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.
22,592 documents
This note describes a stock screen based on three chart conditions: amplitude above a threshold, a rising price base, and a rounded-bottom pattern. It presents the pattern as a way to find stocks whose lows are gradually moving higher, while filtering out…
This article surveys six implementation choices that shape equity factor strategies: selecting proxy measures, constructing portfolios, combining factors, allocating among them, trading, and managing risk. It argues that one factor can be represented by…
This brief template note describes a stock-screening strategy that selects companies using natural-language queries. Its example combines a small-market-cap screen with recent net inflows from large investors, a prior period of contracting volume, and a…
This report reviews market conditions relevant to Chinese quantitative equity strategies in July 2022. It tracks the number of unstable factors as a proxy for how supportive conditions may be for strategy excess returns. The report says this count stayed…
The document outlines a stock screen that selects companies in a designated metaverse category, requires a positive recent return, and excludes firms listed for less than a chosen period. Its rationale is to combine thematic exposure and recent performance…
This post describes a rule-based screen for Chinese main-board stocks. It selects shares with turnover between 3% and 12%, a daily gain above 1%, and a circulating share count no greater than 5.5 billion. The stated intent is to combine trading activity,…
This paper summary examines whether historical trading data can predict the next month’s cross-sectional returns of Chinese A-shares. It describes a dataset of 108 stock characteristics from 1997 to 2019 and compares traditional econometric methods with six…
This Chinese-language stock screen combines an amplitude threshold above 1, a ranking among the five highest stocks by the day’s auction amount, and a rising K value from the KDJ indicator. Its rationale is to find stocks with larger price movement and…
This Chinese-language screening note describes selecting stocks with turnover between 3% and 12%, an external-to-internal trading volume ratio above 1.3, and a reversal pattern called “fan bao.” The stated rationale is to find active stocks where buying-side…
The document presents daily portfolio rebalancing as a Markov decision process. An agent selects nonnegative weights for Dow 30 stocks, normalized to sum to one, using a state that combines a rolling covariance matrix with MACD, RSI, CCI, and ADX indicators.…
This stock-selection post proposes combining three filters: an increase in holdings share above a stated threshold, weekly MACD above its zero line, and at least two limit-up sessions within a stated lookback period. The accompanying explanation treats these…
This forum post asks how to use pyramiding in a strategy that combines a higher-level long signal with lower-level entry and exit signals. The author wants to add long entries whenever the smaller-scale long condition occurs while the larger long condition…
The document studies a definition of an industry leader based on analyst coverage and strong links between a stock’s fundamentals and those of its industry peers. It describes the selected stocks’ general characteristics and examines whether their price…
This article introduces XGBoost as a machine-learning method for quantitative stock selection using price and volume factors. It explains boosting as a process that adds weak learners in sequence, then contrasts AdaBoost’s reweighting of misclassified…
This research summary examines whether stock return synchronicity—the degree to which a stock’s returns move with common factors—signals more or less information in prices. The conventional view treats high synchronicity as evidence of less firm-specific…
This post presents a Chinese stock-selection screen using three main conditions: RSI below 65, float market capitalization between 5 billion and 10 billion yuan, and positive net buying attributed to major participants during the opening auction. Its Python…
This post describes a Chinese A-share stock screen combining daily price range, company size, profitability, turnover activity, and a weekday filter. It selects stocks with a daily high-to-low range of at least 1%, market capitalization up to 10 billion…
The document raises a portfolio-construction question about using a stock-ranking model to select both ends of its predictions: stocks with the highest factor scores and stocks with the lowest scores. The proposed idea is to hold the two groups together as a…
The proposed Chinese equity screen focuses on stocks classified in the metaverse sector. It ranks them by the day’s auction amount, keeps the top five, and requires the prior day’s main-fund-flow measure to be positive. The post presents this as a way to…
This stock-selection method screens Chinese equities using three filters: a 14-period RSI below 65, a ten-day price gain above zero but below 35%, and a ranking by individual-stock popularity. The accompanying Python example additionally considers stocks…
This analysis compares analyst forecast data from two Chinese financial data providers, examining report coverage, forecast accuracy, and the usefulness of forecast-related factors in stock selection. One provider is described as covering more stocks, while…
A forum post asks how to retrieve popular industry indices through iWencai and pass them into a variable. The author describes a query combining industry indices with weekly KDJ conditions and recent main-fund-flow filters, then reports that a Python query…
This article introduces support vector machines for classification and regression, then applies them to A-share stock selection. It explains the maximum-margin principle for linear SVMs, slack variables for imperfectly separable observations, and kernel…
This tutorial explains how support vector machines classify data by finding a boundary with a wide margin, and how slack variables allow some classification errors in noisy data. It introduces kernel methods as a way to handle nonlinear boundaries by…