SEC vs. DeFi - Part II: TheDAO Rises

INTRO

Well, it’s been a hell of a week in cryptolaw: the SEC proclaimed to Unikrn, “ALL YOUR TOKENZ ARE BELONG TO US,” an SEC Commissioner scolded the SEC for doing that, Uniswap dropped $UNI, Coinbase insta-listed it and cryptolawers were #triggered, along with others far and wide, each raising thinly veiled accusations and defenses about whether UNI is or is not a security. Stephen Palley seized upon the moment to tweet the SEC a cri de coeur I can certainly relate to:

Why was UNI so triggering? I have no idea. Similar DeFi governance/fee tokens were being launched all summer, so UNI was probably just “the straw that broke the camel’s back.” And, after all, with the prominence of Uniswap in the world of DeFi, it’s a pretty big straw.

That being said, UNI has marginally better optics than some other governance token schemes: Distributing a huge chunk of UNI to verifiable, past, organic Uniswap users was a galaxy-brained move by Uniswap to give its true-blue users—the ones who didn’t have to be bribed to use it—a voice in protocol governance. The distribution method might seem a little arbitrary, but could be argued to be quite fair. (Of course, this assumes that insiders with advance knowledge of the September 1 cut-off didn’t use their edge to ninja-mine. According to some, we should treat the mere possibility of such ninja-mining as a certainty and have a higher standard of fairness. I can respect that level of paranoia, but I guess I’m a little more easily appeased).

Unlike user rewards, “liquidity mining” / “farming” distribution schemes incite inorganic demand. By offering farming, the system designers are basically bribing degens to start using, or increase their use of, the relevant system. Users rush to lock up their capital in order to receive the new (potentially very valuable) governance tokens “for free”. But it’s not really for free, because to get the tokens, users are giving the system and its creators something valuable in return: liquidity. (As an aside, that is (probably) an “investment of money” for Howey purposes).

Uniswap distributed UNI through both user rewards and liquidity mining. But if they had only done the user rewards, I think their legal position that UNI are not securities would be stronger, albeit still far from bullet-proof.

Anyway, this post is not (mainly) about UNI. I mention UNI only to say: the timing of this “Part II” of my SEC vs. DeFi series couldn’t be better! Because, as promised, I’m going to “steel-man” the SEC’s potential arguments that DeFi governance/fee tokens are securities.

Why am I doing this? Well, because I can. And because lawyers are no fun. This is why we can’t have nice things. I promise to say mean things about the un-enlightened SEC in Part III, and if you follow me on twitter than you’ve already heard my spiel.

FLASHBACK: THE SEC’S DAO REPORT

The SEC’s first major statement on blockchain tokens was its 21(a) report on TheDAO in the summer of 2017. People have short memories, but unlike much of what came from the SEC afterwards, the DAO report was not about ICO-style “utility tokens”—it was about governance / fee tokens. Thus, it has a whole lot of relevance to what people are doing in DeFi right now.

The SEC’s reasoning for why DAO tokens were securities was both simple and persuasive. The SEC argued, element-by-element, that the DAO tokens satisfied the Howey test. For completeness, the SEC also addressed the potential argument that the DAO could be a “general partnership”—which, if true, would mean that the DAO tokens were not securities. A “general partnership” was impossible among DAO participants because, although the DAO’s technology facilitated the trustless escrow of funds and digital voting, it had not solved the “collective action” and “free rider” problems so notorious in governance. The smaller DAO token holders were just too numerous, dispersed and ineffectual to exercise informed, effective power over the DAO—thus, the DAO was almost certain to be dominated by its creator (Slock.it), the “curators” appointed to whitelist proposals, and DAO token whales:

The holders of DAO Tokens stood to share in the anticipated earnings from these projects as a return on their investment in DAO Tokens. In addition, DAO Token holders could monetize their investments in DAO Tokens by re-selling DAO Tokens on a number of web-based platforms. . .Slock.it did create and maintain online forums on which investors could submit posts regarding contract proposals, which were not limited to use by DAO Token holders (anyone was permitted to post). However, DAO Token holders were pseudonymous, as were their posts to the forums. Those facts, combined with the sheer number of DAO Token holders, potentially made the forums of limited use if investors hoped to consolidate their votes into blocs powerful enough to assert actual control. . .The DAO Token holders’ pseudonymity and dispersion diluted their control over The DAO [and resulted in] an arrangement that bears little resemblance to that of a genuine general partnership . . .

