are they way they are before proposing to change them.

Shuttling between FUBARs, it’s back to GameStop!

Last week there were House hearings regarding the GameStock saga. As is usual with these things, they were more a melange of rampant narcissism and political posing and outright stupidity than a source of information. Everyone had an opportunity to identify and then flog their favorite villains and push their favorite “solutions.” All in all, very few constructive observations or remedies came out of the exercise. I’m sure you’re shocked.

Here are a few of the main issues that came up.

Shortening the securities settlement cycle. The proximate cause of Robinhood’s distress was a huge margin call. Market participants post margins to mitigate the credit risk inherent in a two day settlement cycle. Therefore, to reduce margins and big margin calls, let’s reduce the settlement cycle! Problem solved!

No, problem moved. Going to T+0 settlement would require buyers to stump up the cash and sellers to secure the stock on the same day of the transaction. Almost certainly, this wouldn’t result in a reduction of credit in the system, but just cause buyers to borrow money to meet their payment obligations. Presumably the lenders would not extend credit on an unsecured basis, but would require collateral with haircuts, where the haircuts will vary with risk: bigger haircuts would require the buyers to put up more of their own cash.

I would predict that to a first approximation the amount of credit risk and the amount of cash buyers would have to stump up would be pretty much the same as in the current system. That is, market participants would try to replicate the economic substance of the way the market works now, but use different contracting arrangements to obtain this result.

I note that when the payments system went to real time gross settlement to reduce the credit risk participants faced through the netting mechanism with daily settlement, central banks stepped in to offer credit to keep the system working.

It’s also interesting to note that what DTCC did with GameStop is essentially move to T+0 settlement by requiring buyers to post margin equal to the purchase price:

Robinhood made “optimistic assumptions,” Admati said, and on Jan. 28, Tenev woke up at 3:30 a.m. and faced a public crisis. With a demand from a clearinghouse to deposit money as a safety measure hedging against risky trades, he had to get $1 billion from investors. Normally, Robinhood only has to put up $2 for every $100 to vouch for their clients, but now, the whole $100 was required. Thus, trading had to be slowed down until the money could be collected.

That is, T+0 settlement is more liquidity/cash intensive. As a result, a movement to such a system would lead to different credit arrangements to provide the liquidity.

As always, you have to look at how market participants will respond to proposed changes. If you require them to pay cash sooner by changing the settlement cycle, you have to ask: where is the cash going to come from? The likely answer: the credit extended through the clearing system will be replaced with some other form of credit. And this form is not necessarily preferable to the current form.

Payment for order flow (“PFOF”). There is widespread suspicion of payment for order flow. Since Robinhood is a major seller of order flow, and since Citadel is a major buyer, there have been allegations that this practice is implicated in the fiasco:

Reddit users questioned whether Citadel used its power as the largest market maker in the U.S. equities market to pressure Robinhood to limit trading for the benefit of other hedge funds. The theory, which both Robinhood and Citadel criticized as a conspiracy, is that Citadel Securities gave deference to short sellers over retail investors to help short sellers stop the bleeding. The market maker also drew scrutiny because Citadel, the hedge fund, together with its partners, invested $2 billion into Melvin Capital Management, which had taken a short position in GameStop.

To summarize the argument, Citadel buys order flow from Robinhood, Citadel wanted to help out its hedge fund bros, something, something, something, so PFOF is to blame. Association masquerading as causation at its worst.

PFOF exists because when some types of customers are cheaper to service than others, competitive forces will lead to the design of contracting and pricing mechanisms under which the low cost customers pay lower prices than the high cost customers.

In stock trading, uninformed traders (and going out on a limb here, but I’m guessing many Robinhood clients are uninformed!) are cheaper to intermediate than better informed traders. Specifically, market makers incur lower adverse selection costs in dealing with the uninformed. PFOF effectively charges lower spreads for executing uninformed orders.

This makes order flow on lit exchange markets more “toxic” (i.e., it has a higher proportion of informed order flow because some of the uninformed flow has been siphoned off), so spreads on those markets go up.

And I think this is what really drives the hostility to PFOF. The smarter order flow that has to trade on lit markets doesn’t like the two tiered pricing structure. They would prefer order flow be forced onto lit markets (by restricting PFOF). This would cause the uninformed order flow to cross subsidize the more informed order flow.

