Source views: Source JSON(ish) on GitHub (VSCode)   Doc views: Document (&k=r00t): Visual Print Technical: OpenParameters Xray
it may be necessary for a user to convert Bitcoin that it has mined into a real currency or another convertible virtual currency, either because the seller of the goods or services the user wishes to purchase will not accept Bitcoin, or because the user wishes to diversify currency holdings in anticipation of future needs or for the user’s own investment purposes;[59]> in undertaking such a conversion transaction, the user is not acting as an exchanger, notwithstanding the fact that the user is accepting a real currency or another convertible virtual currency and transmitting Bitcoin.[60]>
# The SAFT Project: Toward a Compliant Token Sale Framework =
### October 2, 2017 =
Juan Batiz-Benet =
Jesse Clayburgh =
Marco Santori =
**Abstract:** Blockchain protocol tokens, or simply “tokens” are digital assets used in connection with decentralized services, applications, and communities (collectively, “token networks”). As of this writing, dozens of such networks are in use world-wide, with many more in development. Bitcoin and Ethereum are the most notable examples. Token networks may bring about positive paradigm shifts to computing, finance, law, government, and more. Tokens leverage computation and cryptography to represent consumptive goods (known as “utility tokens”) or replacements for traditional investments (known as “securities tokens”). =
The public token sale, colloquially known as an “Initial Coin Offering,” is a powerful new tool for creating decentralized communities, kickstarting network effects, incentivizing participants, providing faster liquidity to investors, and forming capital for creators. In these sales, network creators sell an amount of the network’s tokens at a discount to users, investors, or both. Some token sales take place when or after the token network is launched, as a means to disseminate some fraction of the token supply to early users. Other token sales happen long before the token network has genuine functionality; so-called “direct token pre-sales” are sold at greater discounts with the goal of financing the development of the network and its launch. =
Purchasers in these direct presales tend to expect profit predominantly from the seller’s efforts to create functionality in the token. As such, these sellers may unintentionally be selling securities, and may have failed to comply with several U.S. laws. =
We propose a path toward a new, compliant framework called the Simple Agreement for Future Tokens, or “SAFT”. Together with the publication of this paper, we launch the SAFT Project—a forum for discussion and development of the SAFT framework. =
The SAFT is an investment contract. A SAFT transaction contemplates an initial sale of a SAFT by developers to accredited investors. The SAFT obligates investors to immediately fund the developers. In exchange, the developers use the funds to develop genuinely functional network, with genuinely functional utility tokens, and then deliver those tokens to the investors once functional. The investors may then resell the tokens to the public, presumably for a profit, and so may the developers. =
The SAFT is a security. It demands compliance with the securities laws. The resulting tokens, however, are already functional, and need not be securities under the Howey test. They are consumptive products and, as such, demand compliance with state and federal consumer protection laws. =
To be sure, public purchasers may still be profit-motivated when they buy a post-SAFT utility token. Unlike a pre-functional token, though, whose market value is determined predominantly by the efforts of the sellers in imbuing the tokens with functionality, a genuinely functional token's value is determined by a variety of market factors, the aggregate impact of which likely predominates the “efforts of others.” Sellers of already-functional tokens have likely already expended the “essential” managerial efforts that might otherwise satisfy the Howey test. =
Beyond the securities laws, the SAFT framework elegantly navigates the money services and tax laws, and addresses the significant policy concerns with the direct token presale alternative. =
A copy of a draft SAFT is attached to this whitepaper as Exhibit 1. The SAFT Project website is available at [www.saftproject.com](saftproject.com). =
---- =
## Table of Contents =
- Introduction to Token Sales =
- U.S. Legal Concerns with the Direct Token Presale Model =
- Policy Concerns with the Direct Token Presale Model =
- The SAFT =
- Policy Benefits of the SAFT =
- Limits of the SAFT Framework =
- The SAFT as a Project =
---- =
## Introduction to Token Sales =
We believe, without overstatement, that token-based networks hold the potential to create value on an order of magnitude similar to that of the internet. The token sale plays a critical role in unlocking the value inherent in these networks. Sellers, investors, and users alike could stand to gain tremendously from this innovation if it is broadly adopted.[1] For developers, a token sale could jump-start network effects at the earliest and most vital stages of a software project’s development. For investors, tokens mean earlier liquidity and broader investment cases. Users likely stand to benefit the most inasmuch as they may participate directly in the creation and growth of the value of a network. Critically: If executed as set forth in this whitepaper, a token sale can permit users to participate financially in that creation and growth without taking on significant enterprise risk. =
Because tokens transform open networks into open markets, all participants can benefit from a liquid, transparent, middleman-optional secondary market for anything that can be digitized. On-chain transactions make systemic market risks more transparent to analysts and regulators. Add to that near-instant settlement times, immutable transaction records, network resiliency, and all the other benefits of underlying blockchain architecture. =
Though this whitepaper focuses on U.S. law, token-based networks are not a U.S. phenomenon, and they are certainly not a Silicon Valley phenomenon. One of the effects of the rise of the token network is that a Silicon Valley presence is no longer required, or even beneficial, to raising investment capital or generating a dedicated community. Some of the most successful token sales, by amount raised, have been orchestrated by decentralized teams with members around the globe. Because we analyze U.S. law, we focus on the effects of token sales on the U.S., but the policy issues we address are borderless. We hope that participants in the SAFT Project (proposed in further detail in the final section of this whitepaper) can work to expand the legal and policy discussion beyond the U.S. perspective. =
There are two categories of tokens in circulation today that are relevant to this discussion. The first kind of token is meant to serve simply as a highly-liquid substitute for a traditional security like corporate stock, a limited partnership interest, and the like. These so-called “securities tokens” can offer obvious benefits over the traditional securities infrastructure. For example, they can be traded, cleared, and settled nearly instantaneously, leaving an indelible chain of title.[2] Today, though, these benefits are overshadowed by regulatory and policy uncertainty—especially when used in the public markets.[3] This whitepaper does not discuss securities tokens in detail. =
Instead, we focus on so-called “utility tokens.” This category of blockchain tokens contains assets that do not purport to replace legacy financial services products. They are designed to offer intrinsic utility that powers a decentralized, distributed network that delivers to the users of the network a consumptive good or service. For example, when their networks are functional, some tokens will act as currencies, like bitcoins do. Some will act as staking or betting mechanisms, membership rights, or loan collateral. Some will simply act as cryptographic “coupons” redeemable for mundane goods and services like bags of ground coffee or boxes of razor blades. =
With a particular emphasis on legal and policy effects, we describe the often-used “direct token presale” model, which involves the direct sale of a utility token to the public prior to the functionality of the token, and compare it to the SAFT framework, which only produces functional tokens. Our analysis of the direct token presale is critical to understanding why the SAFT is a preferable alternative. =
### The Direct Token Presale =
In a typical[4] direct token presale[5], a software development team forms a legal entity, typically a corporation or a limited liability company if under U.S. law. Prior to the creation of any functional token, the entity accepts money from public, retail purchasers in exchange for a right to the token. The developers are, to greater or lesser extents, skilled in applied cryptography, computer programming, or both. =
The network itself may be a new decentralized ledger with its own native token or it could be comprised of a smart contract built “on top of” an existing decentralized ledger, like ERC20 tokens programmed with Solidity[6] and built “on top of”[7] the Ethereum ledger. The code that gives rise to the network is usually, but not always, open source. In the case of ERC20 tokens, the smart contract component of this code must necessarily be open source because of the open, transparent nature of the blockchain on which it must run. =
Prior to the direct token presale, the entity has usually produced a technical whitepaper to postulate the feasibility and value of the unbuilt network. It has released the whitepaper for peer review, which occurs primarily via email, online or offline meetings, social media networks, or private online chat rooms. It has also generally engaged in extensive public relations campaigns to market the direct token presale. These marketing efforts include the creation and dissemination of websites, banner advertisements, press releases, and collaboration with media outlets. Thus, a great many of the purchasers in a token sale are unaccredited, retail investors. =
However, entities offering token sales that raise large amounts of capital generally do not rely solely (or even mostly) on these broad solicitations. They typically spend months pitching and negotiating with professional investors to seed these sales.[8] Very often, these professional investors meet the SEC’s accredited investor definition requirements.[9] By some informal accounts, funds from accredited investors make up between 60%-80% of the total funds raised in a direct token presale. =
