Wetware Docs

A systematic quant desk, issued as a token, that funds itself from its own trading and pays its profit back to holders — every hour, entirely on-chain.

α.What Wetware is

Wetware is a systematic trading desk expressed as a single token. There is no fund manager to call and no subscription form to sign. Holding the token is the position; the desk trades a rules-only book on Lighter and returns its profit to the bearer each epoch.

β.Why it matters

Alpha is usually a story told after the fact. Wetware inverts that: the exposure is live, the fills are on-chain, and the payout is mechanical. You carry the actual book, not a narrative about it, and the thing pays for itself instead of relying on a raise or a treasury you have to trust.

γ.The self-funding loop

Six steps run every hour with no human in the path. Fees the token generates become the margin the desk trades, and the desk's profit flows back to the same holders who generated the fees.

  1. 1.
    Fees accrue

    A fixed 4% of every token buy or sell is skimmed at the source and pooled in the fee contract.

  2. 2.
    Harvest

    On each hourly epoch a keeper harvests the accrued fees from the contract.

  3. 3.
    Swap

    Harvested fees are swapped to USDG, the desk’s settlement asset.

  4. 4.
    Deposit

    USDG is deposited into the trading account on Lighter as working margin.

  5. 5.
    Size

    The engine sizes new exposure inverse-vol against the fresh margin, respecting caps.

  6. 6.
    Distribute

    Realised profit above the high-water is apportioned pro-rata and paid to holders.

δ.The strategy

The core book is a trend ensemble over three lookbacks — 168h, 720h and 2160h — blended into a single target weight per market. A funding-carry tilt of α = 0.25 nudges weights toward positive carry. The blended signal is:

w_raw = Σ_k β_k · sign(EMA_fast_k − EMA_slow_k) · |z_k|
w_i   = w_raw · (1 + α · carry_i),   α = 0.25
        where carry_i = funding_i · side_i

ε.Sizing

Positions are sized inverse to their own volatility toward a 30% annualised portfolio vol target. Per-market weight is capped at 3× the target, gross at 5×, and no single market may exceed 10× its inverse-vol baseline.

size_i = (target_vol / vol_i) · w_i
target_vol = 30%   ·   caps: 3× target / 5× gross / 10× per-market

ζ.Funding overlay

On top of sizing, a funding overlay leans the book toward markets that pay to hold and away from markets that cost to hold. It never flips a trend signal — it only scales conviction within the bounds the caps allow.

η.The open-interest edge

Divergence between open-interest momentum and price momentum is the desk's cleanest edge on Lighter's deepest books. Backtested, out-of-sample:

Sharpe
4.01
APR
+93%
Max DD
−3.7%
Vol
23.5%

θ.Distributions

Each hourly epoch closes with a cumulative Merkle root committed on-chain. Holders claim against the root, or a gasless relay pushes payouts for them in USDG. Because the root is cumulative, a missed epoch never loses funds — the next successful claim settles everything owed.

ι.Settlement without flattening the book

Profit is measured against a high-water mark, so the desk distributes realised gains above the prior peak without being forced to close healthy positions. The book keeps compounding; only the surplus leaves.

κ.Verify it on-chain

Every fill, harvest and distribution is a public transaction. The live book on the home page is the same account, so what you see streaming is what settles. Nothing is claimed that the chain does not show.

λ.Risk & controls

Leverage is bounded by the sizing caps, a −5.0% daily drawdown trips the kill-switch and flattens risk, and a stale holders-poster raises a health alarm rather than paying against uncertain state. Systematic does not mean riskless — the token can lose value, and nothing here is investment advice.

μ.Addresses

Token ($WETWARE)coming soon
Fee collectorcoming soon
Desk / trading accountcoming soon
Distributor (Merkle)coming soon

Wetware is experimental software and a volatile on-chain asset. Backtested figures are hypothetical, out-of-sample where noted, and not a promise of future results. Nothing in these docs is investment advice.