Category Archives: Uncategorized

Balloons, Retail Order Auctions, Citadel and Antitrust, IEX and FTX

I am going to try to write a bit more casually here, seeing as this is a blog. Please feel free to discuss anything that looks wrong in the comments section or over email with me.

Balloon Shootdowns

It seems unlikely that an HFT would use an unmaneuverable balloon. They’d get blown off their useful path pretty quickly, and even if it were economically viable on some routes to float expendable communications relays into the air every hour or so, it’d create an irresponsible amount of litter. Nevertheless, how long until an F-22 shoots down a high-frequency trader? I wouldn’t be surprised if HFTs were operating maneuverable balloons, balloons tethered to buoys, or drones that might be near some of the reported balloon incidents.

Retail Order Auctions

There’s quite a hubbub over the SEC’s proposal for retail orders to be sent to auctions at market centers, where some order information would be shown to the public via auction messages, allowing low and mid-latency traders to compete over their execution price. I will admit that running auctions so often makes me nervous — they generally have an elevated potential for mishaps compared to order books. So some of the criticism is justified. But the near-unanimity about the proposal seems like too much. If this proposal isn’t workable, then retail flow should probably be restructured another way. The industry catchphrase that “retail has never had it better” is true, but that doesn’t mean that it can’t get better still.

Additionally I think the current system is unsustainable, and may eventually result in degraded execution quality for retail traders. Citadel is in danger of gaining a monopoly on retail order flow. Virtu (which, in keeping with a certain tradition, questions the SEC’s good faith) appears to be losing the ability to compete at the top level. Some new entrants could potentially give Citadel more competition, like Jane Street. But I have my doubts that new entrants can survive unless they’re willing to lose a lot of money for multiple quarters as they accumulate volume, not only because of the regulatory and legal overhead, but also because Citadel has enormous market color on retail orders and the market overall. Which brings me to the next topic.

When Does a Market Participant’s Dominance Become Anticompetetive?

This question is a thorny one. Even in a completely open market, a participant with high market share could have enough additional color from its own trading activity that it’s very hard for newcomers to compete. A new entrant might have access to all public data, but they cannot match what the dominant trader sees, which learns a lot about market conditions from its own fill information. Now, there’s nothing wrong with traders compliantly using information from their own orders — every trader’s order flow is path-dependent — but I think everyone would agree that it’d be a problem for one entity to have 99% market share (double-counted). In such a case, the entity would not only get color from their filled and unfilled orders’ particular details, but would know the direction of order flow of the overall market. There’d also be serious financial stability problems resulting from such high concentration.

In any case, US equities and other electronic markets are not completely open and transparent, although they’re far more than they used to be. One of the biggest examples of that is access to retail flow. When a market has sufficient barriers to entry, the dominant participant can offer an inferior product and still maintain its position. I’m not saying Citadel is doing so right now, but the danger is clearly there.

Flash Boys, 9 Years Later

Hard to believe it’s been nearly a decade. The eponymous “boys” of IEX preached several tenets of exchange design. How would the characters in the book judge today’s IEX?

Tenet 1: “Fading liquidity” is a huge problem created by electronic trading, creates a “rigged” market, and an exchange should do as much as possible to prevent it.
IEX today: Has special order types that use advance information — skipping the speedbump — so resting orders can fade when incoming, contra-side order flow is judged to be unprofitable. IEX offers this functionality on displayed orders too, which the CEO committed not to do in 2016 because “that’s what started this whole journey for us” (footnote 10).

Tenet 2: “Fading liquidity” is in large part caused by high-frequency traders sniffing out large, routable orders before they can reach their destination markets. An outbound speedbump is necessary for an exchange to prevent this form of fading.
IEX today: Has eliminated its outbound speedbump.

Tenet 3: Exchanges should have simple, easily understood order types and mechanics.
IEX today: An alphabet soup of complex order types.

Tenet 4: Market data fees contribute to a “rigged” market.
IEX today: Charges for market data.

A pretty lackluster performance. So it’s no surprise that people paranoid about markets aren’t interested in “Flash Boys” anymore, preferring darker conspiracy theories. But somehow, IEX seems be involved there too! IEX appears to have cross-invested with FTX, planning some kind of joint venture to help unrig the financial system. [1] The details are murky (naturally, IEX used to preach increased disclosure), but it seems like IEX may have engaged in some kind of share swap with FTX. [2] In any case, while this venture says a lot about IEX’s judgment and governance, it doesn’t necessarily say much about the executives’ honesty. People who assume fraud is behind every regretted transaction can make easy targets for actual fraudsters.

And who better to call when trying to differentiate a seller of genuine snake-oil from an outright swindler? Evidently, Michael Lewis. Appearing with Arthur Hayes at a crypto conference, Michael Lewis said Brad Katsuyama introduced him to Sam Bankman-Fried and asked for Lewis’s opinion on whether IEX should proceed with its FTX deal. Lewis jokingly recalls how he told Katsuyama yes, then got interested in profiling SBF. Now, maybe this story isn’t told completely accurately, for instance it could be that Katsuyama had already decided to do the deal and was hoping to get Lewis interested in promoting FTX. That theory is compatible with a report that IEX had such a good time with FTX that it wants to try collaborating with another crypto exchange.

But what a twist! After the smoke has cleared from “Flash Boys,” its legacy is finally emerging: a sequel and an influential Apple TV show about the largest financial fraud since Madoff. And Michael Lewis not only had a “front row seat” to the scam, he was a tool in SBF’s reputation-laundering campaign that intended to restructure our financial system. Anyway, maybe Lewis will appreciate the delicateness of how he tells a story about a con-man who used stolen money to nearly buy off the world’s power centers. Or maybe he’ll broadcast unfounded conspiracy theories to a hundred million people. Given his record, I won’t deny feeling nervous.

[1] I’m obviously being slightly tongue-in-cheek — but only slightly. A big argument of “Flash Boys” is that regulators help rig markets, sometimes unwittingly. IEX’s CEO said that their joint venture with FTX was focused on helping “shape” regulation, drawing from their experience doing so with equity markets. It’s not a big stretch to think this “shaping” aimed to help FTX with its desire to replace the whole financial system with something they could claim is “unrigged.”

[2] My (admittedly crude) understanding is that exchanges are required to disclose all 5%+ owners on their Form 1 Exhibit K (p7). It appears that IEX has not done so yet. I don’t know why.

A Quick Note About the FCA Occasional Paper on “Latency Arbitrage”

Aquilina, Budish, and O’Neill (ABO) today published a paper on the “HFT Arms Race” with the headline that estimates a “latency arbitrage tax” on global investors of $5B/yr. The paper shows some interesting statistics on a unique dataset, raw messages sent to the LSE which are captured and timestamped by one of LSE’s TAPs. Unfortunately, I don’t see how the activity measured in the paper resembles anything approaching a “tax”. I’ve only had a short time to read the paper, so I apologize if I’ve made any errors. And the paper does not necessarily reflect the views of the FCA.

Essentially ABO measure messaging activity that occurs in short bursts, particularly batches of marketable orders — possibly accompanied by cancel-requests — where some of the later orders or cancel-requests fail because an earlier message succeeded. E.g. if 3 marketable orders (from different trader IDs) are sent in short succession when the order book only has the quantity available to fill one of them, then that is identified as a “race” which the first order wins and the later 2 orders lose. Similarly, a “race” is also identified if the LSE receives marketable orders shortly before it receives cancel requests for the matched resting orders. In this case, the failed cancel-requests are the losers of the “race”. The authors provide event counts, volume, and estimated profit for race-winners — where they vary the duration window to identify a “race”, as well as other parameters such as race-type. The headline number of a 0.42bp “tax” is the race-winners’ estimated profits as a share of total volume, where the race window is set to 500us.

Figure 5.1 is a histogram (truncated at 500us) of the time-differences between the first successful message in a race, and the first message to fail (either by being too late to cancel, or trying to trade with liquidity that’s already gone). It’s no surprise that the distribution is highly-skewed, with a mode of 5-10us and a 90th percentile of about 200us. And Figure 4.1 shows reaction-times on the LSE for a different type of event, which the authors use to estimate LSE latency at around 29us. Tables 5.5, 5.6, and Figure 5.3 also show the midpoint price trajectory after race events. The authors do a very nice sensitivity analysis that shows how the numbers vary if you change the race window and other parameters. It’s important to note that the longer the race window, the more it will blend messages stemming from different market events. The 500us window selected could help include failed messages sent across Europe on slow networks, but it also includes ~15 roundtrip times from the LSE — enough time for order books on correlated products to react and refill multiple times. One noteworthy measurement is in Table 6.3, which gives an idea of the volume that market-makers perhaps wish they could avoid by being faster: the approximate “profit” that marketable orders make from trading with someone who tried-and-failed to cancel within 50us. This number is very low, about 0.03bp, an indicator of a pretty efficient market where market-makers are not getting “picked-off” very often.

Now, it’s very interesting to see statistics on these types of events, but I don’t see how they can be considered a “tax” in any way. All they show is that trading is competitive, and that traders may send messages even if there’s a low probability of success. As a real-world example, consider America’s favorite holiday, Black Friday. Perhaps your local toy store is selling the new Nintendo for $100 off, but they only have 10 available with the discount. You know it’ll be a long shot, but you really want to catch some Pokemon, so you get up early and head to the store. You get there just in time to be 10th in line. A bunch of people arrive a few minutes later, but they all have to pay full price.

Did you levy a tax on them? It’s hard to see how. You got there first and you got the reward. Your time and transportation cost were arguably wasted. But in financial markets, investments in network infrastructure which ensure the law of one price have significant economic value. I suppose Nintendo could forbid discounts, and maybe everyone else could get 0.004% off, but Nintendo gets to decide how to sell their wares. It’s a silly analogy, but maybe stocks aren’t so different. If a shareholder prefers to disengage from this type of race, she can sell her stock in an auction if she likes. And the paper appears to suggest that a way to avoid the “tax” is to have everyone trade in auctions, referring to the batch auction proposal from Budish, Cramton, and Shim (BCS):

BCS showed that even information seen and understood by many market participants essentially simultaneously — e.g., a change in the price of a highly-correlated asset or index, or of the same asset but on a different venue, etc. —creates arbitrage rents too. These rents lead to a never-ending arms race for speed, to be ever-so-slightly faster to react to new public information, and harm investors, because the rents are like a tax on market liquidity. BCS showed that the problem can be fixed with a subtle change to the underlying market design, specifically to discrete-time batch-process auctions; this preserves the useful function of algorithmic trading while eliminating latency arbitrage and the arms race.

Of course, auction prices are still partly set by proprietary traders, sometimes in a latency-sensitive way. The BCS proposal claims to eliminate “latency arbitrage” by having auctions every 100ms or so, with only price priority and ties broken randomly. To further prevent “latency arbitrage”, BCS proposes no pre-auction transparency. Normal exchange auctions have a long period of transparency before the final match occurs, in order to attract liquidity and ensure a sensible price. This transparency often takes the form of a partly displayed order book (e.g. on many futures exchanges), or indicative pricing and order imbalances (e.g. on many equities exchanges). Pre-auction transparency does make trading somewhat latency-sensitive, but exchanges in the real world know that having an auction blind can result in chaotic pricing — even when auctions occur once a day on highly liquid products. Holding blind auctions so frequently that they have only 0 or 1 orders of natural liquidity, and where traders are rewarded with priority for aggressive pricing, is a recipe for disaster. Prop traders would probably make a lot of money off of the ensuing chaos, but the uproar might be so intense that few of them would recommended it.

