Token Lockups & the Telegram Case

Note: This has been republished, with a new preface containing some important context and corrections, from a 2020 article still available at https://lex-node.medium.com/sec-v-telegram-three-deeper-takeaways-423b197f76d6

Preface (New Material 1.12.2025)

Recently we have again been getting a lot of questions on the legal ramifications of ‘token lockups’. This is a widely misunderstood topic, and, from a U.S. law perspective, it is almost impossible to understand without reading Judge Castel’s opinion holding in favor of the SEC in SEC v. Telegram, summarized below.

Most tokens are not ‘intrinsic’ securities, but transactions in such tokens may be covered by the securities laws in various ways nonetheless, because the tokens were sold at some point in connection with a securities financing. The Telegram case highlights this.

Now, as background, you need to understand that every securities transaction must be either registered (like an IPO) or exempt from registration. The most typical way to be exempt is if the security is being offered and sold privately to accredited investors who intend to hold it as an investment long-term. There are various rules that codify different ways of doing this—most famously, Rules 506(b) and 506(c) from the SEC’s Regulation D, which is how many people get the impression that there is a ‘rule’ that ‘only accredited investors can invest’. That rule, however, is both non-exclusive and oversimplifying, as not only must the investors be accredited, but they must also be buying for the long term, as true investors rather than ‘flippers’ (i.e., underwriters or brokers).

Now basically in Telegram, as we explain below, the court held that when venture capital investors buy a security that converts into a token, special issues apply, due to how liquid and broadly distributed tokens are. The court found in Telegram that the usual indicia that a venture capital investor intends to hold a security long term and therefore qualifies as a buyer for a Regulation D exemption (such indicia usual being a representation about long-term holding in the purchase agreement, combined with a minimum one-year holding period for the security itself) were insufficient. This meant that the initial fundraise (a sale of SAFTs or similar agreements by Telegram to accredited venture capital investors) itself was never exempt in the first place, and thus the securities laws were violated in that initial raise. Because the distribution of GRAM tokens was an integral part of that securities law violation, the distribution of GRAMs was enjoined (fancy word for prohibited).

The way the ‘legal market’ has responded to this is by imposing post-TGE lockups on SAFT tokens to create better ‘proof’ that investors really do want to hold long term. A period of at least one year post-TGE is typical, because this tracks with Rule 144’s minimum holding period for securities. However, this does not mean the token itself is a security or that transfers in the token can be/must be Rule-144 compliant overall—in fact, due to various other requirements in Rule 144, they basically never are.

As you will see, I was a bit pessimistic in my initial reaction to the case—I said that SAFTs were dead. The real solution, as it turned out, was not to kill SAFTs and similar agreements, but simply to impose longer holding periods on the investors. Thus, a SAFT holder will typically be subject to a ‘double holding period’—at least one year for the SAFT itself, and then at least one year for all tokens that the SAFT converts into. This is still best legal practice in the U.S. if you are trying to conduct a Reg-D-exempt offering as it better demonstrates the long-term holding intent of the investors, and thus makes the fundraise itself more likely to be deemed exempt from registration.

Is this unfair? Are tokens subject to heavier burdens than normal securities? Yes and no. On paper a two-year overall holding period for token investors looks bad; however, in practice, private securities markets are highly illiquid and opaque. Although their legally mandated holding period is only one year, most VCs who buy traditional securities are practically forced to hold them for 3+ years. In contrast, a crypto VC can usually be sure of selling their tokens at a reasonable valuation after the recommended two-year period described above. Thus, the circumstances of the two markets are simply very different—for traditional securities in early-stage companies, it is almost a certainty that the VC intends to hold long-term—by contrast, this is not so with crypto VCs. Thus, it is somewhat understandable the two markets would have different rules under the securities laws, despite this sounding unfair on paper.

Introduction

On March 24, 2020, Judge P. Kevin Castel of the Southern District of New York granted the SEC’s motion for a preliminary injunction against Telegram’s distribution of GRAM tokens for its new TON blockchain to investors.

As a U.S. corporate/securities attorney, I have been deep in the trenches of the “token vs. security” debate since 2017, advising clients on token issues and participating in numerous public and semi-public debates on the topic with attorneys, regulators and entrepreneurs. The purpose of this article is to outline the most critical takeaways from Castel’s opinion in light of the rich history of the issue.

The case has been widely covered in the media. For more background check out the National Law Review’s summary and CoinDesk’s extremely thorough coverage by Anna Baydakova and Nikhilesh De.

