The document runs four pages. Almost nobody will read past the third. That is where the strike price lives.
Forward Industries — the Nasdaq-listed company that recapitalized itself in 2025 into a Solana treasury vehicle backed by Galaxy Digital, Jump Crypto and Multicoin Capital — has mounted a board challenge at SkyAI, the AI infrastructure firm whose listed equity has become one of the more reflexive instruments in the crypto-adjacent complex. The stated dispute is the company's equity incentive plan. The actual dispute is who gets to define the denominator.
Proxy fights are normally about control. This one is about measurement. When a company's enterprise value leans on a token, an unvested option pool stops behaving like a compensation cost and starts behaving like a derivative written on a narrative. The equity plan is not a compensation document; it is a control document. Whoever defines the plan defines when the story gets repriced, and at what strike.
Strip the language of the proxy down to the mechanics and the shape is familiar. A shareholder of record — here, Forward Industries — has taken a position against a slate and against a proposal. The proposal is an equity incentive plan: a pool of shares reserved for employees, directors and advisers, usually with an evergreen provision that automatically replenishes the reserve each year, usually with vesting measured in years, and usually with an exercise price derived from a trailing average of the share price.
Standing alone, that is the most routine item on any annual meeting ballot. Shareholders approve pools of 10 to 15 percent of fully diluted shares all day long. Institutions vote for them because the alternative — paying cash to people who could be paid in equity — is worse for the income statement and worse for alignment. Approval rates for uncontested plans routinely clear 90 percent.
Two things make this one different: the identity of the counterparty, and the composition of the issuer.
Forward Industries is no longer a design and manufacturing company. Since its recapitalization it has become a listed wrapper around a token treasury, a structure in which equity is a leveraged claim on a balance sheet of digital assets, and in which the share price is therefore a function of two variables — the token price, and the premium the market assigns to the wrapper. That premium, the multiple over net asset value, is the entire business model. A treasury vehicle trading at a discount to its own holdings has no reason to exist. It cannot raise accretive capital. It cannot issue equity to buy more of the asset. It becomes a slower, more expensive, more heavily regulated version of simply holding the token.
SkyAI sits on the other side of that boundary. Its equity story is AI infrastructure: compute, orchestration, and increasingly the payment rails on which autonomous agents transact. It is the kind of company that can plausibly justify a token, a treasury, and a listed equity at the same time, because its product touches both machine-to-machine settlement and conventional enterprise contracts.
That dual identity is what makes an equity plan a live wire. If the operating thesis is that agents will transact at high frequency and low value — my own 2026 modeling assumed a 300 percent increase in transaction count alongside a 50 percent decline in average ticket size — then the long-run economics of the business depend on throughput, not on price appreciation. An incentive plan keyed to the share price, in that world, pays people for the wrong variable.
I have spent enough time inside custody reviews to be wary of wrappers generally. When I built the custodial due diligence framework for a Miami fund's 2024 spot ETF allocation, the exercise was not about price at all: it was about who holds the asset, under what legal conditions it can move, and what happens to the claim if the entity in the middle stops functioning. The same discipline applies to a listed treasury vehicle. The shares are not the asset. They are a claim on an entity that holds the asset, and the claim is only as good as the governance surrounding it.
There is nothing illegal about any of this. That is the point. The tension between a shareholder's influence and a board's discretion is not a bug in the system; it is the system. The interesting question is which of the two parties is actually exposed to the risk it claims to be protecting against.
Start with the accounting, because every crypto-adjacent issuer runs three ledgers at once and they rarely agree.
The first is the GAAP ledger: shares outstanding, option expense, treasury holdings marked to market. The second is the token ledger: circulating supply, unlock schedule, emissions. The third is the incentive ledger: who gets paid, in what instrument, on what schedule, against which benchmark.
Equity incentive plans are where the first and third ledgers are stitched together, and the stitch is almost always a trailing average of the share price. Token vesting schedules are where the second and third are stitched together, and that stitch is almost always a calendar. History does not repeat; it rhymes in code. The four-year token cliff and the four-year employee vest are the same instrument wearing different regulators.
When I audited Paragon Coin's ERC-20 contract in 2017 — 45,000 lines of Solidity, read by hand — the vulnerability I found was not in the arithmetic. The transfer function computed exactly what it was written to compute. The flaw was an assumption: that nobody would ever call it with an input the author had not imagined. An integer overflow is not a math error. It is a failure of imagination wearing the costume of math.
Equity plans fail the same way. The schedule is sound. The vesting curve is sound. The share reserve calculation is sound. What is not sound is the assumption embedded in the strike: that the price at the reference date is a neutral observation. It is not. It is a chosen input, selected by a compensation committee, at a moment of the board's choosing, from a window the board can influence. The math was sound; the trust was the variable.
Proxy advisors publish their methodology, and the methodology is mechanical: share reserve as a percentage of outstanding, three-year burn rate, presence of an evergreen, presence of performance vesting. Nothing in that rubric asks whether the incentive is tied to the right variable. It cannot ask, because the rubric is built to be applied across thousands of issuers simultaneously.
Run the arithmetic on a representative pool, because the headline number misleads in both directions. Suppose a company reserves 12 percent of fully diluted shares for incentives with a 4 percent annual evergreen. Over five years, absent buybacks or a rising share price, roughly 30 percent of the original float migrates to employees and advisers. That sounds catastrophic and usually is not, because the company is expected to grow into the pool — the denominator expands, the percentage holds roughly stable.