More generally, there is a tradeoff between immediacy and liquidity. If you want to have liquid auctions, they probably can’t be very frequent. Donier and Bouchaud suggest that for batch auctions to improve liquidity they should occur on the order of once per hour. I think most traders like auctions — BCS is right about that. And since we’re discussing pet market structure reforms with no chance of happening, I may as well suggest one that would give us more auctions and could save something of far more significant value: people’s time. The industry currently spends a lot of time keeping markets open when they probably don’t need to be. We could reduce that time, which currently has many professionals in the office for 11 hours a day. Stock exchanges could hold 2-4 sessions per day, each around 30-60 minutes long and overlapping with market hours in other major financial centers. This schedule could encourage working from home and a more clear demarcation of business hours between offices in different timezones.

I enjoyed seeing the statistics in ABO and wish regulators would publish more statistics on raw market data. The “tax” claim seems a bit strange though.

Related Party Transactions In ETH?

Related party transactions are a common feature of pump-and-dumps. Nefarious issuers may sell securities to their fiduciaries, affiliates, or family — without disclosure to other buyers. These schemes can create the illusion that a security has a broader and deeper market than reality. Thin floats also make it easier for manipulators to create price momentum.

Because of their pseudonymity, it’s hard to rule out related-party transactions and wash trading in crypto-assets. Usually investors only have founders’ words to rely on. In the case of Ethereum, some of their words are troubling. To be clear, this post is speculative and I have no evidence of Ethereum founders acting nefariously. There’s also nothing necessarily wrong with fiduciaries trading their own security, as long as it’s properly disclosed. [1] But if I were an investor or regulator, I’d want to know more about some of the facts below.

Founders Potentially Bidding in Their Own ICO

Laura Shin relayed a rumor that Ethereum cofounder Joe Lubin purchased around 10-20% of the ETH ICO:

Rumored to be one of the top buyers in the Ethereum crowdsale, Lubin, who had been funding ConsenSys with his stash of Bitcoins, says he began selling some of his Ethers last year to fund the firm’s development…
Crypto-industry insiders believe he may be the single-largest holder of Ether, with valuations as high as $10 billion, but Lubin insists he’s been selling.

The Times’s Danny Fortson later reported that numerous founders had traded in the ICO:

The tokens sale went off without a hitch, selling at about 30 cents apiece and raising $18m. The founders all bought large holdings in the newly created currency.

If Shin’s rumor and Fortson’s reporting are correct, then ETH insiders may have bought a large portion of the offering. How could they have hidden such large purchases from the market? Prior to the offering, Lubin explained that Ethereum would allow purchasers to hide their identities, and encouraged large traders to split their orders over many pseudononymous accounts:

Q: Will there be a limit on the amount that a person can invest in Ethereum?

Lubin: A person can buy from any number of different identities. We may limit the size — the unit size of a sale — just to make it easier to disguise. Let’s say if you’re a whale and you want some privacy, you can buy 50,000 units, just so nobody scares people with an enormous purchase. So if you are a whale, if you’re planning on investing several billion US dollars worth, then you can do that with multiple identities. We will ask for real-world identities in the form of an email address, just so that we can make sure that everything works smoothly… But we won’t be requiring it. So you can create a pseudonoymous email identity and actually purchase from it.

Prior Activity

Lubin says he started “a hedge fund or a set of hedge funds,” before cofounding Ethereum. I can’t find any details about this “set” of hedge funds, except Fortson’s description in The Times:

He worked as a software engineer before ending up on Wall Street, where he managed IT projects for Goldman Sachs’s wealth management division. The job piqued his interest in finance. He left Goldman and founded and ran a few hedge funds, which used a controversial trading strategy that took advantage of mutual fund price inefficiencies.

Could that indicate that Lubin’s hedge funds engaged in mutual fund timing? I don’t know, but it might fit with his above statement about disguising order flow with multiple pseudonymous accounts. Mutual fund timers have employed similar schemes.

Inside Information


William Hinman, the SEC’s Director of Corporate Finance recently said that he believes Ether is not currently a security:

[P]utting aside the fundraising that accompanied the creation of Ether, based on my understanding of the present state of Ether, the Ethereum network and its decentralized structure, current offers and sales of Ether are not securities transactions. And, as with Bitcoin, applying the disclosure regime of the federal securities laws to current transactions in Ether would seem to add little value.

If ETH ownership is (or was) more concentrated than he thought, would that change his view?

Many ETH investors are hoping for further development from the Ethereum team, including dividends via the promised fork(s) that enable sharding, Casper FFG, proof-of-stake, etc. [2][3] These efforts (and their promotion) are not just performed by the Ethereum Foundation, but also ConsenSys — which Laura Shin reports is partly funded from Lubin’s personal sales of ETH on the secondary market. So Ethereum founders may be trading tokens while possessing special information about future dividends and their own forward-looking statements. Again, there may be nothing nefarious going on here, but the community might be interested in seeing founders’ trading activity. If Ether is a security, then this disclosure is probably mandatory. But even if securities laws aren’t applied to ETH, the market could potentially demand that information.

Paranoia can help people navigate the traditional financial system. Though usually participants’ darkest anxieties don’t keep them from losing anything except a good night’s sleep. The crypto financial system has the opposite problem. Seasoned crypto participants imagine wild conspiracies plotting to take their money [4], and yet sometimes their conspiracy theories aren’t crazy enough. It may seem far-fetched that ETH is closely-held, but investors would be naive not to ask.


[1] The remaining fragments of the 2014 ETH ICO announcement committed founders to a lockup period:

The fiduciary members will be vested over a 12 month period and cannot divest more than 1/3 of their position per year thereafter

But Vitalik Buterin said in 2017:

[P]resent-day me does NOT endorse everything that we did in the 2014 presale. For example the lack of lockup periods for founders was not good imo.


[2] Forks are often equated to dividends.


[3] E.g., Lubin in March 2018:

One year, with respect to Ethereum: we’re going to replace the proof work algorithm with a much more efficient proof of stake algorithm. So that’s going to give us greater scalability.

In May:

Roubini: Five years ago, Vitalik Buterin said ‘We’ll find the solution to scalability, right. It’s going to be proof-of-stake.’ Five years later, you’re telling me it’s going to come now. It hasn’t been there. So that problem of scalability with decentralization, security has not been resolved. And will never be resolved.

Lubin: I can show you the code.

Roubini: Come on. You’ve been saying these things for five years. Come on, five years. And we don’t have it yet.

Lubin: We’ll go back to the office after — I’ll show you the code [crosstalk]

Roubini: It’s vaporware.


[4] At the other end of the spectrum, there are also retail traders who assume that crypto markets have the same protections as regulated securities exchanges. (They don’t.)

If ETH Isn’t a Security Then Nothing Is

Ethereum tokens (ETH) are often considered commodities but not securities, like Bitcoins. [1] But I think that ETH meets any common sense definition of “security” and that the ETH ecosystem needs to be closely examined by securities regulators. I’m not a lawyer, and not the right person to be making this argument, but it seems important and urgent — so I hope you’ll forgive any errors. [2]

Ethereum’s Announcement

The Ethereum Project was announced on the bitcointalk.org forum in January 2014 by cofounder Vitalik Buterin. The announcement solicited “investors and evangelists” who would be protected by a “fiduciary team,” and linked to an announcement blog post titled “Ethereum: Now Going Public,” which said:

We will be able to develop fully functional and robust Ethereum clients with as little as 500 BTC funding… To that end, we would like to be able to invest a large quantity of funds into securing top-notch talent for improving the security and scalability of the Ethereum network itself, but also supporting a robust Ethereum ecosystem

An archived snapshot of Buterin’s announcement (taken weeks later) commits the “fiduciary team” to certain trading restrictions, and contains fragments of the planned “IPO and Money Supply of Ethereum”:

The goal of our IPO is to endow an organization, collection of contributors and strategic partners with enough funding to ensure both the Ethereum protocol and the Ethereum ecosystem are bootstrapped to a point of critical mass. We also seek to develop a new type of venture capital

Their whitepaper similarly says:

The issuance model will be as follows:

Ether will be released in a currency sale at the price of 1000-2000 ether per BTC, a mechanism intended to fund the Ethereum organization and pay for development that has been used with success by other platforms such as Mastercoin and NXT. Earlier buyers will benefit from larger discounts. The BTC received from the sale will be used entirely to pay salaries and bounties to developers and invested into various for-profit and non-profit projects in the Ethereum and cryptocurrency ecosystem.

Definition of a Security

In its order halting the Munchee ICO, the SEC stated:

An investment contract is an investment of money in a common enterprise with a reasonable expectation of profits to be derived from the entrepreneurial or managerial efforts of others.

Investment contracts are securities in the US.

The ETH “IPO” is obviously a security by this definition. But there are signs that Ethereum sought legal advice and changed their model. [3] Buterin’s final pre-offering announcement says:

Ether is a product, NOT a security or investment offering… The Terms and Conditions of the Ethereum Genesis Sale, and the Ether Product Purchase Agreement, are the only authoritative documents regarding the sale. Any statements that we have made up to this point are null and void. The ethereum.org website, the blog posts since Taylor Gerring’s recent howto and other posts we have made on various forums including our forum and reddit are to be treated as informal communications designed to keep the community updated on what is happening with Ethereum.

It’s reasonable to let people change their minds after seeking legal advice, provided they haven’t done much yet. Ethereum insiders though, had aggressively marketed the sale as a securities offering beforehand. [4] If they later changed their minds, they seem to have made little effort to correct the public’s perception that ETH would be an investment contract. And the offering documents themselves continue to describe a securities offering [5]:

Parties may be interested in purchasing ETH… to support the development of the Ethereum Platform.

The ETH offering was also not an unregistered SAFT. A SAFT is an untested structure that could fund a coin’s development with a securities offering, and later delivers a finished coin. A coin initially funded by a SAFT *might* not be a security, if “entrepreneurial or managerial efforts” truly do cease after the coin is distributed. When ETH officially launched, the foundation specifically said that it would continue developing Ethereum using funds raised from ETH sales. [6] The foundation’s entrepreneurial and managerial activities continue today.

So Ethereum entities made a cursory effort to claim that ETH isn’t a security, but before, during, and after the offering, they made it otherwise clear that ETH is a security.

ETH’s Flow of Funds

The following is just my general impression from reading ETH documents, and may have errors.

The initial offering raised around 30k BTC, which I believe went to a variety of entities controlled by the Ethereum Foundation. Most of these BTC have been distributed to individuals or sold.

Approximately 6M ETH were retained by the foundation’s treasury, some of which were later sold into the secondary market. As of January 2016, 2.3M ETH were still held by the foundation, with the remainder sold to fund operations or “distributed to its rightful owners”. [7]

Another 6M ETH were “gifted” to early contributors. Some of these “gifts” have been sold into the secondary market. [8]

Economics of ETH Ownership

ETH may be a security, but it is clearly not equity or debt. Normal equity gives owners exposure to all of a corporation’s unencumbered property. ETH gives owners exposure only to the project’s intangible property. Since the project is open source, it probably has limited intellectual property (though it does own trademarks). So, the majority of ETH’s value is derived from its brand, platform, and network.

