TickTrader: Simulated α on the Tape

Desk parameters

Observable information flow

The desk is an open measurement channel. Quotes, orders, and notebook text share one refresh surface. Attention collapses onto a finite observation vector; the fill kernel never sees wallet balances or settlement rails.

(1) o_t = (q_t, g_t, ρ_t, n_t, E_t, Π_t)

Here q_t is the quote map on S, g_t the gap field, ρ_t round-trip cost, n_t the notebook state (versioned notes), E_t paper equity, Π_t cumulative paper PnL. o_t is readable; it does not authorize on-chain transfer.

Notebook entries are an annotated path on the same filtration as the tape. A note such as “no arbitrage edge on the tape · v5” is an element of n_t, not a cash flow.

Quote discovery and gap measure

For each name s ∈ S let q_s be the on-chain mid (or last) and bid_s a reference bid on the same surface. Basis-point gap:

(2) g_s = ((q_s − bid_s) / bid_s) · 10⁴ (bp)

Round-trip friction ρ_s aggregates advertised crossing cost on the displayed book. Discovery is the joint map (q, g, ρ); it does not imply inventory capacity.

(3) D_t = (q_s, g_s, ρ_s)_{s ∈ S}

The desk’s strategy kernel π reads D_t and emits intended orders. Intent stays inside the paper measure until F applies.

Simulated fill operator

An intent is a tuple u = (s, side, size). The fill operator F maps intent and local quote into a paper fill price; chain cash remains zero:

(4) x = F(u; q_s), cash_chain = 0

Slippage in the paper world can be written as a signed displacement from quote:

(5) x = q_s · (1 + ε(side, ρ_s))

ε is a desk-internal model of simulated impact. No transfer event is emitted. Positions update only in paper state.

(6) θ_{t+} = θ_t + Δθ(u, x)

Paper equity and α

Mark paper holdings θ against quote map q. Equity:

(7) E(t) = c_paper(t) + Σ_{s ∈ S} θ_s(t) · q_s(t)

Cumulative paper PnL relative to an initial seed equity E(0):

(8) Π(t) = E(t) − E(0) + Σ_{k ≤ t} R_k

R_k denotes realized paper PnL on closed paper trades. Instantaneous paper α is a path functional of Π, not a claim on chain balances:

(9) α̂(t) = Π(t) / E(0)

Public panels expose E, Π, closed-trade count, and win rate when defined. Missing win rate means zero closed paper trades in the current window.

Decoupling from real capital

Let P be the paper state (θ, E, Π, n) and K any real capital measure on-chain. The published operating constraint:

(10) ∂K / ∂Π = 0, ∂P / ∂K = 0

Simulated α does not mint settlement; settlement does not resize paper inventory. Coupling exists only in the UI: one page renders o_t for both discovery and paper accounting.

Public surface

• Desk: https://www.ticktraderai.io/

Eqs. (1)–(10) are a formal reading of the public desk. Live E, Π, and quote fields are instantaneous; no backtest series is asserted here. Paper trading only—simulated fills on live on-chain quotes, no real funds.