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 article explains a GBDT-plus-logistic-regression method for classifying stock direction. Gradient-boosted decision trees capture nonlinear feature relationships; each tree maps observations to leaf nodes, which are encoded as indicator variables.…
This research note describes a tactical asset allocation method that converts the tone of company and macroeconomic news into a slower-moving signal. News sentiment scores are aggregated weekly, then filtered with a two-sided cumulative sum (CUSUM) procedure…
This document proposes screening stocks for turnover between 3% and 12%, an external-to-internal trading volume ratio above 1.3, and positive price-to-earnings. It presents turnover as a measure of market activity, the volume ratio as an indication of buying…
This note introduces a smoothed variation of the range-weighted exponential moving average. It attributes the smoothing approach to a method discussed by Lee Leibfarth in work on adaptive price zones. The stated aim is to reduce noisy changes in the…
This brief indicator note describes an RSI variant whose calculation allows users to choose both the lookback length and the smoothing method. It contrasts this flexibility with the conventional RSI description given in the source, which uses exponential…
This report presents machine learning as a data-driven approach and distinguishes supervised learning, which maps labeled inputs to outputs, from unsupervised learning, which seeks structure in unlabeled data, and reinforcement learning, which learns actions…
This report introduces machine learning as data-driven learning and outlines its three broad forms: supervised, unsupervised, and reinforcement learning. It describes how computing advances, larger datasets, and accessible software libraries have helped make…
This report reviews the weekly returns and directional changes of 16 equity factors. For the week covered, a three-month price momentum or reversal measure led with a reported return of 1.45%, while the 24-day Relative Strength Index returned 0.43%. All…
This research summary investigates the A-share small-cap premium, comparing its strength across market-cap segments, industries, and global markets. It reports a pronounced size effect across the broader A-share market and within CSI 500 constituents, but a…
This brief overview introduces genetic programming as an evolutionary approach to generating quantitative investment factors. It describes starting with basic data inputs and mathematical operators, then iteratively evolving candidate expressions through…
Real Motion expresses an equity’s closing price as a percentage deviation from a reference moving average. The calculation divides the close by the average, subtracts one, and scales the result by 100. This centers the indicator at zero when price matches…
The document describes using genetic programming to discover stock-selection factors from intraday price and volume data. Factor expressions are represented as trees; successive candidate populations are produced through selection, crossover, and mutation.…
This indicator combines a Rainbow Moving Average with an RSI calculation, then smooths and transforms the result. Its configurable inputs are the Rainbow Moving Average period, RSI period, smoothing period, and overbought and oversold levels. The document…
The document outlines an end-to-end portfolio construction framework that embeds a differentiable convex optimization layer within a neural network. Raw asset data passes through the network for feature extraction and combination, then an optimization layer…
The document answers a data-preparation question: how to retain the time of day from a timestamp column while omitting its calendar date. It describes applying a datetime formatting operation to each value in a pandas column, creating a separate time field,…
This document catalogs built-in factor datasets for Chinese A-share equities. The listed fields span price history, returns, volume, trading value, turnover, active money flows, valuation ratios, financial statement measures, analyst forecasts, shareholder…
The indicator described here tracks three definitions of an unfilled price gap and displays the remaining distance to closure as histogram bars. Daily gaps measure the space between a session’s close and the next open. Intraday gaps compare the shadows of…
This research summary reviews several empirical findings drawn from macroeconomic and equity data. It reports that principal components built from 132 monthly macro series across 15 categories produce eight factors that explain a stated share of the next…
This report describes combining event-driven signals with a multi-factor stock selection model for Chinese equities. It first recommends estimating event-related abnormal returns with cross-sectional regressions that control for industry and market-cap…
This essay challenges common claims about government money creation, bank reserves, public debt, quantitative easing, inflation, interest rates, and the Federal Reserve. It argues that commercial bank lending creates much of the money supply, reserves remain…
This stock screen combines three conditions: a daily high-to-low range above 1% of the prior close, an opening price within 5% of the 10-day moving average, and a positive return below 35% over a 10-day lookback. The post presents the conditions as a way to…
The document describes an experiment for comparing randomized trade sequences, focusing on whether the direction of the first position affects the outcomes. It contrasts alternating sequences that begin with a buy and those that begin with a sell, with the…
This stock screen selects shares with price amplitude above 1, a listing history longer than one year, and turnover between 2% and 9%. The post frames amplitude and turnover as measures of trading activity and listing age as a simple stability filter. It…
This Chinese stock-screening example combines three filters: daily price amplitude above 1, prior-day actual turnover between 3% and 28%, and 2021 revenue greater than 1.1 times 2018 revenue. The intended idea is to select stocks showing trading activity…