It’s funny how times have changed. When the DAO report came out, we (lawyers working on blockchain things) all said— “Yeah, sure, of course the DAO tokens were securities—the DAO was basically a giant venture capital fund sitting on Ethereum! But, SEC, tell us something we don’t know! What about utility tokens?! Surely they are much less like securities! They present more ‘novel’ issues and and are harder to analyze, right?! Our clients paid us to say this. Kek.”

Now we basically have DeFi projects distributing new DAO tokens daily—tokens that entitle you to dividends and voting rights and thus are a whole lot like the prototypical securities, share of stock. And, I have to say—it’s really just way beyond weird to me how it suddenly seems so not obvious to everyone that such tokens are securities! And how utility tokens seem to everyone more obviously like securities! Four years ago, everyone thought the exact opposite: They would put things in their utility token memos like “because there are no voting rights or rights to receive dividends, these tokens [probably, subject to five pages of CYA qualifications] are not securities. That’ll be $100,000, please!” Now the memos say things like “everyone can vote on proposals and the money comes from a smart contract, so it’s [probably, subject to 20 pages of CYA qualifications] not a security, yay! That’ll be $200,000, please!”

So, WTF, right? Was there some radical change in securities law between 2017 and now I didn’t read about? No, there wasn’t. The law hasn’t changed (if anything, it’s only gotten worse)—everyone has just completely lost their minds. People really, really like money. And lightspeed financial hacking is dank af.

Thus, armed with ridiculously over-optimistic misinterpretations Hinman’s “sufficient decentralization” speech, these money-liking, innovation-loving DeFi partisans are milking fairly minor distinctions between new “decentralized governance” projects and TheDAO for every last drop of securities law arbitrage goodness they are worth—knowing that it’ll likely be many months before the SEC even starts catching up to them. By then, they’ll almost certainly be rich and famous enough or a16z-backed enough to litigate against the SEC for years, if need be, to defend themselves and get off with a light, EOS-style penalty.

SOME REMINDERS ON SECURITIES LAW CANONS OF CONSTRUCTION

Too often, lawyers summarizing U.S. federal securities laws for a casual audience just state the rules and tests, but fail to situate them in the broader hermeneutic* context to which they belong. (*Hermeneutics: the theory and methodology of interpretation). This context is crucial because the text of statutes and rules are often by themselves ambiguous—in many cases, on purpose—and thus one must know the applicable rules of interpretation to predict how a judge or regulator will apply them.

The securities laws are considered “remedial legislation” and thus should be interpreted broadly with the aim of applying them in situations where their purposes, such as investor protection, would be served.[1] Therefore, U.S. federal courts tend to adopt the principle that the label given to a contract, scheme or transactions does not determine whether it is a security. Instead, courts will assess all relevant circumstances “in terms of their substance (the economic realities of the transaction), rather than their form” and decide whether the securities laws apply based in large part on whether the policy concerns of the securities laws are implicated.[2] One of the major policy objectives of the securities laws is “compelling full and fair disclosure relative to the issuance of ‘the many types of instruments that in our commercial world fall within the ordinary concept of a security.’” (Howey). That doesn’t sound so bad or unreasonable, does it?

The U.S. securities laws are also supposed to be non-waivable.[3] This means that parties are not permitted to directly or indirectly agree by contract or otherwise that the securities laws do not apply when they otherwise would. Because of this critical feature, if a particular way of structuring an investment would, if the securities laws were not applied to it, create a huge loophole in the securities laws—effectively enabling parties to “opt-out” of the securities laws by adopting that structure—then the securities laws should apply to that structure; otherwise the “no-waiver” principle would be violated!