Prior to the direct token presale, some portion of the tokens is “allocated” to the developers’ entity and some to the developers individually as a type of compensation for the developers’ efforts. Some networks allocate a portion of the tokens to be “mined” as a reward for users who run the network software and maintain the network’s operation. The entity then sells rights to some portion of its tokens to the public, collecting in exchange anywhere from a few hundred thousand dollars to hundreds of millions of dollars.[10] =
In this way, the direct token presale is a sale of a utility token before it can function on the network or offer its intended utility. Some developers simply record each sale and promise to deliver a token when the network is built and functional. Some developers, immediately upon the execution of the sale, create and issue a pre-functional token – one that cannot yet function as intended on the network beyond the limited ability to be issued and traded on a secondary market exchange. Once the tokens exist and are tradeable, the purchasers and the entity are free to trade the tokens on any exchange that might list them. =
The entity then builds the network. It uses the proceeds of the token pre-sales to fund development, including developers’ salaries, office rent, utility bills, and the like. Though the entity is rarely obligated to its purchasers to build the network, almost all have embarked upon the task nonetheless. As of the date of this writing, it is too early to tell whether these projects trend toward completion or abandonment. That said, the great majority of them seem, outwardly, to be working diligently toward completion. =
One of the critical characteristics of the direct token presale model, for purposes of our later comparison to the SAFT Project, is that, in the typical direct token presale, tokens are sold widely to the public at a time when purchasers are still reliant on the developers to build a functional network. In a SAFT, no pre-functional tokens are ever created or sold, let alone released to the public and traded on exchanges. =
### The Nonprofit Twist =
Some developers form a nonprofit foundation based in a non-U.S. jurisdiction as part of the token network development process. Developers opting for this approach do so for a variety of reasons. Many represent real benefits to the project. For example: neutral governance of a decentralized network; mediation between stakeholders; or investment in long-term research endeavors that might inure to the benefit of the network as a whole. This approach has significant appeal, since the nonprofit structure can establish the founders’ responsibility to the community in a legally binding way. In some cases, foundations can even obligate themselves to periodic audits which ensure the foundation’s resources are allocated according to the nonprofit’s mandate. Some of these foundations accept genuine donations from donors who do not expect any pecuniary return in exchange. =
Some developers, though, use this nonprofit twist as a vehicle for direct token presales. Money accepted by the nonprofit entity in exchange for the tokens is characterized as a “donation” to the foundation, even though the “donors” expect to receive tokens in return. The commercial documents evidencing the transaction of course disclaim any obligation of the foundation to actually deliver any tokens to the donor. Instead, they represent that the network’s community will decide whether to execute the smart contract that would deliver the tokens. Nonetheless, the economic reality of the arrangement is very often a purchase and a sale. =
The benefits of this structure for direct token presales, by and large, are merely perceived benefits. For example, developers may believe that they can avoid U.S. securities, money services and tax laws by executing the direct token presale from an offshore entity. They may believe that they can avoid U.S. securities laws by characterizing the direct token presale as a “donation” campaign. They may believe that the local government is “friendly” to cryptocurrency, and thus offers an opportunity to conduct regulatory arbitrage from both U.S. and other countries’ securities laws. They may believe the sale will offer certain tax advantages under local law. =
A comprehensive comparison of the various international and nonprofit models to the SAFT framework is outside the scope of this whitepaper. In sum, though, based on our independent research and analysis, we believe that the SAFT offers a superior alternative to these perceived benefits. =
## U.S. Legal Concerns with the Direct Token Presale Model =
The direct token presale implicates a number of U.S. state and federal laws. Sadly, in many cases, direct token presales do more than merely implicate them—they can violate them. We limit our discussion to the most pressing of these laws from a risk exposure perspective: the federal laws governing securities, money services, and taxation.[11] =
### The Federal Securities Laws =
It is illegal to offer or sell securities in the United States unless the offer and sale are exempt under the federal securities laws or made pursuant to an effective registration statement filed with the SEC.[12] One kind of a security under federal law is an “investment contract.”[13] =
We believe that the “investment contract” analysis is the best fit for analyzing most direct token presales under the federal securities laws.[14] Though the law is broad enough to capture any “contract, transaction or scheme,”[15] the utility token by itself, once issued, imbued with genuine functionality and circulating on its network, rarely possesses qualities that would satisfy the requirements for an investment contract. However, in many cases, the contracts used to sell these so-called “utility tokens” (even when couched in terms of a “donation”) can be securities despite their promised utility. As we set forth in the following sections, direct token presale agreements executed prior to network launch are especially vulnerable to being deemed a security. =
#### The Howey Test =
Courts since the 1930s have generated significant analysis of what is meant by the term “investment contract.” The Supreme Court in its 1946 decision in SEC v. W.J. Howey Co.,[16] provided the seminal definition of that term. An investment contract was a “contract, transaction or scheme whereby a person invests his money in a common enterprise and is led to expect profits solely from the efforts of the promoter or a third party.”[17] Many courts in the succeeding seventy-one years have further expounded on each of the constituent parts of this test, now known as the Howey test. =
Courts often break the Howey test into four prongs to determine (i) whether there exists an investment of money, (ii) whether there exists a common enterprise, (iii) whether there exists an expectation of profits, and (iv) whether the expectation of profits is solely from the efforts of others. If all prongs are satisfied, then a contract, scheme, or arrangement passes the Howey test and constitutes a security. If any one of the prongs is not met, the arrangement fails the Howey test and there is no security.[18] A comprehensive analysis of the application of the federal securities laws to any particular direct token presale is outside the scope of this whitepaper.[19] However, we do note some of the more significant direct token presale trends in each of the following sections, organized by prong. =
###### Investment of Money =
A direct token presale typically requires participant purchasers to deliver a government currency or another digital asset such as bitcoin or ether in order to receive tokens in exchange. Although the original Howey test clearly references “money,” more recent decisions have held that an investment of blockchain tokens such as bitcoin,[20] or even an investment of labor,[21] can satisfy this prong. Indeed, even donations can satisfy this prong.[22] As a result, this prong of Howey is often met by direct token presales. =
###### Common Enterprise =
Direct token presales often admit of a common enterprise. Courts are split on what is the correct threshold for finding the existence of a common enterprise. The majority of courts apply the so-called horizontal commonality test.[23] Under this approach, a common enterprise exists where multiple investors pool assets and share together in the profits and risks of the enterprise.[24] A minority of courts instead apply the vertical commonality test. There are two variations on the vertical commonality formulation. Under the narrow vertical commonality variation, a common enterprise exists where the fortunes of the investors are bound up with the actual fortunes of the promoter or issuer of the security.[25] Under the broad vertical commonality variation, a common enterprise exists where the fortunes of the investors are bound up with the mere efforts of the promoter or issuer.[26] =
Some direct token presales satisfy one or more of the variations on this prong, and some do not. Take horizontal commonality, for example. The developers have laid the groundwork for meeting this prong when (i) tokens are fungible, (ii) the entity pools all the money raised from selling the tokens, and (iii) the entity uses the pooled funds to build the network.[27] Vertical commonality is rarer. To be sure, token purchasers might rely on the efforts of the developers to create the network, but that fact might support the “efforts of others” prong of the Howey test, not the broad variation of the vertical commonality prong, in which the fortunes of the investors must be bound up with the efforts of the issuer. Likewise, narrow vertical commonality is rare, since the purchasers’ profit from the token sale is rarely dependent upon the ultimate profitability of the developers or their entity. The value of a truly decentralized network is decoupled from the financial success of the original developers. Moreover, the mission of many developers’ entities is to expend all of its resources to develop an open, permission-less network that acts as a public good, slowly and expectedly entering insolvency as it does so. =
###### Expectation of Profits =
An expectation of profit generally means expected capital appreciation resulting from the development of the initial investment or expected participation in earnings resulting from the use of investor funds.[28] There can be such an expectation where expected profits “come primarily from the discount below the current price” of a commodity.[29] There is no expectation of profit where a purchaser is motivated primarily by the desire to use or consume the item purchased.[30] =
How much expectation of profit is permissible before the arrangement satisfies this prong? In United Housing Foundation v. Forman,[31] a purchaser of shares in cooperative housing almost certainly expected to sell the shares for more than the purchase price. So, without a doubt, that a profit motive is present is insufficient. Still, Forman can teach us more: It stands to reason that the purchaser likely would not have purchased the shares at all if he expected to lose money or merely break even upon resale. After all, what purchaser would buy a home knowing that it would be underwater when he decided to sell? Even if profit was a necessary outcome of the transaction for a prospective purchaser, it would be insufficient to satisfy this prong of Howey. To satisfy this prong, the purchaser’s expectation of profit must predominate the expectation of using the thing purchased.[32] =