The segmentation of order flow may make prices on lit markets less informative. Although the default response among finance academics is to argue that more informative is better, this is not generally correct. The social benefit of more accurate prices (e.g., does that lead to better investment decisions) have not been quantified. Moreover, informed trading (except perhaps, ironically, for true insider trading) involves the use of real resources (on research, and the like). Much of the profit of informed trading is a transfer from the uninformed, and to the extent it is, it is a form of rent seeking. So the social ills of less informative prices arising from the segmentation of order flow are not clearcut: less investment into information may actually be a social benefit.

There is a question of how much of the benefit of PFOF gets passed on to retail traders, and how much the broker pockets. Given the competitiveness of the brokerage market–especially due to the entry of the likes of Robinhood–it is likely a large portion gets passed on to the ultimate customer.

In sum, don’t pose as a defender of the little guy when attacking PFOF. They are the beneficiaries. Those attacking PFOF are actually doing the bidding of large sophisticated and likely better informed investors.

HFT. This one I really don’t get. There is HFT in the stock market. Something bad happened in the stock market. Therefore, HFT caused the bad thing to happen.

The Underpants Gnomes would be proud. I have not seen a remotely plausible causal chain linking HFT to Robinhood’s travails, or the sequence of events that led up to them.

But politicians gonna politician, so we can’t expect high order logical thinking. The disturbing thing is that the high order illogical thinking might actually result in policy changes.

One metaphor that might shed some light on how seemingly small events can have cascading–and destructive–effects in financial markets is to think of the financial system as consisting of borgs programmed to ensure their survival at all costs.

One type of borg is the clearinghouses/CCP borg. The threat to them is the default of their counterparties. They use margins to protect against these defaults (thereby creating a loser pays/no credit system). When volatility increases, or gap risk increases, or counterparty concentration risk increases–or all three increase–the CCP Borg responds to this greater risk of credit loss by raising margins–sometimes by a lot–in order to protect itself.

This puts other borgs (e.g., Hedge Fund Borgs) under threat. They try to borrow money to pay the CCP Borg’s margin demands. Or they sell liquid assets to raise the cash.

These actions can move prices more–including the prices of things that are totally different from what caused the CCP Borg to raise margins on. This can cause increases in volatility that triggers reactions by other Managed Money Borgs. For example, these Borgs may utilize a Value-at-Risk system to detect threats, and which is programmed to cause the MM Borg to reduce positions (i.e., try to buy and sell stuff) in order to reduce VaR, which can move prices further, triggering more volatility. Moreover, the simultaneous buying and selling of a lot of various things by myriad parties can affect correlations between prices of these various things. And correlation is an input into the borgs’ model, so this can lead to more borg buying and selling.

All of these price changes and volatility changes can impact other borgs. For example, increases in volatilities and correlations in many assets that results from Managed Money Borgs’ buying and selling will feed back to the CCP Borgs, whose self-defense models are likely to require them to increase their margins on many more instruments than they increased margins on in the first place.

This is how seemingly random, isolated shocks like retail trader bros piling into heavily shorted, but seemingly trivial, stocks can spill over into the broader financial system. Borgs programmed to survive, acting in self-defense, take actions that benefit themselves but have detrimental effects on other borgs, who act in self-defense, which can have detrimental effects on other borgs, and . . . you get the picture.

This is a quintessential example of “normal accidents” in a complex system with tightly coupled components. Other examples include reactor failures and plane crashes.

I note–again, reprising a theme of the Frankendodd Years of this blog–that clearing and margins are a major reason for tight coupling, and hence greater risk of normal accidents.

I note further that it is precisely the self-preservation instincts of the borgs that makes it utterly foolish and clueless to say that creating stronger borgs with more powerful tools of self-preservation, and which interact with other borgs, will reduce systemic risk. This is foolish and clueless precisely because it is profoundly unsystemic thinking because it views the borgs in isolation and ignores how the borgs all interact in a tightly coupled system. Making borgs stronger can actually make things worse when their self-preservation programs kick in, and the self-preservation of one borg causes it to attack other borgs.

Why do teenagers in slasher flicks always go down into the dark basement after five of their friends have been horribly mutilated? Well, that makes about as much sense as a lot of financial regulators have in the past decades. Despite literally centuries of bad historical experiences, they have continued to try to make stronger, mutually interacting, borgs. Like Becky’s trip down the dark basement stairs, it never ends up well.