The biggest “intuition pump” I use when advising clients about the applicability of the securities laws is to reason by reductio ad absurdum—if the result of not applying the securities laws to a particular set of transactions would open a gaping loophole in the securities laws relative to their intended purposes and scope, then a reasonable judge, bearing in mind the policy goals of the securities laws and their nature as remedial legislation that is supposed to be construed broadly, should hold that the securities laws do apply to those transactions! Too many lawyers get bogged down in splitting hairs over the exact wording used in prior court opinions while missing this very obvious and, in my view, correct way of reasoning about securities laws. Is it circular? Superficially, yes. But the threshold question of whether the securities laws apply is a policy-driven one, not a narrow analytical inquiry. Failing to think in a policy-driven way about securities laws is failing to understand them and often results in mistaken conclusions.

Finally, it is worth noting the general trend of U.S. securities case law—which is that the scope of our securities laws has gotten broader, not narrower, over time, as a result of judicial interpretation. For example, the Howey court said that an investor must have an expectation of profits “solely from the efforts of others,” but subsequent case law expanded the test and counted situations where an investor relies “predominantly” on such efforts or where such efforts are “the undeniably significant ones.” There is nothing to stop a future judge from doing the same thing with regard to tokens, and I believe the Telegram case—which, under a more technical way of reasoning, Telegram should have easily won, but actually lost—provides a shining example. Judges are not supposed to blindly stick to past securities law tests—when they are examining new facts, they can modify those tests to apply the securities laws to the new facts if the relevant policy goals—investor and market protection against certain types of risks—are appropriately served by doing so.

WHY (LIKE IT OR NOT) ALMOST ALL GOVERNANCE TOKENS PROBABLY START AS SECURITIES

To credibly argue that nearly all governance/fee tokens are securities, all the SEC needs to do is dust off the DAO Report and make roughly the same arguments. The cosmetics of recent DeFi launches are different from TheDAO—the underlying substance is not.

Let’s just take a hypothetical DeFi governance token—GOV—and analyze it.

• GOV is connected with some type of DeFi smart contract system, Liquinator, in which you lock-up other tokens in some useful way and receive rewards for providing this liquidity.

• In addition to the normal rewards for locking up liquidity, for a limited time, liquidity providers will also receive GOV in proportion to how much liquidity they commit.

• GOV has convertible value initially because there is a Uniswap pair for it against ETH, and eventually because there are other pairs and it even gets listed on some centralized exchanges where it can be bought and sold for fiat.

• Liquinator and GOV were created by a company, Liquid Inc., with plenty of venture-capital funding and lots of talented employees. The VCs have both equity in Liquid Inc. and the right to a matching percentage of GOV.

• The founders and employees of Liquid Inc. all have GOV awards that vest based on continued service to Liquid Inc. over some period of years.

• GOV holders have the “right” (but not really) to vote on all future protocol upgrades, on the use of fees/taxes paid by users for using the protocol, and potentially other matters. Honestly, these governance rights/powers are quite open-ended, because there is no document or contract describing what rights or powers GOV token holders have or do not have and there is nothing binding the people—still employed by Liquid Inc.—with the most expertise and information to obey them or refrain from using their own GOVs to overrule and abuse the minority GOV holders. Rather, GOV is just called a “governance token” and some vague language is used about how “users now govern the protocol.” But no one actually knows for sure what the rules of this governance are or what it means to govern “a protocol,” particularly one that is free open-source software.

This is a generic but roughly typical set-up for most of the DeFi projects that have added governance tokens in recent months. It enables me to sorta-kinda talk about all these projects without really talking about any of them, which is what I want, because I’m not here to create legal problems for any specific project and also don’t want to provide them or any of their token holders with legal advice

Now, to start with, let’s just apply the Howey test to GOV:

1. Is there an “investment of money”? Yes! Because they will get GOV and GOV has value, many persons who would not commit liquidity to the system otherwise, or who would otherwise commit less liquidity, will now change their behavior in order to earn (more) GOV. Locking-up capital is a form of providing value (and incurring an opportunity and risk cost), and is thus an “investment of money” for the purpose of obtaining GOV. (See the DAO Report, quoting Uselton v. Comm. Lovelace Motor Freight, Inc., 940 F.2d 564, 574 (10th Cir. 1991): “[T]he ‘investment’ may take the form of ‘goods and services,’ or some other ‘exchange of value’.”).

   Furthermore, many people will also undoubtedly buy GOV on Uniswap or other cryptocurrency exchanges—these persons will even more clearly be making an “investment of money” in GOV. For more detail on this reasoning (including why even “completely free” distributions of GOV would represent an “investment of money”), see our prior issue, Debunking Securities Law Myths About Tokens.

