Streetwise Professor

Today was one of the most epochal days in the history of oil trading, which is saying something. The front month May contract–which expires tomorrow (21 April, 2020)–(a) settled at a negative price of -$37.63, (b) declined $55.90 from the previous day’s settlement, and (c) exhibited a trading range of $58.17, which is about 3.5x Friday’s settlement price.

Even these eye-popping numbers don’t tell the full story. The last traded price was -$13.10. Note that the settlement price is based on the volume weighted average price during the last two minutes of trading, so an average price of $-37.63 in the last two minutes and a last traded price of -$13.10 means prices moved more that $25/bbl in these minutes.

And we’re not done yet! As I write at 1841 CDT, the price is up to -$5.00–an increase of $32.63.

That there’s what they call volatility, folks.

To put these numbers in perspective, the largest trading range on the any day of the last three trading days of CL contracts from 2000-2019 is $26.65, a day around the time of the financial crisis and the aftermath of Hurricane Ike (October 2008 contract): here’s what I wrote about that event. The median intra-day range on the last three days is $1.6, the mean is $1.5, and the standard deviation is $1.47. So we are around a 40 standard deviation event here.

The largest daily price change during the last 3 trading days is $16.37, with a median of $.73 and a standard deviation of $1.45.

The calendar spread is also extreme, settling at $58.06 between the June and the May. Meaning that if you had storage, you could get paid to take delivery, sell it forward, and lock in that $58.06 (net of what the storage costs you). I guess you could call that the megacontango.

All in all, a historically unprecedented day.

The proximate cause of these wild gyrations, and unprecedented negative prices is, of course, the collapse of demand and the looming exhaustion of storage space, including at the delivery point of Cushing, OK. But although this is a necessary condition for today’s events, it is not a full explanation.

The storage issue has been known for weeks, and discussed intensely. It had been priced in to a considerable degree: contango was already at a historically high level. What information about the availability of storage arrived between Friday and today? Unlikely to be anything that could cause such chaotic price movements.

The likely cause is the difficulty of liquidating about 100,000 open contracts (100 million barrels!) in such extreme technical conditions. It is plausible, and indeed likely, that strategic behavior–perhaps rising to the level of manipulation–is the major cause of how prices moved today against the background of conditions that were widely known on Friday.

Let me start out by noting that something similar, though not as extreme, occurred during the demand collapse and associated flooding of storage during the Financial Crisis. As I documented here, the expiries of the January, February, and March 2009 WTI contracts saw what were then historically unprecedented price collapses. So did other US grades of oil. Here’s a picture from the linked document:

The big downward spikes in the front month-back month spreads correspond with the days around expiry.

How does strategic behavior/market power/manipulation play into this? The model of short manipulation in my 1996 book (only $169 paperback–buy two!) and 1993 J. of Business article formalizes the argument, but the intuition is fairly straightforward. Manipulation exploits frictions and bottlenecks. (My article/book refer to “frictional manipulations.”). There is now a huge friction/bottleneck in Cushing–constrained storage. This bottleneck makes the demand curve for crude at Cushing extremely inelastic, and means that the movement of even small excess quantities of oil into that location will cause prices to decline dramatically.

In these conditions, a trader, or a group of traders with modest-sized short positions can exercise market power by delivering even a small amount of oil over and above the quantity that should flow to Cushing. This drives down the price and allows the trader or traders to cover his (their) position(s) at artificially low prices.

In this situation, the storage bottleneck is the gasoline, the exercise of the market power is the match. With 100,000,000 barrels of open long positions needing to liquidate, given the storage constraint, the resulting conflagration can be epic.

This is, at this stage, a hypothesis. It is a possible explanation of the beyond extreme movements observed today. Under the circumstances, it is a very plausible explanation, and one that deserves scrutiny. And given the amount of money that changed hands today (~$6 billion on a mark-to-market basis) I’m sure that it will get it.

The only parallel I can think of is the onion market in 1955, when the movement of a couple of superfluous carloads of onions into Chicago, and delivery thereof against futures, caused the price to crash below the cost of the bags that they had to be delivered in. There was no demand for the onions (being perishable, and people eat only so many hot dogs), so many of the excess plants ended up getting dumped into the Chicago River. (Which, in 1955, probably improved the water quality.) (Another irony being that Chicago means “stinky onions” in the Miami-Illinois Indian language.)

In 1955, demand was inelastic because onions are perishable (i.e., they can’t be stored). In a way, the lack of storage space makes oil perishable. Even if that analogy isn’t perfect, the economics are the same: an economic constraint (the non-storability of, or the lack of storage space) leads to extremely inelastic demand that makes short market power manipulation possible.

Tomorrow is the last trading day for CLK20. Strap it up! It’s going to be a wild ride.

What a helluva few weeks it’s been, eh boys and girls? By way of post mortem (hopefully?) rather than prediction, here’s my take.

Under “normal” circumstances, two factors drive asset valuations: expectations of cash flows, and the rate at which investors discount those cash flows. COVID-19–Wuhan Virus, to call it by its proper name–has has profound influence on both.

WV has caused a major aggregate supply shock, and an aggregate demand shock, and these amplify one another. The aggregate supply shock stems from shutdown of productive capacity due to social distancing. And people who aren’t working aren’t earning and aren’t spending, hence the aggregate demand shock.

These developments obviously reduce the income streams from assets (e.g., corporate profits). That’s a negative for stocks.

As an aside, these factors defy traditional policy prescriptions. Monetary and fiscal policy are focused on addressing aggregate demand deficiencies, i.e., trying to move demand-deficient economies (where demand deficiencies arise from price rigidity and nominal shocks) back to the production possibilities frontier. Supply shocks shrink the PPF. Pushing the PPF back to its normal state in current circumstances is a function of public health policy, and even that is likely to be problematic given the huge uncertainties (that I discuss below) and the dubious competence of government authorities (which I discussed last week).