Direct token presales often satisfy this prong. In most direct token presales today, purchasers seem predominantly motivated by a desire to profit from the transaction. Most seem to hope to sell the token at a higher price than they paid. Some commentators have ignored this trend, or justified it by arguing that the developers’ promise of future functionality is a substitute for a genuinely consumptive motive on the part of the purchaser. We take up this issue in depth in the following sections comparing pre-functional token sales to sales of already-functional tokens. =
###### From the Efforts of Others =
Assuming the purchaser’s expectation of profit predominates any consumptive motive, this prong considers the source of that expectation. It asks “whether the efforts made by those other than the investor are the undeniably significant ones, those essential managerial efforts which affect the failure or success of the enterprise.”[33] An analysis of this prong in the token sale context requires a deep understanding of the rights, powers (and, sometimes, obligations) attendant to the token in question. Sadly, this prong and the prior prong are often conflated or collapsed into one another. We take up this issue in the following sections, through the lens of token functionality. =
#### Already-functional Utility Tokens Are Unlikely to Pass the Howey Test =
Sellers of already-functional utility tokens have very strong arguments against characterization as a security: Such tokens rarely satisfy both the “expectation of profits” and “from the efforts of others” prongs of the Howey test. =
Generally speaking, there are two categories of purchasers of already-functional utility tokens. First, there are purchasers who buy tokens to actually use them—as network fees, membership coupons, value-staking mechanisms, currencies, etc. Their consumptive desires predominate their profit-seeking motives. Second, there are purchasers who buy a token expecting to profit merely from resale on a secondary market. These purchasers might have an “expectation of profit” under Howey, but it is usually not predominantly “from the efforts of others.” Sales to purchasers in both of these two categories fail the Howey test. =
The first category is self-evident. A profit motive simply does not predominate the transaction. The second category is more controversial. Critics of sales in this category might argue that the expectation of profit from resale on a secondary market is just speculative activity seeking capital appreciation. These critics might cite myriad federal court decisions holding that an expectation of mere “capital appreciation” on a secondary market, is sufficient to satisfy the Howey test.[34] =
This oft-repeated criticism does not stand up to scrutiny. At heart, the criticism collapses the “efforts of others” prong into the “expectation of profit” prong. It does so by relying on decisions which do not actually turn on the secondary market appreciation issue, and do not analyze it in much depth. Decisions that do so repeatedly hold that an expectation of profit from the mere increase in value on a secondary market is not from the “efforts of others.” In Noa v. Key Futures, for example, a case involving a forward contract for silver bars, the Ninth Circuit found no expectation of profits from the efforts of others because once the purchase of silver bars was made, the profits to the investor depended primarily upon the fluctuations of the silver market, not the managerial efforts of Key Futures.[35] SEC v. Belmont Reid, a case involving a forward contract for gold coins, held similarly because profits to the coin buyer depended primarily upon the fluctuations of the gold market, not the managerial efforts of others.[36] In another case involving a futures contract for sugar, a federal court in New York held the presence of a speculative motive on the part of the purchaser or seller did not, on its own, evidence the existence of an investment contract.[37] =
To be sure: Gold, silver, and sugar are different from tokens in important ways. For example, in the case of a token sale, the seller may continue to improve the network and the secondary market price of the token may appreciate as a result. This characteristic is not shared by precious metals or sugar. So should utility tokens really be treated similarly? =
For already-functional utility tokens, we think so. Because there is no central authority to exert “monetary policy,” the secondary market price of a decentralized token system is driven exclusively by supply and demand. Supply and demand can be due to a variety of factors. One of those factors could be the efforts of the development team creating the token’s functionality; but once that functionality is created, any “essential” efforts have by definition already been applied. It would be difficult to argue that any improvement on an already-functional token is an “essential” managerial effort.  =
Furthermore, the market effect of a mere improvement on an already functional utility token is likely dwarfed by the multitude of other factors that act on it. For example, the value of a token that powers a decentralized market for buying and selling graphics processing power in real time would likely fluctuate depending upon, among many other factors, the retail or wholesale availability of high-powered professional graphics cards.[38] The value of a token that entitles a token holder to one box of the seller’s razor blades might fluctuate with the popularity of beards in the company’s target markets.[39] The value of a token that permits users to store encrypted passwords conveniently on a blockchain might increase when a high-profile data breach is announced or decrease when a major keylogging botnet is disabled.[40] Indeed, the value of bitcoin, another (mundane example of an) already-functional utility token, often fluctuates with changes in global geopolitical instability. The forces that could affect supply and demand for a functional utility token are countless. Supply and demand for functional tokens are affected by a variety of forces that determine the price on a secondary market—just like demand for gold in the commodity cases. It is no coincidence that blockchain tokens have been referred to as “digital gold.”[41] For functional utility tokens, mere price appreciation on a secondary market is not a substitute for the essential managerial efforts of others. =
Indeed, for the purposes of the Howey analysis, the secondary market itself is a red herring. When a token purchaser resells a token on an exchange platform for more than the purchase price, it is not the exchange platform that created the price difference. To the contrary, the market is merely the venue where the token purchaser executes the sale. The relevant inquiry is not: “Did the purchaser expect profit from the sale on a secondary market?” It is: Of all the myriad forces affecting the secondary market price, do the “efforts of others” predominate or do the aggregated myriad forces predominate? In the case of the graphics-processing token, the predominant force one day could be the supply of rare metals required for fabricating graphics chips or the announcement of the next installment of a popular video game franchise. The next day, it could be a workers’ strike at a major chip fabricator in Taiwan or the invention of a new compression algorithm. For an encrypted-password-storage token, it could very well be a data breach one day, a new law expanding (or limiting) data privacy the next day, or the announcement of a new hacking tool on the following day. For the razor blade token, it can be a trend among celebrities to don beards or the invention of an at-home laser hair removal process. These are the same kind of market forces and mechanics contemplated by the gold, silver, and sugar cases. The potential forces affecting any particular functional token are countless. It would be difficult to establish that, against this backdrop of global supply and demand dynamics, the developers’ efforts still stand out as the “undeniably significant” effect on the price after the developers have expended their “essential” efforts and delivered a functional product. =
Thus, an already-functional utility token is less likely to be a security for two independent grounds. First, it is more likely that purchasers have bought them to use them (since, unlike pre-functional utility tokens, they can be used immediately to satisfy imminent needs). Second, purchasers who buy them with an eye toward profit upon resale can expect those profits to be determined by a variety of market factors that predominate the efforts of the seller in updating the token’s functionality. =
There are, of course, limits to this position. Not all already-functional utility tokens will fail the Howey test simply by virtue of being functional. A later section of this whitepaper sets forth some of the limits on this position and, therefore, on the SAFT framework itself. Still, **for an already-functional utility token, a great variety of forces can predominate the effects of the “efforts of others” on the purchaser’s “expectation of profit.”** In the case of a pre-functional utility token, though, there is often just one force that clearly predominates. We take this up in the next section. =
#### Pre-functional Utility Token Sales Are More Likely to Pass the Howey Test =
Potential plaintiffs (whether private or governmental) seeking to assert that a token is a security, might argue that the “efforts of others” prong is satisfied whenever the developers sell utility tokens prior to the network launch and prior to genuine utility (aside from mere tradability). They might take the position that the purchasers of these pre-functional utility tokens in such a scenario necessarily rely on the managerial and technical efforts of the developers to realize value from their tokens. =
For many pre-functional utility tokens, there is merit to this position. As stated in the prior section, the secondary market price of an already-functional utility token is determined by a great variety of factors. Likewise, the secondary market price of a pre-functional utility token could also be determined by a great variety of factors. However, the application of the technical and managerial efforts of the seller is likely the predominant factor in the price of a pre-functional utility token until it transitions to being a functional utility token. The purchasers of a pre-functional utility token are, by and large, reliant on the efforts of the seller to develop functionality. These sellers have not yet expended their “essential” efforts. Those efforts are still required to deliver functionality, and therefore profit. =