In this article, by “cryptocurrencies” I will be referring to what previously used to be called “utility tokens” — public blockchain-based tokens with no express payment or stockholder-style voting rights. GRAMs are arguably an example of such a cryptocurrency, or in any event were treated by Castel as such since he did not base his opinion on any of the staking features of GRAMs. Of course, it should go without saying that different legal considerations may apply to tokens with voting and/or revenue-earning features that are taken into account.

Takeaway #1: The “Functionality/SAFT School” and “Manner of Offering School” are Dead

In late 2017, when it became obvious that cryptocurrencies frequently implicate the securities laws, securities attorneys and others still disagreed on the details. At the risk of oversimplification, we can summarize the different views as if they came from four different “schools of thought.” Two of the most influential schools of thought were what I call the “Functionality/SAFT School” and the “Manner of Offering School,” which are discussed in this section and I view as clear losers in the Telegram case. The other two schools of thought will be discussed in the next section, and the Telegram case is more ambiguous about them, so both live to fight on in future wars.

• The “Functionality/SAFT School” believed that promises to deliver cryptocurrencies before the underlying technology is built and deployed — i.e., before the cryptocurrencies are “functional” — constitute securities, because the pre-buyers were supplying risk capital to the developer team and could profit from that team’s efforts by the increase in value between the date of pre-purchase and the date of network launch. However, once the cryptocurrencies were “functional” they would not implicate the securities laws anymore, being mere products or digital commodities. This view was best represented by the SAFT Whitepaper jointly published by Protocol Labs and Cooley LLP.

• The “Manner of Offering School” believed that cryptocurrencies are intrinsically commodities or products, but that marketing them with words like “investment,” “profit,” “increase in value,” etc. would cause them (or certain transactions in them) to be regulated as securities by creating a reasonable expectation of profit on the part of buyers, contrary to their normal nature as non-securities. If issuers marketed cryptocurrencies consistently as products or currencies and avoided all insinuations of future upside potential, that would be enough to cause cryptocurrencies to be treated the way they should be: as products or digital commodities, not securities. This view was best reflected in aspects of The Brooklyn Project’s Consumer Token Framework and ConsenSys’s Proof-of-Use Approach.

I view the Telegram decision is the last nail in the coffin for both these schools of thought:

1. The fact that the TON blockchain would be functional when GRAMs would be issued to SAFT holders and sold on secondary markets had zero impact on Castel’s reasoning and opinion. Castel reasoned that despite GRAMs’ functionality and potential usefulness as a mere virtual currency, GRAMs could still continue to appreciate substantially in value over time based on Telegram’s ongoing entrepreneurial efforts to promote GRAMs and the TON blockchain, through integration with the Telegram app or otherwise. Moreover, buyers of GRAMs could still have a reasonable expectation of such future appreciation on such basis due to Telegram’s large holdings of GRAMs, the incentives it had set up for TON developers and its reputational issues around the fact that Telegram, as a technology company that independently has a very popular messaging app, is associated publicly with GRAMs and the TON blockchain.

2. Prior to the injunction hearing, Telegram publicly disavowed all future efforts to support the value of GRAMs in any way, shape or form. If the “Manner of Sale” school were right, these disavowals would mean that when GRAMs would hit the market, no reasonable person could expect to profit on GRAMs from Telegram’s future entrepreneurial efforts, because there would be a lack of marketing that says people will profit on GRAMs from Telegram’s efforts — in fact, there would be abundant marketing expressly to the contrary. However, Castel’s decision to grant an injunction against the distribution of GRAMs and the way he reasoned about the ongoing objective alignment between Telegram by virtue of holding GRAMs and future hypothetical GRAM buyers strongly imply that “manner of sale” is not dispositive. Rather, Castel’s opinion is best read to imply that even in the absence of any express “investment-style” marketing , the securities laws can continue to apply if the issuer has knowable objective reasons for continuing to make such efforts.

Takeaway #2:We Don’t Know if or when Cryptocurrencies Represent Securities in the Secondary Market

Castel’s opinion does not directly help us decide when cryptocurrencies themselves represent securities or should be regulated as such in secondary market transactions. Thus, it does not help us decide the often fierce debate between the two other main schools of thought on cryptocurrency securities law issues, which I summarize as follows (again, at the risk of some oversimplification):

• The “Investment Contract Literalism School”: This school of thought assumes that an “investment contract” is literally a contract between a securities issuer and a securities purchaser pursuant to which the purchaser has a reasonable expectation of profits from the entrepreneurial efforts of the issuer. When cryptocurrencies are sold pursuant to such a contract , the securities laws apply — not to the cryptocurrencies themselves, but to the contractual transaction. However, when the cryptocurrencies are later resold without such a contract, the securities laws do not apply because the new buyers don’t have a contract with the issuer. Even if, for some reason, the new buyers expect a profit from the issuer’s ongoing efforts, it doesn’t matter, because, in the absence of a literal contract with the issuer, the expectation is unreasonable — they have no rights against the issuer and should therefore expect no benefits from the issuer. The best representation of this view can be found in “Ain’t Misbehavin’: An Examination of Broadway Tickets and Blockchain Tokens” by Lewis Cohen.