Parts of the crypto community may argue that this model is fundamentally new to markets. It doesn’t feel that way to me. Many of the most successful companies in the world derive their value from intangible assets.

Even under the modern securities disclosure regime, valuing intangible property can be quite difficult. Market history is littered with the bodies of roll-up companies, and there are indications that ETH could be in this category. As far as I can tell, the biggest use of ETH is to fund ICOs, many of which are themselves platform “companies.” [9] This usage of ETH appears to have been intended from the beginning. In an illuminating interview (promoted by Ethereum) one month before the ETH ICO, Vitalik Buterin described his ideal types of Ethereum-powered platforms — gambling and unregulated fundraising:

Danaylov: If our viewers and listeners were to take a single message from our one hour conversation with you today — like the most important thing that you’d like them to remember — what would that be?

Buterin: Yeah. I would say it really boils down to this idea that Ethereum is not a currency… You can use it [decentralized consensus methodology] for so much more, and I think now is actually the time when we’re starting to move away from ‘Oh, let’s have a decentralized currency for decentralized currency’s own sake.’ And actually realize that there’s all sorts of these different applications where you can actually do so much that you can’t even do in the existing paradigm at all… One of the examples that I always like to bring up is actually gambling. So in 2006, if you wanted to make a gambling site, it would be extremely difficult… In 2012 it got a bit easier, because you got Bitcoin. And with Bitcoin, you can actually do all the currency part without most of the security, without the payment processing, and so forth. In 2014, guess what the gambling sites of the future are going to look like? They’re just going to be Ethereum contracts in 15 lines of code: write them once, click compile, click ‘push to blockchain,’ ‘push to blockchain’ returns to you an Ethereum address, paste the Ethereum address onto a forum, and bam, anyone can gamble.

Danaylov: Well, well, I can see how the casinos are going to love you for that. But I’m personally one of those crazy people who believe that it’s good that it’s hard to get involved into gambling — online or offline.

Buterin: It’s an example, alright. The idea is that you’re taking this industry — it’s actually very bad for consumers right now — because it’s very easy for gambling sites themselves to cheat… Gambling is such a nice example because it’s this sort of pure financial service.

Danaylov: It’s not my favorite example. So let’s do one better than that…

Buterin: So, a good example might be Kickstarter — is another one. So Kickstarter — once again, it’s a great idea — you can actually grab funding together from thousands of different people. People can start these large, well-funded projects without having to go to a VC. It’s good. But at the same time, it’s also inefficient… All the other sites take a pretty good cut too. And at the same time, they’re also geographically limited. There’s a lot of countries I’m sure you can’t really access them at all, right?

So, ETH is supposedly valuable because it’s a platform for other platforms (some of which are, also, platforms for other platforms). It is perhaps more of a roll-down than roll-up. Regardless, the danger of this model makes it doubly important for market gatekeepers (e.g. lawyers, brokers, and exchanges) to do their jobs and ensure securities they help sell are reasonable and compliant.

ETH Brokers and Exchanges

The SEC has made it clear that cryptoexchanges must be registered as ATSs or national securities exchanges if they enable exchange-like trading of securities. That is, if ETH is a security, every major American cryptovenue may be in violation of the law. [10] I don’t know what the penalties would be for violating these (well-known) rules, but they might be higher for venues that understand ETH is a security. Coinbase is potentially one such example.

Coinbase, in marketing ETH, pointed to its active developer community and leadership. [11] Even a casual inspection would have revealed that ETH’s developers and leaders are partly funded through ETH sales. [12] So, either Coinbase didn’t do the most basic diligence before helping retail investors purchase ETH, or they had good reason to believe that ETH was a security.

Update: It appears that Coinbase did recognize that ETH development was financed by ETH sales from the primary offering and later sales in the secondary market. Coinbase also explicitly described ETH fundraising as analogous to traditional securities offerings. This history is particularly interesting in light of Coinbase’s recent Congressional testimony.[13]

We Still Need Securities Regulation

It is easy to get frustrated with our regulatory regime. In one view, unethical participants ought to be deterred by markets’ self-policing mechanisms, so why should we burden ethical participants with all this red tape? It’s been nearly a century since the Great Depression — maybe humans have gotten smarter. Maybe we don’t need post-depression rules anymore.

Cryptocurrencies have been a fascinating natural experiment on the consequences of a complete rollback of regulation. So far, the results are frightening. Perhaps the cryptomarket would eventually learn to police itself, but it increasingly looks like there’d be unacceptable carnage on the way. It’s time to end the experiment before too many people get hurt, and make cryptocurencies follow the same rules as everybody else. Then we can let the market determine their value.


[1] Peter Van Valkenburgh of the Coin Center says:

They [The SEC] have also begun an important process of drawing a distinction between “initial coin offerings” that are securities on the one hand, and decentralized cryptocurrencies on the other, which don’t fit the bill and are more suitably regulated as digital commodities such as digital gold (bitcoin), digital fuel (ether), or digital real estate (filecoin/storj).


[2] This post contains no advice to buy, sell, or trade securities, commodities, virtual currencies, derivatives, or other financial instruments.


[3] Ethereum cofounders Joseph Lubin and Stephen Tual, 3 months before the offering:

We have done even more work on the legal front. We have been working, and continue to work with lawyers in different countries to structure the pre-sale of ether so that it is not seen by securities regulators as an offering of securities. We believe that selling ETH, a fuel that powers applications on the Ethereum platform, does not constitute sale of a security, be we hope to achieve more certainty before starting the sale.


[4] A small sample of times when ETH was described as an investment contract before the offering:

In January 2014, Vitalik Buterin responded to questions on reddit:

Q: Why will these coins [founder grants] be spendable after one year and not immediately?
Buterin: To prove to the world that we are not a pump and dump and are interested in Ethereum for the long term.

Q: What benefits do you have of personally owning pre-mined coins?
Buterin: What benefits do we have? We get profit if Ethereum succeeds; this is obvious. What benefit do you have? The assurance that the five originators of the project are heavily incentivized to do everything in their power to make as high quality a product as possible and not abandon it.

Q: Why create a non-profit?
Buterin: To give fundraiser participants the assurance that they have legal recourse if we run away, and to prevent double taxation (ie. we want to only pay taxes at the point when we spend the money on salaries and bounties, not several times before that)

Q: Why use a fundraiser?
Buterin: To get funds to develop the protocol and applications around the protocol.

Buterin in March 2014:

Q: What’s to prevent this altcoin from developing the same benefits that Ethereum has?

Buterin: Absolutely nothing. We’re even handicapping ourselves, just because everything we do is completely open source. If you’re investing in Ethereum, I would say you’re investing in the team and the community to a large extent.

Joseph Lubin, in May 2014 (2 months before the offering, and 1 month after he stated [3] that ETH was not a security) :

People and businesses are interested in purchasing ETH to power their own business applications, to make use of business applications offered by other service providers, to trade on forthcoming exchanges, or to speculatively hold for future sale to people and businesses. ETH may be purchased in the Genesis Sale (details forthcoming, please watch this space), on forthcoming 3rd-party exchanges and ATMs, and on exchanges that are implemented as DApps on Ethereum.

When purchasing ETH in the Genesis Sale, the buyer is supporting the development of the product, just as with a kickstarter campaign.


[5] Employees and volunteers were to develop the ETH platform using proceeds from the offering and future sales in the secondary market:

The Ethereum Platform is being developed primarily by a volunteer contributor team — many of whom will be receiving gifts of ETH in acknowledgement of their dedication — and will continue to be developed on a volunteer basis by some developers as well as under a more formalized contracting or employment relationship for other developers. The group of developers and other personnel that is now, or will be, employed by, or contracted with, Ethereum Switzerland GmbH (“EthSuisse”) is termed the “Ethereum Team.” EthSuisse will be liquidated shortly after creation of genesis block, and EthSuisse anticipates (but does not guarantee) that after it is dissolved the Ethereum Platform will continue to be developed by persons and entities who support Ethereum, including both volunteers and developers who are paid by nonprofit entities interested in supporting the Ethereum Platform.

The “intended uses of revenue” included development, outreach, research, and an allocation to a non-profit foundation. Page 4 of the DevPlan also describes their plans for 2 non-profit entities, a short-lived for-profit corporation, and an SRO.

I’m not aware of any progress towards the SRO in the ensuing 4 years. It’s common in the cryptoverse to hear about SROs, but rare to actually see one.


[6] Immediately after the launch announcement, on the Ethereum blog:

The Ethereum Foundation is a not for profit (‘Stiftung’) organization registered in Switzerland, and has the purpose of managing the funds that were raised from the Ether Sale in order to best serve the Ethereum and decentralized technology ecosystem.


[7] I’m not sure, but I think that some of the foundation’s ETH allocation was distributed to employees as part of the “developer purchase program.” E.g. Buterin:

I received ~553,000 in the genesis premine and ~150,000 from the developer purchase program. I currently hold ~630,000 ETH.


[8] E.g. Buterin in 2016:

I’ve sold about a quarter of my ETH.


[9] And CryptoKitties.


[10] It’s especially important to determine the legal status of ETH trading if there are going to be ETH futures and ETFs. Cofounder Joseph Lubin predicted that ETH futures would soon launch, and ETFs would follow — in December 2017:

Q: When do you expect that we could see an Ethereum future out there?
Lubin: It’s… I don’t want to really…
Q: Months?
Lubin: I think months would be a long, long time.
Q: Weeks?
Lubin: We’ll see.


[11] When Coinbase began trading ETH, they recommended a blog post from cofounder and market professional Fred Ehrsam:

To learn more about Ethereum, please visit our Ethereum FAQ page and co-founder Fred Ehrsam’s medium post “Ethereum is the Forefront of Digital Currency”.

Ehrsam wrote:

Ethereum’s core development team is healthy while Bitcoin’s is dysfunctional
Vitalik Buterin, the creator of Ethereum, has shown early promise as the leader of an open source project. He seems both comfortable as a community and technical leader…
Meanwhile, the core development team in Ethereum is focused. This is evident from the Ethereum blog. When I started reading it, it was everything I found myself thinking about for the present and future of Bitcoin but didn’t see being discussed much: scaling the network, the viability of proof of stake, how to create a stable digital currency, what a blockchain based company (DAO) would look like, amongst other topics. These are very ambitious ideas and some won’t work. But some probably will work, and they will be important…
What is very real, though, is the possibility that Ethereum blows past Bitcoin entirely. There is nothing that Bitcoin can do which Ethereum can’t. While Ethereum is less battle tested, it is moving faster, has better leadership, and has more developer mindshare….


[12] Miners also perform a service for Ethereum in exchange for new ETH tokens. It is not so different from a company paying contractors with newly-issued equity. Perhaps mining is a business activity funded by Ethereum issuing and selling securities?