For example, at the pre-functional stage, the price of an encrypted-password-storage token is probably more dependent upon the seller’s creation of the token’s fundamental underlying network than it is on a transitory event like a data breach. After all, the data breach could be done and forgotten by the time the password-storage token could actually be used to prevent it. Likewise, the waxing and waning production cycles of graphics-card hardware would be unlikely to be the predominant effect on demand for a token that could not be traded for graphics processing power anytime soon. If nobody is demanding the token to put it toward its intended purpose, what does it matter if the alternative product is more or less available this week? Indeed, these factors are much less important to the “expectation of profit” than the efforts of the seller. At least prior to the token receiving its intended functionality, the price of the token is quite likely to be predominated by one thing: the efforts of its seller in developing the token’s essential functionality. As such, **if the purchaser has an “expectation of profit” from the purchase of a pre-functional token, that expectation is very likely to be predominantly from the “efforts of others.”** =
The chief legal concern with direct token presales under the securities laws is that many of them, for the above reasons, might pass the Howey test for an investment contract. The SAFT framework, as set forth in more detail below, never results in a pre-functional utility token. =
#### The SEC’s Position =
The SEC has not yet publicly spoken on utility tokens. However, on July 25, 2017, the SEC issued an Investigative Report on the application of the federal securities laws to the Decentralized Autonomous Organization (DAO) and the blockchain token associated with it (DAO Token).[42] It marked the SEC’s first official pronouncement on token sales. As such, it bears inclusion in this whitepaper. =
The DAO was a smart contract on the Ethereum blockchain that operated like a venture fund. According to the SEC, the DAO organization sold DAO Tokens to purchasers in exchange for ether. The ether was pooled, and then DAO Token purchasers would vote on a menu of investments to which the DAO would apply portions of the pooled funds. The purchasers would then share in profits from the investments pro rata according to their DAO Token holdings. The DAO Tokens operated like limited partnership interests. They were, using the terminology of this whitepaper, securities tokens. =
In its report, the SEC took several positions consistent with the arguments in this whitepaper, including: =
- The Howey test for an investment contract applies to blockchain token investments.[43] =
- Whether a token is a security is based on the facts and circumstances of each token.[44] =
- Securities laws reach sellers in foreign jurisdictions, especially when the token is offered or sold to U.S. persons.[45] =
- Exchanges listing securities tokens are subject to the securities laws and must, among other things, register with the SEC.[46] =
The DAO Token was not a utility token, and the SEC did not speak on utility tokens specifically. Since utility tokens now make up a significant proportion of aggregate token offerings, we await and expect further guidance from the SEC on the issues addressed herein. =
### The Federal Money Services Laws =
Federal law makes it a crime for anyone to knowingly conduct, control, manage, supervise, direct, or own all or part of a money transmitting business which is not licensed under state and federal law, referred to as an “unlicensed money transmitting business.”[47] We believe that some direct token presales may be characterized as unlicensed money transmitting businesses. =
We focus our analysis on federal regulation promulgated by the Financial Crimes Enforcement Network (“FinCEN”).[48] FinCEN, a bureau within the U.S. Department of the Treasury, oversees registration of money transmitting businesses. Through its regulations, FinCEN requires registration of a “money services business,” a term that is defined to include “money transmitters.”[49] A “money transmitter” can be either: (i) a person that provides money transmission services, or (ii) any other person engaged in the transfer of funds.[50] Money transmitter status is a matter of facts and circumstances, and explicitly excludes several categories of persons, including, in relevant part, anyone that only provides the medium for money transfers (such as a communications network),[51] and anyone that accepts and transmits funds solely to facilitate the purchase of goods or services.[52] FinCEN has explicitly stated that this exclusion applies to persons brokering the sale of commodity contracts or similar instruments.[53] =
FinCEN has published guidance to clarify whether a person dealing in cryptocurrency (which it terms “convertible virtual currency” or “CVC”), would fall under the definition of a money transmitter. In its guidance, FinCEN has stated that users of CVC are not money transmitters, but those who both issue and redeem CVC (administrators) and those who exchange CVC for either fiat or other CVC (exchangers) who accept and transmit funds as a business would be deemed money transmitters.[54] =
In further rulings, FinCEN has taken the position that a user is not a money transmitter if it is buying and selling CVC to make investments on its own account.[55] Specifically, FinCEN has stated that non-money transmission activities in the context of CVC include “(in the case of a corporate user) making distributions to shareholders.”[56] In addition, FinCEN has explicitly stated that investments for a company’s own account, even CVC investments, do not constitute money transmission activity.[57] =
Of particular importance to tokens, FinCEN originally took the position that “a person that creates units of convertible virtual currency and sells those units to another person for real currency or its equivalent is engaged in transmission to another location and is a money transmitter.”[58] In a subsequent ruling, however, presumably presented with additional data about the mining industry, it then added an important caveat to this position. It stated that “how a user obtains a virtual currency may be described using any number of other terms, such as ...’mining,’ ‘creating,’ …[or] ‘manufacturing’”; that: =
> it may be necessary for a user to convert Bitcoin that it has mined into a real currency or another convertible virtual currency, either because the seller of the goods or services the user wishes to purchase will not accept Bitcoin, or because the user wishes to diversify currency holdings in anticipation of future needs or for the user’s own investment purposes;[59] =
and finally, that: =
> in undertaking such a conversion transaction, the user is not acting as an exchanger, notwithstanding the fact that the user is accepting a real currency or another convertible virtual currency and transmitting Bitcoin.[60] =
Thus, FinCEN now arguably takes the straightforward position that a person can create a CVC and then sell it on its own account without being a money transmitter. The potential application of this reasoning to tokens is obvious. Yet, FinCEN has consistently maintained that “[a]n administrator or exchanger that (1) accepts and transmits a convertible virtual currency, or (2) buys or sells convertible virtual currency for any reason is a money transmitter.”[61] Thus, there is a risk that direct token presales violate the money transmission laws.[62] We analyze this issue in more depth in our discussion of the SAFT. =
### The Federal Tax Laws =
Tokens, whether CVC or other kinds of blockchain tokens, are generally treated as “property” for U.S. federal income tax purposes.[63] Consequently, proceeds from a token sale (whether pursuant to a SAFT or a direct presale) are taxable to the entity[64] selling the tokens.  The entity will generally recognize gain in the sale equal to the difference between the proceeds raised from the sale and the tax basis the seller has in the tokens. The tax basis of the seller is, of course, usually zero for a newly-formed entity engaging in a direct token presale. Other taxes, such as sales taxes and the alternative minimum tax may also apply. =
As a result, entities that are organized as corporations formed in the United States selling tokens will generally bear a heavy tax burden on the proceeds from the token sale. Typically such sellers should expect a combined federal and state tax rate between 35% and 50%. This income can be offset with operating losses (if any) incurred in the year of the token sale (or prior to the year of sale, to the extent carried forward). In addition, the seller of tokens may be able to carry back to the year of sale operating losses incurred in the two years subsequent to the token sale to offset the income incurred in the year of sale. Because sellers in direct token presales realize all the income from a token sale in a single tax year (often in a single day), they often will not be able to generate sufficient losses in the two subsequent tax years to fully offset the income from the token sale. =
Many prospective sellers of tokens, when confronted with the probability of a heavy U.S. tax burden resulting from a token sale, have considered incorporating “offshore” (e.g., setting up an entity in the Cayman Islands, Switzerland, Estonia, Singapore, or some other low-tax jurisdiction) in order to avoid U.S. taxes on the token sale. The rules regarding offshore sales of tokens are beyond the scope of this paper; however, simply offering tokens through an offshore affiliate will, in most cases, not reduce (and may increase) the U.S. tax incurred in connection with a token sale. As such: Direct token presales, from a tax perspective, generally do not allow the seller to effectively manage its tax liability. The SAFT, as set forth in more detail below, offers greater opportunities for tax management under the current tax regime as we understand it. =
## Policy Concerns with the Direct Token Presale Model =
The law contains within it a distinction between investor protection and consumer protection. Consumer protection laws are tailored to protect against risks to purchasers of goods and services. Investor protection laws are tailored to protect against risks to investors in securities. Both kinds of laws control for the risk of loss, but are calibrated differently to (i) suit the quality and degree of the risks inherent in a particular business arrangement, as well as (ii) minimize the collateral damage that needlessly burdensome regulations can cause to innovation. =
