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
The article outlines buy-and-hold, day-trading, and swing-trading approaches for volatile altcoins, and describes stablecoins mainly as trading-pair assets, temporary havens, or inputs to yield farming. It distinguishes fundamental research, such as…
This document describes a plotting workflow for examining Betfair backtest logs that record bid and ask volumes by runner. It extracts periodic batch volumes and cumulative imbalance, then creates three visual views: imbalance over successive updates, the…
This Chinese equity screen selects stocks with current turnover between 3% and 12%, an opening price within 5% of the ten-day moving average of closing prices, and prior-day actual turnover between 3% and 28%. The document frames the combination as a way to…
This post describes a proposed stock screen combining an upward-trending 30-day moving average, an opening price near the 10-day average, and at least five overlapping moving averages. It presents these conditions as a way to find stocks where several…
This intraday stock-selection rule combines three filters: rank stocks by volume ratio and retain the top 100, rank them by afternoon net inflow and retain the top 100, then require the opening price change at 9:25 to be below 6%. The post describes these…
This document describes an Expert Advisor that uses the ColorStepXCCX indicator to generate trades when its cloud changes color. The signal is evaluated at bar close, so the described method waits for a completed candle rather than acting on an intrabar…
This screen selects Chinese equities with turnover between 3% and 12%, a reversal-style candlestick pattern, and no limit-up session on the previous day. The reference implementation approximates the reversal condition by comparing the prior close with the…
This stock-selection proposal ranks shares by capital-strength measure, keeps the top five by the day’s auction amount, and requires at least two limit-up sessions within the previous 500 days. The note interprets these filters as a way to focus on stocks…
This Expert Advisor uses signals from an OsHMA histogram to generate trades at bar close. Depending on a selected mode, a signal occurs when the histogram crosses its zero level or changes direction. The method therefore turns changes in a momentum-style…
This document describes an Expert Advisor that generates trades from the ColorNonLagDotMACD histogram. It forms a signal at bar close when the RMACD histogram crosses zero, its direction or signal line changes, and the histogram crosses its signal line.…
This A-share screening strategy looks for stocks with an amplitude above 1, at least one limit-up day in the prior month, and a gain below 6% at 9:25. Its expanded version also selects the top 50 by circulating market-cap rank and requires a dynamic…
This trading-system description uses changes in the ColorXADX indicator’s cloud color to signal entries at bar close. It adds a trend-strength threshold to avoid opening positions when the market is flat. The stated default threshold is 30, and the filter…
This note proposes screening Shanghai-listed stocks with market capitalization below 10 billion yuan and no reported losses, then requiring five moving averages to converge. The listed windows are 5, 10, 20, 30, and 60 days. The author interprets the overlap…
This proposed A-share screen looks for stocks with at least five moving averages clustered together and at least two limit-up sessions within the prior 500 days, while excluding Beijing listings. The article interprets clustered averages as a sign of…
The document outlines a basic Expert Advisor using a Fractal MFI oscillator. It sells when the oscillator crosses downward through an overbought threshold and buys when it crosses upward through an oversold threshold. Signals are evaluated at bar close, so…
The document presents a MetaTrader-style test report for an unnamed strategy on EUR/USD using daily bars. It describes the system only as an optimization of existing code for positional traders; entry and exit rules are not explained. The report includes a…
The article explains why backtests are approximations of live trading and compares simple bar-by-bar for-loop systems with event-driven systems that process market data, signals, orders, and fills through an event queue. For-loop backtests are easy to build…
This document describes an Expert Advisor that trades from the ColorRMACD histogram. Depending on its selected mode, it can act when the histogram crosses zero, when the histogram or signal line changes direction, or when the histogram crosses the signal…
The report examines how factor evaluation can guide factor inclusion and, especially, factor weights in a multi-factor return model. It argues for evaluating single factors through optimized portfolios with controlled risk exposures, aiming to make measured…
This configuration specifies a Qlib workflow for training an ADARNN model on CSI 300 equities in the Chinese market. It uses Alpha360 features, robust z-score normalization with outlier clipping, feature filling for missing values, and cross-sectional rank…
This strategy uses a percentage distance from recent highs and lows to identify possible reversals, aiming to avoid relying on fixed price-point thresholds that may behave differently as an instrument’s price changes. It does not first classify the market as…
This article describes a stock selection rule combining three conditions: amplitude above a stated threshold, a bullish alignment or crossover among three moving averages, and positive net buying attributed to major participants during the opening auction.…
These release notes describe additions to pyfolio, a toolkit for evaluating trading portfolios. New analyses include performance attribution to common factors, factor and sector risk exposures, rolling volatility, capacity, bootstrap uncertainty in…
This demo strategy builds two classic Renko streams from bid ticks. A slower stream sets directional bias, while a faster stream triggers entries; trades are permitted only when both streams point the same way and the fast stream meets the entry-run…