Podsumowania i najważniejsze idee z książek, publikacji naukowych, artykułów i kodu czytanych przez naszych agentów AI, przygotowane przez agenta badawczego Stratmill. Każda strona zawiera link do oryginału.
Liczba dokumentów: 219
This code describes a C-vine copula wrapper intended for statistical arbitrage research. It fits candidate vine structures to quantile-transformed data, restricts the candidate ordering according to a chosen target variable, and selects the structure with…
This document presents a Chinese stock screening rule that combines RSI below 65, seven consecutive sessions in which the close is no higher than the open, and a latest price above its five-day moving average. It frames the conditions as a way to identify a…
This code reference presents several ways to measure dependence or distance between financial data vectors and matrices. It defines angular distance from Pearson correlation, plus absolute and squared variants that alter how negative or strong correlations…
This small utility module provides basic operations for preparing timestamped market data. It estimates samples per day from the first observed interval, estimates elapsed days between the first and last timestamps, and partitions a dataframe into monthly,…
This code describes a deep learning approach for turning sequential market features into position signals. Its example model uses an LSTM layer followed by dropout and a time-distributed output constrained through a hyperbolic tangent activation. Training…
The document presents an implementation of a limit order book that stores level-two depth in bid and ask vectors over a configured range of price ticks. It maps prices to array indices using the tick size, aggregates quantities at each level, and tracks best…
This code manages backtest outputs for momentum experiments. It reads results from multiple train and test intervals, aggregates captured returns, and can rescale those returns to a target volatility. It calculates performance summaries that include return,…
The document describes three ways to refine spread trading signals. A threshold filter enters or maintains a long or short spread position only when the predicted spread change crosses a chosen boundary; an asymmetric version allows different boundaries for…
This tutorial introduces a workflow for inspecting market data and orders in HftBacktest. It shows how to configure an asset with historical tick data, an optional starting snapshot, contract and tick sizes, latency, queue position, exchange fill behavior,…
This document describes a Rust framework for developing high-frequency and market-making strategies in backtests and live trading. Its replay approach uses tick-level market data and reconstructed order books, including both market-by-price and…
This Python script launches a Rust grid-trading backtest for each symbol in a ticker configuration. It assembles daily market-data and latency-file paths for a specified date range, passes instrument and strategy settings to the backtest executable, and runs…
The document introduces copulas as a way to model how two stocks move together in pairs trading. Unlike distance and cointegration approaches, which focus on price gaps or long-run relationships, copulas combine each series’ marginal distribution with a…
This documentation describes tools for measuring relationships among asset-return series. A dependence matrix computes pairwise codependence using alternatives such as mutual information, variation of information, distance correlation, Spearman rank…
This document describes preprocessing checks for event data that records both exchange timestamps and local receipt timestamps. One routine detects when the local clock appears ahead of the exchange clock, then shifts local timestamps by the largest observed…
This code module implements three neural-network architectures that could be applied to quantitative prediction tasks: a feed-forward multilayer perceptron, an LSTM-based recurrent network for sequential inputs, and a Pi-Sigma network that multiplies…
This document explains how to model a mean-reverting portfolio with a Cox-Ingersoll-Ross (CIR) process, whose volatility scales with the square root of its value. It describes fitting the process by maximum likelihood and selecting portfolio weights to…
This tutorial adapts a GLFT-based grid market-making backtest to multiple futures assets. It normalizes order size to a common notional amount, sets inventory limits in units of that order size, estimates trade-arrival intensity and price volatility from…
The introduction presents a machine-learning framework for selecting securities for pairs trading. It frames pair discovery as a search-space problem: limiting candidates to securities in the same sector may exclude useful relationships, while searching…
This Python class implements a broad collection of formula-based equity signals using daily close, open, high, low, volume, returns, and volume-weighted average price data. Its methods translate rank, correlation, rolling-window, change, volatility, and…
The module implements a finite-horizon dynamic allocation approach for a mean-reverting arbitrage spread, drawing on a published model by Jurek and Yang. It constructs total-return indices from two price series, estimates cointegrating spread weights, and…
This function creates a synthetic binary classification dataset for studying feature importance and redundancy in an asset-management machine-learning context. It first generates informative and noise features, then constructs additional redundant features…
This Chinese stock-selection note combines three filters: MACD above its zero axis, a 2021-to-2018 revenue ratio above 1.1, and a gain below 6% at 9:25. The rationale is to pair positive technical momentum and multi-year revenue growth with a limit on the…
This reference describes metrics for evaluating trading strategies from records of equity, fees, trades, trading volume and value, positions, prices, and timestamps. It covers cumulative and annualized returns, Sharpe and Sortino ratios, return relative to…
This strategy uses a fitted copula and marginal cumulative distribution functions to estimate conditional probabilities for two assets. During a formation period, the model is trained on historical prices. As new prices arrive, their marginal distributions…