The pandemic nature of WV also makes it the systematic shock par excellence. It hits everyone and every asset class, and cannot be diversified away. A big increase in systematic risk results in a big increase in risk premia, meaning that the already depressed expected cash flows on risky assets get discounted at a higher rate, leading to lower valuations.

A lot higher rate, evidently. Why? Most likely because of the extreme uncertainty about the virus. Data on how infectious it is, how many people have been infected, the fatality rate, how it will be affected by warmer weather, etc., are extremely unreliable. In other words, we know almost nothing about the salient considerations.

This is in part due to lack of testing, and to inherent defects in the testing: those who get tested are disproportionately likely to be symptomatic, exposed, or hypochondriacal, leading to extreme sample selection biases. The tests are apparently unreliable, with high rates of false positives and false negatives. The RNA tests cannot detect past infections. It is in part due to the novelty of the virus. Is it like influenza, and will hence burn out when temperatures warm? Or not?

Another major source of uncertainty is due to the fact that the initial outbreak in China was covered up by the evil CCP regime. (Which now, in an Orwellian twistedness that only totalitarian regimes can muster, is boasting that it will save the world. And which is blaming the United States for its own abject failures. Which is why I insist on calling it the Wuhan Virus–so go ahead, call me a racist. IDGAF.) Thus, data from Ground Zero is lacking, or wildly unreliable. (Ground One–Iran–is equally duplicitous, and equally malign.)

This huge uncertainty regarding a major systematic factor leads to even greater discount rates–and hence to lower stock prices.

And then there is the truly disturbing factor. These textbook causal channels (lower expected cash flows, higher discount rates) have in turn caused changes in asset prices that force portfolio adjustments that move us into the realm of positive feedback mechanisms (which usually have negative effects!) and non-linearities. This represents a shift from “normal” times to decidedly abnormal ones.

When some investors engage in leveraged trading strategies, big price moves can force them to unwind/liquidate these strategies because they can no longer fund their large losses. These unwinds move asset prices yet more (as those who placed a lower valuation on these assets must absorb them from the levered, high-value owners who are forced to sell them). Which can force further unwinds, in perhaps completely unrelated assets.

Not knowing the extent or nature of these trading strategies, or the degree of leverage, it is virtually impossible to understand how these effects may cascade through the markets.

The most evident indicators of these stresses are in the funding markets. And we are seeing such stresses. The FRA-OIS spread (known in a previous incarnation–e.g., 2008–as the LIBOR-OIS spread) has blown out. Dollar swap rates are blowing out. The most vanilla of spreads–the basis net of carry between Treasury futures and the cheapest-to-deliver Treasury–have blown out. Further, the Fed has pumped in huge amounts liquidity into the system, and these alarming spread movements have not reversed. (One shudders to think they would have been worse absent such intervention.)

One thing to keep an eye on is derivatives clearing. As I warned repeatedly during the drive to mandate clearing, the true test of this mechanism is during periods of market disruption when large price moves trigger large margin calls.

Heretofore the clearing system seems to have operated without disruption. I note, however, that the strains in the funding markets likely reflect in part the need for liquidity to make margin calls. Big margin calls that must be met in near real-time contribute to stresses in the funding markets. Clearinghouses themselves may survive, but at the cost of imposing huge costs elsewhere in the financial system. (In my earlier writing on the systemic impacts of clearing mandates, I referred to this as the Levee Effect.)

The totally unnecessary side-show in the oil markets, where Putin and Mohammed bin Salman are waging an insane grudge match, is only contributing to these margin call-related strains. (Noticing a theme here? Authoritarian governments obsessed with control and “stability” have a preternatural disposition to creating chaos.)

Perhaps the only saving grace now, as opposed to 2008, is that the shock did not arise originally from the credit and liquidity supply sector, i.e., banks and shadow banks. But the credit/liquidity supply sector is clearly under strain, and if parts of it break under that strain yet another round of extremely disruptive knock-on effects will occur. Fortunately, this is one area where central banks can palliate, if not eliminate, the strains. (I say can, because being run by humans, there is no guarantee they will.)

Viruses operate according to their own imperatives, and the imperatives of one virus can differ dramatically from those of others. Pandemic shocks are inherently systematic risks, and the nature of the current risk is only dimly understood because we do not understand the imperatives of this particular virus. Indeed, it might be fair to put it in the category of Knightian Uncertainty, rather than risk. The shock is big enough to trigger non-linear feedbacks, which are themselves virtually impossible to predict.

In other words. We’ve been on a helluva ride. We’re in for a helluva right. Strap it tight, folks.

A recent RFS paper by Brogaard, Riggenberg, and Sovich purports to show that the rise of commodity index investing has had adverse effects on the real economy. Like most of the papers that analyze index investing, this one is seriously flawed and does not support the broad conclusions it asserts.

Like virtually all of the papers in this literature, it relies on crude before-after comparisons based on some magic year (2004? 2005? 2006?) during which commodity index investing became important. Even assuming that the rise in index investing was a sort of exogenous shock, this crude method of attributing causation has problems. After all, a lot of other stuff happened after, say, 2004. The rise of China as the predominate force in commodity markets, for instance.

Perhaps this problem would be less troublesome were commodities assigned randomly to “treatment” (part of an index) and “non-treatment” (non-index) groups. But this is definitely not the case. There are systematic differences between index and non-index commodities, so it is difficult to know whether the effects documented in Brogaard et al are due to a correlation between these systematic differences and the performance measures they utilize.