[13] In November 2016, after Coinbase listed ETH, Ehrsam said on the Coinbase blog:

And therein lies a key point in the early stages of protocol tokens. We have yet to hit a situation where the combination of funds set aside in the initial sale and tokens held back have run out — yet. This is a close equivalent to running out of cash and the options pool in a typical company. Ethereum, for example, hasn’t hit this bump yet because 1) it’s young and 2) the price of ETH vs every fiat currency has, more or less, been on the rise, making the ETH the Ethereum Foundation set aside last longer. As a result, we’ve never seen the idea of multiple fundraises as is common with startups. Running out of funds for future development seems likely, as work on these protocols is never really “done”, although they may mature and change less over time…

This model is new and largely untested. The most important question is whether or not it will be considered a security under US securities law…

Most groups who have raised funds through a token sale create some kind of entity to handle the funds raised. The most common structure at the moment is setting up an (often Swiss) foundation. This foundation has some governance, much like the corporate governance you’d see in a normal company, which determines both the initial and ongoing distribution of funds raised. This is what Ethereum has done, more or less.

A month earlier, Ehrsam made similar remarks:

If you want to make this analogous to kind of regular companies in the world, you can think about the proceeds — whether it’s Ether or dollars from the pre-sale — as like the cash that a company might raise. And then the tokens that they hold back as kind of like the options pool, more or less. One thing I think is interesting that we haven’t seen a whole lot of yet in the world is: (And this happens at companies and startups all the time) What happens when those initial funds from the initial fundraise run out? Right? There’s kind of this model at least so far — and these are all really early projects — of we just do one fundraise and hopefully that just funds the thing in perpetuity. I’m guessing that that will not continue forever. With Ethereum, one of the main reasons it’s worked so far is: just like a successful company, the Ether that the foundation holds have appreciated quite significantly in value — meaning they could continue to pay for development more and more and more. So one thing I’m quite interested in is: what happens in the future, when those initial funds that we held back start running out and we want to continue funding the development of these protocols.

In February 2017, Coinbase continued to give the impression that ETH was a security:

Raising Capital

Lastly, one of the major use cases for Ethereum is decentralized fundraising from a global network of investors. Crowdsales lower the barrier to entry for developers working on high risk projects. Since Ethereum launched in July 2015 we have seen unprecedented amounts of funds raised for decentralized applications through crowdsales. Ethereum itself was funded through a crowdsale that raised $18 million in bitcoin and a project called The DAO raised $160 million.

However, Coinbase testified in March, 2018:

Coinbase is very deliberate about which tokens it supports on the exchange. Currently, the exchange supports only four assets (BTC, ETH, LTC and BCH). Part of the reason we trade only those four assets is that each has been determined by regulators to be a virtual currency and therefore not a security. We are studiously avoiding listing tokens that could be determined to be securities because we are not currently licensed to trade securities and cannot take the risk of inadvertently trading an asset that is later found to be a security…

At Coinbase, we have worked to bring clarity to the issue of what assets we can support since 2016. To help potential market participants, we published our Digital Asset Framework to provide transparency about how we consider listing new assets. A key factor in our framework analysis is a determination that the potential new asset is not a security under U.S. law.

Some Questions for Crypto-Exchanges


1. Does your exchange engage in proprietary trading? [1]

A) If so, what is the nature of the prop-trading? Does it occur on your own exchange, other exchanges, or both?
B) Are there information barriers between client data and prop-trading decisions? Are customers ever front-run?
C) Does prop-trading have the same access to the exchange as other customers?
D) Are exchange operational decisions, like pausing (or not pausing) matching due to IT issues, ever influenced by profit or loss by your prop-trading?
E) Who oversees exchange prop-trading and ensures it is non-manipulative? Is oversight segregated from the rest of exchange operations?

2. Does your exchange offer securities trading functionality?

A) If so, is the exchange registered and approved by the appropriate regulators? For example, if the exchange is in the US and matches Ether (likely a security in the US [2]), is it registered as an ATS, National Securities Exchange, OTC market, Broker-Dealer, or otherwise?
B) Are the securities classified as penny stocks, NMS stocks, Regulation A stocks, or another type of stock?
C) What standards determine the securities that the exchange matches?
D) Does margin trading of securities comply with applicable regulation? [3]
E) Does the exchange report suspected insider trading to the appropriate regulators? Are exchange employees monitored to prevent them from insider trading?

3. For exchanges offering execution services such as Coinbase, how is the execution price determined?

A) Is the customer executed at the worst price over a lookback period? What is the lookback period, and can it vary?
B) Do you reject orders via last look? Is it symmetric? Does it comply with market norms? [4]
C) Are details like the above, and any additional spreads, fully disclosed to customers?
D) If you offer execution services for securities such as Ether, do they comply with best-execution requirements? Are you registered as a broker? Do your retail customers understand any risks of using your services?

4. Do clients have equal access to the exchange?

A) Can every client access the same market data feeds and order entry gateways? At similar latencies?
B) Do any clients receive different fees than the public fee schedule?
C) When the exchange has IT issues, do any clients get preferential access? E.g., when the exchange is offline for some clients, does it ever make efforts to stay online for favored clients?
D) What safeguards are in place to ensure client information isn’t leaked to other clients?
E) If there is a margin deficiency, theft, or other shortfall, how are losses distributed and who oversees the process?

5. What mechanisms are in place to prevent and police manipulation?

A) Such as: momentum ignition, denial-of-service attacks, triggering margin liquidations, collusion, spoofing, and settlement-price/benchmark manipulation.
B) Are margin levels set such that manipulative traders or hackers (outside developed market jurisdictions) can’t force liquidations?
C) What are the penalties and enforcement mechanisms? Are instances of manipulation disclosed to clients?
D) If the exchange lists derivatives that depend on other exchanges’ pricing, do you communicate with those exchanges about potential manipulation and their own prop-trading? Do you have confidence in their capacity to give you reliable and complete information? [5]

6. Does the exchange disclose and investigate security breaches?

A) E.g. if the exchange suffers a suspected DOS attack, when and how is that disclosed to customers?

7. Does the exchange guarantee that its customers are not buying stolen or laundered coins? If stolen coins were bought on the exchange by an innocent customer, and reclaimed, will the customer be compensated?


[1] It is somewhat reassuring that the Administrator of the CME’s Bitcoin settlement price, Crypto Facilities, says it does not engage in (at least some forms of) proprietary trading. In an interview with Bitcoin Futures Guide, the CEO said:

It [the Turbo 50x leverage product] also does not change the risk profile of Crypto Facilities as we are not a counterparty in the trades on our platform.


[2] The SEC “has determined that DAO Tokens are securities under the Securities Act of 1933.” You can read Ether’s offering documents and come to your own conclusions.


[3] I believe Kraken allows margin trading for retail customers of Ether and Tether at levels of up to 5x.


[4] I have heard anecdotally that, when the price moves in the customer’s favor, Coinbase may reject executions with error messages like “the exchange rate updated while you were waiting” — but may not always do such rejections when the price moves against the customer.


[5] The CME’s settlement price mechanism may be modified in “unusual and extreme circumstances.” The Administrator (Crypto Facilities) is responsible for making recommendations in such circumstances, and requires the approval of the CF Member of the Oversight Committee (who I think is currently the CEO of Crypto Facilities), and one of the CME’s representatives on the committee.

In the Bitcoin market, I’m not sure how unusual “unusual and extreme circumstances” are.

Do Bitcoin Markets have a Conversion Tort Problem?

In many jurisdictions, a buyer of stolen goods does not acquire their legal title. This principle, “nemo dat“, often applies regardless of whether the buyer knows the goods were stolen. The rightful owner can sue the current holder and take possession.

Bitcoins (and other cryptocurrencies) are well-known to be stolen in large quantities. Bitcoin thieves are usually hard to identify or outside the reach of developed markets’ courts, so victims believe they have no recourse. But the Bitcoin blockchain allows coins to be traced with high precision, and it seems possible for stolen coins to be returned once they re-enter the regulated domain. I’m not a lawyer and don’t understand the relevant intricacies, so this post is just intended to start a discussion. I may be wrong, but buying coins of unknown provenance could be a major risk for Bitcoin holders and intermediaries.

Pro-rata Tracking

All transfers between Bitcoin addresses are recorded and publicly accessible. In principle, this audit trail should make it easy for victims to track stolen coins and reclaim them if they enter the custody of an identifiable entity. There are, however, a few complications. Bitcoin transactions can have multiple inputs and outputs, so if only some inputs are stolen, it may be difficult to determine the portions of the outputs that should be returned.

Bitcoin is frequently called “digital gold,” and perhaps that analogy is useful for tracking ownership across multi-party transactions. As a toy example, say that Tom steals 1 bar of gold from Val. Tom then melts Val’s gold, combines it with another bar of gold, and sells the two resulting bars to unsuspecting buyers, Bob and Barbara (who buy 1 bar each). How should we determine how much of Bob’s and Barbara’s purchases belong to Val? I think the intuitive choice is that each of them should return half a bar. This pro-rata allocation seems fair even though we don’t know which of the individual gold atoms were previously Val’s, or how thoroughly Tom stirred the melted gold.

In some circumstances there could be documentation that suggests an alternate allocation. Suppose Tom steals a bar from Val, then entrusts it to a gold broker, who melts it with other bars. The broker then sells a bar of the blended gold to Bob on Tom’s behalf, recording a trade between them, and giving Tom the cash. Even though “Bob’s” bar doesn’t contain the same gold atoms as Val’s did, perhaps a sensible outcome is for Bob to return the entire bar to Val?

Applying the same intuition to Bitcoin, we could determine which current holders should return which fractions of their holdings, to which victims. The default flow of ownership would be pro-rata, unless there’s documentation proving otherwise — such as trades facilitated by an exchange with client coins in its custody. [1]

Legal Venue

Every jurisdiction deals with this liability differently. Many don’t follow “nemo dat” at all, instead using versions of the “market overt” rule. [2] Other venues only return stolen property if it’s tangible. [3] The statute of limitations can vary dramatically between venues. Another question is to what extent a “merchant exemption” applies to misappropriated coins. [4]

So, could lawsuits proceed in a victim-friendly venue like New York? You can argue that Bitcoin is inherently pan-jurisdictional. Bitcoins’ value derives from the consensus of participants on the network, who download the blockchain to confirm that coins are authentic and not part of a fork. If some coins were left out of the version downloaded by NY residents, those coins would be worth much less. Thus, every coin’s value is dependent on NY activity. [5]

Potential Consequences for Market Participants

If the above reasoning applies, then victims have a good chance of recovering their stolen coins. It’d be logistically complex, but I can imagine a group of victims demanding that major exchanges, custodians, and intermediaries return the appropriate amounts, if coins with certain address-chains come into their control.

If enforced by a powerful court, demands like this could have serious effects on the Bitcoin market. Many people would have coins — that they innocently bought — seized. These people might then sue their counterparties and exchanges. Depending on the number of stolen Bitcoins circulating, some exchanges could go bankrupt (if they have assets within the reach of the relevant court). OTC market makers, which may account for over half of volume [6], could have massive liabilities to their customers.

Bitcoins would have different values depending on their transaction history. [7] Coins considered “suspicious” [8] or sold without a guarantee from a well-capitalized counterparty [9], might be worth much less than freshly mined ones. In an extreme case, the market could cease to function until the provenance of enough Bitcoins is publicly known. And perhaps future transactions would have to be fingerprinted and authenticated though an ancillary blockchain.