• The “Investment Contract Functionalism School” Under this approach, investment contracts are typically not literal contracts, but rather are implied-by-law contracts. The law creates these ‘virtual contracts’ whenever there is a transaction that functionally meets the criteria of the Howey test, including when someone buys a cryptocurrency on an exchange. The issuer is a ‘party’ to this ‘virtual’ investment contract by virtue of being the entrepreneurial focus of a Howey scheme. By analogy, consider a paper stock certificate: it is literally just a piece of paper, but because it functions as a representation of the stock, all transactions involving that piece of paper will be regulated as securities transactions rather than paper sales. Like a paper stock certificate, it is almost as if a centralized cryptocurrency simply is a security because practically all transactions in it will satisfy the Howey test. Eventually, based on sufficient decentralization, the Howey factors will cease being met, and the token will cease functioning as or representing an investment contract. That could also happen with a stock certificate — the company could dissolve, and the certificate could revert back to being a meaningless piece of paper. This view is set forth in detail in my “Open Letter to Commissioner Hester Peirce Regarding an SEC Token Safe Harbor” and aspects of it are also reflected in William Hinman’s speech “When Howey Met Gary (Plastics)”.

Judge Castel expressly refused to decide one way or another whether GRAMs would represent securities on the secondary market. Instead, he based his grant of an injunction on the view that the SAFTs were securities and “the resale of Grams into the secondary public market would be an integral part of the sale of securities without a required registration statement.” Thus, Castel’s reasoning does not contradict or disprove either the Investment Contract Literalism School or the Investment Contract Functionalism School. However, I believe the overall flavor of the opinion and the reasoning used in it is more consistent with the Investment Contract Functionalism School.

Takeaway #3: Regulation D and the Rule 144 Safe Harbor Do Not Work for Pure Cryptocurrencies

This is the most surprising and controversial result of Castel’s opinion. For effect I am intentionally stating it in the strongest and most provocative possible terms.

The SAFT buyers were all accredited investors and had purchased their SAFTs more than a year prior to the planned conversion of the SAFTs and distribution of GRAMs. Ordinarily, this would mean the sale of the SAFTs satisfied Rule 506(c) of Regulation D and that, even if the GRAMs themselves were securities, they should sellable on the secondary market by the SAFT holders because the 12-month holding period of Rule 144 would be satisfied via “tacking” of the SAFT holding period onto the token holding period.

However, Castel’s opinion clearly contradicts this reasoning and implies, at a minimum, that he viewed the Rule 144 safe harbor as being unavailable. The question is: why? The opinion is not completely clear on the reasoning — it does not even mention Rule 144 — but seems to be based on the view that notwithstanding possible technical Rule 144 compliance, the SAFT buyers were nevertheless acting as underwriters and therefore “the integral scheme” of GRAM sale and distribution never complied with the securities laws, even at the stage when the SAFTs were originally sold:

Everyone needs to pause and appreciate what a surprising result this is. It is very hard to reconcile with standard transactional practice in private stock financings. Typically, in Regulation D transactions, issuers get a flat rep from the investors that they are not acting as underwriters, and also sign the investors up to covenants requiring them not to transfer the securities at least until it is possible to do so within the Rule 144 safe harbor (usually requiring as 12-month lock up and a legal opinion to the effect that Rule 144 is available after the 12 months). This is generally considered sufficient. Before now, issuers have not been tasked with reading the minds of investors to determine if they really are underwriters or not in their heart of hearts. However, in this case Castel found that Telegram had no valid exemption because it failed to use reasonable due diligence to determine that the SAFT buyers were not underwriters. Why the difference here?

Although Castel’s reasoning is not perfectly clear, I have a theory: I believe it is tacitly based on the perception that pure cryptocurrencies — having no dividend or other payment rights — are prima facie absurd for traditional investors like venture capital funds to hold for a long period. Unlike when a VC is buying stock in a private placement for a pre-public Uber or Facebook, or a pre-M&A WhatsApp, with cryptocurrencies there is no pot of gold at the end of the rainbow. There is no natural “exit event” to HODL for — no IPO, no big acquisition by a strategic acquirer, nothing. Instead, pure cryptocurrencies are Ponzi-like in that they primarily become more valuable by more people buying them on the secondary market. Therefore, I think, in Castel’s view, it really beggars all belief that the SAFT buyers intended to hold their GRAMs for a long period as true investments. In other words, Castel must think it is readily apparent simply from the dynamics involved in cryptocurrency price increases, together with the discount schemes that Telegram set up for the investors, that the investors intended to sell the GRAMs very quickly from the start, and thus were acting as underwriters, rather than true investors all along. Therefore, Telegram did not exercise reasonable diligence in confirming that the investors were not underwriters, and no Regulation D exemption was available, even for the SAFTs.

