01
Market Makers
Post two-sided quotes around mid. Lean against inventory, refuse the one-side hard cap, rebalance aggressively as their book grows.
A real-time equities and options exchange powered by
autonomous agents, dynamic liquidity, and realistic market microstructure.
Every ticker is traded by an independent population of synthetic agents. Their net flow each tick feeds back into the price step, so the market is order-driven — not just noise.
Post two-sided quotes around mid. Lean against inventory, refuse the one-side hard cap, rebalance aggressively as their book grows.
Uninformed buy / sell flips. Provide the baseline liquidity floor and the constant background hum of order flow.
Momentum followers. Buy when recent return is positive, sell when negative — magnitude scales with the size of the move.
Fade deviations from a personal rolling mean. Threshold-gated so they only fire on real extensions, not chop.
Anchor to a private fair-value estimate that drifts toward mid. Heterogeneous priors produce persistent disagreement.
Buy the dip past their drawdown threshold, sell the rip past their rally threshold. Re-anchor windows after firing.
Each tick is one trading second. The mid-price evolves through a stochastic differential equation with three feedback channels — drift, endogenous volatility, and order-flow impact — plus a Poisson jump term.
Volatility mean-reverts to a baseline σ0, but absorbs shocks proportional to flow magnitude, market-maker stress, and illiquidity. The leverage term λ amplifies sell-side impact, reproducing the empirical down-move asymmetry.
Aggregate agent order flow moves price inversely to instantaneous depth Lt. Liquidity heals toward L0 at rate κL, but is suppressed exponentially by flow magnitude, volatility, and stress — so the book thins precisely when it is needed most.
Discontinuous price shocks arrive at a Poisson-clocked rate λJ with normally-distributed sizes. Captures the fat-tail behaviour — earnings prints, macro surprises, headline risk — that pure Brownian motion misses.
The asset universe consists of approximately 60 equities and ETFs, all traded simultaneously in a fully synchronised environment. Prices, volumes, and cross-asset correlations evolve in real time on a single shared tick.
The underlying assets are calibrated from empirical market data — drift, volatility, skewness, kurtosis, jump intensity, rolling volatility distributions per symbol, and sector-based correlation structures — ensuring realistic statistical behaviour across instruments.