Regardless of whether any particular direct token presale violates the letter of these laws, direct token presales, as a model for early-stage finance, may violate the spirit of these laws.[65] Two examples are often discussed. First, in a direct token presale, unaccredited investors purchase pre-functional tokens. Thus, they take on enterprise risk—the risk that the project fails without producing a functional network. Second, and at the same time, they are asked to purchase based on caveat emptor—buyer beware—and so they have little recourse if the project fails without producing a functional token. We believe that for token sales to become a viable market standard, their transactional models must mature to the point where they satisfy both black letter law and the policy goals that drive the law. =
It has not been clear, however, how the laws should apply to achieve these goals. Consumer protection laws do a good job of protecting consumers, but are inadequate to control for enterprise risk. They control for product and service risks.[66] Similarly, investor protection laws do a good job of protecting against enterprise risk, but when misapplied, may prevent new and innovative products from coming to market. =
Applying mere consumer protection laws (such as Federal Trade Commission regulations) throughout the token sale process is probably insufficient. For good reason, we rarely ask product purchasers to risk managerial failure, technical mistakes or capitalization inadequacies when they buy a new widget. Yet, direct token presale sellers ask their purchasers to do this. Consumer protection laws don’t offer the right remedies for all token sale losses. =
Applying investor protection laws (like accreditation rules, numerical investor limits, and registration requirements) throughout the token sale process is just as troublesome. The investor protection laws, when improperly calibrated, can prevent capital formation, stifle innovation by preventing new financial models from accessing market resources, and shut out the economically disenfranchised.[67] Indeed, investor protection laws, in addition to all the good that they do, can come at a significant cost. These unquantifiable costs are only amplified in high-growth areas like blockchain technologies. Worse yet, the harms investor protection laws might mitigate are greatly attenuated at the early stages of this growth because those who stand to lose are primarily enthusiasts who understand the risks of harm better than anyone. This will, of course, change with time, but this is a particularly salient issue for token sales today. =
So what policy goals should predominate in token sales? Should the laws we apply to tokens protect against consumer risks or investor risks? We believe the answer is: both, but each at the appropriate time. =
## The SAFT =
The SAFT is a framework which seeks to navigate the federal securities and money-transmitter laws, provide greater flexibility for tax management. It further seeks to apply both investor protections and consumer protections, but timed to minimize their negative impacts and maximize their positive effects. The SAFT framework works for tokens which are not themselves independently securities. That is to say, it works for utility tokens, not securities tokens. The SAFT would have little or no beneficial effect for a DAO Token-like arrangement, for example. =
The SAFT is based on the Y Combinator Simple Agreement for Future Equity, or “SAFE,” which has been widely used to finance early stage companies for many years.[68] Both the SAFE and the SAFT memorialize an exchange of investment capital in an early stage or developing company for the right to something of value in the future for the investor—preferred stock in the case of the SAFE, and functional utility tokens in the case of a SAFT. =
In short, the SAFT provides investors with the right to fully-functional utility tokens, delivered once the network is created and the tokens are functional. The SAFT is very likely a security, namely an investment contract. Once the tokens have been imbued with utility and are genuinely functional, the SAFT investors’ rights in the SAFT automatically convert into a right to delivery of the tokens. For the now-functional utility tokens, there is a very strong argument that the tokens themselves are not securities. The same should apply to any ultimate sale of the tokens to retail purchasers, whether by the SAFT investors or by the seller. =
Still, a SAFT is not a SAFE, and a SAFT is not “safe.” Indeed, no court, regulator, or taxing authority has yet interpreted the SAFT framework, nor can the SAFT claim the transactional history and ubiquity of the SAFE. =
### The SAFT Transaction =
Here is an example SAFT transaction. =
**Step 1:** Developers publish their whitepaper, incorporate a Delaware corporation called Developers Inc., and secure commitments from accredited investors. =
**Step 2:** Developers enter into a SAFT with the accredited investors relying on the exemption set forth in Rule 506(c)[69] of Regulation D of the Securities Act, and the accredited investors transfer funds in the amount of $15 million to Developers Inc. The SAFT offers investors a discount on the final token sale and is a security, so the developers file a Form D with the SEC disclosing the sale.[70] =
**Step 3:** Developers Inc. uses the proceeds to develop the network into a product that provides genuine utility to its users. =
**Step 4:** Developers Inc. launches the network and delivers the tokens to the investors. The investors (and potentially Developers Inc.) begin sales of the token to the public, either directly or through exchanges. =
We take up each step in more detail below. =
#### Step 1 in depth =
At the outset, the developers in our example have little more than a whitepaper in hand. The network is not built and the whitepaper is not yet peer-reviewed. They take the time between Step 1 and Step 2 to offer the whitepaper for peer review and perhaps publish a bit of code. Critically, they spend this time pitching experienced, accredited investors. They might also engage in extensive public discourse, defending the whitepaper and even advertising an upcoming token sale. As a result of this marketing, advertising and public discourse, the SAFT transaction might rely on Rule 506(c) of the Securities Act which allows for general solicitation of investors, but requires that the offering must be limited, in the end, only to verified accredited investors.[71]  They incorporate using a simple Delaware corporation. Though not necessary, they might even take advantage of the new blockchain amendments to the Delaware General Corporation Law that would allow the entity to itself incorporate directly on a blockchain.[72] They do not seek donations for a foundation to distribute the tokens, nor do they try to avoid U.S. securities, tax, or money services laws. =
#### Step 2 in depth =
By Step 2, the developers have secured $15 million in commitments from U.S. accredited investors. The total amount of the commitments is arbitrary, but subject to important tax considerations set forth in Step 4. After confirming through affirmative representations (or verification, as required) by investors that they are accredited, Developers Inc. enters into a SAFT with the accredited investors. Developers Inc. also files a Form D with the SEC. The SAFT is an investment contract whereby investors purchase the right to receive tokens in the subsequent network launch. It exchanges Developers Inc.’s promise to deliver tokens upon the launch of the network for the investors’ promise to immediately pay. The SAFT is, at heart, a forward contract, but for tokens.[73] =
#### Step 3 in depth =
The duration of Step 3 will for most developers be a long period of time. In this step, Developers Inc. uses the proceeds of the SAFT sale in Step 2 to develop the network. The company pays for rent, utilities, software engineering, and any additional items required to launch the planned network. The characteristics of the network will dictate the use to which Developers Inc. puts the investors’ funds, and the great variety of possible models will mean a great variety of potential budgetary details. =
Still, the ultimate goal of this step is the same regardless of the developers’ architectural goals: develop the network into an genuinely functional product. By this we mean that by the end of this step, the network and the token must be genuinely useful such that they are actually used on a functional network. This means that any future agreement to purchase the tokens—and indeed the tokens themselves—should fail the Howey test and fall outside the definition of a security. More on this in Step 4. =
#### Step 4 in depth =
At the beginning of Step 4, Developers Inc. delivers the tokens to the investors and the investors’ rights in the SAFT are extinguished. The investors can achieve liquidity in their investment and Developers Inc. can, if the developers so desire, achieve operating revenue by selling some of the now-functional tokens directly to the public. =
The SAFT framework mitigates several risks that would typically endanger a public token sale at this stage. =
First, the SAFT framework elegantly navigates the federal securities laws: The SAFT itself is likely a security, and all requirements of the securities laws can be complied with in a summary fashion.  As set forth in Step 3, the tokens themselves are not, and never were, securities.[74] At minimum, the sale of tokens to the public fails the “expectation of profit” prong of the Howey test. To be sure, some purchasers may still purchase the tokens primarily to resell them on a secondary market for a profit. For the reasons set forth in the above section on already-functional utility tokens, this should not endanger the model. Developers Inc. has, in our example, already expended its “essential” efforts to create a functional network within which token holders can use a genuinely functional token. Now, after Step 4, any purchaser who expects profit from increase in secondary market value must expect that increase from the great variety of factors that affect markets for commodities. With the seller’s essential efforts already expended, it is unlikely that these “efforts of others” would predominate over that great variety. Moreover, since the tokens are not securities and the SAFT is non-transferrable, the investors do not, merely by purchasing the SAFT, risk being deemed underwriters if they resell their tokens.[75] =
Second, the SAFT framework elegantly addresses the money services laws: Even assuming FinCEN takes the position that the tokens are a CVC (which is not a fait accompli for all tokens), investors and the company are probably not money transmitters. To the contrary, when the company issues the SAFT (or the tokens) or when the investors resell the tokens, they are almost certainly trading on their own account. They exercise investment discretion as to the timing and terms of sale of their tokens. If Developers Inc. desires to sell further tokens to the public, it too might argue that it is trading on its own account pursuant to its investment discretion. The stronger argument is that it was only selling its mined tokens to the public. Thus, it did not both accept and transmit tokens. It created the tokens *ab initio* and at most only transmitted them, which is insufficient for “exchanger” status.[76] =
