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.
16,761 documents
This stock-selection method combines a daily price-range filter, a history of limit-up moves, and revenue growth. It looks for stocks whose high-low range exceeds one percent of the previous close, with at least two qualifying limit-up events in the…
This guide organizes U.S. equity research inputs into five source groups: company filings, investor relations materials, government macroeconomic releases, trading calendars and halt notices, and public checks on brokers or advisers. It treats regulatory…
This guide surveys ten cryptocurrency trading bot platforms, describing their strategy types, free-plan restrictions, and usability. It introduces bots as software that connects to exchanges and executes predefined rules, then compares offerings that include…
This post proposes screening shares using RSI below 65, a stated circulating market value between 5 billion and 10 billion yuan, and a rising weekly red-bar condition. Its explanation associates the RSI and chart condition with technical weakness and upward…
This Chinese-language post describes an equity screen combining three conditions: a 14-period RSI below 65, three consecutive sessions in which the close is below the open, and an appearance on the prior day’s securities activity leaderboard. The proposed…
This document introduces a series on building a multi-currency Expert Advisor that coordinates several trading strategies in one system. The design aims to make parallel strategies easier to manage on a shared trading account. The articles describe a…
This Expert Advisor bases its trading decisions on two moving averages, with periods of five and four, on a five-minute chart. It assigns separate take-profit and stop-loss levels to buy and sell positions. The decision conditions are evaluated only when a…
The document describes the delivery of historical and current market and blockchain datasets through cloud storage, including data in a format suited to analytical workflows. The described coverage includes decentralized exchange trades and liquidity,…
This documentation explains why a trading strategy may need its own view of the current date and time. The strategy's clock reflects the simulated point in time during a backtest and the relevant time during live trading. This matters when historical logic…
This description outlines an Expert Advisor that enters when the ColorZerolagHLR indicator cloud changes color. It forms a signal at a bar close when an oscillator crosses its signal line. The strategy depends on two compiled indicator files, which must be…
RSIOMA is presented as an RSI indicator calculated from a smoothed price series, with an additional smoothed signal line intended to help identify entries and the prevailing trend. The indicator is displayed in a zero-centered symmetric form. Its…
The strategy first classifies market conditions with a linear-regression slope statistic, treating readings below a stated threshold as a no-trend phase. During that phase, a smoothed stochastic turning upward from a low level triggers a long stop entry…
This tutorial introduces a basic neuron model, describing weighted inputs, a threshold, and an activation function. It explains why a smooth sigmoid function is used in practice in place of an idealized step function, then applies a neural network to…
This MetaTrader expert advisor trades from the LeManTrendHist indicator. A signal is generated when a bar closes and the indicator histogram changes direction, so decisions are tied to completed bars rather than intrabar moves. The advisor requires the…
This Chinese stock-screening proposal selects companies classified in the metaverse industry when price is near its 10-day moving average and at least one session in the prior 25 trading days gained 10% or more. The article presents the average-price…
This example describes a daily SPY strategy that assigns market analysis and order decisions to separate AI agents. The research agent compares the latest completed daily close with its 20-bar average and reports the date, observed prices, evidence for and…
This document describes an Expert Advisor that trades signals from the CronexMFI oscillator. It generates a signal when a bar closes and the indicator cloud changes color, so decisions are tied to completed bars rather than intrabar color changes. The…
This BigQuant support exchange explains a possible source of confusion in backtests: the platform’s `hold_days` setting defines a build-up period during which the strategy can buy, and does not itself require each purchased stock to remain in the portfolio…
The document explains a two-stage model that uses gradient-boosted decision trees to create features for logistic regression. The trees split and combine the original inputs; each sample is then represented by a binary vector marking the leaf it reaches in…
This strategy forms a long-short equity portfolio by ranking stocks on trailing returns, buying the strongest group and shorting the weakest. The ranking uses a twelve-month lookback while skipping the most recent month, a common way to reduce the influence…
The document explains how to define and run a Qlib research workflow through a YAML configuration and the `qrun` command. It lays out the main stages: loading and preprocessing market data, training and applying a model, then analyzing forecast signals and…
The strategy uses the iBands indicator to define an upper and lower price channel. It takes a contrarian position when price crosses beyond a channel boundary: a break above the upper band triggers a sell, while a break below the lower band triggers a buy.…
This research summary examines why stocks can earn excess returns before earnings announcements. It proposes that investors and analysts resolve some uncertainty ahead of the report, supporting price gains before the announcement itself. The study measures…
This algorithm describes a weekly long-short strategy among large U.S. equities. It first filters for liquid stocks, then selects the largest companies by market capitalization. From that group, it buys the ten stocks with the weakest returns over the prior…