Crypto vs. SEC Part 2

In June 2023, the SEC sued Coinbase alleging that some of the tokens listed on the Coinbase exchange were securities and by not following securities’ law, Coinbase was acting illegally. This series of posts proposes crypto asset taxonomies that make it easier to apply the four Howey Tests, on which rests the definition of securities. The previous post built a taxonomy for the investment of money criteria . The current post covers the common enterprise test.

Howey Test 2: Common Enterprise

There are three interpretations of the common enterprise criteria that have been applied by US courts:

Horizontal Commonality

Consider a group of individuals who have invested money to acquire crypto tokens:

Horizontal commonality can be tested by categorizing crypto tokens using the following flowchart:

Category A:

Bitcoin (BTC) is an example of this category. As discussed previously, the money spent to acquire new BTC is electrical power and hardware. No doubt, a publicly listed crypto mining company like Riot Blockchain is pooling funds of investors to build a BTC mining operation. But in that scenario, it is the shares of Riot Blockchain where horizontal commonality exists and not BTC.

Category B:

In this category, money is pooled but the economic value obtained is not derived from pooling the purchasers’ funds.

An example of this is BTC mining pools. Since the costs of BTC mining are high, individuals can pool together their computational power. The spoils are distributed among pool participants based upon a pre-agreed formula. We can compare two scenarios, one where funds are pooled to buy an apple orchard and the second is a mining pool:

In the first case, the amount of apples produced is strongly correlated with the amount of funds pooled (larger lot size, more fertile land, better fertilizers etc.). However, the number of BTCs awarded per 10 minutes is algorithmically determined, regardless of how much money (i.e. computational power) is pooled (though the probability of mining success may increase with relative advantages in computational power).

The equivalent of BTC mining pools on Ethereum are ETH staking pools. Instead of computational power, stakers have to deposit their existing holdings of ETH. Since the minimum staking size is 32 ETH, most people cannot afford to stake. A staking service allows people to pool their ETH. The total amount of ETH that can be earned within any slot is independent of the amount pooled (though the share of ETH earned is proportional to amount staked).

The SEC has alleged that Coinbase’s staking service are unregistered securities. Whether the courts will consider horizontal commonality to exist in Category B is unknown. The nuance here is that while money is pooled, the act of pooling does not impact the economic value generated.

Category C:

Ether (ETH) is an example of a token that falls within this category. While new bitcoins can only be acquired via mining, there are two possible sources of new ETH tokens: pre-mined ether and ongoing issuances.

Pre-mined ETH refers to the 72 MM ETH tokens created at inception in 2014. These tokens were partly granted to founders and the Ethereum Foundation, and partly sold to investors via a crowdsale. Investors paid in BTC and received ETH in return. The money raised was used to build the Ethereum platform. The ETH issued to the Ethereum Foundation has been sufficient to fund ongoing R&D because of ETH price appreciation.

The economic value of ETH today is at least partially derived from the pooling of investor funds received at the time of the crowdsale. However, the pooling occurred only once. Hence, two conflicting legal interpretations are possible:

1. The presence of horizontal commonality has a time frame attached to it.

2. Horizontal commonality is permanent because the impact of the original pooling is permanent

Category D

It is possible that purchaser’s funds are being pooled and the act of pooling is creating economic value, but the economic value so created is not shared among the pool participants.

Consider Yuga Labs, an entity that sells non fungible tokens (NFTs) the most popular of which is The Bored Ape Yacht Club (BAYC). BAYC is a collection of 10,000 ape cartoons (refer link at the NFT marketplace OpenSea). Each cartoon is a unique combination of parameters (like fur color, type of hat worn) that were randomly assigned using an algorithm. Some of these combinations are considered rare and have sold for millions of dollars. Each NFT carries a different price based on demand-supply of that specific variant. While it can be argued that Yuga Labs pools the money received from NFT purchasers to run its operations, the wide dispersion in the prices fetched by the NFTs it sells, indicates that the purchasers do not share the attendant risks and rewards.

A more piquant example of this comes from the decentralized exchange Uniswap. Uniswap allows users to become market makers by contributing their crypto assets to a pool against which traders can trade. In return, the these depositors, called liquidity providers (LPs) earn a share of the trading fees paid by traders. The LP can specify terms beyond which she will not trade. Since each LP’s choice is unique, the trading position is represented by an NFT, an example of which is shown below:

The chart frame above on the left panel is a visual representation of a trading position and is created using generative art. The visual parameters of the chart- like the steepness of the curve, the beginning and end points of the curve, the color distribution- are unique to each user’s position. Uniswap, thus pools the money (crypto assets) contributed by LPs into a Liquidity Pool but each LP’s financial outcome is unique.