Ah. The performance measures. That raises yet more issues. The paper claims that firms with exposure to index commodities experienced lower returns on assets (i.e., operational income/assets) post-indexation than their non-index-exposed counterparts.

So what is the implicit model of the market in which these firms operate? If the presumption is that the markets are relatively competitive, why would you expect profit margins to change over a period of several years? Even assuming that indexation increased costs via the channels posited by the authors, in equilibrium this would lead to (a) an initial decline in profits, (b) exit (meaning fewer assets and lower output), and (c) a return of profits to the competitive level. Meaning that actually looking at assets or output would be a better measure of how indexation affects cost than profit margins.

If, conversely, the presumption is that these markets were imperfectly competitive prior to the indexation shock, the fall in profit rates could be a good thing, rather than a bad thing. It could indicate that indexation resulted in more intense competition/lower market power.

This is particularly interesting, given that this was a period in which measured profit rates were rising generally in the economy, a phenomenon which some (not I, but some) attribute to declining competition. Now, there are big problems with that literature (problems that the late great Harold Demsetz identified in the 1970s, but which modern scholars have forgotten), but take it at face value. Declining profits/margins could be interpreted as a signal of increased competitiveness and efficiency. A feature, not a bug.

And via a channel exactly contrary to that posited by Brogaard et al. They posit that indexation reduced the informational efficiency of commodity prices. But an increase in informational efficiency would tend to reduce market power–and reduce margins/profits. Note that many firms HATE the introduction of futures, or increased trading of futures, of their main inputs or outputs. Precisely because futures trading increases the informational content of prices which undermines an information asymmetry that generates profits for major producers and consumers.

Which brings me to that information channel. Brogaard et al measure informativeness using the autocorrelation in commodity returns.

Really?

They basically find that autocorrelations went up, which they interpret to mean that commodity futures markets became weak form inefficient (or further from weak form efficiency).

Again: really?

Taking their argument literally, it means that speculators were able to eliminate predictable movements in prices prior to indexation, but weren’t able to do so afterwards. This stretches credulity to its limits. After all, “financialization” of commodities also saw the entry of banks and hedge funds into commodity trading. We’re supposed to believe they left easy money on the table?

This purported reduction in price efficiency raises another issue. When I read about return autocorrelations I think time-varying expected returns. So the Brogaard et al result suggests that index commodity returns became more time-varying after indexation became a thing.

Arguably the primary impact of indexation was to integrate more closely commodity prices and stock and bond prices, and in particular, ensure that systematic risk was priced more consistently across commodities and financials. We know that equities and bonds have time varying returns. It is plausible that the documented increase in time variation in expected commodity returns reflects this integration, and is yet another feature rather than a bug.

If so, commodity prices that more accurately reflect the pricing of risk should lead to better investment decisions, not worse decisions. Further, these better investment decisions need not be associated, over several years, the operational performance measures they employ.

Indeed, along the lines of a paper that I wrote a few years back, indexation improves the allocation of risk and reduces the cost of commodity price risk to firms. In equilibrium, this would lead to lower rates of return. Which is exactly what Brogaard et al find.

Pursuing this further, I note that the authors do not use sensitivity to commodity prices as a means of determining whether firms are exposed to commodity prices, on the basis that hedging can reduce price exposure. Yes, but consider the following. Hedging is costly. Hedgers frequently pay a risk premium to speculators. This leads them to hedge less, which leads to high exposure. If indexation reduces hedging cost (as my paper implies), at the margin, firms will hedge more, and bear less risk. This reduces expected profits (which incorporate compensation for risk).

This has at least two implications. First, the average profit rate of firms exposed to hedgeable commodity prices (which are precisely the commodities that are included in indices) should decline post-indexation: this is consistent with the Brogaard et al finding. Second, firms should hedge more–leading to lower exposure to commodity prices. Brogaard et al do not test this latter implication. (Admittedly, this raises complications. Firms may limit commodity exposure in ways other than hedging, and a lower cost of hedging can lead to a substitution of hedging for these other means, leading to offsetting effects which reduce the commodity exposure impact of cheaper hedging.)

The upshot of all this is that the conclusions that the authors advance–namely, that indexation resulted in poorer real outcomes in terms of performance and investment–are not supported by their empirical findings. That is, there are alternative hypotheses that are consistent with the empirical findings that are diametrically opposed to their hypothesis.

This paper is well done within the analytical confines that its authors select. But that’s exactly the problem with this literature. The analytical confines exclude important channels of cause and effect. Most importantly, these confines make implicit assumptions about the degree and nature of competition that either make their results problematic, or mean that their results have diametrically opposed implications to those of the chosen analytical framework.

Meaning that the debate on the effects of indexation are still very much open, and that finance and economics scholars have largely failed to devise reliable tests to distinguish indexation-is-bad hypotheses from indexation-is-good hypotheses. Which is precisely why the debate is still open.

On Thursday, the CFTC voted along party lines to approve a proposal on position limits. The party line vote reveals a salient fact: the proposal represents a virtual abandonment of the Commission’s earlier proposals (2011, 2013, 2016). Indeed, virtually all of the features that I criticized in the earlier proposals are gone, and the current proposal largely mirrors the recommendations of the Energy and Environmental Markets Advisory Committee that I served on (before being uninvited by current EEMAC chair Dan Berkovitz). (More on EEMAC below.)

Most importantly, limits outside the “spot month” (which is actually just a few days for some commodities) for energy and metals commodities are gone. Good riddance. They remain for nine legacy ag futures contracts (corn, cotton, and the like), but the any-and-all limits have been expanded substantially.

