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 study examines holdings reported by institutions using the Stock Connect channels into mainland China, comparing static position and daily increase factors. It reports that Morgan Stanley Hong Kong Securities had stronger historical information…
The article outlines ways to adapt portfolio construction to crypto markets, whose skewed returns and extreme losses can make standard mean-variance assumptions unreliable. It recommends evaluating downside risk with measures such as Sortino, drawdown, VaR,…
The report proposes left- and right-tail conditional value-at-risk factors from intraday stock returns. For each stock, it describes measuring the average returns beyond the lower and upper 5% quantiles, then averaging daily values across a 20-trading-day…
This research note summarizes evidence that changes in the wording and structure of company filings can contain information about future operating performance and equity returns. The underlying study compares the text of SEC annual and quarterly reports with…
This report summary examines how market and style conditions may affect equity market-neutral strategies. Such strategies select stocks using fundamental or price-and-volume signals, then hedge broad market exposure with index futures or short selling. Their…
This Chinese A-share screening proposal combines daily turnover between 3% and 12%, a positive but sub-35% ten-day gain, membership in a robotics concept group, and circulating market value below 10 billion yuan. The accompanying Python sketch uses market…
The article explains why a factor’s strong historical returns may come from an unintended exposure to market capitalization or liquidity. It uses large-order sell value as an example: because the measure is an absolute amount, it tends to be larger for…
This research summary describes an equity selection process built from daily price and volume data. Genetic programming searches for candidate factors, using RankIC to favor linear relationships or mutual information to seek nonlinear ones. To focus on…
This document describes an AI-assisted value strategy organized as three roles. A research agent reviews company reports and current prices to identify a small group of businesses with steady profits, durable competitive advantages, and prices judged…
This article examines how the length of an AI model’s training window can affect an equity strategy. It compares manually rolled long windows spanning several years with shorter windows ranging from a month to a few years, and recommends evaluating each…
This research summary examines whether market conditions and macro variables can predict the next month’s long-short reversal-factor returns in Chinese equities. It identifies market funding sensitivity, turnover, volatility, and bid-ask spread as…
This brief strategy note describes selecting stocks whose top ten tradable shareholders include a social security fund, based on the reported shareholder list. It pairs that ownership condition with a maximum holding period of ten days and identifies an…
This tutorial explains how an AlphaModel fits a model from a prepared AlphaDataset and produces prediction scores for a chosen data segment. It describes a shared interface for fitting, prediction, and model details, then distinguishes training, validation,…
This stock-selection proposal combines technical filters with reported earnings growth. It looks for RSI below 65, year-over-year growth in net profit attributable to parent-company shareholders above 20% and no more than 100%, and MACD above its zero axis.…
The document describes a Chinese stock screen that selects shares with price amplitude above 1 and RSI below 65, while excluding stocks listed on the STAR Market. It gives formulas for amplitude, based on the high, low, and previous close, and for RSI using…
This note describes a reproduction of a Guosen Securities quantitative research report on daily momentum factors that combine individual stock and industry returns. Its central idea is to use the relationship between a stock’s performance and its industry’s…
This document describes a rules-based screen for listed Shenzhen main-board shares. It combines a minimum trading-range amplitude with a circulating market capitalization above 10 billion yuan, a price-to-earnings ratio from zero to 29.01, and a…
The report proposes combining company-level stock selection with industry allocation for Chinese equities. Its coupling model uses each stock’s historical beta to its industry as a dynamic weight: more industry-sensitive stocks receive more influence from…
This summary of a Changjiang Securities study explains a stock-selection factor based on the balance between active and passive buying. Active buying occurs when an investor’s order trades against displayed sell orders; passive buying places a limit order…
This research summary argues that factor signals weaken over time, so holding a monthly portfolio unchanged can leave later-period capital exposed to less informative signals. It proposes measuring a factor’s decay speed through the half-life of its…
This forum post outlines the basic workflow of a multi-factor analysis backtest: sort stocks into groups by factor value, buy according to a rebalancing schedule, calculate each group’s returns, and report performance measures. It also notes that…
This article summarizes research on the relationship between U.S. industry portfolio returns and the VIX, an options-derived measure of expected stock-market volatility. It describes analysis of 49 industry portfolios, comparing return characteristics and…
This research summary reviews a study of how U.S. industry returns respond to market uncertainty measured by the VIX and its change. Using 49 industry portfolios over a historical sample, the study compares return distributions and return spreads across VIX…
The post outlines an A-share stock screen combining a turnover-rate band of 3% to 12%, a market capitalization ceiling of 10 billion yuan, and a requirement to exclude loss-making companies. It adds a filter for selected ownership or enterprise types, though…