National Stolen Property Act

Is it possible that it’s a crime to knowingly sell (or transmit) stolen Bitcoins? I have no understanding of this, but if you’re worried you possess stolen coins and want to sell them, you may want to consult a lawyer first. The National Stolen Property Act could apply to intangible goods, and I don’t want to hear about any readers of this blog going to jail.

Bringing Transparency to Bitcoin

Established institutions are trying to make Bitcoin a more trustworthy asset. That requires knowing that marketable coins have similar values and are legitimately owned, as well as reducing the incentives of crime. There’d be some pain involved in reversing past thefts, but it may be better for the Bitcoin community to deal with these issues sooner rather than later.


[1] Some Bitcoin exchanges may not keep good records of counterparties to trades they facilitated. Perhaps the allocation method should be pro-rata in these cases, i.e. anyone with Bitcoin in the custody of such an exchange has their coins partly tainted when stolen coins are deposited at the same exchange address. This may feel unfair to the innocent depositors at the exchange, but it does seem a lot like a shady gold broker combining legitimate with stolen gold, and returning the blended product to innocent customers. Customers may feel bilked when a victim of theft reclaims gold that the customers believed was theirs, but that’s the fault of the broker, who should compensate them.

This method makes particular sense in the extreme case of a Bitcoin exchange that intentionally loses records, which is arguably similar to a “mixer.” In the gold analogy, a “mixer” would melt together gold from many sources, and return the same quantities to the sources. You can certainly question the motivations of mixer-users, but even those acting in good faith probably have an idea that they might be acquiring tainted coins.


[2] “Market overt” generally grants the title to stolen goods to an innocent buyer. See, e.g., Schwartz’s footnote 15 and Appendix for a brief overview of jurisdictions that follow “market overt.”


[3] Bitcoins are generally considered electronic and intangible, but some people print them out on sheets of paper.


[4] In the US, if a merchant sells goods entrusted to it in the “ordinary course of business” — whether authorized or not — the purchaser acquires the title. So perhaps if a Bitcoin exchange sold client coins to an innocent buyer on the exchange, without transferring the proceeds to the client, the buyer would acquire the title to the coins (in the US). But, if an exchange absconded with client coins and sold them later, outside the normal course of business, the victim would still own the title. Likewise for coins stolen from an exchange by a third party.

This treatment could have important consequences for the missing Mt. Gox coins, worth several billion USD at current prices. We still don’t know the full story behind those coins, but it seems likely that the victims own the title in the eyes of many jurisdictions.


[5] Similar arguments have allowed lawsuits concerning allegedly misappropriated art to proceed in NY.


[6] Bobby Cho of Cumberland, a DRW subsidiary and large OTC market maker, says that:

On any given day, exchanges are reporting anywhere from $200 million to $250 million traded in a 24-hour period. OTC trading is not reported anywhere, but I’d imagine that the OTC market is larger than that number.


[7] JP Koning discusses the notion of fungibility as it relates to “nemo dat.” That is, if “nemo dat” applied to legal tender, identical units of currency would have different (and hard to ascertain) values depending on their history. If only one jurisdiction applies “nemo dat” to Bitcoin, then its value could become unstandardized everywhere.


[8] Especially high-risk coins might include the Silk Road coins auctioned by the US Marshals. The US Government said that Ross Ulbricht “is the only person or entity reasonably believed to be a potential claimant to the Computer Hardware Bitcoins.” I don’t know how civil forfeiture works, so maybe this doesn’t matter at all, but isn’t there at least a chance some of those coins didn’t belong to him? Ulbricht was convicted of charges related to drug-dealing, hacking, and money-laundering. I’m not sure it’d be a surprise if something in his possession had been stolen.

GBTC, a Bitcoin-holding trust listed on OTC Markets, bought 48,000 BTC from this source. Cumberland reportedly bought 27,000 BTC. Tim Draper bought around 30,000 of “these bitcoins with a tainted past,” and deployed them at market maker and exchange Mirror. Exchange itBit, which will form part of the settlement price for CME Bitcoin futures, bought 13,000 BTC.


[9] I don’t know if OTC market makers keep large cash reserves, but perhaps some clients patronize them because they offer more reliable protections in case they accidentally sell illegitimate Bitcoins. I can imagine a situation where clients purchasing Bitcoin require market makers to post margin as an additional protection; and conversely market makers could require margin when buying coins from clients.

A Curious Feature of IEX Auctions

IEX is busy setting up its listing business and has designed an auction mechanism. If I understand IEX’s documentation properly — and I might not [1] — their auctions may have some pretty weird functionality. [2]

IEX’s auction documentation says:

For the Opening/Closing Auction, non-displayed buy (sell) orders on the Continuous Book with a resting price within the Reference Price Range will be priced at the Protected NBB (NBO) for the purpose of determining the clearing price, but will be ranked and eligible for execution in the Opening/Closing Auction match at the order’s resting price.

Now, I’m not easily surprised by complex exchange logic, but this got my attention. Again, I may have misunderstood something, but if I haven’t, this logic could be controversial with long-term traders once they familiarize themselves with IEX auctions.

Take IEX’s “Example 1” from their “CLEARING PRICE EXAMPLES”:

The Closing Auction Book includes the following orders:
 LOC order to buy 1,500 shares with a limit price of $10.10; and
 LOC order to sell 1,000 shares with a limit price of $10.10.
Shares are maximized at $10.10; therefore
 1,000 shares would execute at the IEX Official Closing Price of $10.10.

My understanding is that a hidden order submitted after these LOC orders could gain priority without providing a better price. If the NBBO were 10.09/10.13, and someone submitted a hidden bid at $10.11 100ms before the auction, I believe IEX’s auction logic would still price the auction at $10.10 [3] — but the hidden order would be executed in place of the LOC bid. This may feel “unfair” because the hidden order was submitted long after the LOC order and had virtually zero risk of moving the matching price (because the hidden bid is priced at the $10.09 “for the purpose of determining the clearing price”). And unlike LOC (or MOC) orders, the hidden order does not have a “no-cancel” period, so could be canceled even 1ms before the auction.

“Example 2” in “PRIORITY OF EXECUTION EXAMPLES” provides a slightly more complex illustration of this behavior. [4]

Opaque, Non-Competitive Auctions

Under certain conditions, this functionality gives traders the opportunity to simultaneously gain priority while offering a less competitive price, a rare confluence in normal markets. One reason why traders submit LOC/MOC orders (and displayed orders) is because they want to announce their intentions to attract liquidity. IEX’s hidden order logic shouldn’t change that. But other auction traders only announce their intentions because they want priority. This latter group may find the advantages of hidden orders hard to resist.

Here is a potentially problematic example:

  1. After the final opportunity to submit LOC/MOC orders, a closing auction has a big sell imbalance: there is 1 LOC buy at $10.00 for 1,000 shares, and 1 LOC sell at $10.00 for 10,000 shares.
  2. In the remaining 10 seconds of the day, traders will compete to trade with the sell imbalance. On IEX, the only way they can do so is by using displayed and undisplayed orders.
  3. Maybe 10ms before the close, the NBBO is $10.05/$10.07 (with all displayed quantity on IEX).
  4. The bidder at $10.05 realizes they can get a better price and better priority in the auction by deleting their displayed order and resubmitting a hidden buy order at $10.06. They do so, and the NBBO is now $10.04/$10.07.
  5. The bidder at $10.04 does the same thing. Likewise those at $10.03, $10.02, and $10.01. The NBBO is now $10.00/$10.07.
  6. The auction occurs at $10.00. The hidden orders are filled completely, but the LOC bid at $10.00 doesn’t receive any execution.[*]

I can’t imagine that either LOC side would be happy in this circumstance. The LOC sell order was filled at a non-competitive price. And the LOC buy order was skipped-over in favor of hidden orders that were submitted after it, but effectively at the same price. This process could also negatively affect market stability by causing the NBBO to widen at the most important and volatile part of the day. [5]

Now, I don’t think this sort of outcome will happen most of the time. But it certainly could happen some of the time. And even when it doesn’t, traders in the final second of the day may play a guessing game, trying to determine whether it’s worth submitting a displayed order on top of suspected hidden orders. (In our example, a trader might want priority over the hidden orders and submit a displayed bid at $10.06.)

I’m also not sure this functionality was an oversight. IEX says it exists “to protect the anonymity of resting non-displayed interest.” And “IEX Auctions were strategically designed after extensive research and informal discussion with various market participants.” But IEX did not find it appropriate to employ this functionality for auctions occuring after volatility pauses or halts, perhaps because those auctions can be chaotic and have a particular need for transparency. The thing is, some days the open and close can also be volatile, and those are the days when it’s most important for exchanges to run a smooth auction. [6]

To be clear, I wouldn’t say that this feature is scandalous. The behavior (I think) is fully disclosed, and professional traders should always read the manual before using any exchange. But issuers are probably not equipped to do that. And if IEX wants its auction to be “simple,” I’m not sure this is the right approach.

 


[1] It’s hard for me to completely understand IEX’s documentation. It has some clear errors in it — such as the “CLEARING PRICE EXAMPLES,” which say “Each example below assumes the Protected NBBO is $10.09 by $10.11 at the time of the Closing Auction,” but “Example 3” has displayed limit orders on IEX at $10.12 (buy) and $10.13 (sell) — so something is not quite right. Even aside from errors in the documentation, I may have made errors interpreting it.


[2] I don’t mean to write so much about IEX, but functionality like this has not gotten the attention it deserves.


[3] I think the auction would actually occur at $10.09 in this case, were it not for the rule:

If more than one price maximizes the number of shares that will execute, resulting in an auction price range, the Reference Price is set to the price at or within such range that is not lower (higher) than the most aggressive unexecuted buy (sell) order.

Which I think guarantees that even if the LOC bid at $10.10 is skipped-over and doesn’t receive an execution, it will still affect the auction price. I suppose this is better than performing the auction at $10.09, but it will still frustrate the LOC-sender who missed their execution and moved the closing price.


[4]


The Regular Market Continuous Book Contains the following orders:
 Midpoint Peg order to buy 2,500 shares with a resting price of $20.20.
The Closing Auction Book includes the following orders:
 LOC order to buy 500 shares with a limit price of $20.19; and
 LOC order to sell 2,000 shares with a limit price of 20.18.
For purposes of determining the clearing price, the Midpoint Peg order is priced to the Protected NBB ($20.19), but remains ranked and eligible to execute at its resting price;
Accordingly, shares are maximized between $20.18 and $20.19 resulting in an auction price range, and $20.19 is the only price within such range that is not below the price of the most aggressive unexecuted buy order; therefore
 2,000 shares would execute at the IEX Official Closing Price of $20.19;
 The Midpoint Peg buy order would receive an execution of 2,000 shares;
 The LOC sell order would receive an execution of 2,000 shares; and
 The LOC buy order would not receive an execution, because the LOC sell order is fully filled after matching with the Midpoint Peg buy order with superior priority.


[5] It also seems potentially risky to have the auction price depend other exchanges’ order books. If somebody submits a 100-share order on another exchange, they could potentially move the IEX open or close by multiple ticks. Even ignoring the possibility of manipulation, it might not be the best idea for an important cross to be sensitive to small orders that aren’t eligible to participate in that cross.

