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 report applies a parametric portfolio policy approach to equity factor timing, using external variables to adjust factor weights. It tests the approach on Chinese A-share data and compares two ways to use information: time-series models and…
This indicator identifies and labels a broad set of bullish and bearish candlestick formations on a price chart, placing colored arrows near detected patterns. Examples include doji, hammer, engulfing, morning and evening stars, three-candle formations,…
This document presents a QuantStats tear sheet for an automated strategy labeled “orb-plain,” compared with SPY over January 4–9, 2026. It reports return and risk statistics, including a 0% total return for the strategy, a 0.67% maximum drawdown, a 0.55…
This report describes a Chinese equity selection factor called CPV, built from the relationship between intraday prices and trading volume. It combines three characteristics of daily price-volume correlation: its average level, variability, and trend. The…
This teaching example ports a close-based trend strategy into JavaScript and extends it across multiple cryptocurrency futures symbols. It calculates an exponential moving average of typical price as a midline, computes ATR, and places upper and lower bands…
This report develops Chinese equity money-flow signals using tick data to reconstruct order sizes. It argues that conventional fixed cutoffs miss meaningful institutional activity, tests 44 absolute amount thresholds, and finds that lower cutoffs can better…
The study applies machine learning to estimate an optimal rebalancing frequency for pairs trading. It uses minute-level prices for 50 large cryptocurrencies from Binance during 2022 and 2023. For each asset pair, a simulated pairs-trading algorithm…
The document proposes a Chinese A-share stock screen combining a relative strength index below 65, a ratio of external to internal trading volume above 1.3, and a share price below 12. It presents the rules as a way to combine a technical condition, a…
This forum post raises a factor-research reproducibility problem: a user reports that returns from reproducing four factors differ substantially from the platform’s published results. The suggested comparison holds the factor and backtest period constant,…
The discussion addresses a practical gap between signals based on closing prices and orders that may be impossible to execute at a daily price limit. A trader may see a buy signal when a security opens at its upper limit, or a sell signal when it opens at…
The article tests whether Monte Carlo simulations of stochastic differential equations can forecast stock prices, directions, or severe losses. It selects the simulation curve that best fits historical data, then evaluates it on later observations, arguing…
The post proposes an A-share stock screen centered on a reported increase in holdings above 5% and a relationship between price change and net large-order flow. It also refers to 2021 performance, then adds market capitalization above 10 billion and a…
This A-share screening idea looks for stocks classified in the metaverse sector that appeared on the previous day’s trading activity leaderboard and whose intraday low is 4% to 5% below the previous day’s high. The article frames the screen as a way to find…
This Chinese-language post describes a stock-selection screen focused on the metaverse concept. It requires the 30-day moving average to be rising and filters stocks using a measure combining the prior turnover rate with the ratio of today’s auction volume…
This one-hour Hang Seng strategy uses RSI thresholds to enter long or short positions, with an ATR-based filter that requires volatility relative to price to exceed a set level. It also restricts entries to a specified UK-time window and disables order…
The document presents a Chinese A-share screening rule that selects stocks with a daily high-low range exceeding one percent, at least one limit-up session in the prior 25 trading days, and a high ranking on a large-order net-volume measure. The combined…
This document describes a long-only S&P 500 strategy that uses Bollinger %B to identify a pullback during an established uptrend. The indicator expresses the close’s position between the lower and upper Bollinger Bands. Entries require the close above its…
The document explains how a multi-strategy backtest module combines strategies. It applies user-set weights to the return series from each strategy’s backtest and aggregates those weighted series into a single performance curve. The weight therefore applies…
This document describes a latency interface for high-frequency trading backtests, separating the delay from submitting an order to exchange processing from the delay between exchange processing and receiving a response. A constant model assigns fixed values…
This support note investigates why a portfolio log showed a sale on one date while the holding still appeared in positions and no sell signal was visible until the following day. The explanation identifies the test universe’s end-date setting: it extended…
This study explains combinatorially symmetric cross-validation (CSCV) as a way to estimate the probability that a strategy selected as best in a backtest will rank poorly across alternative partitions of that history. The backtest period is divided into…
The article explains naive Bayes classification and applies it to a stock selection strategy. The classifier estimates the probability of each outcome from historical labeled observations, assuming features are conditionally independent and equally…
This stock-selection proposal combines turnover between 3% and 12%, a positive ten-day gain below 35%, and a weekly close crossing above its 30-week moving average. The turnover band is intended to select actively traded shares, the return filter limits…
This study examines a database of 486 patent-derived factors for Chinese listed companies and tests their usefulness in stock selection. It organizes patent information by type, procedural status, metric, and lookback period. Measures include patent counts,…