The conflict of interest, removed
Ask a prop trader why they distrust prop firms and the answer comes back the same way every time: the firm pays out of its own pocket. Your profit is its expense. Everything traders complain about in this industry follows from that one fact, and this protocol is built to invert it.
The problem, stated plainly
A conventional prop firm has two revenue lines and one cost line. Evaluation fees are the revenue. Funded payouts are the cost. The firm is the counterparty to every account it funds, so a trader who passes and withdraws is, on the firm's own books, a loss.
Nobody has to be dishonest for that to bite. It bites through ordinary business decisions. Rules written tight enough that most accounts breach one. A support queue that slows down as a payout gets larger. A risk-management clause invoked when a trader gets too good. One more discretionary review step in the withdrawal process than it needs. Each is defensible on its own, and together they are a firm behaving rationally toward the fact that its best customer is its largest liability.
The industry's answer has been to promise it doesn't happen: review scores, payout screenshots, a public "we've paid $X million" counter. All of it is a claim, made by the one party with a reason to make it.
What this protocol does instead
A firm operator here is paid by the payout, out of the payout, at the moment it settles.
When a funded trader withdraws, the firm's treasury spends SOL buying its own token on its own and delivers those tokens to the trader (Funded accounts & payouts covers the full four-layer waterfall). The trader's locked profit split, 80% at a healthy firm, goes to the trader. The remaining 20% is the stakeholder share, and the operator's cut of a payout lives entirely inside it.
No payout, no stakeholder share. No stakeholder share, no operator cut. The channel does not exist until a trader wins.
Where the money actually goes
Every Tier-1 payout distributes the stakeholder share six ways. These are the on-chain defaults, read from the compiled program:
| Destination | Share of the stakeholder share |
|---|---|
| (SOL, carved before the buy) | 40% |
| Firm operator () | 30% |
| Buyback and burn ($FIRMA) | 10% |
| Firm treasury reserve ($FIRMA) | 10% |
| pool ($FIRMA) | 5% |
| pool ($FIRMA) | 5% |
Worked through on a $10,000 trader profit at an 80/20 split:
| Leg | Amount |
|---|---|
| Trader | $8,000 |
| Universal Treasury Pool | $800 |
| Firm operator | $600 |
| Burned | $200 |
| Firm treasury reserve | $200 |
| No-risk stakers | $100 |
| Backstop stakers | $100 |
The operator earns $600 on a trader's $10,000 win, worth 6% of the gross profit, paid in the firm's own token rather than in cash. Had that trader failed instead, the number is zero.
Two details in that table matter more than they look. The operator's share is denominated in $FIRMA, so an operator cannot take a payout cut without taking a position in their own firm. And the operator's leg is three times the burn leg and three times the treasury reserve, which makes paying traders the single largest use of a stakeholder share.
Four more loops pointing the same direction
The payout is a buy. The SOL that funds a payout buys $FIRMA on the firm's own curve, so a payout is buy pressure on the token the operator holds 10% of at launch. Bigger payouts are bigger buys.
The payout burns supply. A tenth of every stakeholder share is burned outright against a fixed 1,000,000,000 supply whose mint authority was revoked at deploy. Supply falls as the firm pays. Nothing else on the protocol burns $FIRMA.
Payouts raise the score that sells evaluations. The public weights payout reliability at 25 and win attainability at 20, out of 100. A firm where more traders genuinely pass and get paid promptly scores higher, and the score sits on the firm's storefront where buyers read it. An operator cannot write this number (How the happiness score works).
A well-run firm gets a raise. As a firm's treasury fills, the operator's share of every evaluation fee climbs automatically, from a 6% to 14% baseline by tier up to 13.6% to 20% at the top zone. The thresholds are platform constants that no operator can redraw (How the treasury health split works).
What an operator can and can't change
The six ratios above are adjustable on-chain by the firm owner, inside bands the protocol fixes: the operator's own share can move between 10% and 50%, the burn never below 3%, staking between 5% and 30%, the treasury reserve up to 20%. Every firm starts on the same defaults, 30% owner and 10% burn among them, and most never move.
One practical caveat, so you don't go hunting for a switch that isn't there: this is an on-chain instruction the owner's wallet signs, and the operator console does not surface it yet. Today it takes a direct call against the program rather than a button.
Note what that dial does and doesn't touch. It redistributes within the stakeholder share only. A trader's profit split is locked into their account's on-chain rules at purchase and can never be raised or lowered by a later configuration change, so an operator who dials their own share from 30% up to 50% takes it from the burn, the stakers, and the reserve, never from the trader. And the higher they set it, the more of their income depends on payouts happening at all.
Where the alignment stops
Everything above is a mechanism, so its limits are mechanical too. Stated plainly rather than glossed:
Evaluation fees are still revenue regardless of outcome. An operator earns 6% to 20% of every evaluation fee whether that account passes or breaches. The honest claim is narrower than "the firm only makes money when you win": the firm makes money both ways, and the payout channel is the only one that grows with your success rather than shrinking.
A firm on emergency liquidity earns its operator nothing. When a payout draws on the backstop pool or the Universal Treasury Pool instead of the firm's own treasury, the stakeholder split doesn't run at all. The full drawn amount goes to the trader, and the operator's cut is zero. The alignment inverts under stress, in the trader's favor.
A stressed firm pays a smaller split. The risk engine cuts the reference split to 70% at Warning and 60% at Critical. That is solvency management rather than appetite for trader failure, and it applies to accounts sold during the stressed window, never retroactively to a split already locked (How the risk engine thinks).
The execution engine is a pure simulation. No external counterparty takes the other side of your trade. Alignment here is economic, built into how a payout is distributed, and it is not a claim that the engine itself is rooting for you. What keeps the engine honest is a separate mechanism: seeded, reconstructible fills and an internal fidelity audit that found no directional bias on verified templates (How fills are actually fair).
Non-verified templates carry extra slippage terms. On Standard templates, firm-level slippage scaling and the risk engine's slippage multiplier can apply on top of the base fill. Both are off on verified templates, which is why the no-bias finding above is stated for verified mode. Both scale with order size and firm-wide stress rather than with who the trader is, and which mode you're buying is disclosed before you pay.
None of this has had a third-party audit yet. The programs are internally red-teamed and devnet-proven, and no independent security firm has reviewed them (Security & audit status).
Check it yourself
The split above is the default value of a struct in a program you can read. The source is published to a public verified-build mirror, the payout instruction distributes the stakeholder share in a single transaction, and every payout a firm has ever made is a permanent transaction on Solana that you can open and add up without asking the firm anything.
That's the part worth keeping. In this industry a firm's payout record is a marketing claim. Here it's an address. Verify it yourself covers reading protocol state directly.