Third, as compared to the direct token presale, a SAFT transaction offers greater tax management opportunities, which could result in a lower overall tax burden. The funds raised in Step 3 are taxable as income to Developers Inc.,[77] but the SAFT is generally intended to be structured and taxed as a forward contract. If the SAFT so qualifies as a forward contract for tax purposes, then the transaction’s first taxable event does not occur until the tokens are delivered to the investors and the SAFT terminates.[78] Thus, Developers Inc. could have an additional year of tax losses to be carried back to the year of the taxable event. In fact, the company could achieve a tax-neutral transaction if the $15 million raised is less than or equal to Developers Inc.’s: =
(i) net operating losses incurred developing its network prior to delivery of tokens in the year of delivery, plus =
(ii) those incurred during the 2 years subsequent to token delivery, plus =
(iii) those carried over from previous years.[79] =
In this way, sellers that do not subsequently sell functional tokens to the public may obtain an additional tax year in which to incur losses to offset income from a token sale. This could result in a tax neutral sale similar in result to the sale of equity in a venture model. For sellers who do subsequently sell to the public, the primary benefit of the SAFT from a tax perspective is not that it eliminates taxable income, but that it staggers income, providing a longer period to capture operating loss carrybacks (i.e., the seller will raise a relatively smaller amount of proceeds under a SAFT before committing to a larger public token sale and thus will be more able to generate tax losses to offset the income from the token sale). =
The duration of Step 3 could be months or years. It is a careful balance between the amount of money raised from investors, the budgetary needs of developing the network, and the ability to expend the funds raised without incurring excessive tax liability. Since the SAFT is an investment contract with the right to receive a digital asset in the future, the contract may remain “open” for a variable length of time to simulate “vesting” periods for investors and further extend the time to triggering taxable income. =
**It is important to note that the tax treatment of a SAFT is not clear under current U.S. tax law and it is possible that the Internal Revenue Service may assert that the taxable event takes place at the time of sale rather than the time of delivery.**  In such case, the tax treatment of a SAFT transaction will be no more efficient than a direct token presale, but also not any less efficient. =
The U.S. tax laws, as they apply to token sales, may be suboptimal. They may benefit from future reconsideration or even revision, and we welcome additional input from commentators as part of the SAFT project. As they exist today, though, the SAFT is likely the best tool for managing them. =
## Policy Benefits of the SAFT =
At a high level, the goal of the SAFT framework is to ensure the appropriate laws apply to the appropriate steps and elements of a token sale transaction. When the developers sell the SAFT to investors, they sell a security. Thus, the investor protection laws, which seek to protect investors, should apply. When the tokens are ultimately distributed, though, the risks that the investor protection laws seek to mitigate are, by and large, absent. The network is functional, and the developers have already expended the “essential” efforts to imbue it with functionality. Thus, the risks of enterprise failure have subsided, and application of the investor protection laws that seek to mitigate those risks would be excessive. The broader consumer protection laws should apply instead. These laws provide the right protection for the product-based risks that now predominate the token. Accredited investors should take on investment-related risks where they arise out of a growing enterprise, and the public should take on consumer-related risk where they arise out of a working product. =
In our opinion, this approach best calibrates the investor protection laws to minimize suppressive effects on innovation and maximize entrepreneur opportunity. In this sense, it harmonizes SEC’s dual mandates of investor protection and capital formation. It does so by (i) concentrating the enterprise-related risk with accredited investors sophisticated enough to appreciate such risk and liquid enough to absorb it; and by (ii) allowing widespread public purchase when tokens present primarily product-related risk. =
The SAFT framework is attractive for another reason: It reflects what is actually happening in the market today. That is, it takes a descriptive rather than a prescriptive approach. Few (if any) direct token presales are actually raising tens of millions of dollars from the public in a few minutes. In reality, the developers behind these sales today spend months preparing materials, arguing the value proposition, pitching professional (usually accredited) investors and seeding their public sale with interest or even commitments. Following the SAFT framework would not demand that the industry discard its core values. =
A final policy benefit of a robust SAFT framework: Potentially eliminating the impetus behind the mass exodus of crypto developers to foreign jurisdictions. To be sure, the rise of token networks is a global phenomenon. Development teams are not always – or even often – based in Silicon Valley, or the U.S. This is an unequivocally positive expression of decentralized principals, and a circumstance we celebrate. In many cases, though, projects based originally and organically in the U.S. have made the difficult decision to relocate the project to non-U.S. jurisdictions like Switzerland. This seems to be due primarily to questions around the applicability of the U.S. securities, money services or tax laws. =
Put simply, this is a forfeiture of intellectual capital - a categorical loss for the U.S. and other countries like it, which seek to lead the world in technology innovation. This exodus is a significant problem for creators, for investors, and for users, too. Creators see themselves forced to leave, uprooting their lives and spending precious resources relocating themselves. Investors seeking to fund these networks are usually excluded with the rest of the U.S. citizenry. Early adopting users are shut out entirely. This has significant second order effects: The economies and businesses of other countries are exposed to these technologies first, and will begin to wield greater advantages over the U.S. economy and its businesses. Worse, businesses who relocate may find their new homes less friendly to their business models once those jurisdictions develop comprehensive token policies. Well-developed domestic standards for token sales, like the SAFT framework, could provide the certainty and comfort that keeps innovators at home. =
## Limits of the SAFT Framework =
At heart, the SAFT approach bridges treacherous regulatory waters between a whitepaper and a functional network MVP. If the network is already functional, or more likely, the company has the financial resources to complete development of the network without additional capital, the SAFT probably does not improve the company’s regulatory position. So long as the network and token are genuinely functional, similar arguments in defense of a public token sale exist regardless of whether the company used a SAFT to arrive at the public token sale. =
Furthermore, the SAFT framework is not very useful to non-utility tokens that are themselves securities when sold to the public. A token representing equity in a for-profit corporation or a limited partnership interest in a fund, for example, is a security (or potentially a security derivative) regardless of whether the seller used a SAFT to arrive at the public token sale. The SAFT would not have aided the DAO Token sellers in complying with any applicable securities law, for example. =
Ultimately, not every already-functional utility token will fail the Howey test. For these tokens, using a SAFT will not aid in navigating the securities laws. Howey is not a black-and-white metric for security status. It is a highly variable facts-and-circumstances test.[80] Some utility tokens, due to their particular facts or circumstances, may pass the Howey test despite being already-functional. How might this happen? In at least three ways. =
First, a seller might weaken its defense against the third prong of Howey by selling tokens predominantly to purchasers who could never put the token to its intended use. This weakens the token’s position as a non-security because it eliminates the consumptive use defense entirely. That is, it concedes the “expectation of profits” prong. For example, consider a network that is only usable by members of a particular industry, like the apparel trade. Perhaps the network ensures genuineness of an article of clothing by tracking its provenance from fabric mill to design house to distributer to retailer, and the token acts as a unique per-item identifier.[81] If the token’s sellers sold that token to the public at large, it would be highly unlikely that members of the public bought that token predominantly to use it. After all, a relatively small portion of the token-buying public operates fabric mills or apparel distributorships. Without more, a plaintiff could argue that token buyers made their purchases predominantly to profit, satisfying the third prong of Howey. =
Second, a seller might weaken its defense against the fourth Howey prong when the seller significantly over-promises in its sales materials. In such a circumstance, the seller’s efforts to imbue the token with greater utility might still predominate the variety of other market forces acting upon the token’s price. A profit-seeking purchaser might predominantly rely upon the efforts of the seller, even post-functionality, where the seller makes bold promises of developing more sophisticated functionality beyond that present at issuance. Purchasers might rely on those promises and expect to profit from the resulting increase in functionality, thus satisfying Howey’s final prong. Everyday sales language and speculation about future applications of the underlying technology should not endanger the token’s status as a non-security. After all, a great many software products are pre-sold to the public with promises of updates and ongoing improvements. Nearly all of those products can be resold on secondary markets, and they should not be construed as securities. To avoid the over-application of the securities laws, bold promises should only transform a token into a security when the effects on the market of the promised functionality clearly and demonstrably exceed the effects of all other factors. =