2. Is there a “reasonable expectation of profits”? Again, yes! The smart contract system is popular and collects usage fees. These fees grow over time. The user base itself is likely to grow over time—in part thanks to the incentive of getting GOV, but also potentially due to other factors we’ll discuss below. All the governance tokens that have been launched so far have done so with the idea—hinted at or easily inferred—of eventually enabling the token holders to vote on whether or not all or a portion of those fees should go to governance token holders—and of course such token holders will vote yes! As explained in the DAO report, “‘[P]rofits’ include ‘dividends, other periodic payments, or the increased value of the investment.’” Just the prospect of receiving some portion of accumulated Liquinator fees alone—never mind various others we will delve into under “efforts of others”—is enough to justify a “reasonable expectation of profits” from acquiring GOV.

3. Is there a “common enterprise”? Of course there is! There is a common enterprise: (1) among all GOV holders (if GOV price/value increases, they all benefit—this is called “horizontal commonality”); and (2) between GOV holders and Liquid Inc. (because that corporation, its team and its investors all hold huge GOV bags, so when GOV price/value increases, they all benefit—moreover, since GOV is a tool for incentivizing use of their product, they similarly benefit from GOV value increase through that route—this is called “vertical commonality”).

4. Are the reasonably expected profits “from the efforts of others”? This one is trickier and where, I suspect there is the most room for reasonable debate. But, in my opinion, the answer is “yes”. The quickest (and admittedly possibly also the cheapest) way I can illustrate the underlying intuition is by showing the results of some twitter polls I conducted yesterday:

   

   If people would be “justifiably shocked and outraged” when a new & improved version of Liquinator is announced which uses a different governance/fee token than GOV, this demonstrates their “reasonable expectations”: They expect that future efforts of Liquid Inc. and its affiliates will aim to drive value to GOV.

   This expectation is also reasonable because: (1) Liquid Inc., its investors and team hold mammoth amounts of GOV; (2) as shown by the poll, people would be furious at the team (and possibly sue them) if they launched a new derivative of the protocol without GOV; and (3) the new version would essentially be a competitor to the old version, dramatically reducing the upside potential of GOV if it were not tied to the fee rewards accruing from the old version—a fact which people expect would lead the secondary market value of GOV to “plunge” and harm all GOV holders, including the Liquid Inc. team and investors with their huge bags of GOV. Using common sense, I think we can see why GOV holders expect future efforts from Liquid Inc. and its affiliates and why that expectation is perfectly reasonable.

   Even this, though, is understating the level of reasonable expectations GOV holders should have about the future efforts of the Liquid Inc. team. Liquid Inc. will undoubtedly do a lot more than just make a new version of the protocol. Its personnel will market the system, enter into commercial agreements to encourage enterprise adoption of it, come up with incentive schemes to grow the user base, maintain, update and pay for hosting for the “front-end” (fancy word for website) that is optimized for interacting with the smart contract system, etc.

   Taken together, these reasonably expected efforts—not the efforts of a random small GOV holder dropping a new proposal in a forum from time to time—are the “undeniably significant ones” that will tend to drive the biggest and most reliable and most sustainable upside swings to GOV. As the SEC stated in the DAO Report, “Even if an investor’s efforts help to make an enterprise profitable, those efforts do not necessarily equate with a promoter’s significant managerial efforts or control over the enterprise.” At an early stage, the probable impact of Liquid Inc.’s efforts on GOV’s value far outweighs that of all other investors combined.

   I could probably write 50 pages just on this “efforts of others” prong. For example, I would love to elaborate on why it is absurd to argue, as some seem inclined to, that merely because the GOV holders have to vote that they should receive fees from the protocol, instead of that being decided by Liquid Inc., that the GOV holders are providing their own “entrepreneurial efforts’” which defeat this prong of the Howey test. But, in this format, the quick/dirty method of capturing people’s intuitions about how they expect the development team to treat their token in the future is probably about the best I can do.

“SUFFICIENT DECENTRALIZATION” REDUX

A fair question to ask at this point would be— “But, Gabe, what if there is no new version and no reasonable reason to expect it? What if it’s a “set it and forget it” system that just happens to accrue fees? And what if there is no centralized marketing but just random GOV holders promoting it on Twitter and 4chan?”