[6] NYSE learned this the hard way after Rule 48 probably contributed to the market disruption on Aug. 24, 2015.

[*] I have edited this example because I made an error in the original version. The LOC bid in the original version was at $10.10. In that circumstance, the LOC bid at $10.10 *would* get fully filled at $10.00 (because it’s more aggressively priced than the hidden orders), in contrast to what happens to the LOC bid at $10.00 in our example.

 

Liquidity Disincentives and IEX

Exchange fee-schedules are probably the most boring topic in market structure. But there are occasional exceptions that are almost interesting, like IEX’s new fee. In essence, the new fee will charge liquidity takers an extra 30 mils if the “Crumbling Quote Indicator” is on at the time IEX completes an execution. [1][2] Most exchanges segment order flow with incentives; IEX is attempting to carve out a niche by using disincentives.

With the fee, IEX seems to be deliberately targetting traders they view as harming their market. John McCrank relays the following quote from IEX’s Eric Stockland:

If we can stop HFT (high-frequency trading) market makers from getting picked off by predatory HFT strategies, the byproduct should be more liquidity and a better experience for folks interacting on IEX

And IEX’s filing makes clear that the fee targets “Only 13 Members… likely to be engaging in a deliberate strategy to target resting orders at soon to be stale prices.”

We have discussed the peculiarility of an exchange penalizing trades at prices that are “soon to be stale,” a phrase that borders on oxymoron. Volume and price-changes are inseparable in a healthy market: traders buy when they think the price is about to increase, and prices tend to increase after buyers initiate trades. These processes are natural features of markets, and ordinarily can be prevented only via transaction-blocking price controls. So I find IEX’s goal of “[p]reventing someone from anticipating an NBBO change” a very odd one.

Part of me suspects that what IEX really hoped to do was extend the last-look feature of its peg orders to displayed liquidity, but doing so might’ve been too problematic even for them. Brad Katsuyama told Matt Levine:

“People have asked us why not make D-peg a displayed order. That becomes very challenging… within the context of displayed liquidity — seeing something and having it fade on you — we’ve been very cognizant that’s what started this whole journey for us. We don’t want to contribute to that.”

So, instead of allowing displayed orders to fade “during periods of quote instability,” IEX is opting to charge takers the legal maximum.

Now, IEX certainly knows the identity of the 13 members that would pay this fee. Stockland says that the fee is targeted at “predatory” HFT strategies [3], and designed to help other HFT strategies. An exchange is free to make a good-faith judgment about what types of HFT they want to encourage or discourage, but IEX continues to use emotionally-charged language. I hope that their feelings were not a factor in their determination.

In any case, if the fee succeeds and market-makers quote more displayed liquidity on IEX, then the mix of traders paying the fee will likely change. I strongly suspect that the Thor router, as described in “Flash Boys,” would pay this fee on most of its eligible orders. If Thor swept Nasdaq’s book, it’s very likely that IEX would determine that the quote is “crumbling” by the time Thor’s order exited the IEX speedbump. Thor seems to be designed for click-traders, so it could delay sending its orders to Nasdaq and Bats by an additional 350us, preventing IEX from realizing the quote is about to “crumble” until after its execution. But many smart order routers and execution algos trade during periods of market activity, and cannot delay their orders without experiencing slippage. [4]

Displayed Liquidity and Listings

IEX desperately needs to increase displayed liquidity on their market. In principle they may not mind being a predominantly dark venue, but if they want to attract listings they will need to convince issuers that they can realistically facilitate price discovery. To be frank, the current level of pre-trade transparency on IEX is an embarrassment and any issuer choosing to list there is taking a serious risk. I suspect IEX knows this and the new fee schedule is an genuine attempt to incentivize displayed quoting.

That doesn’t mean it will work though. IEX believes that their refusal to pay rebates is the reason their exchange is mostly dark. [5] Another possible purpose of the new fee is to increase support for a reduced access fee cap. The fee and any copycats may annoy market participants that are currently indifferent to the access fee cap. Reducing the current cap from 30 mils would cut the rebates exchanges can afford to pay. The result, if IEX’s claim is true, would be to improve their displayed market share. I’m a bit skeptical: I suspect that IEX struggles with displayed liquidity because their last-look-functionality allows dark orders to preferentially trade with low-alpha aggressors, leaving abnormally high-alpha flow to hit their lit quotes.

Charging Traders For Ex-post Market Moves

The new fee depends on the state of the market after an order is sent. If the access fee cap were higher, this would enable pricing that is economically equivalent to last look. But even with the current cap, there are some weird possibilities. If this fee is allowed, would there be anything stopping an exchange from levying a fee that depends on the market a millisecond in the future? What about a second? Or a day? An exchange could, for instance, charge 30mils extra if a trade is followed by a favorable price movement larger than 0.1%. With the revenue generated, it could pay a huge rebate to resting orders that are blown-through (by say, 1%) during volatile moments, compensating market makers for their losses. This could make the exchange’s order book thicker during high-volume periods, resulting in market share gains. It’s easy to see how such pricing could backfire, with the exchange having to make larger-than-expected payments if a tail-event occurs on the thickened book. And there is something uncomfortable about paying traders for being wrong. But it’s an intriguing possibility that issuers might like.

An exchange could also try the opposite, paying high-alpha traders for correct predictions and penalizing noise-traders. Most likely, noise traders would move to other venues, leaving only orders that are sent by confident participants. This would add another tier to the venue segmentation spectrum. Under the current regime, low-alpha flow is disproportionately directed to internalizers and wholesalers, forcing high-alpha traders to interact with each other on lit exchanges, the centers of price discovery. Perhaps a few ultra-high-alpha traders would be attracted to this new type of venue of last resort, particularly when prices are in transition. I doubt it’d be wise for a major exchange to introduce this type of pseudo-sin-tax, but perhaps a smaller venue like PSX or EdgeA will try it out.

I’m usually skeptical of complex exchange logic, but it’ll take time to see the consequences of this type of pricing. IEX’s proposal shows that they are starting to think like their competitors do, and they’ve spun it in a way that no other exchange could. It may not be to the liking of long-term traders, but there’s no doubt that IEX is gradually becoming more sophisticated. [6]


[1] The filing says the fee will be applied when “Taking Liquidity During Periods of Quote Instability, as defined in IEX Rule 11.190(g)” if the volume of such executions exceeds “5% of the sum of a Member’s total monthly executions on IEX [and] at least 1,000,000 shares during the calendar month, measured on an MPID basis.”


[2] Matt Hurd is not a fan of the new pricing. For more details on the “Crumbling Quote Indicator,” see his analysis and update on the logistic regression IEX uses to assess whether a quote is “crumbling.” The regression uses non-delayed market data, which arrives at IEX 350us later than traders’ orders.


[3] IEX discloses that the largest such trader would have paid about $120k (footnote 17) under the new fee schedule for all of June. Assuming, generously, that these frowned-upon trades made 30mils, then their profit would be about $120k/month.


[4] IEX appears to think that anyone hit by the fee, presumably including these SORs, “degrade the qualtiy of the market.” And “ideally, the fee will be applied to no one, because participants will adjust their trading activity to account for the pricing change.” I could be wrong, but I suspect many brokers disagree.


[5] IEX executives have repeatedly referred to rebates as “kickbacks.” Matt Hurd argues that if a payment is a kickback, then so is providing a service for free:

IEX fail to recognise that zero pricing for lit at IEX would meet their own description of a kickback as a kickback is a remunerative, not monetary, exchange. Giving something away for free, such as an order execution, might also be considered a kickback; as may something for a discount. So if rebates are kickbacks, then IEX is also offering kickbacks. It becomes a matter of degree.

 


[6] It’s worth reflecting on “Flash Boys,” which opined:

Creating fairness was remarkably simple… They would pay no kickbacks to brokers or banks that sent orders; instead they’d charge both sides of a trade the same amount: nine one-hundredths of a cent per share (known as 9 “mils”). They’d allow just three order types: market, limit, and Mid-Point Peg

IEX has come a long way since then.

The Shape of Supply and Demand Curves in Rapidly Clearing Markets

A central challenge in economics is understanding how price affects the quantity of supply and demand, a relationship often assumed to be approximately linear. But there are markets where this notion of linearity, sometimes called “elasticity,” may not hold. In a paper that deserves more attention, Donier and Bouchaud show that supply/demand curves of rapidly clearing markets (with a Brownian price process) have an average shape that is locally quadratic, with no linear term.

Here is a classical illustration of how total supply and demand tend to vary with price:


Total demand decreases — and total supply increases — monotonically with price. The curves intersect at p*, which is the volume-maximizing, market price. [1] Near the market price, supply and demand vary linearly with changes in price.

If we “clear” the market, so that the supply on offer at p^* trades with the demand at p^*, then the curves will look like:


Which is the same as the first plot, just shifted downwards by the quantity traded. The curves are still linear near p*. And if we zoom in, the average supply/demand curves will look like [2]:


But, Donier and Bouchaud [3] show that markets with certain features will actually have *average* supply/demand curves that are locally quadratic. I.e.:


In this regime, a market order will have price impact that scales with the square-root of its size, on average.

Their result also raises the question whether some markets operate at saddle points in their production and consumption curves, which may look like:

Where W_{saddle} is the rough width of the “saddle zone,” the region where the curves are predominantly quadratic.


This strikes me as qualitatively different from the classical economic picture. It also makes sense intuitively: when an asset is volatile, it’s difficult to know the exact price where supply and demand balance. Donier and Bouchaud don’t speculate on the size of W_{saddle}, but there is at least the possibility that their results apply to a wider range of prices than expected. If any real markets have a large W_{saddle}, estimating their price elasticities would be difficult or impossible. It could also explain the ennui of financial markets — where headline-generating price moves have little effect on real-world supply and demand. [4]

The Donier-Bouchaud Model

Donier and Bouchaud (and previous co-authors) use a reaction-diffusion model to obtain this result. Briefly and roughly:

  1. New buy orders are created (in time interval dt) with probability \omega_+(y), where y = p - p^* is the difference between the price of the new order (p) and the market price (p^*). Sell orders are created with probability \omega_-(y).
  2. Existing orders are canceled with probability \nu_{\pm}(y).
  3. The market clears with periodic time-interval \tau — when buy and sell orders with crossing prices are matched, and removed from the market.
  4. The underlying process of the price p^* is Brownian.

When \tau is small, they show that supply and demand curves are locally quadratic, for any reasonable \omega_{\pm}(y) and \nu_{\pm}(y).

Of course, real-world markets do not instantly clear crossed meta-orders, even when markets trade in continuous time. For example: a trader might intend to buy $100M of stock at any price below $100/share, while another trader intends to sell $100M of stock at any price above $90/share. The two traders might dribble out their order-flow over weeks, instead of instantly trading with each other at a price between $90 and $100 per share. [5]

Nonetheless, it is conceivable that some markets behave as if they’re in the small \tau limit. In the Bitcoin market, traders may be less inclined to hide their intentions than in traditional markets, and the visible order book might represent the true levels of supply and demand near the market price. The authors present the average displayed supply and demand for Bitcoin in Figure 6, which is very close to a quadratic function for prices within 2% of the clearing price (where cumulative supply/demand is typically ~400k BTC). So electronic markets’ “saddle zones” may be about as wide as their daily volatility, which doesn’t seem surprising; few oil producers are going to increase drilling because the price went up by 1%.

