In an effort to imitate tax law, cryptolaw introduced two new acronyms over the last week:
- DUNAs, or decentralized unincorporated nonprofit associations; and
- BORGs, or cybernetic organizations.
Some US tax musings on both...
DUNAs. The DUNA, conceived by Miles Jennings (of a16z crypto) and tax lawyer David Kerr, gives DAOs with > 100 members the ability to contract, sue, and pay taxes (i.e., legal personhood) while according members limited liability. Wyoming adopted a DUNA law effective 7/1/24.
TAXES. DUNAs are intended to be taxed as domestic corporations for US tax purposes. That means they:
- Pay US corporate tax;
- Need to withhold on payments to non-US tokenholders if treated as dividends; and
- Have info reporting requirements.
Let's take a closer look...
CORPORATE TAX. As US corporations, DUNAs would be subject to 21% federal tax on their net income, and possibly state & local taxes too. (Charitable and social welfare DUNAs might be eligible to apply for tax-exempt status, but that would be a rare exception.)
WITHHOLDING. The DUNA law prohibits dividends, but contemplates payments treated as dividends for tax purposes. (Confusing, I know!)
DUNAs have to withhold 30% on dividends to non-US holders (with potential reduction to 15% for EU residents willing to provide tax forms).
Dividends don't reduce net income so, after accounting for income tax, the US federal tax hit on income dividended to non-US holders is 44.7%...before accounting for any taxes they might owe in their own jurisdictions!
So...what's a dividend for US tax purposes?
Streamed fees. a16z crypto suggests DUNAs could stream fees to members "in exchange for participation in governance," presumably by staking into a governance module. That might be "compensation" under the DUNA statute, but it would almost certainly be dividends for tax.
Token buybacks. The DUNA law also allows buybacks, which would be dividends for US tax purposes unless they meaningfully reduce a tokenholder's interest in the DAO.
Tax law provides safe harbors for meaningful reduction, but the test is fact specific and applies related-party attribution. Corporations typically require non-US shareholders to complete questionnaires before doing share buybacks to determine whether they have to withhold.
INFO REPORTING. DUNAs could face penalties if they fail to comply with tax info reporting, including:
- Form 5472 & Form 1120 Schedule G: Certain large direct or indirect holders.
- Form 1099: Dividends and comp to US people.
- Form 1042: Dividends to non-US people.
TAX ADMINISTRATION. DUNAs have "administrators" to perform ministerial tasks, which a16z crypto's commentary suggests include "signing tax returns." Tax admin isn't easily automatable, so it's worth considering what that might mean in practice.
Corps hire tax accountants, respond to their questions, take positions when law is unclear (as it often is in crypto), interact with IRS auditors, determine whether and how to settle disputes with the IRS, & hire service providers to do withholding and info reporting. While we don't yet have a DUNA in the wild to see how it would all work in practice, it's likely many DUNAs will find it necessary to vest their administrators with significant discretion over those actions, plus access to the DAO treasury to carry them out.
REACTIONS. a16z crypto has indicated that, "where appropriate," it will require its portcos to commit to use DUNAs to decentralize governance. It's worth asking when a DUNA would be "appropriate." Below are my thoughts, which I caution are purely tax-driven.
tldr: A DUNA may be appropriate where tokenholders:
(a) Conduct substantial business activities (not including voting) on behalf of the DAO from within the United States; OR
(b) Expect US law, not code, to govern disputes.
Let me explain...
US tax law views a JV for profit as an entity, even absent a legal entity. While some DAOs (eg, big protocol DAOs) might be viewed as parameter selectors, not JVs, that argument is probably lost if someone is conducting business on behalf of the DAO.
Once you've conceded a DAO is an entity, you should ask if the entity is in a "US trade or business" (USTB). A USTB means US-based profit-making activity regularly conducted, but does not include voting.
Generally, non-US corps are subject to 44.7% US tax on net income connected to a USTB. By contrast, as discussed above, US corps are subject to only 21%, with an add'l 30% on the remainder imposed thru withholding only on foreigners and only when the corp pays dividends.
