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
This research roundup describes several quantitative finance studies. One classifies equity trades by their short-term co-occurrence with other trades and standardizes associated order imbalances into conditional order imbalance measures. These measures…
This tutorial explains how to convert a discretionary futures idea into a specification a program can follow. Its five elements are signal timeframe, entry rules, exit rules, position size, and behavior in edge cases. It emphasizes separating directional…
This overview of a Chinese securities firm’s research describes a shared AI platform for financial data, model development, GPU computing, backtesting, performance analysis, and deployment. It also discusses deep reinforcement learning for stock selection…
This document introduces a proposed study of Chinese stock price-limit events. It says the work explains the limit-up mechanism, classifies stocks that reach the upper price limit, and examines returns after those events. The stated research direction is to…
This guide explains how FreqAI trains and deploys adaptive machine-learning models in live or dry trading and in historical backtests. Live operation can retrain models as capacity permits, use the latest trained model for predictions, and apply limits on…
The document addresses a stock-selection model whose chosen stocks and reported backtest returns change when the backtest end date is moved forward by about twenty days. The explanation is that the training data differs across the runs. A model trained on a…
This article outlines a workflow for creating custom equity-selection rules from platform factors. It illustrates price-location features based on recent highs and lows, then combines return, volume, and market-relative conditions into a screening rule. The…
This study builds separate XGBoost models to predict next-day Ethereum volatility from activity in the USDC, USDT, and DAI lending ecosystems. Features include lagged borrowing, repayment, and flash-loan measures, plus seven-day averages. It uses time-series…
The article argues that robust quantitative trading depends less on choosing a factor or model in isolation than on a repeatable development process. It outlines a sequence: form a market-based hypothesis, analyze whether candidate factors distinguish…
This MetaTrader expert advisor applies a simple perceptron to three forex symbols: EURUSD, GBPJPY, and AUDNZD. At a synchronized new bar, it evaluates selected values from the Acceleration/Deceleration oscillator and combines them with four parameter-derived…
The document describes a deep learning model intended to estimate a company’s next-quarter operating revenue year-over-year growth. It frames the forecast as a potential input to value-oriented investment decisions and says the target corresponds to roughly…
This document is a brief summary of a research presentation on changes in quantitative investing. It outlines a model that combines a linear factor component with a nonlinear function of factor exposures, leaving a residual term. It says that residual…
This forum post questions the target-return calculation in a one-day equity strategy’s labeling code. The strategy is described as buying near the open on one day and selling at the close of the next, yet the label reportedly divides a later session’s high…
This report summary describes an index-timing model that applies wavelet analysis to filter price series before using a support vector machine to predict direction. It compares daily and weekly data and different model settings for the Shanghai Composite and…
This brief strategy outline combines a machine-learning stock ranking model with market-level and individual-stock exit rules. It proposes training a StockRanker on price, volume, and valuation factors, then applying three conditions to the CSI 300 index as…
The report compares machine learning approaches with traditional linear methods for predicting stock-level excess returns from alpha factors. Rather than predict excess return directly, it forecasts return dispersion with an AR(1) model and predicts…
This configuration defines an equity prediction workflow for China’s CSI 300 universe. It applies robust feature normalization and missing-value filling, drops labels with missing values, and ranks labels cross-sectionally. An ordinary least squares linear…
This summary of a Huatai Securities research report discusses using alternative target labels and ensemble learning to improve machine-learning stock selection. It reports comparing models built with different labels across multiple training-window lengths.…
This document introduces a Python library for engineering technical-analysis features from financial time series, using price and volume fields such as open, high, low, close, and volume. It catalogs indicators across volume, volatility, and trend…
This meetup Q&A discusses several practical questions in quantitative research. It explains that traders can label chart patterns for AI training in principle, but visual labels may vary with chart scale and between annotators. A model can only learn useful…
This lecture listing introduces policy-gradient methods, which learn a policy directly, and actor-critic methods, which pair a policy learner with a value estimator. The description highlights the actor-critic combination as a way to use value predictions to…
The document outlines methods for investigating suspicious cryptocurrency activity through transaction graphs, address clustering, and tracing funds across chains. It explains that investigators may infer common control from shared transaction inputs, but…
This brief strategy note outlines an equity selection approach using a random forest regression model. It adds features such as beta and technical indicators, including MACD and Williams %R, and predicts a return-style target calculated from the difference…
This report compares several Chinese language models as assistants for discovering equity liquidity factors over a specified historical period. Each model proposes data inputs and a formula built from trading volume, turnover, amount, or price. The author…