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.
4,194 documents
This proposed Chinese stock screen combines three conditions: at least five moving averages overlap, the stock has had more than two limit-up sessions within ten days, and its 2019 dividend payout ratio exceeds 25%. The article names five average windows,…
The strategy selects equities with an RSI below 65, 2021 revenue greater than 2018 revenue by a ratio above 1.1, and positive net profit in the first quarter of 2021. It then proposes taking the first N qualifying stocks and holding them for one year. The…
This Chinese stock screen combines clustered moving averages with a historical revenue comparison. It describes selecting stocks with at least five overlapping averages based on 5-, 10-, 20-, 30-, and 60-day periods, alongside a revenue ratio requiring 2021…
This stock-screening proposal selects Chinese equities with turnover between 3% and 12%, circulating market value between 5 billion and 10 billion yuan, and a condition described as the beginning of a major upward move. It presents the turnover and…
The document explains how factor-based Smart Beta products can serve investors seeking diversified equity exposure and can act as underlying holdings in fund-of-funds portfolios. It describes the approach as using passive strategies that increase exposure to…
This Chinese-language account describes a learning project that builds a China stock selection strategy with five ranked factors and a two-day forward price label. It trains a Ridge regression model, standardizes the inputs, and periodically retrains before…
This report examines the predictive accuracy of a structured multi-factor risk model for China A-share portfolios, building on earlier work on style-neutral portfolio optimization. It reports that a bias test found statistically significant evidence related…
This study examines the low-volatility effect in Chinese A-shares using price amplitude as a volatility proxy. The original amplitude factor has negative stock-selection ability, but its performance is unstable: the reported monthly IC and rank IC are…
The document describes a stock selection screen that combines a turnover range, a listing-year filter, and a price-to-earnings ceiling. Its stated logic selects stocks with turnover between 3% and 12%, an IPO year of 2021, and P/E below 20. It also gives an…
This Chinese-language forum exchange addresses a data handling issue that can arise after computing regression coefficients with an OLS factor expression. The questioner finds that beta values appear in DataFrame cells as array-like objects rather than as…
This research explores shared analyst coverage as a way to represent links between Chinese A-share companies. It extends direct links to indirect, multi-step connections, arguing that these can reveal less visible information spillovers, especially for…
The document summarizes two strands of research. One examines analyst-run mutual funds across multiple fund companies, reporting that analyst-managed funds outperformed comparison funds on risk-adjusted measures and that stronger analyst skill was associated…
The document describes two measures of active fund manager ability based on whether a manager’s holdings or portfolio changes resemble those of historically capable peers. The holdings measure assigns stocks a quality score from the weighted skill estimates…
This stock-selection approach combines four filters: daily amplitude above 1, a listing age greater than one year, market capitalization below 10 billion yuan, and positive net income in each of the most recent four quarters. The accompanying sample workflow…
This strategy ranks constituents of the CSI 300 by their trailing 12-month dividend yield and selects the ten highest-ranked stocks. It updates the target list on the first trading day of each month, sells names that leave the selection, and buys newly…
The article explains recurrent neural networks and LSTM gates, then outlines an A-share stock-selection example that predicts five-day returns from sequences of seven factors. It describes missing-value handling, outlier filtering, standardization, and…
This screening method targets Shenzhen main-board stocks using a rising 30-day moving average, an amplitude threshold, and valuation ranges. The stated criteria are amplitude above 1%, price-to-earnings ratio from 0 to 29.01, and price-to-book ratio from 0…
The document describes a China equity index enhancement approach that combines a multi-factor return forecast with portfolio constraints. Candidate factors span size, valuation, growth, profitability, technical signals, liquidity, and volatility. Symmetric…
The document outlines an A-share stock screen combining a turnover-rate band, a market-cap ceiling, and a positive-profitability condition with a threshold for the day’s increase in holdings. Its rationale is that the holdings-change measure may help…
The document describes a Chinese equity screen that combines a robotics concept classification with a circulating market capitalization below 10 billion yuan, a turnover range, and positive returns. Its stated final rule tightens the return requirement to…
This post outlines a historical screen for Shenzhen Main Board stocks in 2021. It selects shares with amplitude above 1, a price-to-earnings ratio from 0 to 29.01, and a price-to-book ratio from 0 to 3.11. The approach combines a basic activity filter with…
This overview surveys a broad set of quantitative approaches, describing their basic mechanisms and the market settings or investor capabilities they may suit. It covers trend following, machine learning, arbitrage and market neutral methods, factor…
This document describes a Chinese equity screen combining three filters: price amplitude above 1, a 2019 dividend ratio above 25%, and prior-day turnover above 8%. It presents the combination as a way to find stocks with strong recent activity and…
This Chinese-language research abstract describes a study of order sizes in stock-level tick data. It reports a Pareto-like pattern: small orders occur more often than large orders, with order-size frequencies declining according to a power law. The authors…