That scenario is what I would describe as “sufficient decentralization.” It could occur if Liquid Inc. ran out of money and dissolved and the relevant team members no longer had service-based vesting to get additional GOV and started going off and doing new things. Some of the team might even keep working on Liquinator improvements, but on a much more modest, less well-capitalized, less formally coordinated scale—similar to how unpaid or low-paid volunteers contribute to other free open source software projects.

In such a “sufficiently decentralized” context, the SEC would have much weaker arguments that there is a reasonable expectation of profits from the “efforts of others” and would probably have little interest in arguing that GOV is a security. When things reach this stage, there just isn’t an un-performed “investment contract” affecting the value of GOV anymore: There is no person or group of affiliated persons like Liquid Inc., from whom investors should want the extensive disclosures mandated by the securities laws. This is the type of dynamic William Hinman meant to capture in “When Howey Met Gary (Plastic)” when he coined the term “sufficient decentralization.”

However, in Part III, I will surface a somewhat obscure line of cases relating to the sale of fractional interest in life insurance contracts—viaticals—which the SEC staff is studying in connection with DeFi. These cases acknowledge that pre-launch “entrepreneurial efforts of others” may count as the relevant “efforts of others” under Howey. Using such cases, the SEC staff could argue—though I do not know if they ultimately will argue—that even “set it and forget” smart contract systems having a fee-distribution token are covered by the securities laws.

AREN’T GOVERNANCE TOKENS GENERAL PARTNERSHIP INTERESTS AND THUS NOT SECURITIES?

An interest in a “general partnership” is presumed not to be a security. A DAO or smart contract system governed by token voters has many of the superficial properties of a partnership. Like a general partnership, the unincorporated association of GOV holders lacks executive officers or a board of directors, and is supposed to decide important matters by holders personally voting their interests rather than delegating decision-making to fiduciaries. Does this mean that the Liquinator system is run by a general partnership, the voting and economic interests in which (GOV) are not securities?

No, GOV tokens are not interests in a general partnership and GOV token holders are not general partners. General partnerships are small (generally 100 person or less) organizations in which the partners have equal access to information, a roughly level governance playing field and the ability to easily coordinate with one another. The situation with GOV could not be more different.

The situation with GOV holders is very similar to that of DAO token holders as described in the DAO Report:

The voting rights afforded DAO Token holders did not provide them with meaningful control over the enterprise, because (1) DAO Token holders’ ability to vote for contracts was a largely perfunctory one; and (2) DAO Token holders were widely dispersed and limited in their ability to communicate with one another.

These facts diminished the ability of DAO Token holders to exercise meaningful control over the enterprise through the voting process, rendering the voting rights of DAO Token holders akin to those of a corporate shareholder. Steinhardt Group, Inc. v. Citicorp., 126 F.3d 144, 152 (3d Cir. 1997) (“It must be emphasized that the assignment of nominal or limited responsibilities to the participant does not negate the existence of an investment contract; where the duties assigned are so narrowly circumscribed as to involve little real choice of action … a security may be found to exist … . [The] emphasis must be placed on economic reality.”) (citing SEC v. Koscot Interplanetary, Inc., 497 F.2d 473, 483 n. 14 (5th Cir. 1974)

In comparison to ordinary GOV holders, Liquid Inc. and its investors and employees have enormous information, competence and resource advantages, combined with a huge stack of GOV tokens. These persons are also extrinsically affiliated with each other through equity holdings in Liquid and contracts they all have signed—more on that topic in a future edition—which enables them to more easily coordinate with one another than the typical GOV holders can coordinate with other GOV holders. These factors provide the makings of what in Delaware corporate law would be referred to as a “controlling shareholder” or “control group.”

Furthermore, there are no rules of the road that explain to GOV holders exactly what rights they have and how to exercise them. I presume that Liquid Inc. owns the Liquinator trademarks and branding, and that it still has some moral rights and copyright in the Liquinator code, albeit, in the latter case, subject to an open source license. If GOV holders are really in control, shouldn’t Liquid Inc. have signed a contract that says Liquid Inc. will apply the Liquinator trademarks and branding to whatever the collective of GOV holders decides is “the real Liquinator”? Shouldn’t it also have signed agreements not to do anything affecting Liquiator which the GOV holders have not approved? But Liquid Inc. and its team have signed no such contracts, and thus it is very unclear what it is that the GOV holders even have governance power over. It is more like they are making non-binding, advisory, symbolic votes on various matters—not like they are full partners who govern the project.