Latent Liquidity as First-Passage Time

Donier and Bouchaud’s result seems to be a general feature of Brownian price processes, and doesn’t depend much on the specifics of the model. The spirit of their model raises the question whether there’s a connection between the marginal supply/demand at a given price, and the time required for the market to move through that price. That is, perhaps latent liquidity has properties similar to first-passage time statistics.

A quadratic supply/demand curve is equivalent to marginal supply/demand varying linearly with price. By definition, cumulative supply (S(y)) at a price y away from the “true” price is just the sum of marginal supply (\rho_{S}) available up to that price: S(y) = \int_{0}^{y} \rho_{S}(y') dy'. [6] \rho_{S}(y) may also be called the volume of latent sell orders available at price y.

One way to reach Donier and Bouchaud’s result is by assuming that \rho_{S}(p) builds with time, after the market price moves through p. [7] To be clear, the following model is different and less sophisticated than what Donier and Bouchaud did, but I think it’s a good way to capture the intuition.

As an illustration, consider what happens to the latent order book after the clearing price drops instantly from p_0 to p_1. At first, the supply on offer between p_0 and p_1 will be zero:



Afterwards, inside this new gap, latent sell orders will start to build. Let’s assume that \rho_{S} grows as a function only of the time t(p) since the price dropped through p: \rho_{S}(t(p)). [8] The precise form of the function \rho_{S}(t) doesn’t really matter.

If the market doesn’t move after that initial price drop, time t_1 later, the latent order book will have replenished. Between p_0 and p_1, there will be quantity \rho_{S}(t_1) on offer:

To calculate the expected value of \rho_{S}(t(y)), we need the probability density of the time since the price last passed y away from the present price: \mathbf{p}_{y}^{LPT}[t]. For a time-reversible process, this distribution is the same as that of the first-passage time, \mathbf{p}_{y}^{FPT}[t].

For continuous-time Brownian motion, the first-passage time distribution is well-known:

\mathbf{p}_{y}^{FPT}[t] = \frac{y}{\sqrt{2\pi \sigma^2 t^3}} e^{-y^2 / (2 \sigma^2 t)}

For y \ll \sigma \sqrt{t}, this is linear in y:

\mathbf{p}_{y}^{FPT}[t] \approx \frac{y}{\sqrt{2\pi \sigma^2 t^3}}

Which gives an average marginal supply curve that’s linear in y:

\mathbf{E}_{t}[\rho_{S}(t)] = \int_{0}^{T} \rho_{S}(t) \mathbf{p}_{y}^{FPT}[t] dt \propto y

Where T is the total time that the market has been operating.

Thus, the expected cumulative supply is quadratic in y: \mathbf{E}_{t}[S(y)] = \int_{0}^{y} \mathbf{E}_{t}[\rho_{S}(y')] dy' \propto y^2.

The model should fail at sufficiently large t, when it has been a long time since the price last reached its current level. E.g., if the price of oil rises past its high for the year to $60/bbl, then we’d expect the level of marginal supply near $60 to reflect the real-world economics of oil extraction. So clearly, the model shouldn’t work for t=1yr in the oil market. But if oil breached only its high-of-the-day, perhaps the marginal supply would just be a mechanical function of that duration. We could argue that the model will start to fail when t is long enough for businesses to react to new highs/lows, which should be about the typical time between business decisions (t_{InterDecision}). [9] In that case W_{saddle} \sim \sigma \sqrt{t_{InterDecision}}, which could be quite large for illiquid markets.

Discretizing the Process

If \tau > 0, the market clears every period \tau in a batch auction, and the price process becomes a discrete-time random walk. An infinitely long \tau should recover the uncleared, classical supply/demand curves at the top of this post. So, as \tau increases, we expect a transition from the quadratic supply of Brownian motion to a linear regime.

To get the average supply curve for discrete markets, we need the first-passage time distribution. When a random walk has price steps that are independently drawn from a symmetric, continuous probability distribution with finite second moment, its first-passage time PDF is asymptotically: [10]

\mathbf{p}_{y}^{FPT}[n] \approx  (\frac{1}{2 \sqrt{\pi n^3}} + \frac{y}{\sqrt{2\pi \sigma_{step}^{2} n^3}}) e^{-y^2 / (2 \sigma_{step}^2 n)}

Where n is the number of steps the random walk has taken, and \sigma_{step}^2 is the variance of each step’s price movement. The approximation is valid in the limit n \to \inf with \frac{y}{\sqrt{n}} finite. The step count is related to continuous time via n = \frac{t}{\tau}. And if the underlying process is Brownian, \sigma_{step} = \sqrt{\tau}\sigma.

When the price process has a typical step size (\sigma_{step}) that’s small compared to the distance from the market price (y), then the second term dominates and \mathbf{E}_{t}[\rho_{S}(t)] is identical to the continuous Brownian case. That is, cumulative supply varies quadratically with price when \tau \ll \frac{y^2}{\sigma^2}.

When the process is heavily discretized, y is small compared to \sigma_{step} and the first term dominates, which will be approximately constant in y. Thus the marginal supply will be constant, and the cumulative supply linear in y.

This result is the same as Donier and Bouchaud’s. In fact, if we expand \mathbf{p}_{y}^{FPT}[n] to first order in \frac{y}{\sigma_{step}\sqrt{n}} (near the market price), then we get:

\mathbf{E}_{n}[\rho_{S}(n)] \approx L(y + u_0 \sigma \sqrt{\tau})

Where L is some constant obtained from integrating out t, identified as a measure of liquidity by Donier and Bouchaud. And u_0=\sqrt{\frac{1}{2}} \approx 0.71 is a constant not terribly far from the one obtained by Donier and Bouchaud (u_0 \approx 0.82). [11][12]

Thus, slowly-clearing markets — which are heavily discretized — may not have a saddle zone. [13]

Supply/Demand Curves when the Price Process is a Lévy Flight

The above asymptotics apply to a broad class of random walks if their variance is finite. But markets can have price fluctuations with fatter tails, particularly on shorter timescales. A Lévy flight of index 0 < \alpha < 2 has price increments (x_{t} = p_{t} - p_{t-\Delta t}) with divergent variance and a power-law tail: \mathbf{p}[x] \sim \frac{1}{|x|^{\alpha + 1}}.

The first passage times of a Lévy flight have asymptotic PDF: [14]

\mathbf{p}_{y}^{FPT}[t] \sim \frac{y^{\alpha / 2}}{t^{3/2}} for long t. [15]

This distribution gives, on average, a cumulative supply curve S(y) \sim y^{(2+ \alpha) / 2}. And a market order will have price impact \mathcal{I}(q)=S^{-1}(q) \sim q^{2 / (2+\alpha)}. As an example, \alpha=1.5 would correspond to a rather “jumpy” market, and would have S(y) \sim y^{1.75} and \mathcal{I}(q) \sim q^{0.57}. [16]

Supply/Demand Curves for a Sub-Diffusive Price Process

The volatility of a sub-diffusion increases with timescale more slowly than the volatility of an ordinary Brownian motion. For example: \sigma_{\Delta t}^2 \sim \Delta t^{\gamma} . When \gamma = 1, the volatility scales in the usual way for Brownian motion: linearly with the timescale. When 0 < \gamma < 1, the process is a sub-diffusion. A sub-diffusive market is mean-reverting in the sense that a price fluctuation is likely to be reversed in the future. Because sub-diffusive markets have “memory,” they’re considered “inefficient.” [17]

The first passage time distribution of a sub-diffusion is asymptotically: [18]

\mathbf{p}_{y}^{FPT}[t] \sim \frac{y}{t^{1 + \gamma / 2}}

This has linear price-dependence like ordinary Brownian motion. So the cumulative supply is again quadratic, and market impact is again square-root.

Certain types of “efficiency” can lead to square-root price impact. But if this model is approximately accurate, then “inefficient” markets like sub-diffusions can also have square-root impact.

Update: Benzaquen and Bouchaud just examined a reaction-diffusion model for sub-diffusions. They show that the latent order book is locally linear (eq. 10), like in the crude first-passage analysis here. For quickly executed meta-orders, they show \mathcal{I}(q) \sim \sqrt{q}. But for slow meta-orders that give latent orders more time to react mid-execution, they get \mathcal{I}(q) \sim q^{1-\gamma / 2}.

Order-of-Magnitude Scaling of Impact in Diffusions and Sub-Diffusions

I find this result interesting because it appears, at first glance, to contradict the simplest, order-of-magnitude “derivation” of the square-root impact law. But on closer inspection, I think order-of-magnitude logic is consistent with sub-diffusions and ordinary diffusions having similar impact scaling.

If a market is Brownian, its price changes will scale like \sigma \sqrt{\Delta t}. One view of price discovery is that a fraction (of order 1) of those price changes come from traders’ impact. Thus, a meta-order’s impact will roughly scale with the square root of its duration: \mathcal{I} \sim \sigma \sqrt{t_{OrderDuration}}. Market-wide volume roughly accumulates at a constant rate in time (V), so a meta-order of size q will last for a duration t_{OrderDuration} \sim \frac{q}{V}. This gives \mathcal{I}(q) \sim \sigma \sqrt{\frac{q}{V}}.

Now, if a market is sub-diffusive then its price changes scale more slowly with time: \Delta p \sim \Delta t^{\gamma / 2} . If everything else in the above argument were the same, then we’d get \mathcal{I}(q) \sim q^{\gamma / 2} . But there’s no reason to expect that volume in a sub-diffusive market should occur at a constant rate in clock-time.

Sub-diffusions can be reformulated as processes with independent increments, where the time between steps is variable with a long tail. [19] In this formulation, steps may correspond to trading activity — with each step roughly having volume V_{step}. A meta-order would then last for N steps, with N \sim \frac{q}{V_{step}}. The step lengths of this sub-diffusion have finite variance \sigma_{step}^2, so we can follow the same argument as in the Brownian case, with the step count replacing time: \mathcal{I}(q) \sim \sigma_{step}\sqrt{N} \sim \sigma_{step}\sqrt{\frac{q}{V_{step}}} — which has the same square-root scaling.

Conclusion

Market-behavior is crowd-behavior, and a crowd can be more predictable than the individuals inside it. Donier and Bouchaud — and the loose interpretation in this post — show that that when markets are mechanical and boring, they will have cumulative supply and demand curves that are roughly quadratic, on average. Now, lots of us do not think markets are boring! Even if many markets do operate at saddle points, they probably do not have very wide “saddle zones.” But if any do, a small price change might have virtually zero effect on production and consumption, and a large price change might have a giant effect. That’s different from the standard intuition.


[1] This picture is obviously heuristic. Supply/demand curves are well known to have many shapes, may be non-monotonic, and may have more than one price that locally maximizes the exchanged volume.


[2] The supply and demand curves near p* should (probably) be symmetric around p*, if we average over some ensemble of situations in a given market.