The mechanism only works if the share price rises for reasons unrelated to the pool itself. If the stock climbs because the narrative improves, the pool becomes a transfer from existing holders to insiders that is invisible in the period it is granted and enormous by the time it is exercised. If the stock falls, the pool reprices in practice — through refresh grants, exchanges, and new plans — and the transfer happens anyway.
The asymmetry is the point. Efficiency is the enemy of resilience. A pool calibrated for a growth environment becomes a permanent claim in a flat one, and a flat market is exactly what we have. Chop does not reward optimism. It rewards positioning.
Now layer in the token.
A treasury vehicle's equity is a levered claim on a volatile asset. When the wrapper trades at a premium, the equity is a claim on the premium as well as the asset. That premium is not a floor. Liquidity is not a floor; it is a horizon. It moves toward you or away from you depending on whether capital is entering or exiting the structure, and the market that sets it is not the market that sets the token's spot price.
We are watching the decay of leverage in this corner of the complex — not the violent deleveraging of 2022, but the slow compression of premium multiples as capital rotates out of wrappers and back into the underlying. My 2020 framework for DeFi yields made the same point in a different venue: when the return is funded by emissions rather than revenue, the yield is a countdown, not an income stream. In that regime the equity of a treasury company stops being a growth instrument and becomes a duration instrument. Its value depends on how long the premium persists.
Which is why a governance fight over an equity plan is not a side quest. If you are a treasury vehicle whose currency is your own premium, the ability to influence the equity structure of an adjacent company is an optionality play. It costs a proxy solicitation. It potentially buys a seat at the table where token strategy, treasury policy and the incentive plan are all set.
I added a standing chapter called Regulatory Arbitrage Risk to every macro outlook I have written since 2022, after deconstructing TerraUSD's collapse in a fifty-page paper that the SEC later cited in enforcement actions. The lesson of that paper was not that algorithmic stablecoins were fraudulent. It was that the structure's equilibrium depended on a jurisdictional gap — conduct legal in one venue and not another, plus a leverage market that could be tapped without a balance sheet.
The same gap exists here, and it is wider than most equity analysts assume. Token compensation sits outside the disclosure regime that governs equity compensation. A grant of tokens to a contributor does not trigger a Form 4. It does not appear in a proxy table. It does not require a shareholder vote. The equity plan — the visible instrument — carries the full disclosure load.
So when a shareholder challenges an equity plan at a company that also issues tokens, the fight is structurally incomplete. They are litigating the ledger that can be seen while the ledger that moves faster goes unmentioned.
Here is the measurement that should matter and almost certainly does not appear in the proxy.
In the agent economy I modeled last year, the binding constraint on infrastructure companies is neither capital nor compute. It is unit cost per machine transaction. A network whose settlement cost is a fraction of a cent can absorb millions of micro-payments. A network whose cost is five cents cannot absorb anything below a dollar. Every AI infrastructure firm touching payments is optimizing against that threshold whether it admits it or not.
If a plan paid out against cost-per-agent-transaction, or against gross margin on machine payments, the incentive would align with the actual business. If it pays out against share price, it aligns with the market's current narrative about the business. Those two things have been correlated for a while. Correlation is the smoke; divergence is the fire. They will not stay correlated forever, and when they separate, a price-linked plan will have paid for a story the operators cannot reproduce.
The obvious reading of this challenge is that a shareholder is doing its job — standing between a board and a dilution machine.
The less obvious reading is that the challenger is not a passive guardian. It is an operating company with its own capital structure, its own premium to defend, and its own reasons to want influence over an AI-adjacent issuer. Proxy fights are expensive. Boards do not pick them for sport. When a treasury vehicle whose currency is its own listed equity challenges another company's board, the plausible motives are strategic: access to compute, access to a payment rail, access to a counterparty, or leverage in a transaction that has not been announced.
There is also a harder version of the argument, the one that makes people uncomfortable.
An equity pool that is too small is a company that cannot hire. In a market where the scarce input is human capital, and where the most mobile engineers can be paid in tokens by a foundation, in stablecoins by a protocol, or in equity by a startup, a board that shrinks its pool to satisfy a shareholder is choosing efficiency over resilience. The shareholder wins the vote and loses the team. Then the narrative dies — not because the ledger bled in the accounting sense, but because nobody was left to write the next entry in it.
And both sides are campaigning to an audience that largely will not show up. Retail holders of crypto-adjacent equity vote at single-digit rates. The decisive ballots belong to index funds, quant funds and proxy advisors. A fight framed as shareholders versus management is, mechanically, two institutions arguing over a recommendation.
That is not a scandal. It is worth stating plainly, because it means the outcome of this challenge will be determined by a governance methodology, not by a mandate.
Three signals will tell you how this resolves, and none of them is the headline.
First, the reference date for the strike price. If the exercise price is benchmarked to a trailing window that includes the current consolidation, the pool is being struck near a local low, and the value transfer is larger than the headline percentage implies. If the board moves to a forward-looking or performance-gated structure, the fight was worth having.
Second, the evergreen clause. A fixed pool is a one-time negotiation. An evergreen is a perpetual claim that renews without a vote. If the evergreen survives, the challenge failed regardless of who sits on the board.
Third, whether the performance metrics are denominated in a token the company can influence. A plan gated on token price, at an issuer that can shape token price, is not an incentive. It is an authorization.
The vote itself is a lagging indicator. By the time proxies are counted, the structure has already been set. What matters is whether equity holders of a token-adjacent issuer still know what their shares are a claim on: the cash flows of a business, or the residual of a narrative. In a sideways market, that question does not get answered. It just gets more expensive to ask.