Thus, if a DAO is a tax entity and most of its income is connected to a USTB, a US corp generally yields a lower tax bill than a non-US corp. In any event, if members want US law (not foreign law or code) to govern, the deemed tax entity generally defaults to domestic.
There might still be good tax reasons not to use a DUNA, eg:
- If only some income is connected to a USTB, a foreign corp (eg, a Cayman foundation DAO wrapper) would be subject to US tax on only that income, not all income. (It can also block USTB income in a US sub.)
- A DUNA might have to collect tax forms to comply with info reporting. Foreign corps don't have the same info reporting requirements. DAOs might decide it's worth the extra tax to avoid doxxing members.
- If members are US and doxxed, a tax partnership might make more sense. Partnerships don't pay entity-level tax, but partners are taxed on the DAO's income on a pass-through basis. Members might prefer a more traditional entity like an LLC to house a tax partnership.
In short, while DUNAs represent another tool for the web3 structurer's toolkit, they are no panacea and could require significant centralization and KYC. Founders should consult with tax and regulatory counsel before adopting - or committing to adopt - any legal structure.
BORGs. On Monday, MetaLex Labs (i.e., Gabe Shapiro and Alex Golubitsky) published a white paper on BORGs (https://metalex.substack.com/p/the-metalex-whitepaper?r=a1p1d&utm_campaign=post&utm_medium=web&triedRedirect=true). While the legal community (including Miles Jennings) appears to be viewing BORGs and DUNAs as an either-or proposition, I think that's myopic.
WTF IS A BORG? The white paper is largely a cypherpunk manifesto, with lofty assertions like: "MetaLeX’s core innovation will be to separate law from the State." While that's the sort of thing that got me into crypto, it's not very...actionable. At least not today.
What I find more exciting for the near term is MetaLex's intent to build off-the-shelf parameterized entity-multisig hybrids - legally wrapped multisigs whose rules maximally contour what the signers can do while limiting their liability.
For a sense of how this might work, read the grants BORG case study in Part 1A, which envisions smart contract-based rules that strictly control the number and amount of grants that can be made in each epoch while the entity's org docs specify permissible grant purposes.
Using BORGs might obviate the need to wrap some DAOs in legal entities at all. Instead, any business activities could be cordoned off in their own entities. For US tax purposes, the DAO is either a mere voting arrangement (a "tax nothing") or a foreign corp with no USTB.
Curve DAO offers an example. As I've previously explained, if an entity, Curve is likely a foreign corp with no USTB (see https://www.friedfrank.com/uploads/siteFiles/Publications/Decentralized%20Autonomous%20Organizations%20_%20Decentralized%20Law.pdf).
But Curve does have an emergency multisig that can kill pools, subject to a Curve DAO override. Ideally:
- The multisig is standardized so a front end can describe its parameters in plain English.
- The parameters are narrowly tailored.
- The multisig is legally wrapped to protect the signers from liability and to enable a supervisor to remove signers who go rogue.
But in some situations, there's just no separating a putative DAO from an active business. There, it might be necessary to legally wrap the DAO. Here's where BORGs and DUNAs might intersect...
First, instead of handing administrators the keys to a wallet and subjecting them only to a legal obligation not to rug, a DAO could pair that obligation with carefully coded multisigs - one for accountant fees, one for taxes, etc., and subject excess expenditures to a vote.
Second, maybe it makes sense to separate the tax admin function from the DUNA itself. A DAO-adjacent BORG could be created whose sole purpose is to hire and interface with accountants. Service expenditures in excess of some amount each year could require a DAO vote.
CONCLUSION. The legal landscape for DAOs is nascent. You should be skeptical of anyone who suggests all DAOs are treated alike or one legal structure fits all. We're lucky to have people like Miles Jennings, David Kerr, Gabe Shapiro, and Alex Golubitsky moving the conversation forward.