Those who follow cryptocurrencies will know that an initial coin offering (ICO) refers to the issuance of digital tokens to finance a blockchain-based venture. They will also know that the legal status of tokens is an ongoing discussion among market participants and regulators.
ICOs have attracted policymakers' attention because they are growing in number and size. For example, late last year Filecoin, a venture promising to reduce the costs of data storage, raised $257 million in a token offering. A Filecoin token gives its holder the right to buy storage services on the Filecoin platform.
Besides being a means by which users purchase goods and services on blockchain platforms, tokens are the reward to those users who help to effect transactions. In the case of Bitcoin, the gains to these 'miners' come from their role in validating exchanges of bitcoin. Filecoin miners, for their part, are rewarded for making abundant storage space available.
How do blockchain ventures price tokens in an ICO? Do they have a financial incentive to mislead buyers? How do venture capitalists, who invest in ventures with an expectation of future profits, differ from people who acquire tokens because they expect to use the goods and services to which tokens entitle them?
A new NBER working paper by MIT's Christian Catalini and Joshua Gans from the University of Toronto supplies answers to these questions by studying a two-period model of a token offering, covering the ICO stage during which interested buyers acquire tokens, and a market stage at which the platform goes live.
Among its findings:
1. Token prices reveal buyers' valuation of tokens' usefulness.
The near-term objective of an ICO is to raise funds for a venture. But buyers bid for tokens on the basis of the gains they expect to make from using them. For example, those who took up the Filecoin offering did so in the expectation that the tokens would be valuable as means to buy data storage. Thus an ICO reveals buyers' valuation of the venture's business proposition.
Without a claim against future profits, and because the tokens on offer have no established exchange value, a token buyer's gain comes solely from his or her ability to enjoy the goods and services that the platform promises to make available. That's why the issues in an ICO are known as 'utility tokens' and why many analysts have opposed their designation as securities.
It's true that a token buyer could re-sell the token, thereby perhaps deriving a gain — although many ventures, to stay on the safe side of the SEC, have followed the SAFT (Simple Agreement for a Future Token) model of issuance, which forecloses secondary trading until the platform is live. But even if there is secondary trading, the ultimate source of a token's value — the reason the final purchaser will pay good money for it — is the claim it offers against a particular good or service.
2. ICO funding is less certain than equity funding. But successful ventures funded by an ICO have higher returns than when financed with equity.
There are benefits to funding a venture by issuing tokens. The venture is able to attract not just funds but future users of the platform. As we saw above, success in doing so is a measure of whether the venture is likely to become popular.
Yet utility tokens don't entitle their holders to any profits, just the goods and services that the token promises to buy. Catalini and Gans show, as a result, that ICOs may sometimes fail to raise enough funds even when they are predicted to be viable, whereas equity financing, in their model, always raises sufficient money to fund viable ventures.
This finding helps explain why many ventures, including Filecoin, have venture capital investors in addition to raising funds through token issuance. VC investment helps to cover the fixed costs of development and increases the probability that viable ventures will come to fruition. On the other hand, equity-funded ventures yield lower returns to developers than ICO-funded ones. A better chance at success doesn't come free of charge.
3. Commitment has financial value.
Much recent discussion of ICOs has involved the incentives of the venture. With so many crypto-asset scams in the press, some have wondered whether it might be the technology itself that encourages deception.
Catalini and Gans' paper suggests otherwise. If the venture cannot, for example, commit to a predictable schedule of token growth (or no growth) over the lifetime of the platform, then the value expectations of prospective buyers will change. Uncertainty about platform governance thus makes it less likely that an ICO will raise sufficient funds. The doomed launch of the Venezuelan Petro last month underscores the problems that unreliable token offerings face in raising capital.
The venture must inspire confidence that the tokens it issues will remain the only medium to purchase goods and services on its platform. If dollars can subsequently buy one access to the platform as well as tokens, then the latter have no utility value.
This finding explains why ICO white papers — see, for instance, the one issued by developers of Filecoin — go to great lengths to explain the mechanics of their platform. Some ICOs, it is true, lured investors with the promise of large gains while giving little detail about the function of the proposed platform. But can those poor buyers be said to have had rational expectations about anything?
4. Network effects are virtuous (for the platform).
If users derive value from other people's joining a blockchain platform, i.e. there is a network effect, then they will be willing to pay more for ownership of tokens. In the case of Filecoin, there is a benefit to those buying storage space from more miners' being involved, namely better pricing through more competitive bidding by those selling storage space.
But if token buyers have rational expectations and all benefit from network effects, then Catalini and Gans show that no individual buyer will be able to pay any less for the tokens than their true valuation. As the authors put it, unfavorable expectations about demand are not sustainable.
Skeptics may quibble with the paper's simplified model, particularly its assumption of rational expectations from token buyers. Given cryptocurrency volatility, isn't irrational exuberance perhaps a more plausible assumption?
Still, Catalini and Gans' conclusions reveal much about the incentives and trade-offs facing blockchain ventures. To succeed, they must commit to predictable governance and clearly show the gains to be had from ownership of the token.
Perhaps most relevant for policymakers, the paper's findings suggest that utility tokens should not be designated as securities according to the Howey test. In Catalini and Gans' model, the tokens are not an investment for profit but an advance purchase of goods and services on a platform. In legal terms, tokens are commodities, not securities.
Whether regulators will be persuaded by these arguments is an altogether different question.