Some may disagree with the implicit anti-HODL assumptions being employed here. But, if I am reading the tea leaves right here, and regulators and other judges end up agreeing with Castel in similar circumstances, the previously popular Regulation D + Rule 144 combo for selling pure cryptocurrencies is dead in the water. In light of this result, I will now be hard pressed to advise a client that they can sell SAFTs — or even just cryptocurrencies in their raw state— in reliance on that one-two punch.

Instead, I will now have to advise clients that there must be a more robustly evidenced basis for believing that the accredited investors truly want to hold the tokens for a long period. As part of that, I might advise them that they should sell a more complex token with some real back-end value, rather than a pure cryptocurrency. If the token has real investment features like the right to receive payments, vote on software upgrades and/or use of funds, etc. —those facts could be helpful in distinguishing the token from GRAMs and pure cryptocurrencies, and could, in theory, be more consistent with the view that the initial investors are not underwriters. But let there be no mistake — there is no quick-fix silver bullet to address this issue. The bottom line is that much more care will now have to taken to look at the motives of token investors, and projects will have to consider longer lock-ups and other measures to support the existence of a true investment intent on the part of early token buyers.

Post-publication update: There are some other possible ways of reconciling Castel’s reasoning with Rule 144:

• Some SAFT buyers did transfer their SAFTs in secondary transactions during the first 12 months after they were sold, which would violate Rule 144’s conditions — and even though the terms of the SAFT prohibited such transfers, Telegram knew they were happening and seemingly failed to try to stop them from happening. Perhaps this makes Rule 144 unavailable, in some sense, not only for those particular SAFT buyers, but for all of them. Or perhaps it is evidence of a broader failure by Telegram to confirm lack of underwriting intent on the part of all the SAFT purchasers. However, one would think that if Castel were basing his order primarily on this fact, that he would have emphasized it — he did not.

• Rule 144 is “is not available to any person with respect to any transaction or series of transactions that, although in technical compliance with Rule 144, is part of a plan or scheme to evade the registration requirements of the Act.” The overall GRAM distribution scheme could be seen as such an evasion scheme, and thus Rule 144 may be unavailable. However, if this is Castel’s reasoning, it seems rather circular to me and ultimately just seems like another way of saying that Telegram should have known from the start that the SAFT buyers were really underwriters even if they technically complied with Rule 144.

• Rule 144 is not available except if the issuer has done a proper private placement in the first place. Telegram committed some minor technical mistakes regarding its Form D filing for Regulation D, and was somewhat wishy-washy as to whether it was relying on Regulation D or the statutory provision it is based on — Section 4(a)(2). Thus, perhaps Castel decided that because the private placement exemption was not properly secured, Rule 144 could not be relied upon either. However, this seems to be very different from Castel’s express reasoning — which related to the SAFT buyers being underwriters rather than any independent technical problems with Telegram’s exemption.

• Rule 144 applies differently to “affiliates” of the issuer. In the hands of such an affiliate, the SAFTs would be not only “restricted securities” but also “control securities” and would generally not be transferable even after 12 months. Could it be that large SAFT buyers became affiliates of Telegram by virtue of purchasing SAFTs convertible into large amounts of GRAMs and that therefore the SAFTs and associated GRAMs were “control securities”? This would be an interesting way of reasoning about the issue, but I assume this is sufficiently technical and non-obvious that if Castel were thinking this way, he would have said so. He did not.

Conclusion

Although I always predicted Telegram would lose and should lose this case, I was surprised by the way it lost the case. I believed that Judge Castel would need to determine whether GRAMs represented securities on the secondary market, and thus would need to look at whether an average secondary buyer would have a reasonable expectation of profits from Telegram’s future efforts in a common enterprise. I still believe this would have been a better, more coherent way to tackle the case — and it would have resulted in the same outcome.

However, because Castel was reluctant to explore the secondary market issues, we are now left with a much more muddled result that throws into doubt many years of standard practice under Rule 144 and may further discourage blockchain entrepreneurs from trying to comply with securities laws at all. So my overall verdict is: Right result, wrong reasoning. Nevertheless, the decision represents progress in the evolution of cryptolaw and will hopefully discourage the long-derided SAFT model with its built-in gains for wealthy venture capitalists at the expense of ordinary retail investors.