The rule expands hedging exemptions beyond the prior proposals, and in doing so meets the objections of companies like Vitol. Interestingly, the proposal tidies up the definition of a “bona fide hedge” and makes explicit, rather than implicit, the principle that bona fide hedges are solely for the reduction of price risk.

The Commission did eliminate the “risk management” hedging exemption for swaps dealers, based on an interpretation of Congressional intent and a reading of statute that limits hedging to the management of risk of physical commodity positions. On principled grounds, I object to this. A swap dealer buying an oil swap from an E&P firm facilitates the hedging of a physical position, and hedging that swap via the futures markets serves a classical risk transfer function. A dealer selling an index swap to a pension fund isn’t hedging a physical risk, but it is still serving a risk transfer function and the distinction between physical commodity hedges and non-physical hedges is rather Talmudic.

The practical effect is unknown. In terms of index swaps, most swap dealers are out of the nearby contract when an index (e.g., GSCI) rolls, which is well before the spot month for energy and metals that make up the bulk of most indices. Hedges of the ag portion of these swaps could be affected by the any-and-all limits and the elimination of the risk management exemption, but the dramatic increase in the size of these limits may well greatly reduce any impact. A dealer hedging a swap with payments based on final settlement prices of say NYMEX crude or natural gas could be impacted by the elimination of the exemption, but the spot month limits may be large enough to cushion the impact here as well.

The most interesting feature of the proposal is its rather tortured attempt to address the “necessity” issue that derailed previous proposals in court.

The most important aspect of this is that it appears that the Commission has essentially conceded that a necessity finding is, well, necessary. That raises the issue of the criteria for establishing necessity.

One criterion could be that a limit is necessary only if the risk of speculation causing unwarranted price fluctuations is sufficiently great.

An alternative criterion is that a limit is necessary as long as the risk of unwarranted price fluctuations exists at all, if the contract is important enough.

The Commission took the latter approach, and limited its limits to commodities it deemed were sufficiently important (measured by volume and open interest) so that any unwarranted price fluctuation could lead to impairment of price discovery and risk transfer on a large scale. The closest that the Commission came to taking likelihood of disruption into account is its restriction of the limits to physical delivery contracts that could be cornered or squeezed. This is a logical problem (cash-settled contracts give rise to manipulation too) but this is of secondary importance. But it could be read to limit the Commission’s interpretation as to the source of unwarranted fluctuations to market power manipulation, which would be a good limitation indeed.

A sufficient statistic to infer that the Commission conceded much in its necessity finding is Dan Berkovitz’s freak out on the issue in his dissent.

As a manipulation-related aside, I will note that the spot month limits are justified by the notion that a position in excess of deliverable supply is necessary to execute a market power manipulation (i.e., a corner or squeeze). I have some recent research (which I’ll post and write about soon) showing that this may be a sufficient condition, but not a necessary one. Meaning that the limits will not be sufficient to eliminate market power manipulation.

The recent proposal, assuming it is finally approved as a rule in something resembling its current form, represents the end of a saga that has had a major influence on my life. I began writing about the speculation issue when it became a source of renewed political controversy in 2006. I wrote my first major post in response to a Senate Permanent Subcommittee Report (large authored by Dan Berkovitz) on oil speculation in August 2006.

As oil prices spiraled upwards in 2007 and 2008, I wrote more about the issue, and gained more notoriety. This resulted in my testifying before the House Ag Committee in July 2008 (a day or two before oil prices reached their all time high) and led to a WSJ oped.

Then the Financial Crisis happened, and I focused more on clearing issues, with periodic forays into the speculation debate. But Frankendodd included a provision on speculative position limits in commodities, and the CFTC rolled out a proposal in 2011.

I wrote a comment letter on the proposal. That letter (and others I wrote subsequently) were sufficiently important that in the final rulemaking and in later proposals it or other things I’ve written were cited dozens of times (my name gets 50 hits in the 2016 proposal).

But the impact of the letter on my life went beyond that. Gene Scalia–son of Justice Antonin Scalia, and now Secretary of Labor–retained me to write a declaration criticizing the inadequate cost-benefit analysis in the proposal (it being something required under the law.)

Perhaps most importantly, Blythe Masters at J. P. Morgan liked it, and called to tell me so. She then proposed that I write an analysis of the systemic risk of commodity trading firms for SIFMA. I did–and came up with the wrong answer. So SIFMA spiked the report. But word leaked out, which prompted Trafigura to retain me to write a study (with a subsequent follow-on study) of the economics of commodity trading firms.

I don’t think I’m exaggerating to say that this study proved to be very influential, perhaps because of the lack of competition: writing on the sector was, and remains, very sparse. I have traveled, lectured, and taught around the world based on people wanting to hear what I wrote about in that piece.

The study was also the hook for the New York Times hit piece on me in December, 2013. See! I took money from evil speculators while writing in opposition to limits on speculation! Never mind that I had been consistently opposed to limits years before, and never mind that Trafigura (and other oil traders) are not speculators and use the futures markets mainly for hedging.

(As an aside, I am convinced, but cannot prove, that Gary Gensler was the moving force behind the piece. After all, why else would the NYT devote front page space to an obscure academic? And under the theory of there-are-no-coincidences-comrade, comments on the 2013 revised proposal were due in January, 2014. So December 2013 was the perfect time to kneecap a gadfly. By the way, Gary, how’s that gig as Treasury Secretary working out. Oh. Right. Well maybe you can chair Hillary’s legal defense fund.)

Asides aside, other than frightening my aged parents this article actually was all for the good. It validated me as an influential voice. It also got many very reputable people to rush to my defense, including Thomas Sowell, one of my long-time heroes.

