This document is still in draft WIP mode. This is not tax advice.
The IRS currently treats all staking rewards as income when received, but this stance poses an existential threat to crypto.
Separation of token from consensus is not a new concept for crypto. The first and still the top blockchain project, Bitcoin, utilizes proof-of-work for its consensus, whereas the BTC bitcoin tokens are not used directly for mining. This is a feature of its security, because energy is abundant (especially equally from the sun), and efficient Bitcoin mining is a highly technical endeavor, whereas the purpose of BTC is for payments among the general population who are luddites in comparison to Bitcoin miners.
The Cosmos whitepaper came out in 2017 and became a sensation because it was the first to provably solve the "nothing at stake" problem of prior proof-of-stake solutions; Cosmos made proof-of-stake a theoretically sound alternative to proof-of-work. We now have a revived Byzantine Fault Tolerance research and implementation industry and about ten thousand unique blockchains partially thanks to the innovations of Cosmos and Tendermint.
In Cosmos, and in its fork AtomOne, the staking token (ATOM and ATONE respectively) are designed to be largely staked, targetting a massive staking ratio of 2/3, or 67%. It is the first provably secure BFT proof-of-stake system, and it also targets the highest staking ratio. And the ATOM token was since the very beggining, pitched as a "staking token" separate in class than those of other utility tokens. The Cosmos whitepaper describes a dual-token model where the ATOM token is the staking token, and other tokens in the Cosmos hub can function as the secondary fee token. The AtomOne blockchain forks Cosmos and provides a specific secondary fee token PHOTON to complement the staking token ATONE.
Ethereum converted from proof-of-work to proof-of-stake in the year XXX and doesn't target a specific staking ratio, but uses a formula to determine the staking inflation rate to incentivize staking. Its staking ratio has settled to about 27%. Unlike Bitcoin or Cosmos, Ethereum uses the same token for both fee payments as well as staking for consensus.
Cosmos's dual token model has real benefits over Ethereum's single token model. The staking token distribution can be different than the fee token distribution, allowing different expectations and incentives. For example, the slashing function of proof-of-stake (for validators that misbehave) only apply to stakers, whereas non-stakers who only hold fee tokens need not be concerned about the risks of slashing. For the most innovation in proof-of-stake dual-token design, see the AtomOne constitution which provides limitations of inflation for the fee token holder with the PHOTON fee token.
Another benefit of the dual-token model is the preservation of sovereignty of the technical core creators and supporters. In a single-token proof of stake blockchain like Ethereum there is a well defined cost to taking over the consensus security of the chain and having complete control over the blockchain. There is a fixed cost to any whale capital investor who accumulates enough voting stake to take control of the chain and even attack it. In contrast in a dual token model the blockchain can only lose control to capital investors (potentially hostile) if the stakers choose to sell their stake. Thus for example if TexasChain created by the State of Texas wanted to issue gold-backed tokens to foreign investors without losing control of TexasChain to foreign powers, it must employ a dual-token model.
Another analogy to convince the reader about the merits of the dual token model can be found in the design of shareholder companies of general partners. Such a company would naturally want to rely on the expertise and insights of the general partners for the execution of its business, which for example might involve the sale of computers like Apple does, or Beanie Babies for a more fungible example. Whereas if the company were controlled by the customers, it probably would not have been as successful. “If I had asked people what they wanted, they would have said faster horses.” (commonly attributed to Ford).
Whether the proof of stake chain employes a single token model or dual token model for its staking security, there is a spectrum of them where the distinguishing variable is the ratio of total tokens staked.
If the staking ratio is very high, the staking economics acts more like a penalty for those who don't stake. When the staking ratio is the highest at 100%, there is in effect no income, because all inflation are distributed pro-rata to the stakers with no change in distribution before and after the staking reward.
If the staking ratio is very low, the staking economics acts more like a reward for those who do stake. When the staking ratio is for example at a low 0.1%, it is in effect income, because there is a tax upon the non-stakers and the proceeds are distributed to the stakers, reflected as change in the distribution.