[3] Borrowing from their earlier work with Bonart, Deremble, de Lataillade, Lempérière, Kockelkoren, Mastromatteo, and Tóth. (e.g. 2014 and 2011)


[4] Even when large price changes barely affect supply and demand, they could still serve an important economic function. The opinions and information of traders are incorporated into prices, which — when markets are transparent — provide useful signals for long-term business decisions.


[5] Table 1 and Figure 13 of Waelbroeck and Gomes indicate that the bulk of institutional equity meta-orders are shorter than a few days.


[6] Demand is similar. For ease, let’s just discuss the supply side.


[7] Latent liquidity isn’t usually observable, so we can’t easily test the hypothesis that latent liquidity at a given price builds as a function of time since the market was at that price. But latent liquidity could be measured via the order book — if a market is heavily financialized, transparent, and dominated by traders who don’t hide their intentions. Arguably, Bitcoin is (or used to be) such a market.


[8] We could add an i.i.d noise term to \rho_{S} and it wouldn’t affect the results. The assumption that \rho_{S} has no explicit price-dependence is similar to Donier and Bouchaud’s example of constant \omega and \nu. This assumption is obviously wrong, but sufficiently close to the market price, it may be reasonable enough.


[9] The reaction time of businesses could be significantly longer than the human decision-making timescale. Especially when the market in question isn’t transparent or mature. E.g., if a farmer switches from growing olives to almonds, it might take her grove several years to become productive again. So the price of almonds may need to reach a multi-year high in order for her to be confident enough to switch crops. But, if the farmer could hedge her future production, she might quickly decide to switch crops after it becomes economical to do so. Perhaps a well-developed futures market reduces the valid range of this model from t \lesssim 1yr to t \lesssim 1day.


[10] See eq. 38 of Majumdar and citations for details. I changed the CDF of eq. 38 into a PDF, and dropped a non-leading term.


[11] Eq. 22


[12] Sparre-Andersen scaling applies to any Markov process with a continuous and symmetric distribution of price movements: For large t, first passage time distributions have to decay like t^{-3 / 2}. So, in order for the expectation over t to converge, we must have \rho_{S}(t) < \sqrt{t} for large t (assuming \rho_{S}(t) is monotonically increasing). That is, the marginal supply at a given price must grow more slowly than the square root of the time since the market was at that price. The connection between latent liquidity and first passage times won’t hold for very long t, but this may provide a loose bound on how quickly latent liquidity can replenish.


[13] According to the model, slowly-clearing supply/demand curves are still dominated by the quadratic term when price movements are sufficiently large (and thus last-passage times long). But the model should fail at very long timescales, when businesses are able to react to price moves. Though, as discussed in [9], the model’s valid timescale could be longer if a market is slow and opaque.


[14] See, e.g., eq. 10 in Koren, et al.


[15] Figure 2 shows pretty rapid convergence to this asymptotic result after a handful of time steps.


[16] Perhaps it shouldn’t be surprising that impact is steeper than in the Brownian case. Heuristically, the Lévy tails make the market more “momentum-driven.” In a Lévy-type market, a trader who initiates a price movement could find that the market quickly moves away from her. The Lévy process used here has independent increments, but we can imagine the independence breaking down in continuous-time, “mid-increment.” I.e., it’s conceivable that momentum-traders could trade in the middle of a timestep.

On a related note, Koren et al. show that the mean leapover length diverges for Lévy flights. The leapover length could be interpreted as the profit made by stop orders, *if* they execute in the middle of a price jump and are able to hit liquidity during tail events. Those are big “ifs,” but the potential to make near-infinite profit may partly explain the popularity of stop orders and short-term momentum strategies. It could also explain why traders are reluctant to post much liquidity far away from the current price.


[17] In a sub-diffusive market with low transaction costs, betting on mean-reversion is a profitable strategy. If you know of any electronic markets like this, let me know.

Even if sub-diffusions are rare in electronic markets, linear combinations of different assets could still be sub-diffusive. The stereotypical “pairs trade” involves a mean-reverting spread between prices of two assets. Also, sub-diffusions may be more common in the broader, off-exchange economy than we’d naively expect.


[18] See the first non-constant term in the series expansion of eq. 30 in Metzler and Klafter (“Boundary value problems for fractional diffusion equations”). Eq. 30 is the survival probability (CDF of the first passage time), the PDF is the time derivative. The approximation is valid for y \ll \sqrt{ K_{\gamma}t^{\gamma} }. K_{\gamma} is the “fractional diffusion coefficient,” analogous to \sigma^2 for \gamma=1.

Note that sub-diffusions are not Markov processes, so Sparre-Andersen scaling [12] doesn’t apply. Liquidity must replenish more slowly than \rho_{S}(t) < t^{\gamma / 2} for large t.


[19] Correlated waiting times may also lead to sub-diffusions in a process with independent increments.

CHX and Four Types of Speed Bump


Speed bumps, the latest fad in market structure, are proliferating. The Chicago Stock Exchange (CHX) recently proposed a new type of speed bump in the US equities market, called the “Liquidity Taking Access Delay” (LTAD). [1] The SEC’s decision whether to approve the LTAD could have a big impact on market structure, possibly more so than the IEX decision.

To understand the consequences of approving LTAD, I think it’s helpful to consider four types of asymmetric access delay: [2][3]

  1. An exchange applies a delay to orders that are accessing resting liquidity. Resting orders themselves may be modified or cancelled without delay. This is the LTAD.
  2. The same as #1, except only resting orders that are non-displayed avoid the delay.
  3. The delay applies to all client messages to the exchange, including cancellations and marketable orders. But the exchange operates algorithmic order-types which adjust their attributes without delay. In particular, a displayed order can be pegged to undelayed price data, while its potential counterparties are subject to the delay.
  4. The same as #3, except only non-displayed algorithmic order-types avoid the delay. This is the IEX speed bump. [4]

Here’s a table of the speed bump combinations, split by which types of resting orders have functionality similar to last look:

Last look Can apply to displayed quotes Can apply to non-displayed quotes only
At discretion of resting order-sender #1 (LTAD) #2
At discretion of exchange algo #3 #4 (IEX)

In my opinion, all four types do more harm than good to market end-users (in the context of Reg NMS) — each allows resting liquidity to fade in some way. But many end-users disagree, and asked the SEC to approve #4. So, now that speed bumps exist and non-displayed peg orders may elide them, which of the other three asymmetric delays should be allowed?

One school of thought is that exchanges have a duty to protect their peg orders to the maximum extent possible. Combining this principle with the new “de minimis” interpretation may imply that speed bumps of types #3 and #4 should be allowed. Healthy Markets takes this view. [5] I think it’s a bad idea for exchange algorithms to have a time-advantage that isn’t offered to traders. Exchanges have neither the incentive nor the expertise to optimize their pricing algorithms. [6] The traditional role of the exchange is to provide a meeting place for buyers and sellers, allowing them to transact at prices of their own choosing. Exchanges offer algorithmic order-types (like pegs) mainly for convenience, and I don’t think there should be any illusion that these order-types are as good as the techniques used by sophisticated brokers and traders. Providing exchange algos with a time-advantage will mean that they out-compete traders, who are otherwise superior in terms of their pricing accuracy, stability under stress, and diversity. Traders and their algorithms do make mistakes, but it’s hard to believe that an exchange algo monoculture can do better. Subsidizing inferior business methods reduces an industry’s productivity, and giving exchange algorithms a structural advantage would be no different.

In my view then, if #4 is allowed, then #2 should be allowed. And, if #3 is allowed, then #1 should be allowed. Since #4 is already approved, the only thing to consider is whether it should be permissible for delays to have asymmetries in how they apply to accessing displayed quotes. The comments mostly do a good job explaining why asymmetries are problematic for displayed market structure. [7] In short, if displayed orders are given extra time to decide whether to consummate a trade, a large number of quotes will be practically inaccessible, even though it’d be prohibited to lock, cross, or trade-through those quotes. [8] Even in markets without these regulations, such as FX, last look causes difficulty. [9] The consequences of combining last look with order protection are unpredictable, but they seem unlikely to be good for long-term traders.

So, if we don’t want Reg NMS order protection to apply to quotes eligible for last look, only options #2 and #4 should be permitted. There is a tradeoff, though. Giving only hidden orders a time-advantage will incentivize dark trading. I doubt that the major exchanges would become as dark as IEX, but the equities market would probably become less transparent.

IEX’s approval is already inspiring a dramatic increase in complexity. CHX might not be the most important exchange, but like IEX, any precedent it sets applies to other exchanges. [10] Restricting speed-bump-asymmetries to hidden orders has drawbacks, but it might be the only way Reg NMS can keep limping on.


[1] Technically, it’s not really new. The LTAD is similar to an old proposal from Nasdaq’s PSX, which was rejected by the SEC.


[2] In order to highlight the essential elements, I’m leaving out some details.


[3] There are other speed bump types of course. For example, an exchange could delay traders from cancelling resting orders, but allow incoming marketable orders to execute without delay. That sort of delay could be used to address complaints about “fading liquidity“, and is similar in some respects to Nasdaq’s tentative “Extended Life Order” and EBS’s “Minimum Quote Life“.

Delays in market data are speed bumps of an entirely different class. These are prohibited (I think) in US equities markets, which require quotes and trades to be published to the SIP without delay. Though perhaps this requirement doesn’t apply to “de minimis” delays?


[4] Algorithmic order-types may also use their undelayed data feeds to trade aggressively with resting orders which can’t be cancelled without going through a delay. IEX’s DPEG does this via its “book recheck” mechanism.


[5] From their comment letter:

Time delays should not apply to an exchange’s ability to price orders on behalf of all participants (i.e. Pegging).

Dave Lauer clarified on Twitter that this principle also applies to displayed peg orders.


[6] CHX’s justification for the LTAD evokes a useful thought experiment. CHX argues that its ETF quotes are victims of “latency arbitrage,” and that the LTAD will prevent this. If the SEC rejects the LTAD, CHX might propose another type of speed bump, where CHX manages displayed orders in ETFs (e.g. SPY, QQQ, etc) by pegging them to undelayed CME market data. Will CHX understand the relationship that these ETFs have to futures markets as well as professional traders do? And these ETFs are just the simple cases — imagine what kind of pegs an exchange might come up with to prevent market makers from being “picked off” on XOM when the price of crude or CVX moves. I’m sure exchanges can think of peg algos that market makers would find very useful, but is that really what we want exchanges doing?


[7] No discussion of the LTAD would be complete without mentioning CHX’s market data revenue sharing program. This post, though, isn’t intended to be a complete discussion. Many of the comment letters (and Matt Hurd’s playful summaries) address this issue.

I found it interesting that Virtu publicly supported the LTAD as soon as it was announced, presumably before they had time to review the filings. Was the LTAD proposed at Virtu’s request? Could Virtu’s CHX profitability rely on market data revenue sharing?


[8] Without using ISOs, that is. If enough displayed orders were granted a last look, then the market would be clogged with inaccessible quotes. Only traders with the legal infrastructure to submit ISOs would be able to navigate the equity market.


[9] This opinion isn’t unanimous. Some long-term traders appreciate last look, like Norway’s SWF, NBIM.


[10] James Angel’s letter argues that CHX should be given the benefit of the doubt just like IEX was. I’m not a lawyer, and I think the LTAD could hurt long-term traders, but some of his points are persuasive.