The article raised its head a few years later when I was serving on EEMAC, and was asked to write the (Frankendodd-mandated) report on the committee’s deliberations. I was the dutiful scribe, and honestly recorded the committee’s adamant opposition to the then-outstanding proposal (which included all the bad features jettisoned in the new proposal).

This caused Elizabeth Warren to lose her [insert vulgar metaphor of your choosing here]. This article in particular cracked me up (and still cracks me up): Why Elizabeth Warren Is On the Warpath This Week.

Well, why was she on “the warpath”? Well–me, now that you ask:

The committee, which was established by Dodd-Frank, has nine members. Though it is supposed to express a “wide diversity of opinion” and “a broad spectrum of interests,” eight of the nine members represent companies or industries with a financial interest in killing the position limits rule, or have a personal financial interest themselves. 

. . . .

The recent inclusion of Craig Pirrong on the committee is perhaps the most flagrant example. Pirrong, who co-wrote the first draft of the report with James Allison, is a professor of finance at the University of Houston, who has been paid by several industry participants and trade groups for his research into commodity speculation. He was also a paid research consultant for the International Swaps and Derivatives Association, the very group that got the initial rule overturned by the courts.

The CFTC report relies mostly on Pirrong’s research and a presentation he made to the committee last year, which did not include the opinion of anyone who believes in the dangers of excessive commodity speculation. In fact, 10 of the 13 witnesses at EEMAC meetings came from industry, two were representatives of CFTC, and the other was Pirrong. The meetings never mentioned that there would even be a final report. [Er, it’s in the law, you knobs.]

As Public Citizen’s Tyson Slocum, the only non-industry committee member and the only one to dissent from the recommendation, points out, Pirrong was not on the committee until after he co-authored the report. Pirrong “is so new to the EEMAC,” Slocum wrote in a minority dissent, “that I only learned he was a member when he was listed as a co-author.” 

This kerfuffle warranted another mention in the NYT, and I’ve been told that Warren used it (and me) in a fundraising pitch.

I’m so proud. One’s enemies are the best comment on one’s character.

What cracks me up is that it wasn’t a right-wing snarkmeister like me that included “Warpath” in the title of an article about Liz Warren. It was the (by then) far-left New Republic. Freudian slip? Whatever, it’s hilarious.

Alas, understandably but not commendably, Commissioner and EEMAC chair Christopher Giancarlo buckled under the political pressure and withdrew the report. But this was almost certainly a non-event: the proposal was dead in the water, and was only salvaged by saving major pieces overboard. And I’ve sailed on.

What a long strange trip it’s been. The speculation debate has had a first-order impact on the arc of my life for more than a decade. Although the Commission proposal will likely put an end to one chapter of that debate, as I wrote in my first piece in 2006, speculation controversy is a hardy perennial, and will no doubt recur the next time some major commodity price spikes or craters. And maybe I’ll be around to draw more fire–and deliver some–when that happens. And until then, I’ll keep Truckin’ on whatever fits my fancy.

Nota bene: I’m not a Dead Head by any means. But if the song fits . . . Well, not the drugs part!

The BIS included a box on the September USD repo spike in a chapter to its Quarterly Review titled “Easing Trade Tensions Support Risky Assets.” The piece lays out many damning dots, but does not connect them. Let me give it a try.

In a nutshell, the BIS report says that as a result of the wind down of the extraordinary post-crisis monetary policy measures there has been a dramatic change in the funding structure in US markets. In particular, the “big four [US] banks” (which the BIS delicately–or is it cravenly?–doesn’t name) have flipped from being suppliers of repo collateral (and hence cash borrowers) to being suppliers of cash (and hence collateral borrowers). Further, other US banks are not viable competitors to the big four, nor are other potential cash suppliers such as money market funds because they have hit counterparty credit limits which have constrained their lending capacity. According to the BIS, these events has made The Big Four Banks Who Shall Not Be Named the “marginal lenders” in the repo market. And mark well: prices are set at the margin.

Further, “leveraged players (eg hedge funds) were increasing their demand for Treasury repos to fund arbitrage trades between cash bonds and derivatives.”

So here are the dots. Recent structural changes have given the Big Four Banks Who Shall Not Be Named a dominant position in the repo market. Their main potential competitors as suppliers of funds are constrained by size or regulation. There has been a large increase in demand for repo funding.

Not even being willing to name the banks (as if their identities are unknown), the BIS does not even draw the blindingly obvious implication of its analysis–that the Four Repo Horsemen have market power. A lot of market power. Are we supposed to believe that (out of the goodness of their hearts, perhaps) they did not exercise it? I didn’t just fall off the turnip truck.

Even the euphemism Big Four is deceptive, for in reality this group is dominated by one bank–Morgan. And of course Morgan has been loudest in its protestations that it really wanted to lend more, but just couldn’t, dammit, because of those cursed liquidity regulations.

The BIS attempts to run cover, and provide some rather lame excuses for the failure to lend more despite the high rates:

Besides these shifts in market structure and balance sheet composition, other factors may help to explain why banks did not lend into the repo market, despite attractive profit opportunities. A reduction in money market activity is a natural by-product of central bank balance sheet expansion. If it persists for a prolonged period, it may result in hysteresis effects that hamper market functioning. For instance, the internal processes and knowledge that banks need to ensure prompt and smooth market operations may start to decay. This could take the form of staff inexperience and fewer market-makers, slowing internal processes. Moreover, for regulatory requirements – the liquidity coverage ratio – reserves and Treasuries are high-quality liquid assets (HQLA) of equivalent standing. But in practice, especially when managing internal intraday liquidity needs, banks prefer to keep reserves for their superior availability.