The vertigo inducing story of GameStop ramped it up to 11 yesterday, with a furore over Robinhood’s restriction of trading in GME to liquidation only, and the news that it had sold out of its customers’ positions without the customers’ permission. These actions are widely perceived as an anti-populist capitulation to Big Finance.

Well, they are in a way–but NOT the way that is being widely portrayed. What is going on is an illustration of the old adage that clearing and settlement in securities markets (like the derivatives markets) is like the plumbing–you take it for granted until the toilet backs up.

You can piece together that Robinhood was dealing with a plumbing problem from a couple of stories. Most notably, it drew down on credit lines and tapped some of its big executing firms (e.g., Citadel) for cash. Why would it need cash? Because it needs to post margin to the Depositary Trust Clearing Corporation (DTCC) on its open positions. Other firms are in similar situations, and directly or indirectly GME positions give rise to margin obligations to the DTCC.

The rise in price alone increased margin requirements because given volatility, the higher the price of a stock, the larger the dollar amount of potential loss (e.g., the VaR) that can occur prior to settlement. This alone jacks up margins. Moreover, the increase in GME volatility, and various adders to margin requirements–most notably for gap risk and portfolio concentration–ramp up margins even more. So the action in GME has led to a big increase in margin requirements, and a commensurate need for cash. Robinhood, as the primary venue for GME buyers, had/has a particularly severe position concentration/gap problem. Hence Robinhood’s scramble for liquidity.

Given these circumstances, liquidity was obviously a constraint for Robinhood. Given this constraint, it could not handle additional positions, especially in GME or other names that create particularly acute margin/liquidity demands. It was already hitting a hard constraint. The only practical way that Robinhood (and perhaps other retail brokers, like TDAmeritrade) could respond in the short run was trading for liquidation only, i.e., allow customers to sell their existing GME positions, and not add to them.

By the way, trading for liquidation is a tool in the emergency action toolbook that futures exchanges have used from time-to-time to deal with similar situation.

To extend the plumbing analogy, Robinhood couldn’t add any new houses to its development because the sewer system couldn’t handle the load.

I remember some guy saying that clearing turns credit risk into liquidity risk. (Who was that guy? Pretty observant!) For that’s exactly what we are seeing here. In times of market dislocation in particular, clearing, which is intended to mitigate credit risk, creates big increases in demand for liquidity. Those increases can cause numerous knock on effects, including dislocations in markets totally unrelated to the original source of the dislocation, and financial distress at intermediaries. We are seeing both today.

It is particularly rich to see the outrage at Robinhood and other intermediaries expressed today by those who were ardent advocates of clearing as the key to restoring and preserving financial stability in the aftermath of the Financial Crisis. Er, I hate to say I told you so, but I told you so. It’s baked into the way clearing works, and in particular the way that clearing works in stressed market conditions. It doesn’t eliminate those stresses, but transfers them elsewhere in the financial system. Surprise!

The sick irony is that clearing was advocated as a means to tame big financial institutions, the banks in particular, and reduce the risks that they can impose on the financial system. So yes, in a very real sense in the GME drama we are seeing the system operate to protect Big Finance–but it’s doing so in exactly the way many of those screaming loudest today demanded 10 years ago. Exactly.

Another illustration of one of my adages to live by: be very careful what you ask for.

Margins are almost certainly behind Robinhood’s liquidating some customer accounts. If those accounts become undermargined, Robinhood (and indeed any broker) has the right to liquidate positions. It’s not even in the fine print. It’s on the website:

If you get a margin call, you need to bring your portfolio value (minus any cryptocurrency positions) back up to your minimum margin maintenance requirement, or you risk Robinhood having to liquidate your position(s) to bring your portfolio value (minus any cryptocurrency positions) back above your margin maintenance requirement.

Another Upside Down World aspect of the outrage we are seeing is the stirring defenses of speculation (some kinds of speculation by some people, anyways) by those in politics and on opinion pages who usually decry speculation as a great evil. Those who once bewailed bubbles now cheer for them. It’s also interesting to see the demonization of short sellers–whom those with average memories will remember were lionized (e.g., “The Big Short”) for blowing the whistle on the housing boom and the bank-created and -marketed derivative products that it spawned.

There are a lot of economic issues to sort through in the midst of the GME frenzy. There will be in the aftermath. Unfortunately, and perhaps not surprisingly given the times, virtually everything in the debate has been framed in political terms. Politics is all about distributive effects–helping my friends and hurting my enemies. It’s hard, but as an economist I try to focus on the efficiency effects first, and lay out the distributive consequences of various actions that improve efficiency.

