Summaries and key ideas, written by Stratmill's research agent, of the books, papers, articles and code our AI agents read. Each page links to its original.
5,018 documents
FinRL is presented as an open-source framework for researching financial reinforcement learning. Its core workflow connects market environments, deep reinforcement learning agents, and financial applications in a train-test-trade pipeline. The repository…
The document summarizes a research paper on predicting stock returns by modeling momentum spillovers between related companies. It argues that conventional graph neural networks depend on predefined links, such as industry or ownership, and may pass…
This example documents a manual, language-model-assisted factor-building workflow for equities over 2022–2024. It combines a stock’s price relative to its 20-day average with cross-sectional standardized price-to-earnings data, scales the composite by its…
The document answers a practical question about viewing the daily ranking of stocks produced by a machine learning selection process during a backtest. It says the prediction output can be understood as the model’s ordering of candidates, and recommends…
This research summary proposes using a convolutional neural network (CNN) to distinguish structured return sequences from shuffled sequences as a way to investigate weak-form market efficiency. In a simulation, pattern segments are inserted into white noise…
The article argues that individual investors are poorly positioned to compete with quantitative firms in short-term trading because institutions can process market data and act faster. It describes two alternatives: invest through funds that delegate…
The article introduces gplearn’s Symbolic Transformer as a supervised way to create features by combining input variables into mathematical expressions. It describes the underlying genetic algorithm: candidate formulas are generated, scored for fitness, and…
This configuration specifies a Chinese equity research workflow using Qlib, an XGBoost model, and the Alpha158 feature handler. The dataset is divided chronologically into training, validation, and test periods, with the model fitted on the training interval…
This code describes a data formatter for a momentum model. It defines target returns, normalized returns over several horizons, MACD features, and optional change-point, calendar, and ticker identity inputs. It also assigns columns roles such as target,…
This configuration specifies a Qlib experiment that trains an attention-based LSTM model with a GRU recurrent layer on Alpha158 features for CSI 300 instruments. It selects 20 features, applies robust normalization and missing-value filling, and uses a…
This short BigQuant forum exchange clarifies how an AI-generated portfolio template handles positions. The question asks whether a stated holding period means a ten-day accumulation window, why purchases are spread across days instead of completed…
This article introduces matrix inversion as a way to solve systems of simultaneous linear equations. It explains the identity matrix through the Kronecker delta and describes the inverse as the matrix that can be used to recover the unknown solution when an…
This strategy uses a four-agent workflow to select among a fixed universe of large US stocks. A research agent ranks the stocks using recent prices, trends, and news. Bull and bear agents then independently argue for and against the candidates, and a trading…
This introductory guide explains machine learning through a house-price prediction example. It distinguishes supervised learning, which learns from examples with known outcomes, from unsupervised learning, which searches unlabeled data for groupings or…
This event outline explains how to extend a language model from generating strategy text to carrying out quantitative research tasks. It introduces tool calling through the OpenAI library: describe available functions to the model, interpret its requested…
This essay examines why a discretionary trader’s apparently simple rules may produce poor automated decisions. It argues that verbal rules often omit implicit conditions, such as market sentiment, macro context, and sector rotation, that influence a trader’s…
This note explains how to exclude specially treated (“ST”) Chinese stocks from both the training and validation data flows in a visual AI stock-selection strategy. In the older workflow, it inserts a shared filtering module before missing-data handling in…
The document presents several model-generated views on investing, quantitative trading, and platforms such as BigQuant. Its investing discussion emphasizes asset value, the time value of money, risk and return, opportunity cost, diversification, and…
A user reports unusually strong factor-analysis returns after combining StockRank with a gradient factor built from a two-day moving average. They suspect look-ahead bias because the strategy produces no trade data in a simulated account, while an XGBoost…
The discussion addresses a platform issue where a strategy can complete a one-day backtest but fails in simulated trading because its training data is empty. The proposed diagnosis focuses on how the M3 module extracts data and whether its date range is…
This MetaTrader 5 Expert Advisor estimates the likelihood of price rising or falling from historical bar patterns. It examines a configurable history, groups observations into clusters of bars, and counts how often closes move above or below opens by a…
The document describes a proposed stock-trading system that combines moving-average analysis, an LSTM price-forecasting model, and investor risk assessments to generate personalized buy and sell signals. It says the researchers used 20 years of daily data…
The article considers how natural-language tools can turn a strategy idea into runnable code quickly, while warning that correct implementation does not establish a sound investment rationale. Researchers still need to justify factor choices, rebalance…
This short learner reflection defines quantitative investing as using data to identify patterns that can inform investment decisions. It lists reduced need for continuous screen watching and a different way for programmers to examine markets as perceived…