Hysterisis? Decaying internal processes and knowledge? Staff inexperience? Complete and utter argle bargle. We’re talking overnight secured lending here, not rocket science structured finance. It’s about as vanilla a banking transaction one could imagine. And LCR provides convenient cover.

The BIS lays out a compelling case that four major institutions have market power in repo. September events in particular are consistent with the exercise of market power, and the alternative explanations are beyond lame. Yet none dare speak its name, or even raise it as a possibility. Not the BIS. Not the Fed. Not the Treasury. Despite the systemic risks this poses.

The Financial Crisis supposedly changed everything. It apparently changed nothing.

The first post I wrote almost 14 years ago was about an electronic trading system (in Japan) being overwhelmed by a avalanche of orders. The main conclusions I drew were: (a) since communications capacity is costly, it is uneconomic to build so much capacity as to make such outages impossible, and (b) pricing of capacity is the best way to ensure that it is utilized efficiently.

What is old is new again. A few weeks ago the CME was strained by a deluge of message traffic in the Eurodollar futures market. The surge was in part the result of something only dimly–if that–grasped 14 years ago: algorithmic trading. More specifically, algorithmic trading combined with the order matching and confirmation protocols of the CME and Eurodollar markets. Eurodollar futures utilize a pro rata secondary priority rule in the absence of a “TOP” order, i.e., an order that improves the best bid or offer. Moreover, the biggest order at the inside market gets informed of executions slightly before smaller orders. This advantage (on the order of 10-20 millionths of a second) provides a valuable edge.

This institutional setup provided incentives for two algorithmic traders to attempt to get a slight edge in quote size. Thus, each would send a message to increase its size when the other had the edge, which induced the prior leader to send a message to increase its size to regain the lead, which induced the other to send a message to send a message to regain the lead . . . which continued until the maximum quote size was reached. This race for priority in the book led to a surge in message traffic which put stress on the CME system.

Matt Levine wondered why the two algos didn’t just keep their orders at the maximum size. My surmise is risk: the larger size you show, the greater the risk. Furthermore, if someone else improved the price and became the TOP order, size didn’t confer an edge, so it makes sense to cut size.

The CME has responded by announcing fines of $10,000, with the threat of cutting off a violator’s trading connection altogether, for violations of a message traffic threshold. This is a rough form of pricing, so why call it a fine? Why not just call it a non-linear message pricing schedule?

I guarantee that a variant on this story will recur. Because capacity is costly, exchanges don’t price it, and rules adopted for other reasons (e.g., priority rules intended to promote quoting in size) provide an incentive to use this unpriced resource.

As I argued years ago, it’s odd that institutions–exchanges–that specialize in providing the platform to allow the pricing of scarce resources don’t find a better way to price their own scarce platform resources.

Standard no-arbitrage derivatives pricing theory rules out certain trading strategies because they are seemingly unrealistic. For example, if the set of trading strategies is unrestricted, the following strategy is guaranteed to make any arbitrarily large profit (pick a number, any number! e.g., $10 quadrillion) prior to some finite time T almost surely–and don’t call me Shirley!: specifically, at time s, hold 1/(T–s) shares of stock in your trading portfolio, and cash out when you’ve made $10 quadrillion.

Obviously, this requires you to hold an arbitrarily large–and arbitrarily expensive–position in the stock. No problem: just borrow! Further, this obviously exposes you to risk of infinite loss: the variance of your portfolio value goes to infinity as you approach T. Again, no problem!: just borrow! The self-financing constraint is satisfied, so dream of how you’ll spend your immense riches.

Well, borrowing infinite sums seems a tad unrealistic, so it is standard to rule out such doubling strategies. One good explanation is in the very good book by J. Michael Steele, Stochastic Calculus and Financial Applications:

Nothing is wrong with the model or with the verification of the self-financing property, yet something has gone wrong. The problem is that our model diverges from the real world in a way that any decent banker would spot in an instant. Just consider what happens as the time gets nearer to T if the investor in charge of the portfolio V(t) has not yet reached his goal. In that case, the investor borrows more and more heavily, and he pours the borrowed funds into the risky asset. Any banker worth his salt who observes such investment behavior will pull the investor’s credit line immediately. The simplest rules of lending practice are enough to prohibit the management of a portfolio by any strategy [a(t), b(t)] like that defined by equations (14.36) and (14.38).

But maybe not, thanks to the miracles of fintech! Apparently a glitch in the Robinhood app has given users in the know “infinite leverage”:

The cheat code was being shared on social media site Reddit, with one trader claiming he took a $1,000,000 position in stock using only a $4,000 deposit. Through Robinhood Gold, the start-up’s subscription service, users can borrow money from the company to make trades. The backdoor was essentially free money and was being called “infinite leverage” and the “infinite money cheat code” by Reddit users who discovered it.

Looks like we can’t rule out “doubling” strategies, and impose a credit constraint restrict trading strategies to what Steele denotes the set SF+ (“self-financing plus”). I guess Robinhood isn’t “a decent banker.” LOL.

This is what happens when you make coders bankers. They blow no-arbitrage pricing theory all to hell.

Looks like imma gonna haveta rework my lecture notes for my PhD derivatives pricing course.

In the Gilded Age, American financial crises (“panics,” in the lexicon of the day) tended to occur in the fall. Agriculture played a predominant role in the economy, and marketing of the new crop in the fall led to a spike in the demand for cash and credit. In that era, however, the supply of cash and credit was not particularly elastic, and these demand spikes sometimes turned into panics when supply did not (or could not) respond accordingly.

The entire point of the Fed, which was created in the aftermath of one of these fall panics (the Panic of 1907, which occurred in October), was to make currency supply more elastic and thereby reduce the potential for panics. In essence, the Fed had one job: lender of last resort to ensure a match of supply and demand for currency/credit, when the latter was quite volatile.

