avellaneda-stoikov

HFTFramework utilized for research on " A reinforcement learning approach to improve the performance of the Avellaneda-Stoikov market-making algorithm "

A Rust library implementing quantitative market making strategies, starting with the Avellaneda-Stoikov model. This library provides the mathematical foundations and domain models necessary for building automated market making systems for financial markets.

Implemented the Avellaneda-Stoikov market-making strategy in an automated trading algorithm. Completed as part of the Optiver Ready Trader Go competition.

Repository for market making ideas

Execution and market-making research platform: Almgren-Chriss, Avellaneda-Stoikov, impact, TCA, calibration, hedging and risk controls in Python, C++26, Rust and Java 25 from one spec

Intraday pairs trading engine using Kalman Filters for dynamic beta estimation and Avellaneda-Stoikov optimal execution.

A fast, embeddable limit-order-book & matching engine in Go — plus a microstructure research harness and a WASM-powered animated explainer. MIT.

Distributed market-making system. Avellaneda-Stoikov strategy with sub-microsecond C++ hot-path (431ns), event-sourced architecture, VPIN toxicity detection, QUIC mesh transport, real-time dashboard. Rust + C++17 FFI.

MFT/HFT & Market Making suite for Hyperliquid. Features SAC/PPO agent that optimizes Avellaneda-Stoikov inventory control, VPIN & Kyle's Lambda toxicity classification, FIFO queue warfare estimation, spoofing counter, predatory liquidity hunting, and tactical funding arbitrage. Simulation & Live WS runner.

Real-time adaptive market-making system using Hawkes processes + deep learning to predict order flow toxicity and avoid adverse selection. Avellaneda-Stoikov + MHLOBT neural network + LOBSTER L3 data.

FPGA HFT trading system with inline AI inference, transformer attention, Avellaneda-Stoikov market making, and full-stack SmartNIC. 19 SystemVerilog modules targeting AMD Alveo UL3524 at 644 MHz. Sub-50ns tick-to-trade WITH neural network in the critical path.

Avellaneda-Stoikov high-frequency market-making model computing optimal inventory-risk adjusted reservation prices and spread quotes.

Avellaneda-Stoikov high-frequency market-making model computing optimal inventory-risk adjusted reservation prices and spread quotes.

ATOMIC MESH — Distributed deterministic HFT market-making engine. Avellaneda-Stoikov strategy with sub-microsecond C++ hot-path (575ns), event-sourced architecture, VPIN toxicity detection, QUIC mesh transport, real-time dashboard. Rust + C++17 FFI. Live on Binance.

Market making engine for Boros (Pendle's on-chain funding rate swap market). Curve bootstrapping over multi-maturity implied APR, Avellaneda-Stoikov quoting on DV01 inventory, off-chain margin/liquidation replica verified against on-chain source, cross-venue and calendar-spread arbitrage signals, independent risk-monitor kill switch. 231 tests.

Short gamma scalping on Deribit BTC/ETH options, delta-hedged via perpetuals. Avellaneda-Stoikov pricing with live vol surface, funding-adjusted sizing, and config-driven risk controls. Kill switch on RV spikes, drawdown breaches, and adverse funding. Asyncio stack: market data, strategy, execution, and risk engine running as independent loops.

Options market making engine for Deribit inverse (coin-settled) BTC/ETH options. Arb-free SVI vol surface, coin-denominated Greeks derived from Deribit's own inverse pricing formula, Avellaneda-Stoikov quoting in vol-space, Whalley-Wilmott delta hedging via the perpetual. Rust workspace, 111 tests.

Event-driven market making: Binance WebSocket, Kalman+imbalance fair value, dual-timeframe vol, Avellaneda-Stoikov adaptive quoting, paper trading engine

Research framework for optimal high-frequency market making with Avellaneda-Stoikov quoting, WRDS TAQ replay backtesting, queue-aware fills, volatility-adaptive spreads, and robust execution/P&L analysis.

Adaptive options market-making & volatility-surface engine for NSE Nifty50: a SIMD C++17 core (Black-Scholes, Heston FFT, arbitrage-free SVI) + a Greeks-constrained multi-strike Avellaneda-Stoikov quoter with VPIN toxicity & jump detection, validated on a no-look-ahead backtester on real 2023-25 data.