Furthermore, single token proof-of-stake chains can only tolerate a low inflation rate (because the targetted class of token users are not stakers), whereas dual-token proof-of-stake chains target a higher staking rate for its security (sovereignty) and to incentivize the distinction into being -- that is, to dissuade the lay crypto user from participating in staking (with risk of slashing).
In the Tezos appeal case [XXX link to] the court published a preliminary response that denies Tezos stakers favorable tax treatment (and instead that earned staking inflation should be counted as income upon the receipt, rather than the sale of those tokens). The basis of this is that the IRS denies that the Tezos stakers have "created" the newly minted staking tokens. The IRS did not state why, merely denying the plaintiff's claims.
This preliminary conclusion might be acceptable if the staking ratio for a proof-of-stake blockchain is low, because most of the staking inflation reward distribution corresponds to a transfer from non-stakers to stakers, and vice versa when the staking ratio is high, less of the staking inflation reward distribution corresponds to such a transfer. For clarification, if staking rewards were like farmers' chickens laying baby chicks, a low staking ratio is like chicken farmers globally paying taxes in the form of hatched chicks to the few staking farmers for the service of staking; where a high staking ratio means most of the hated chicks distributed were paid back to the farmer that actually created them. To see how the math is skewed unreasonably against the high staking ratio farmer, see the blog post "Atom Must Not Be Money" which shows that for a 30% annual staking reward, 8.33% corresponds to a transfer from non-stakers, whereas a whopping (30 - 8.33) = 21.67% corresponds to self creation, which is 72.22% of the total rewards.
This makes the IRS's position on staking reward taxation not only unfair for blockchains that have or target high staking ratios; it makes them ecomically infeasible. Even in a state of stasis with no other economic activity, the IRS through taxes would end up taking the vast majority of a Cosmso or AtomOne staker's net ownership of staking tokens in a matter of decades. And given that dual-token proof of stake chains target a high inflation rate, the IRS is effectively enforcing illegal sanctions on a whole class of sovereign crypto and forcing all crypto to be vulnerable to takeover from capital investors. Not even going public in the seurities markets enforces such a thing. The IRS should not use its taxing authority to stifle home grown innovation in the crypto blockchain space.
Whether or not staking inflation rewards are considered taxable upon reward as income or taxable upon later sale under capital gains/loss treatment, we recommend that the IRS create a special class of crypto tokens called "staking tokens" that is only taxed upon sale of the inflated staking tokens earned by stakers; given that the staking ratio targetted is greater than 50%, and during the periods where the actual staking ratio is greater than 50%. This will allow secure proof-of-stake blockchains to exist within the frame of US tax law, ensuring that the US can continue to innovate in secure proof-of-stake chains and participate in their staking security.
-
internally in implementation when somebody stakes, atoms are converted to a non transferable token called something else. This value is such that it does not change in quantity after staking even after years, unless slashing is applied. For example, 1M atoms staked becomes 1K shares (say). After years with no slashing one still has 1K shares, and only when unbonding does the shares become say, 2M atoms. If we considered these internal representations the tax result would be I believe more favorable — at least capital gains, and only upon unbonding. But this still isn’t fair.
-
staking token inflation rewards should be distinguished from tx priority fee distribution rewards. The latter really is income. This recommendation only pertains to the staking rewards that are not derived from transaction fees.
-
even without IRS clarification as recommended here, developers of proof-of-stake blockchains can make a trivial change to make the total amount of staked tokens remain the same, while reducing the amount of non-staked tokens in non-staked accounts by introducing a global denominator. This mental exercise and real potential workaround demonstrates the illogical nature of the IRS's current position that all staking tokens are to be treated as income.
-
see also https://allinbits.com/blog/nwv-to-prop-848-atom-must-not-be-money/
-- All in Bits, Inc