This week’s repospasm is redolent of those bygone days. Now, the spikes in demand for liquidity are not driven by the crop cycle, but by the tax and corporate reporting cycles. But they recur, and several have occurred in the autumn, or on the cusp thereof (this being the last week of summer).

One of my mantras in teaching about commodities is that spreads price bottlenecks. Bottlenecks can occur in the financial markets too. The periodic spikes in repo rates–not just this week, but in December, and March–relative to other short term rates scream “bottleneck.” Many candidates have been offered, but regardless of the ultimate source of the clog in the plumbing, the evidence from the repo market is that there are indeed clogs, and they recur periodically.

The Fed’s rather belated and stumbling response suggests that it is not fully prepared to respond to these bottlenecks, despite the fact that their regularity suggests that the clogs are chronic. As the saying goes, “you had one job . . . ” and the Fed fell down on this one.

And maybe the problem is that the Fed no longer just has one job, and it has shunted the job that was the reason for its creation to the back of the priority list. Nowadays, the Fed has statutory obligations to control employment and inflation, and views its main job as managing aggregate demand, rather than tending to the financial system’s plumbing.

This is concerning, as dislocations in short-term funding markets can destabilize the system. These markets are systemically important, and failure to ensure their smooth operation can result in crises–panics–that undermine the ability of the Fed to perform its prioritized macroeconomic management task.

One of the salutary developments post-crisis has been the reduced reliance of banks and investment banks on flighty short-term funding. The repo markets are far smaller than they were pre-2008, and the unsecured interbank market has all but disappeared (representing only about .3 percent of bank assets, as compared to around 6 percent in 2006). But this is not to say that these markets are unimportant, or that bottlenecks in these markets cannot have systemic consequences. For the want of a nail . . . .

Moreover, the post-crisis restructuring of the financial system and financial regulation has created new potential sources of liquidity shocks, namely a supersizing of potential demands for liquidity to pay variation margin. When you have a market shock (e.g., the oil price shock) occurring simultaneously with the other sources of increased demand for liquidity, the bottlenecks can have very perverse consequences. We should be thankful that the shock wasn’t a Big One, like October, 1987.

Hopefully this week’s tumult will rejuvenate the Fed’s focus on mitigating bottlenecks in funding markets. Maybe the Fed doesn’t have just one job now, but this is an important job and is one that it should be able to do in a fairly routine fashion. After all, that job is what it was created to perform. So perform it.

The primary reason for my deep skepticism about the wisdom of clearing mandates was liquidity risk. As I said repeatedly, in order to reduce counterparty risk, clearing necessarily increased liquidity risk through the variation margining mechanism. Further, it was–and is–my opinion that liquidity risk is a far graver systemic concern that counterparty risk.

A major liquidity event has occurred in the last couple of days: rates in the repurchase market–the major source of short term funding for vast amounts of trading activity–shot up to levels (around 5 percent) nearly double the Fed’s target ceiling for that rate. Some trades took place at far higher rates than that (e.g., 9.25 percent).

Market participants have advanced several explanations, including big cash demands due to corporate tax payments coming due. Izabella Kaminska at FTAlphavile offered this provocative alternative, which resonates with my clearing story: the large price movements in oil and fixed income markets in the aftermath of the attack on the Saudi resulted in large margin calls in futures and cleared OTC markets that increased stresses on the funding markets.

To which one might say: I sure as hell hope that’s not it, because although there was a lot of price action yesterday, it wasn’t The Big One. (The fact that Fred Sanford’s palpitations occurred because he couldn’t get his hands on cash makes that bit particularly apropos!)

I did some quick back-of-the-envelope calculations. WTI and Brent variation margin flows (futures and options) were on the order of $35 billion. Treasuries on CME maybe $10 billion. S&P futures, about $1 billion. About $2 billion on Eurodollar futures.

The Eurodollar numbers can help give a rough idea of margin flows on cleared interest rate swaps. Eurodollar futures open interest is about $12 trillion. Cleared OTC notional volume (not just USD, but all IRS) is around $80 trillion. But $1mm in notional of a 5 year swap is equivalent to 20 Eurodollar futures with notional amount of $20 trillion. So, as a rough estimate, variation margin flows in the cleared IRS market are on the order of 100x for Eurodollars. That represents a non-trivial $200 billion.

Yes, there are potentials for offsets, so these numbers are not additive. For example, a firm might have offsetting positions in EDF and cleared IRS. Or be short oil and long Treasuries. But variation margin flows on the order of $300 billion are not unrealistic. And since market moves were relatively large yesterday, that represents an increment over the typical day.

So we are talking real money, which could certainly contribute to an increased demand for liquidity. But again, yesterday was not remotely a truly epic day that one could readily imagine happening.

A couple of points deserve emphasis. The first is that perhaps it was coincidence or bad luck, but the big variation margin flows coincided with other sources of increased demand for liquidity. But hey, stuff happens, and sometimes stuff happens all at once. The system has to be able to withstand such simultaneous stuff.

The second is related, and very concerning. The spikes in rates observed periodically in the repo market (not just here, but notoriously in China) suggest that this market can go non-linear. Thus, even if the increased funding needs caused by the post Abqaiq fallout wasn’t The Big One, in a non-linear market, even modest increases in funding needs can have huge impacts on funding costs.

This highlights another concern: inter-market feedback. A shock in one market (e.g., crude) puts stress on the funding market that leads to spikes in repo rates. But these spikes can feedback into prices in other markets. For example, if the inability to fund positions causes fire sales that cause big price moves that cause big variation margin flows which put further stress on the funding markets.

