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 note proposes screening Chinese A-shares for a turnover rate between 3% and 12%, excluding Beijing-listed stocks, and retaining stocks identified as forming a reversal or “reversal package” pattern. The accompanying example describes joining a stock…
This strategy forms a quality score from four accounting measures: accruals, cash flow relative to earnings, debt relative to assets, and return on equity. It favors firms with lower accruals and leverage and higher cash flow and profitability. Stocks are…
This post proposes a Chinese equity screen using turnover between 3% and 12%, an opening price within 5% of the 10-day moving average, and revenue growth: 2021 revenue must exceed 1.1 times 2018 revenue. It frames turnover and the opening-price condition as…
This brief troubleshooting exchange concerns an incorrect regression evaluation result in a BigQuant workflow. The user reports that the input data is correct, and the response identifies the final output step of a custom Python module as the issue: because…
The article describes a risk-model framework for evaluating fund managers beyond headline returns, Sharpe ratio, or drawdown. It decomposes active returns into factor and stock-specific contributions, and examines risk contributions from a fund’s style…
This indicator adjusts an exponential moving average using an R-squared measure calculated from price observations over a chosen period. The measure compares a linear fit of recent prices with their variation, then changes the EMA period through a formula so…
This document describes a synthetic RSI indicator built from three RSI calculations. The combined value is calculated using a method that differs from a simple average, although the exact formula and settings are not supplied. Before the RSI calculations,…
This article collects practical lessons from building supervised machine learning strategies for stock selection. It recommends keeping training and test data separate and splitting financial time series chronologically, since random splits can mix market…
This indicator compares the unmodified closing price with a transformed price calculated from the close and a shifted historical close. The user selects how many bars back to obtain the shifted close. The transformed series applies a base-10 logarithm to the…
The document describes an enhanced Time Segmented Volume (TSV) indicator. It lets users select the price input instead of relying only on closing prices, and choose among volume weighting, a true-range-based pseudo-volume, or no volume weighting. These…
The document describes a stochastic MACD oscillator displayed in a separate chart window. It identifies the indicator as a combination of MACD and stochastic calculations, but does not explain the calculation sequence, signal interpretation, or how traders…
This article summary describes research into time-of-day variation in crypto order-book depth and its implications for execution. Using BTC/FDUSD data, the cited study reports that depth fluctuates substantially over the day, with stronger liquidity near the…
The Wave Period Oscillator (WPO) is presented as a time-cycle oscillator, attributed to a 2018 article in the IFTA Journal. Its described calculation derives a signed period-like value from price and a preceding close, then smooths that value with an…
This indicator description presents a modified Commodity Channel Index (CCI). Instead of scaling price deviation with mean absolute deviation, the described version uses standard deviation. It also applies a Hull moving average to prices before calculating…
The FIR indicator is a finite impulse response filter that smooths an applied price using four observations with symmetric weights. Its calculation assigns weights of one, two, two, and one to the current and three prior price values, then divides the…
This research summary describes a multi-factor tracking framework built around the pricing logic of factors. It examines historical returns for technical, expectation-based, and financial factors using a modified cross-sectional regression approach. To…
The document describes an MT4 indicator based on John Ehlers’ Fisher Transform. It frames the method as a way to reshape price data so that extreme moves stand out more clearly than they may in bounded oscillators such as RSI, Stochastic, or CCI. The…
This document describes a bivariate Frank copula implementation for modeling dependence between two uniform variables. It provides methods to sample paired observations, calculate the copula density and cumulative distribution, and evaluate a conditional…
This Chinese A-share screening idea combines a price-range condition with a volume and turnover measure. It selects stocks whose amplitude exceeds 1 and whose previous-day turnover multiplied by today’s auction volume divided by yesterday’s volume falls…
The Edge Drift Detector compares closing-deal outcomes from a live Expert Advisor with a backtest baseline to assess whether average per-deal performance has weakened. Its central method is a cumulative-sum alarm calibrated by resampling baseline deals, with…
This indicator first smooths a selected price series using a configurable moving-average period and method, then calculates RSI on that smoothed series. It adds floating overbought and oversold levels derived from the recent minimum and maximum RSI values…
This MetaTrader indicator is designed to classify market conditions as directional, mixed, or choppy. It combines Kaufman’s Efficiency Ratio with a direction signal based on price relative to a moving average and the moving average’s slope. The score is…
This BigQuant example addresses a limitation in which its high-frequency feature extraction module calculates within a single day, while a researcher may need features from minute data spanning multiple days. It shows a distributed workflow using FAI: read…
The indicator converts a fan of moving averages into three bounded readings: price versus the fan, the fan’s ordering around a reference average, and the fan’s position relative to its slowest active average. Each score divides signed differences by their…