What are the costs and benefits of short selling? Should the legal and regulatory system take a totally hands off approach even when prices are manifestly distorted? What are the costs and benefits of various responses to such manifest price distortions? What are the potential unintended consequences of various policy responses (clearing being a great example)? These are hard questions to answer, and answering them is even harder in the midst of a white-hot us vs. them political debate. And I can say with metaphysical certainty that 99 percent of the opinions I have seen expressed about these issues in recent days are steeped in ignorance and fueled by emotion.

There are definitely major problems–efficiency problems–with Big Finance and the regulation thereof. Ironically, many of these efficiency problems are the result of previous attempts to “solve” perceived problems. But that does not imply that every action taken to epater les banquiers (or frapper les financiers) will result in efficiency gains, or even benefit those (often with justification) aggrieved at the bankers. I thus fear that the policy response to GameStop will make things worse, not better.

It’s not as if this is new territory. I am reminded of 19th century farmers’ discontent with banks, railroads, and futures trading. There was a lot of merit in some of these criticisms, but all too often the proposed policies were directed at chimerical wrongs, and missed altogether the real problems. The post-1929 Crash/Great Depression regulatory surge was similarly flawed.

And alas, I think that we are doomed to repeat this learning the wrong lessons in the aftermath of GameStop and the attendant plumbing problems. Virtually everything I see in the public debate today reinforces that conviction.

The very first substantive post on this blog, almost 15 years ago, was about a failure of the electronic trading system at the Tokyo Stock Exchange.

Whoops, they did it again!

Apparently believing that misery loves company, Euronext has also experienced failures.

Euronext’s problems seem quite more frightening, because they involve the out-trade from hell: reversing the polarity on transactions:

“It has been identified that some of the 19/10 trades sent yesterday to the CCPs (central counterparty clearing house) had the wrong buy/sell direction”, Euronext said.

Thought you were long? Hahahahahaha. You’re short, sucker!

I hate it when that happens! (Yes, Euronext reversed the trades after it realized the problem.)

The lessons of my “Value of Lost Trade” (“VOLT”) piece still hold. It is inefficiently costly to drive the probability of a failure to zero. Whether exchanges have the efficient probability of failure (or really, the efficient vector of failure probabilities, because there are multiply types of failure) depends on the value of foregone trades when a system is down (or the cost of other types of errors, such as reversing trade direction).

Meaning that system failures will continue to occur, and long after this blog fades away.

Just another day at the Globex, folks. May WTI up a mere $49.88 on its last trading day at the time I write this paragraph, a while before the close. (Sorry, can’t calculate a percentage change . . . because the base number is negative!) That’s just sick. But at least it’s positive! ($12.25. No, $9.96. No . . .) (This reminds me of a story from Black Monday. My firm did a little index arb. We called the floor to get a price quote on the 19th. Our floor guy said “On this part of the pit it’s X. Over there it’s X+50. Over there it’s X-20. I have no fucking idea what the fucking price is.”)

But June has been crushed–down $7.35 (about 35 percent). Now the May-June spread is a mere $.83 contango. That makes as little sense as yesterday’s settling galactic contango (galactango!) of $57.06. (Note that June-July is trading at at $7.71 and July-August at $2.65.

I’m guessing that dynamic circuit breakers are impeding price movements, meaning that the prices we see are not necessarily market clearing prices at that instant.

A few follow-ons to yesterday’s post.

First, the modeling of the dynamics of a contract as it approaches expiration when the delivery supply/demand curve is inelastic, and some traders might have positions large enough to exploit those conditions to exercise market power, is extremely complicated. The only examples I am aware of are Cooper and Donaldson in the JFQA almost 30 years ago, and my paper in the Journal of Alternative Investments almost a decade ago.

Futures markets are (shockingly!) forward looking. Expectations and beliefs matter. There are coordination problems. If I believe everyone else on my side of the market is going to liquidate prior to expiration, I realize that the party on the other side of the contract will have no market power at expiration. So I should defer liquidating–which if everyone reasons the same way could lead to everyone getting caught in a long or sort manipulation at expiration. Or, if I believe everyone is going to stick it out to the end, I should get out earlier (which if everybody else does the same results in a stampede for the exits.)