Yeah. This is what I was talking about.

Today’s events nicely illustrate another concern I raised years ago. Clearing/margining make markets more tightly coupled: the need to meet margin calls within hours increases the potential stress on the funding markets. As I tell my classes, unlike in the pre-Frankendodd days, there is no “fuck you” option when your counterparty calls for margin. You don’t pay, you are in default.

This tight coupling makes the market more vulnerable to operational failings. On Black Monday, 1987, for example, the FedWire went down a couple of times and this contributed to the chaos and the potential for catastrophic failure.

And guess what? There was a (Fed-related!) operational problem today. The NY Fed announced that it would hold a repo operation to supply $75 billion of liquidity . . . then had to cancel it due to “technical difficulties.”

I hate it when that happens! But that’s exactly the point: It happens. And the corollary is: when it happens, it happens at the worst time.

The WSJ article also contains other sobering information. Specifically, post-crisis regulatory “reforms” have made the funding markets more rigid/less-flexible and supple. This would tend to exacerbate non-linearities in the market.

We’re from the government and we’re here to help you! The law of unintended (but predictable) consequences strikes again.

Hopefully things will normalize quickly. But the events of the last two days should be a serious wake-up call. The funding markets going non-linear is the biggest systemic risk. By far. And to the extent that regulatory changes–such as mandated clearing–have increased the potential for demand surges in those markets, and have reduced the ability of those markets to respond to those surges, in their attempt to reduce systemic risks, they have increased them.

I have often been asked what would cause the next financial crisis. My answer has always been: the regulations intended to prevent a recurrence of the last one. Today may be a case in point.

ICE is on the verge of launching Bitcoin futures. The official start date is 23 September.

The ICE contract is distinctive in a couple of ways.

First, it is a delivery settled contract. Indeed, this feature is what made the ICE product so long in coming. The exchange had to set up a depository, the Bakkt Warehouse. This required careful infrastructure design and jumping through regulatory hoops to establish the Bakkt Trust Company, and get approval from the NY Department of Financial Services.

Second, the structure of the contracts offered is similar to that of the London Metal Exchange. There are daily contracts extending 70 days into the future, as well as more conventional monthly contracts. (LME offers daily contracts going out three months, then 3-, 15-, and 27-month contracts). The daily contracts settle two days after expiration, again similar to LME.

The whole initiative is quite fascinating, as it represents a dual competitive strategy: Bakkt is simultaneously competing in the futures space (against CME in particular), and against spot crypto exchanges.

What are its prospects? I would have to say that Bakkt is a better mousetrap.

It certainly offers many advantages as a spot platform over the plethora of existing Bitcoin/crypto exchanges. These advantages include ICE’s reputation, the creation of a warehouse with substantial capital backing, and regulatory protections. Here is a case in which regulation can be a feature, not a bug.

Furthermore, for decades–over a quarter-century, in fact–I have argued that physical delivery is a far superior mechanism for price discovery and ensuring convergence than cash settlement. The myriad issues that were uncovered in natural gas when rocks were overturned in the post-Enron era, the chronic controversies over Platts windows, and the IBORs have demonstrated the frailty, and vulnerability to manipulation of cash settlement mechanisms.

Crypto is somewhat different–or at least, has the potential to be–because the CME’s cash settlement mechanism is based off prices determined on several BTC exchanges, in much the same way as the S&P500 settlement mechanism is based on prices determined at centralized auction markets.

But the crypto exchanges are not the NYSE or Nasdaq. They are a rather dodgy lot, and there is some evidence of manipulation and inflated volumes on these exchanges.

It’s also something of a puzzle that so many crypto exchanges survive. The centripetal forces of liquidity tend to cause trading in a particular instrument to gravitate to a single platform. The fact that this hasn’t happened in crypto is anomalous, and suggests that normal economic forces are not operating in this market. This raises some concerns.

Bakkt potentially represents a double-barrel threat to CME. Not only is it competing in futures, if it attracts a considerable amount of spot trading activity (due to a superior trading, clearing, settlement and custodial platform, reputational capital, and regulatory safeguards) this will undermine the reliability of CME’s cash settlement mechanism by attracting volume away from the markets CME uses to determine final settlement prices. This could make these market prices less reliable, and more subject to manipulation. Indeed, some–and maybe all–of these exchanges could disappear if ICE’s cash market dominates. CME would be up a creek then.

That said, one of the lessons of inter-exchange competition is that the best mousetrap doesn’t always win. In particular, CME has already established liquidity in the futures market, and as even as formidable competitor as Eurex found out in Treasuries in the early-oughties, it is difficult to induce a shift of liquidity to a competitor.

There are differences between crypto and other more traditional financial products (cash and derivatives) that may make that liquidity-based first mover advantage less decisive. For one thing, as I noted earlier, heretofore cash crypto has proved an exception to the winner-takes-all rule. Maybe the same will hold true for crypto futures: since I don’t understand why cash has been an exception to the rule, I’d be reluctant to say that futures won’t be (although CBOE’s exit suggests it might). For another, the complementarity between cash and futures in this case (which ICE is cleverly exploiting in its LME-like contract structure) could prove decisive. If ICE can get traction in the fragmented cash market, that would bode well for its prospects in futures.

Entry into a derivatives or cash market in competition with an incumbent is always a highly leveraged bet. Odds are that you fail, but if you win it can prove enormously lucrative. That’s essentially the bet that ICE is taking in BTC.

The ICE/Bakkt initiative will prove to be a fascinating case study in inter-exchange competition. Crypto is sufficiently distinctive, and the double-barrel ICE initiative sufficiently innovative, that the traditional betting form (go with the incumbent) could well fail. I will watch with interest.