Third, a seller might weaken its defense against the fourth Howey prong when it maintains control over “monetary policy” for a token. In such a case, the fluctuation of the secondary market might not be due to mere supply and demand. It could instead be due to the efforts of the seller. For example, if the seller promises to redeem tokens based on its revenue or performance as an enterprise, the secondary market price of the token could, despite being already-functional, be due primarily to the efforts of the seller in generating revenue. Given the right market conditions, these efforts could be construed as the “essential” or “undeniably significant” factors on the price that might predominate against a backdrop of market factors. =
Additional limits of the SAFT framework surely exist. The possible variations on utility token facts and circumstances are numerous, thus the possible limits to the general rule of utility are numerous. One of the goals of the SAFT Project is to define these limits. =
## The SAFT as a Project =
We view the SAFT project as part of an emerging standard for how blockchain network developers can responsibly innovate. As discussed throughout this whitepaper, the larger blockchain community will benefit from efforts to properly allocate risk and reward between developers, investors and network users. We believe the SAFT is one approach to balancing the risk and reward among these stakeholders, but it is only an initial step. By way of this whitepaper, we hope to spur a conversation within the blockchain community and the legal community on, among other things: =
- the commercial implications of the SAFT framework; =
- the viability of funding open-source software projects through token sales; =
- the application of U.S. state laws to the SAFT; =
- the application of other U.S. federal laws, in particular the commodity derivatives laws, to the SAFT; and =
- the application of non-U.S. laws to the SAFT. =
We are committed to creating many diverse paths for responsible blockchain innovation and we view this whitepaper as our initial contribution to the discussion. To begin, A copy of a draft SAFT is attached to this whitepaper as Exhibit 1. The SAFT Project website is available at www.saftproject.com. =
---- =
[1] A comprehensive defense of token networks is outside the scope of this whitepaper. Influential commentators, though, have already arrived at similar conclusions.  See generally Balaji S. Srinivasan, Thoughts on Tokens, news.21.co (May 27, 2009), https://news.21.co/thoughts-on-tokens-436109aabcbe; Joel Monegro, Fat Protocols, Union Square Ventures (Aug. 8, 2016), http://www.usv.com/blog/fat-protocols. =
[2] Recent amendments to the Delaware General Corporation Law have paved the way for these tokens under state law and for use in the private markets. For more information on the Delaware Blockchain Initiative, see Newly Released Delaware Corporate Law Amendments Would Permit Blockchain Shares, Cooley (Mar. 15, 2017), available at https://www.cooley.com/news/insight/2017/2017-03-15-newly-released-delaware-corporate-law-amendments-would-permit-blockchain-shares. =
[3] For example, on what secondary market might a publicly-issued securities token trade in the U.S. today? A fully-featured secondary market would likely require registration as a securities exchange, and no existing blockchain token exchanges carry such a license. How could the seller satisfy the disclosure requirements of the securities laws? What custodians are qualified to hold securities tokens? What underwriters would support the issuance of a securities token? Given the lack of infrastructure, the questions far outstrip the answers. =
[4] There is of course no “typical” token sale. In this example, we attempt to capture the most common characteristics of various direct forms of token presales. =
[5] Direct token presales are sometimes referred to as Initial Coin Offerings (ICOs). This nomenclature borrows from the Initial Public Offering (IPO). It likewise carries with it much of the optical and regulatory baggage associated with public offerings of securities. Commentators have decried and disavowed it. See generally Zenel Batagelj, ICO 2.0—what is the ideal ICO?, Medium (Nov. 13, 2016), https://medium.com/iconominet/ico-2-0-what-is-the-ideal-ico-ee9d285a8939. =
[6] Solidity is the native programming language of the Ethereum network.  =
[7] See Peter Van Valkenburgh, What does it mean to issue a token “on top of” Ethereum, Coin Center (May 10, 2017), available at https://coincenter.org/entry/what-does-it-mean-to-issue-a-token-on-top-of-ethereum. =
[8] Apocryphal accounts of teams that “raised $XX,XXX,XXX in YY minutes!” are misleading at best. Much of those funds were raised after months of pitching mostly accredited investors, just like typical venture capital financing. The closing of those sales is simply timed to occur concurrently with the public sale. =
[9] See 17 C.F.R. § 230.501(a) for Accredited Investor standards. =
[10] Funds raised are usually in the form of virtual currencies or other blockchain tokens, but are also increasingly in the form of government currencies. =
[11] Many state laws, including securities and money services laws, apply to token sales. Many other federal laws apply as well. Foreign laws are of particular concern to token sales. We regrettably cannot address them either. To that end, see the final section of this whitepaper on the SAFT Project. =
[12] The Securities Act of 1933, as amended, (the “Securities Act”), Section 5, 15 U.S.C. § 77e(a), (c); 77d. =
[13] Securities Act, Section 2(a)(1), 15 U.S.C. § 77b(a)(1). =
[14] Landreth Timber Co. v. Landreth, 471 U.S. 681, 691 (1985) (“[T]he Howey economic reality test was designed to determine whether a particular instrument is an ‘investment contract,’ not whether it fits within any of the examples listed in the statutory definition of ‘security.’”). The SEC has not taken a clear position on the status of utility tokens as securities. As such, the appropriate analysis for new technology is to ascertain the economic reality underlying any business transaction related to that technology. The standard for such analysis is the Howey test, which for over seventy years has “been the appropriate metric for gauging whether, as a matter of economic reality, a business relationship constitutes an ‘investment contract.’” Ruefenacht v. O'Halloran, 737 F.2d 320, 338 (3d Cir. 1984), aff'd sub nom. Gould v. Ruefenacht, 471 U.S. 701 (1985). =
[15] SEC v. W.J. Howey Co., 328 U.S. 293, 298–99 (1946). =
[16] 328 U.S. 293 (1946). =
[17] Id. at 298-99. =
[18] States have adopted alternatives to the Howey test, such as the risk capital test. See generally Moreland v. Dep’t of Corps., 194 Cal. App. 3d 506, 519 (5th Dist. 1987) (explaining the four factors of the “risk capital test” as “(1) whether funds are being raised for a business venture or enterprise; (2) whether the transaction is offered indiscriminately to the public at large (3) whether the investors are substantially powerless to affect the success of the enterprise; and (4) whether the investors’ money is substantially at risk because it is inadequately secured.”). =
[19] For a more comprehensive analysis, with a welcome focus on policy see, for example, Peter Van Valkenburgh, Framework for Securities Regulation of Cryptocurrencies, Coin Center (Jan. 2016), available at https://coincenter.org/entry/framework-for-securities-regulation-of-cryptocurrencies. =
[20] SEC v. Shavers, No. 4:13-CV-416, 2014 WL 12622292, at 5 (E.D. Tex. Aug. 26, 2014). =
[21] SEC v. Glenn W. Turner Enters., 474 F.2d 476, 482 (1973). =
[22] Courts look to the economic realities of the arrangement, not the labels applied to the elements of any particular transaction. See Tcherepnin v. Knight, 389 U.S. 332, 383 (1967) (“Congress did not intend to adopt a narrow or restrictive concept of security in defining that term. . . . [Rather,] security embodies a flexible rather than static principal . . . .” (internal quotation marks and citations omitted)). =
[23] See, e.g., SEC v. SG Ltd., 265 F.3d 42, 49-50 (1st Cir. 2001). =
[24] Revak v. SEC Realty Corp., 18 F.3d 81, 87 (2d Cir. 1994). =
[25] Id. at 88. =
[26] See SEC v. Koscot Interplanetary, Inc., 497 F.2d 473, 479 (5th Cir. 1974); Villeneuve v. Advanced Bus. Concepts Corp., 698 F.2d 1121, 1124 (11th Cir. 1983), rev’d on reh'g on other grounds, 730 F.2d 1403 (11th Cir. 1984). =
[27] In some cases, holders of the token are required to take meaningful action to realize any profit from their purchase, like acting as an “oracle” for real-world events. These token holders are unlikely to satisfy the horizontal commonality standard (among other prongs) because those who take meaningful action profit differently than those who do not. =
[28] United Hous. Found. v Forman, 421 U.S. 837, 854-55 (1975). =
[29] SEC v. R.G. Reynolds Enters., 952 F.2d 1125, 1135 (9th Cir. 1991) (“Investors expected profits to come primarily from the discount below the current price of gold at which the gold ore was being offered [and] were dependent on [the promoter’s] efforts to realize profits from that discount.”). In another case involving a discounted sale of gold coins to be delivered in the future, the Ninth Circuit held that the financial instrument was not a security, but did not focus on the discount; instead the court relied solely on the fact that investors anticipated an increase in the price of gold. SEC v. Belmont Reid & Co., 794 F.2d 1388, 1391 (9th Cir. 1986); R.G. Reynolds Enters., 952 F.2d at 1135 & n.13 (distinguishing Belmont Reid & Co.). =
[30] See Forman, 421 U.S. at 852-53. =
[31] See supra note 28 for case citation. =
[32] See Forman, 421 U.S. at 858 (“What distinguishes a security transaction . . . is an investment where one parts with his money in the hope of receiving profits from the efforts of others, and not where he purchases a commodity for personal consumption . . . .”). Many cases have distinguished between the acquiring an interest for personal use or consumption versus acquiring an interest in a profit-making venture. E.g., Landreth Timber Co. v. Landreth, 471 U.S. 681, 689 (1985) (“Applying the Howey test, we concluded that the instruments likewise were not ‘securities’ by virtue of being ‘investment contracts’ because the economic realities of the transaction showed that the purchasers had parted with their money not for the purpose of reaping profits from the efforts of others, but for the purpose of purchasing a commodity for personal consumption.”); SEC v. SG Ltd., 265 F.3d 42, 53 (1st Cir. 2001) (“The Supreme Court has recognized an expectation of profits in [situations] to be contrasted with transactions in which an individual purchases a commodity for personal use or consumption.”); Bronstein v. Bronstein, 407 F. Supp. 925, 930 (E.D. Pa. 1976) (distinguishing between ventures with “an eye towards personal use or consumption of the underlying interest [and for] the purpose of acquiring an interest in a profit-making venture”). =
[33] SEC v. Glenn W. Turner Enters., 474 F.2d 476, 482 (9th Cir. 1973). =