In these situations, anything can happen, and the process of coordinating expectations and actions is likely to be chaotic. Cooper-Donaldson and Pirrong lay out some plausible stories (based on particular specifications of beliefs and the trading mechanism), but they are not the only stories. They mainly serve to highlight how game theoretic considerations can lead to very complex outcomes in situations with market power and inelasticity.

One thing that is sure is that these game theoretic considerations don’t matter much if the elasticities of delivery supply and demand are large. Then no individual can distort prices very much by delivering too much or taking delivery of too much. Then the coordination and expectations problems aren’t so relevant. However, when delivery supply or demand curves are very steep–as is the case in Cushing now due to the storage constraint–they become extremely relevant.

Perhaps one analogy is getting out of a theater. When there are many exits, there won’t be queues to get out and little chance of tragedy even if someone yells “fire.” If there is only one exit, however, hurried attempts of everyone to leave at once can lead to catastrophe. Moreover, perverse crowd dynamics occur in such situations. That’s where we were yesterday.

About 90 percent of open interest liquidated yesterday. That is why today is returning to some semblance of normality–the exit isn’t so crowded (because so many got trampled yesterday). But that begs the question of why the panicked rush yesterday? That’s where the game theoretic “anything could happen” answer is about the best we can do.

About that storage constraint. My post yesterday focused on someone with a large short futures position raising the specter of excessive deliveries by not liquidating that position, thereby triggering a cascade of descending offers until the short graciously accepted at a highly profitable price.

But there is another market power play possible here. A firm controlling storage could crash prices (and spreads) by withholding that capacity from the market. The most recent data from the EIA indicates about 55 mm bbl of oil storage at Cushing. That’s about 80 percent of nameplate capacity (also per EIA.). Due to operational constraints (e.g., need working space to move barrels in and out; can’t mix different grades in the same tank) that’s probably effectively full. Therefore, someone with ownership of a modest amount of space could withhold it drive up the spread. If that party had on a bull spread position . . .

Third, we are into Round Up the Usual Suspects mode:

And first in line is the US Oil ETF. There has been a lot of idiotic commentary about this. They were forced to take delivery! (Er, delivery notices aren’t possible before trading ends.) They were forced to dump huge numbers of contracts yesterday! (Er, they publish a regular roll schedule, and were out of the May a week before yesterday’s holocaust. They also report positions daily, and as of yesterday were 100 pct in the June.)

Not to say that USO can be implicated in hinky things going on in the June right now, but as for May–that dog don’t hunt.

Fourth–WTF, June WTI? Well, my best explanation is that the carnage in the May served to concentrate minds regarding June. No doubt risk managers, or risk systems, forced some longs out as the measured and perceived risk for June shot up yesterday. Others just decided that discretion was the better part of valor. The extremely unsettled positions no doubt impaired liquidity (i.e., just as some wanted to get out, others were constrained by risk limits formal or informal from getting in), leading to big price movements in response to these flows. If that’s a correct diagnosis, we should see something of a bounceback, but perhaps not too much given the perception (and reality) of an extremely asymmetric risk profile, with going into expiry short being a lot more dangerous than going into it long. (This is why expectations about future conditions at delivery can impact prices well before delivery.)

Fifth, on a personal note, in an illustration of the adage that the apple doesn’t fall far from the tree (and also of Merton’s Law of Multiples) my elder daughter Renee completely independently of me used “WTI WTF” in her daily market commentary yesterday. I’m so proud! She also raised the possibility of negative prices some time ago. Good call!

And I finish this just in time to bring you the final results. CLK goes off the board settling at $10.01, up a mere $47.64. CLM settles at $11.57, down -$8.86. The closing KM20 spread, $1.56.

Someday we’ll look back on this and . . . . Well, we’ll look back on it, anyways.

In the aftermath of the last crisis, I played the role of Clearing Cassandra, warning that in the next crisis, supersizing of derivatives clearing would create systemic risks not because clearinghouses would fail, but because of the consequences of what they would do to survive: hike initial margins and collect huge variation margin payments that would suck liquidity out of the system at the same time liquidity supply contracted. This, in turn, would lead to asset fire sales, that would distort asset prices which would lead to further knock-on effects.

I wrote a lot about this 2008-2012, but here is a convenient link. Key quote from the abstract:

The author also believes that the larger collateral mandates and frequent marking‐to‐market will make the financial system more vulnerable since margin requirements tend to be “pro‐cyclical.” And more rigid collateralization mechanisms can restrict the supply of funding liquidity, and lead to spikes in funding liquidity demand that can reduce the liquidity of traded instruments and generate destabilizing feedback loops. 