While the Uniswap example is somewhat arcane, a couple of simpler thought experiments will help clarify the issue:

1. Consider a broker-dealer who offers clients the technology to build their own custom crypto portfolio. Each investor can choose a custom combination of tokens in custom weights. This would effectively be a tradeable NFT version of a separately managed account . Behind the scenes, the broker could be pooling funds to consolidate buy and sell orders on a specific token. But the resulting risks and rewards are not shared between investors.
2. Consider a product that is inverse of loan securitization. Instead of holding a share of a pool of mortgage loans, each investor holds an NFT corresponding to a unique mortgage. The issuer of the NFTs is pooling funds received from NFT buyers for running its operations. But the financial outcomes associated with each mortgage NFT (prepayment, delinquency etc.) is unique.

Category E

This category is relatively straightforward. Stablecoins like Tether and USDC, which are backed by low risk fiat money investments are common examples. The economic value of fiat backed stablecoins is derived from the investment of pooled funds into low risk assets like US Treasury Bills. Hence, all stablecoin holders share the risks and rewards of the underlying investments. Unlike Uniswap above, stablecoins like Tether and USDC do not grant people the ability to customize their fiat asset allocation.

Another example of this category are tokens issued by lending apps like Compound and Aave, which allow users to deposit their crypto tokens into a pool from which others can borrow. The liquidity providers (LPs) in Compound and Aave, are issued a fungible token to record their loan. The fortunes of the LPs are linked.

To conclude, crypto tokens can be categorized into 5 broad groups based on the nature of horizontal relationships between token holders. The risk of a finding of horizontal commonality increases (ceteris paribus) as we move from left to right on the continuum:

Broad Vertical Commonality

Vertical commonality is applied to the relationship between the promoters of a blockchain enterprise and token-holders. Under the broad vertical commonality test, the token holders must rely on the expertise of the promoters for the economic value of their tokens.

To apply these tests, it is useful to identify the governance models used by the blockchain enterprise under the scanner:

Type 1: Decentralized Model

In a decentralized model, no one entity is in-charge. BTC and ETH are examples of tokens managed in the decentralized model. In such decentralized models, there are usually four steps by which change occurs:

Step 1: Proposal

Any stakeholder can propose a change -examples include Bitcoin and Ethereum Improvement Proposals (BIPs and EIPs).

Step 2: Consensus

For changes that are non-controversial, this step is quick. For changes that can generate divergent views, the proponents of change build support through a mix of on-line (e.g., discussion boards) and off-line channels (e.g., conference calls). While votes can occur, the results of the poll are not binding on the community. In any case, the holdings of BTC and ETH are so well distributed that no single entity can influence the outcome.

Step 3: Code change

Changes are made to the copy of the code that is viewed as gold copy by the community (typically at a public repository like Git Hub).

Step 4: Code adoption

In a decentralized network, operators running the network’s nodes (computers) may choose to not deploy code changes. If such code changes are not backward compatible and majority of the nodes disagree to run the new code, the attempted change will fail.

In 2016, a $50MM hack occurred on the Ethereum blockchain. A majority of the Ethereum community agreed to undo the impact of the hack by forcibly altering the transaction history. However, many node operators disagreed with the change on principle and they continued running the version of Ethereum with the unaltered history. This unaltered version of Ethereum became its own blockchain, called Ethereum Classic. Today, Ethereum is a far more successful blockchain than Ethereum Classic. However, this was a market determined outcome rather than a top down diktat.

Broad vertical commonality would mean that investors are relying on the expertise of the promoters. The challenge is that the contributions of Satoshi Nakamoto and Vitalik Buterin as founders of Bitcoin and Ethereum respectively, were mainly at inception. Even if people knew who Satoshi Nakamoto was, he is not running Bitcoin. While Vitalik Buterin unquestionably exerts influence over the direction of the Ethereum platform, he does not run it with any formal control. Hence, the question that the courts would have to address is whether there is a statute of limitations on the existence of broad vertical commonality for such decentralized entities.

This is a hotly debated topic within the crypto ecosystem, whether a token can morph from being a security to not a security.

Type 2: Vote Based Models

In this model, governance occurs via a voting mechanism, with the voting power being derived from the size of token holdings. While token holders may financially benefit from the associated blockchain enterprise, they do not have a formal claim on earnings, like shareholders do. However, token holders are granted the right to propose and/or vote on how the enterprise functions. Sometimes such organizations that lack shareholders, label themselves as Decentralized Autonomous Organizations (DAO).

