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 document describes a rule-based intraday breakout system for the DAX using 15-minute bars. It tracks the day’s high and low from a specified morning start, then sets stop-entry levels inside those extremes once price has moved far enough from them. The…
This report summarizes research on momentum spillovers: stocks may react slowly to news about related firms when investors and analysts have limited attention. It presents shared analyst coverage as a general way to define those links. The measure, called…
The strategy estimates sector return states with a Markov chain, which models each state as dependent on the preceding state. It uses the resulting conditional mean and variance estimates as inputs to a Black–Litterman portfolio process, treating the…
This example applies a linear support vector machine to classify each new bar’s next close-to-close movement as upward, downward, or relatively flat. It builds training examples from historical bar open and close prices, labels each example according to…
This document outlines a trading system based on the UltraWPR indicator. It generates a trade signal at bar close when the indicator cloud changes color. A position can be closed by a pending order or when the cloud switches to a color opposite to the…
The post describes an A-share stock screen combining a 20-day moving average above the 120-day average with a volatility threshold and a nonempty outstanding-convertible-bond name field. It proposes sorting qualifying stocks by the 20-day average, while its…
This document describes ways to use broad-market index data to reduce exposure in an equity strategy. It argues that index signals may be less noisy than signals from individual stocks, then outlines three approaches: moving-average comparisons of index…
This proposed daily stock screen combines amplitude above a stated threshold, at least two limit-up sessions during the prior 500 days, and a rounded or arc-like price pattern. The article treats amplitude and limit-up frequency as signs of market attention,…
This document describes an Expert Advisor that trades signals from the PPO Cloud oscillator. It enters when a bar closes with a crossover between the indicator’s main and signal lines, together with a change in the cloud’s color. The description identifies…
This Chinese-language post proposes a daily screen for A-share stocks, run before market close. It combines MACD above zero as a trend filter, positive trailing-twelve-month price-to-earnings as a valuation condition, and a pattern the post calls the Kute…
The document describes a tool for assessing portfolios of expert advisors (EAs) across three dimensions: correlation between daily profit-and-loss series, the hours and weekdays when strategies trade, and diversity across asset classes. It combines dimension…
The document describes a MetaTrader 5 replay trainer for practicing trades on Renko charts. Users choose a historical interval and brick size, load available broker tick data, then advance the replay manually or one brick at a time. Virtual buy and sell…
This indicator description presents an all-in-one version of Woodies CCI. It emphasizes correct implementation of the system’s trend-change rules, including the required yellow bar, and adds colored slope lines for an exponential moving average and a linear…
This Chinese equity screening idea combines three conditions: a turnover rate between 3% and 12%, appearance on the previous day’s top-trader activity list, and a 15-minute MACD histogram that is contracting. The screen is intended to identify actively…
This report compares a SPY 0DTE options strategy with SPY over a brief backtest covering January 4–6, 2026. It presents standard performance and risk measures, including returns, drawdown, Sharpe ratio, volatility, time in the market, and benchmark…
The document compares TradingAgents, presented as a framework for multi-agent financial research, with Lumibot, described as a Python trading framework that can place agent workflows inside a strategy lifecycle. It outlines how research and debate agents can…
This document describes an Expert Advisor built around the Adaptive Cyber Cycle indicator. It generates a signal when the indicator crosses its signal line at a bar close. The inputs also allow use of related Adaptive CG Oscillator, CyberCycle, or Adaptive…
The article describes a workflow in which a trader states a strategy in everyday language and a tool generates runnable QMT code for backtesting. Its example uses a Chinese stock, buying when the 10-day moving average crosses above the 60-day average and…
This reference catalogs Zipline’s strategy and backtesting interfaces. It covers algorithm setup, market data access, scheduling, asset lookup, order placement and cancellation, and trading controls such as limits on leverage, order count, order size, and…
The document explains why a strategy’s maximum capital capacity cannot be calculated from backtest returns and volatility alone. Capacity depends on the strategy’s logic and trading frequency, the liquidity and order book depth of its instruments, the size…
This article explains how to simplify factor research by converting many distinct raw values into a smaller number of bins. Its example is a daily return factor across thousands of stocks, where nearby values may be treated as equivalent for exploratory…
The article explains how a backtest can look convincing while relying on conditions that may not hold in live markets. It examines three assumptions: orders fill immediately at the signal price, market data forms a continuous sequence, and prices from a…
This article explains how seemingly routine backtesting choices can make a strategy look more reliable than it is. It focuses on three assumptions: orders fill immediately at the modeled price, market data and trading opportunities form a continuous…
This stock-selection proposal looks for shares with a price amplitude above 1%, a claimed signal of main-force control on the prior day, and simultaneous golden crosses among the 5-, 10-, and 20-period moving averages. The accompanying formula expresses the…