Well, the next crisis is here, and these (conditional) predictions are being borne out. In spades.

Here’s what I wrote a few days ago as a contribution to the Regulatory Fundamentals Group newsletter:

In the aftermath of the last crisis of 2008-2009, G20 nations decided to mandate clearing of standardized OTC derivatives transactions.  The current coronavirus crisis is the first since those reforms were implemented (via Dodd-Frank in the US, for example), and this therefore gives the first opportunity to evaluate the performance of the supersized clearing ecosystem in “wartime” conditions.  

So far, despite the extreme price movements across the entire derivatives universe–equities, fixed income, currencies, and commodities (especially oil)–there have been no indications that clearinghouses have faced either financial or operational disruption.  No clearing members have defaulted, and as of now, there have been no serious concerns than any are on the verge of default. 

That said, there are two major reasons for concern.

First, the unprecedented volatility and uncertainty show no signs of dissipating, and as long as it continues, major financial institutions–including clearing firms–are at risk.  The present crisis did not originate in the banking/shadow banking sector (as the previous one did), but it is now demonstrably affecting it.  There are strong indicators of stress in the financial system, such as the blowouts in FRA-OIS spreads and dollar swap rates (both harbingers of the last crisis).  Central banks have intervened aggressively, but these worrying signs have eased only slightly.  

Second, as I wrote repeatedly during the debate over clearing mandates in the post-2008 crisis period, the most insidious systemic risk that supersized clearing creates is not the potential for the failure of a clearinghouse (triggered by the failure of one or more clearing members).  Instead, the biggest clearing-related systemic risk is that the very measures that clearinghouses take to ensure their integrity–specifically, frequent variation margining/marking-to-market–lead to large increases in the demand for liquidity precisely during circumstances when liquidity is evaporating.  Margin payments during the past several weeks have hit unprecedented–and indeed, previously unimaginable–levels.  The need to fund these payments has inevitably increased the demand for liquidity, and contributed to the extraordinary demand for liquidity and the concomitant indicators stressed liquidity conditions (e.g., the spreads and extraordinary central bank actions mentioned earlier).  It is impossible to quantify this impact at present, but it is plausibly large.  

In sum, the post-2008 Crisis clearing system is operating as designed during the 2020 Crisis, but it is unclear whether that is a feature, or a bug.  

It is becoming more clear: Bug, and the bugs are breeding. There have been multiple stories over the last couple of days of margin calls on hedging positions causing fire sales, with attendant price dislocations in markets like for mortgages. Like here, here, and here. I guarantee there are more than have been reported, and there will be still more. Indeed, I bet if you look at any pricing anomaly, it has been created by, or exacerbated by, margin calls. (Look at the muni market, for instance.)

But those in charge still don’t get it. CFTC chairman Heath Tarbert delivers happy talk in the WSJ, claiming that everything is hunky dory because all them margins bein’ paid! and as a result, derivatives markets are functioning, CCPs aren’t failing, etc.

This is exactly the kind of non-systemic thinking about systemic risk that I railed about a decade ago. Mr. Tarbert has a siloed view: he is assigned some authority over a subset of the financial system, sees that it is working fine, and concludes that rules regarding that subset are beneficial for the system as a whole.

Wrong. Wrong. Wrong. Wrong. WRONG.

You have to look at the system as a whole, and how the pieces of the system interact.

In the post-last-crisis period I wrote about the “Levee Effect”, namely, that measures designed to protect one part of the financial system would flood others, with ambiguous (at best) systemic consequences. The cascading margins and the effects of those margin calls are exactly what I warned about (to the accompaniment a collective shrug by those who mattered, which is why we are where we are).

What we are seeing is unintended consequences–unintended, but not unforeseeable.

Speaking of unintended consequences, perhaps one good effect of September’s repo market seizure was that it awoke the Fed to its actual job–providing liquidity in times of stress. The facilities put in place in the aftermath of the September SNAFU are being expanded–by orders of magnitude–to deal with the current spike in liquidity demand (including the part of the spike due to margin issues). Thank God the Fed didn’t have to think this up on the fly.

It also appears that either (a) the restrictions on the Fed imposed by Frankendodd are not operative now, or (b) the Fed is saying IDGAF so sue me and blowing through them. Either way, such liquidity seizure are what the Fed was created to address.