The power bloc formed by Liquid Inc., its employees and its investors is thus precisely the type of power bloc from whom the smaller investors should want the formal requisite disclosures mandated by the securities laws, so that these smaller holders can better protect their investment and monitor the performance of the people they are ultimately trusting to do a good job driving value to the token. As an aside, they should also want something like common-law fiduciary duties from such a control bloc—and I say this as someone who has resisted applying fiduciary duties to developers in other contexts.

Over time, this power bloc could become much weaker and then the governance of the system would look more like a partnership. Then and only then, the securities laws should fall away—that is why the SEC staff talks about “sufficient decentralization”.

There are trends in this direction with some of the new systems—for example, many people are delegating their UNI to Andre Cronje. Over time, assuming he remains unaffiliated with the corporation which launched Uniswap, he and others like him could become independent voices which make Uniswap governance more decentralized and reduce the importance of receiving securities law disclosures from that corporation.

UPDATE: Here is a (rather depressing) post-publication update on the Andre Cronje delegations, which exposes an additional centralizing factor I did not even know about:

But to think this point has been reached at a very early stage—when there is still service-based vesting of UNI for that team, a huge percentage of UNI are owned or controlled by the Uniswap corporation and its affiliates, and the only major pool of funding for additional protocol development is from that corporation’s venture capital—would be premature at best, folly at worst.

Governance tokens per se do not solve the longstanding information asymmetry, collective action, free-rider and conflict of interest problems endemic to governance. Until better software mechanisms are developed for those governance issues, addressing them at scale for the benefit of powerless minorities still requires legal mechanisms of the kind provided by the securities laws and other laws.

POST-SCRIPT & NEXT ISSUES

Well, the Substack editor is now screaming that my post is too long for gmail recipients, and, I must admit, I’m getting slightly bored of this topic anyhow. So, I’ll stop here for now.

Sadly, I have failed to cover about half the ground I intended to in this edition of the cryptolaw newsletter, so that’s going to have to spill over to Part III. There, I’ll finally talk about the viaticals cases I’ve been hinting at the whole time—together with my favorite DeFi securities law hypo, “the owner-less water fountain sitting in the desert.” Are the robots which collect money from this water fountain and bring them back to the robots’ respective owners securities? I don’t think so! The SEC does—maybe—at least if these robots are anything like fractional interests in life insurance contracts!

After that, I’m very eager to move on to non-securities-law topics. To be able to do so, I’ll have to save the “what should the SEC do” stuff I had originally planned for Part III for a future edition. Of course, you can get a pretty good sense of my ideas on the topic here and here anyhow—though my thinking has evolved quite a bit since then and there are now some interesting new angles to explore.

CERTAIN SOURCES

[1]Tcherepnin v. Knight, 389 U.S. 332, 336 (1967): “[W]e are guided by the familiar canon of statutory construction that remedial legislation should be construed broadly to effectuate its purposes. The Securities Exchange Act quite clearly falls into the category of remedial legislation. One of its central purposes is to protect investors through the requirement of full disclosure by issuers of securities, and the definition of security in § 3(a)(10) necessarily determines the classes of investments and investors which will receive the Act's protections.”

[2]United Housing Foundation, Inc. v. Forman, 421 U.S. 837 (1975). See also SEC Paragon Coin Order (https://www.sec.gov/litigation/admin/2018/33-10574.pdf): “In analyzing whether something is a security, ‘form should be disregarded for substance,’ Tcherepnin v. Knight, 389 U.S. 332, 336 (1967), ‘and the emphasis should be on economic realities underlying a transaction, and not on the name appended thereto.’ Forman, 421 U.S. at 849.

[3] See e.g. Elaine A. Welle, Freedom of Contract and the Securities Laws: Opting Out of Securities Regulation by Private Agreement, 56 Wash. & Lee L. Rev. 519 (1999)