Blockchain enterprises that follow a vote based governance model, often issue two types of tokens (some enterprises may not decouple the governance rights and issue only one token, hence the below architecture is by no means universal):

A couple of examples of above:

1. The peer-to-peer lending app Compound issues the C Token to represent the lender’s financial interest in the loan made. Separate from the C Token is a governance token called Comp, which allows the token holder to vote on changes to the Compound platform.
2. The stablecoin issuer MakerDAO’s core token is the stablecoin DAI while its governance token is MKR.

Consider Compound. Robert Leshner and Geoffrey Hayes own voting rights through the governance token COMP. They also exert influence on VC partners who also hold the governance tokens. They are the principal architects of the platform, providing ongoing expertise. They wrote the original white paper in 2019. In August 2022, they launched the third iteration of Compound with a view to making it simpler and more usable. Holders of both the C Token (the core token) and the Comp token (the governance token) are relying, at least to some extent, on the expertise of the founders. Given this context, it is probable that broad vertical commonality exists.

However, it need not be the case that every entity that follows a vote based governance model will have broad vertical commonality. For instance, consider a DeFi app where the following apply:

1. Voting rights are so evenly distributed among investors that founders do not have much direct influence or indirect influence (through institutional investor partners like VCs).
2. Anyone can make fundamental changes to the app as long as they can get sufficient votes on their proposal.

The above scenario will be similar to the decentralized model except that consensus occurs via voting. Hence, we can think of these governance models as a continuum rather than simply three discrete points.

Type 3: Centralized Models

These are centralized entities that conform to a registered form of legal entity (e.g., LLCs, corporations etc.). An example of a centralized entity issuing tokens on the blockchain is the NFT issuer Yuga Labs, which is a Delaware corporation. For centralized entities, there is no crypto specific nuance to vertical commonality.

We can conclude by observing that as one moves right across the below continuum, the likelihood of meeting the Howey Tests on vertical commonality increases (ceteris paribus):

Finally, substance is more important than form. In 2022, Yuga Labs launched a cryptocurrency called ApeCoin, which could be used within the BAYC ecosystem (apps, NFTs, games). The currency was launched via a separate entity called ApeDAO, which Yuga Labs claimed was a separate entity not within its control. Where ApeDAO sits in the above continuum would be determined by the on-ground reality, rather than what a company says.

Strict vertical commonality

Strict vertical commonality requires that the promoters of a blockchain enterprise share the financial risk and rewards with the purchasers of the putative securities. This presupposes the existence of promoters and gets tricky when applied to fully decentralized models. Satoshi Nakamoto’s wallet address owns a lot of BTC (though he has not touched them!). And undoubtedly, Vitalik Buterin, would become richer if the value of ETH goes up. However, applying strict vertical commonality to Bitcoin and Ethereum blockchains suggests an ongoing vertical relationship between their founders and other token holders, which is not the case anymore.

Setting aside the issue of whether Satoshi and Vitalik are promoters, crypto enterprises can be categorized into two groups based on the nature of benefit of accruing to promoters (or more broadly, holders of governance rights):

The most well-known instances of the indirect benefit model are the Bitcoin and Ethereum blockchains. Like the Visa and Mastercard networks, their primary revenue source is transaction fees (after 2140 for Bitcoin). This revenue accrues to crowdsourced network administrators (miners, stakers) and not to promoters (even if we assumed Satoshi and Vitalik are promoters). While there is no direct benefit, there is a correlation between the number of transactions being processed and the price of the tokens.

A similar example is the decentralized exchange Uniswap, where users trade crypto tokens with each other. The main source of revenue is trading fees paid by the trade initiator. However, as of July 2023, the entire fees is paid to the counterparty of the trade and none to holders of its governance token, UNI. The Uniswap example is equivalent to Uber not taking a cut of the fees we pay for a ride! It is hard to imagine Uber’s shares trading at a meaningful valuation if this was so. Despite this, UNI trades with a non-zero value on exchanges. There is no direct revenue accrual to UNI tokens but they derive indirect value as Uniswap’s usage increases.

A contrast is the the governance token of the decentralized trading platform Curve Finance, in which governance token holders get 50% of the trading fees. This is an example of the Direct Benefit model.

Summing up

The following classification taxonomies can help apply the common enterprise criteria of the Howey Test:

1. Horizontal commonality: Categorize entities by whether money is pooled, the resultant economic value is derived from pooling and the risk-rewards are shared between the purchasers who pooled the funds.
2. Broad Vertical commonality: Categorize entities based on the degree of centralized governance with which they are run.
3. Strict Vertical commonality: Categorize entities based on whether promoters benefit indirectly or directly from the revenues earned by the entity.

The next post covers the expectation of profits criteria.