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.
21,023 documents
This Chinese stock screening proposal selects shares with turnover between 3% and 12%, an appearance on the previous day’s 龙虎榜 (a public list highlighting unusual trading activity), and a size measure above 200 million. It also specifies that a stock should…
The SR-RateIndicator is a trend oscillator displayed as a histogram, with values normalized to a range from negative one hundred to positive one hundred. Histogram color indicates trend direction. Two configurable thresholds mark the high and low levels at…
The SpeeD Oscillator combines two measurements related to a moving average: its speed and the distance of price from it. Speed appears as a histogram scaled against its historical average, while price-to-average distance appears as a line measured in…
The document outlines the Chaikin oscillator, a technical indicator built from price and volume data. It first weights volume by the relationship between the close and open relative to the day's high-low range, cumulatively sums that measure, and then takes…
This tutorial describes a Python workflow for comparing two subfunds with the China Securities 1000 Index using weekly net asset values. It calculates rolling correlations over a 20-day window and plots how those relationships change, then compares period…
The document outlines a workflow for building equity strategies that adapt their stock-selection logic to market conditions. It proposes measuring market sentiment with signals such as the number of limit-up stocks and average returns, then adding industry…
This study examines how factor information decay can guide multi-factor stock selection. It defines a factor’s half-life from the time required for its information coefficient (IC), or IC information ratio (IC_IR), to fall to half its initial level. The…
This Chinese market-timing report reviews sector strength, valuation, breadth, moving-average structure, and equity-fund positioning around July 2020. Its summary highlights stronger recent breadth: the share of stocks reaching 60-day highs and the balance…
HVR is described as a trend-strength indicator whose initial input is the ratio of the current bar’s price to the previous bar’s price. The description identifies the basic data relationship used to begin its calculation, but does not provide the remaining…
The article presents a robust version of equal risk contribution (risk parity) portfolio construction. It treats the estimated covariance matrix as uncertain and builds a bounded uncertainty set around the estimate. The optimization then accounts for…
This research summary describes how to build and compare active and passive portfolios of Chinese healthcare-themed funds. For active funds, it argues that screening should use actual underlying holdings, since a fund's label may not reflect a concentrated…
This Chinese-language post describes a daily mainland China stock screen. It filters main-board shares to a turnover range of 3% to 12% and a circulating market value between 5 billion and 10 billion yuan, then ranks candidates using the product of daily…
The document discusses whether stochastic-process mathematics, including Brownian motion, still matters as machine learning becomes more common in quantitative finance. It presents several practitioner views: stochastic calculus has a stronger role in…
This indicator measures the current price’s distance from an intraday volume-weighted average price in standard deviation units. The VWAP and Z-score reset at a chosen start time, and plotting stops at a chosen end time. A smoothing option affects the…
This Chinese-language README catalogs resources for systematic trading research and implementation, including backtesting frameworks, trading libraries, data sources, strategies, books, videos, blogs, and courses. Its listings span multiple asset classes and…
This indicator example applies spectral filtering to a moving average time series. It removes higher-frequency components from the series to produce a smoothed output, and the description says the technique has practically zero latency. It presents the…
The document explains the classic Coppock indicator as a long-horizon signal for potential buying or selling opportunities. Its calculation adds two rates of change, using lookback periods of 14 and 11, then applies a 10-period weighted moving average to…
This equity screen combines three filters: turnover between 3% and 12%, a stock code beginning with 60, and classification in the beverage or alcohol business. The article describes the code prefix as a way to restrict the universe and uses industry…
This indicator description defines an advance-decline measure over a chosen period by subtracting the count of bearish candlesticks from the count of bullish candlesticks. It offers two modes. In non-cumulative mode, the displayed value is the difference for…
This indicator note explains how Average True Range is built from True Range, defined as the greatest of the current high-low range and the absolute gaps between the current high or low and the previous close. It describes an initialization step that…
This overview distinguishes language generation from financial prediction, arguing that strong text generation does not by itself solve markets’ low signal-to-noise ratios, changing patterns, and limited data. It explains GPT’s development and instruction…
The document describes a band indicator that varies Bollinger Bands and Envelopes. It builds the bands from two smoothed moving averages calculated with the linear regression moving average method (LSMA). The two configurable inputs are the LSMA calculation…
This utility finds the maximum high and minimum low within a user-selected range of chart bars. The user specifies a start bar and an end bar; the procedure identifies the candle with the highest high in that interval and returns that high. It performs the…
The document describes an indicator called AIS Central Axis that applies a nonlinear smoothing algorithm related to the HiLo family. It distinguishes the method from the classical version by processing extrema in the input time series. The stated parameter…