[34] E.g., Forman, 421 U.S. at 852 (holding that “profits” in the expectation-of-profits prong can mean “capital appreciation resulting from the development of the initial investment”). =
[35] 638 F.2d 77, 79 (9th Cir.1980). =
[36] 794 F.2d 1388, 1391 (9th Cir. 1986). =
[37] Sinva v. Merrill Lynch, 253 F. Supp. 359, 367 (S.D.N.Y. 1966) =
[38] We are not aware of any such token. This hypothetical example is for illustrative purposes only. =
[39] See supra note 38. =
[40] See supra note 38. =
[41] Bitcoin was the first blockchain token and has often been branded as “digital gold.” E.g., Kyle Torpey, Bitcoin: The “Perfect Gold,” (Jan. 21, 2015), http://insidebitcoins.com/news/bitcoin-the-perfect-gold/29080; see generally Nathaniel Popper, Digital Gold: Bitcoin and the Inside Story of the Misfits and Millionaires Trying to Reinvent Money (Harper 2015). =
[42] SEC Release No. 81207, Report of Investigation Pursuant to Section 21(a) of the Securities Exchange Act of 1934: The DAO (July 25, 2017), available at https://www.sec.gov/litigation/investreport/34-81207.pdf. =
[43] Id. at 11. =
[44] Id. at 10. =
[45] Id. at 16. =
[46] Id at 16-17. =
[47] 18 U.S.C. § 1960. =
[48] While an analysis of state law is outside the scope of this whitepaper, we note as a general matter that state money transmitter law also applies to persons accepting and transmitting funds as a business. For example, the New York “BitLicense” regulations would, on their terms, prohibit nearly all unlicensed token sale activity. Some license applications there have been pending for over two years. This timeline creates a de facto prohibition on token sales in the state. Available exemptions under state law may sometimes track exemptions in federal law, but this is not always the case. =
[49] 31 C.F.R § 1010.100(ff). =
[50] Id. at (ff)(5). =
[51] Id. at (ff)(5)(ii)(A). =
[52] Id. at (ff)(5)(ii)(E). =
[53] 76 Fed. Reg. 43585, 43594 (July 21, 2011) (“[P]ersons that sell goods or provide services other than money transmission services, and only transmit funds as an integral part of that sale of goods or provision of services, are not money transmitters. For example, brokering the sale of securities, commodity contracts, or similar instruments is not money transmission notwithstanding the fact that the person brokering the sale may move funds back and forth between the buyer and seller to effect the transaction.”). =
[54] FIN-2013-G001, Application of FinCEN’s Regulations to Persons Administering, Exchanging, or Using Virtual Currencies (Mar. 18, 2013) at 2-5, available at https://www.fincen.gov/sites/default/files/shared/FIN-2013-G001.pdf. =
[55] FIN-2014-R001, Application of FinCEN’s Regulations to Virtual Currency Mining Operations (Jan. 30, 2014) at 2-3, available at https://www.fincen.gov/sites/default/files/administrative_ruling/FIN-2014-R001.pdf (bitcoin miners are users because they have no obligations to send mine bitcoin to any other person or place for the benefit of another; they mine and use bitcoin solely for the miner’s own purposes and that activity does not involve acceptance or transmission of funds); FIN-2014-R002, Application of FinCEN’s Regulations to Virtual Currency Software Development and Certain Investment Activity (Jan. 30, 2014) at 3-4, available at https://www.fincen.gov/sites/default/files/administrative_ruling/FIN-2014-R002.pdf (investors in CVC are users as they simply make investments for their own account). =
[56] FIN-2014-R001 at 3. =
[57] FIN-2014-R002 at 4. =
[58] FIN-2013-G001 at 5. =
[59] FIN-2014-R001 at 3. =
[60] FIN-2014-R001 at 3. =
[61] FIN-2013-G001 at 3 (emphasis added). =
[62] For an in-depth analysis of the money transmission implications of token sales, see Coin Center’s report on the issue: Peter Van Valkenburgh, The Bank Secrecy Act, Cryptocurrencies, and New Tokens:  What is Known and What Remains Ambiguous, Coin Center (May 2017), available at https://coincenter.org/entry/aml-kyc-tokens. =
[63] In IRS Notice 2014-21 (Apr. 14, 2014) (the “Notice”), available at https://www.irs.gov/irb/2014-16_IRB/ar12.html, the Internal Revenue Service provided guidance regarding the taxation of CVC.  “Virtual currency” is defined as a digital representation of value that functions as a medium of exchange, a unit of account, and/or a store of value.  Virtual currency that has an equivalent value in real currency, or that acts as a substitute for real currency, is “convertible” and is a CVC.  The Notice concluded that CVC is “property” for U.S. federal income tax purposes and not a form of currency for such purposes.  The Notice does not address any virtual currency that is not “convertible”; however, in general the IRS’s logic would likely extend to any token whether or not convertible. =
[64] For purposes of this discussion, we assume that the entity issuing tokens is a fully-taxable corporation, incorporated in the United States.  =
[65] For instance, the spirit of the Federal Securities Laws is investor protection, based on the premise of an informed investor, achieved through full and fair disclosures and to facilitate the exchange of that information to allow for an outlet for capital formation. Yet a direct presale of utility tokens may raise large amounts of capital for a company with minimal disclosures to investors regarding the purpose of the underlying blockchain network or the expertise of the team developing the network. =
[66] For instance, the prohibition against unfair, deceptive, or abusive acts or practices controls for the risk that a product presents undue risk to consumers, or that the product is advertised in a way that dupes consumers, or that the company takes advantage of the information asymmetry between itself and the consumers regarding the product. See 12 U.S.C. § 5536(a). =
[67] See SEC Commissioner Paul S. Atkins, Remarks before the Financial Services Roundtable Lawyers Council 2007 Spring Meeting (May 10, 2007), available at https://www.sec.gov/news/speech/2007/spch051007psa.htm#3. Commission Atkins remarked: “poorly crafted regulations and careless regulators can upset a well-functioning market,” and later stated “[n]ew products may be stifled because they do not fit within a pre-existing regulatory bucket. They may run into a wall of resistance from regulators whose eyes are jaundiced by the constant stream of enforcement cases that they see.”  And in extreme cases “some firms might go so far as to shun innovation.”  See SEC Commissioner Michael S. Piwowar, Capital Unbound: Remarks at the Cato Summit on Financial Regulations (June 2, 2015) available at https://www.sec.gov/news/speech/capital-unbound-cato-summit.html. Commissioner Piwowar urged the Dodd-Frank Act mandated review of the SEC’s accredited investor standards to consider eliminating the “artificial distinctions” between investors noting that if you meet the definitional standards, “then you were in the privileged class and could choose to invest in the full panoply of investments, whether public or private. If not, the government decided that, for your own protection, you were restricted access to these private investments.” =
[68] Safe Financing Documents, Y Combinator (Feb. 2016), available at https://www.ycombinator.com/documents/. =
[69] Other exemptions may be available depending upon the makeup of the investors, marketing methods, etc. =
[70] Form D is used to file a notice with the SEC of an offering of securities that qualifies as exempt under Regulation D. =
[71] See generally SEC Lifts Ban on General Solicitation in Certain Private Placements: What Does it Mean for Entrepreneurs?, Cooley (July 16, 2013), available at https://www.cooley.com/news/insight/2013/sec-lifts-ban-on-general-solicitation-in-certain-private-placements-what-does-it-mean-for-entrepreneurs.  =
[72] Any “tokens” representing stock or otherwise associated with such an incorporation are not a part of this analysis. =
[73] Some commodity forwards are regulated by the Commodity Futures Trade Commission (“CFTC”). For the purposes of this whitepaper, we rely on the CFTC’s position, recently set forth in the well-known Derivabit Order (available at http://www.cftc.gov/idc/groups/public/@lrenforcementactions/documents/legalpleading/enfcoinfliprorder09172015.pdf) that virtual currencies are “exempt commodities.” Tokens probably fit within the CFTC’s definition of “virtual currency” set forth in that order, and the SAFT contemplates “physical delivery” of the tokens as CFTC has previously described the term. Thus, the SAFT transaction is probably not CFTC-regulated. We welcome further analysis by other commentators in the SAFT Project of the treatment of the SAFT under the Commodities Exchange Act and its implementing regulations. =
[74] Indeed, a SAFT seller never needs to create a pre-functional utility token. =
[75] Investors are participants in the distribution of the utility tokens following conversion of the SAFT, but the token is not a security. Though the SAFT is a security, they do not distribute it. The definition of underwriter under the Federal Securities Laws is limited to the participation in a distribution of a security, thus SAFT investors need not fall within the definition or risk exposure associated with being deemed an underwriter.  See Securities Act Section 2(a)(11), 15 U.S.C. § 77b(a)(11).  =
[76] Further, because often the entity creating the token has no special authority to remove tokens from circulation (other than that possessed by any other token user:  by sending them to an address that nobody controls, by refusing to sell any tokens owned), the scenario contemplated here should not make Developers Inc. an “administrator” of CVC as described in FIN-2013-G001. =
[77] A company’s issuance of its own debt or equity for cash is generally a tax-free event. See 26 U.S.C. § 1032.  By contrast, as set forth above, the sale of tokens for cash is fully taxable to the seller, subject to the availability of net operating losses to offset the taxable income. =
[78] A “forward contract” is generally taxed under the “open transaction” theory of taxation with the result that the tax event to the selling party is deferred until the transaction is “closed” and the underlying sold property is delivered. =
[79] In general, corporations may carry forward net operating losses for 20 years and carry back net operating losses for 2 years.  See Section 172 of the Internal Revenue Code of 1986, as amended. =
[80] See SEC Release No. 81207, supra note 42, at 10 (“[T[he U.S. federal securities law may apply to various activities, including distributed ledger technology, depending on the particular facts and circumstances, without regard to the form of the organization or technology used to effectuate a particular offer or sale.“); Id. at 16 (“Whether or not a particular transaction involves the offer and sale of a security—regardless of the terminology used—will depend on the facts and circumstances, including the economic realities of the transaction.”) =
[81] We are not aware of any such token. This hypothetical example is for illustrative purposes only. =