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 brief description introduces a trading indicator based on the generalized autoregressive conditional heteroskedasticity (GARCH) model. It presents the indicator as a way to track changing volatility: readings rise when volatility is high, and changes in…
This module defines three return measures from a series of closing prices. Daily simple return is the percentage change from the previous close; daily logarithmic return is the difference between successive log prices; and cumulative return is the percentage…
This post presents a basic A-share stock screen that selects stocks with turnover between 3% and 12%, excludes Beijing-listed stocks, and requires a 14-period RSI below 65. It frames turnover as a liquidity filter and RSI as a way to identify stocks that may…
This proposed stock screen combines a daily price-range threshold above one percent, previous-day turnover above 60 million, and membership in a selected company type or industry. The article describes the company classification as an additional filter…
This account summarizes a 2021 conference talk on using machine learning to price stocks. A simple example feeds market-wide daily open, high, low, close, and average prices into an LSTM time-series model. The speaker stresses preparing noisy financial data,…
This module describes selecting three partner stocks for each target in a four-stock vine-copula statistical arbitrage framework. It compares four approaches using ranked daily returns: a baseline that sums pairwise Spearman correlations, a multivariate…
Henderson filters are centered moving-average smoothers designed to preserve local cubic trends while reducing short-term irregular variation. Their weights are derived by minimizing the sum of squared third differences in the smoothed series. As a result,…
This Chinese-language forum exchange addresses a data handling issue that can arise after computing regression coefficients with an OLS factor expression. The questioner finds that beta values appear in DataFrame cells as array-like objects rather than as…
This indicator separates traded volume into positive and negative histogram bars according to the direction of each candlestick. A bullish candle contributes a positive bar, while a bearish candle contributes a negative bar. The display offers two ways to…
This post outlines a Chinese equity screen using turnover between 3% and 12%, a daily gain greater than 1%, main-board membership, and a K-line condition described as below 20. The accompanying code example instead checks whether the latest K value is below…
This short Chinese-language forum exchange addresses how to preserve promising results from a quantitative strategy’s hyperparameter search. The example need is to save parameter sets whose score exceeds a chosen threshold. The poster reports that attempts…
The document summarizes two strands of research. One examines analyst-run mutual funds across multiple fund companies, reporting that analyst-managed funds outperformed comparison funds on risk-adjusted measures and that stronger analyst skill was associated…
The document introduces serial correlation, or autocorrelation, as a way to describe dependence between observations in a time series. It reviews expectation, variance, standard deviation, covariance, and correlation, then applies these ideas to time series.…
The document describes a stock screen that requires price amplitude above 1 and RSI below 65, then ranks candidates by capital-flow strength. It defines amplitude from the day’s high and low relative to the previous close, and describes a flow proxy that…
The document presents a quadratic regression indicator that fits a parabola to recent closing prices. It uses a configurable lookback window, set to 100 by default, and calculates the polynomial coefficients from rolling sums of the bar index, its square,…
This indicator presents processed price time series as candlesticks associated with an RSI and volume-based calculation. It uses one upper and one lower threshold, with a colored band between them. Candles that close within the band use conventional green or…
This overview introduces regression as a way to model relationships between predictors and outcomes, then compares seven techniques: linear, logistic, polynomial, stepwise, ridge, lasso, and ElasticNet regression. It describes common use cases and…
This Chinese equities screening idea selects stocks with daily amplitude above 1, positive price-to-earnings ratios, and at least two limit-up moves in the prior 500 days. The article interprets amplitude as a sign of trading activity, positive PE as a basic…
This document explains an oscillator intended to compare institutional and retail activity using the Positive Volume Index (PVI) and Negative Volume Index (NVI). It describes the conventional interpretation behind the comparison: high-volume periods are…
The document discusses ways to turn MACD readings into a trading filter. It distinguishes corrected values from filtered values and cautions that the two should not be combined as though they were equivalent, because their values differ. It outlines three…
The document outlines a framework for allocating across asset classes according to macroeconomic trends. It first identifies whether individual indicators are rising or falling, using historical averages or filtering, then compares asset returns over the…
This Chinese-language research abstract describes a study of order sizes in stock-level tick data. It reports a Pareto-like pattern: small orders occur more often than large orders, with order-size frequencies declining according to a power law. The authors…
This brief notice outlines a composite scoring framework for strategies listed in a marketplace. It assigns weights to three broad categories: performance, risk, and risk-adjusted performance. The listed performance measures are cumulative return and success…
This study compares time-series and cross-sectional standardization of price and volume factors in an A-share stock-selection setting. It uses a dataset spanning 2011–2022 with 98 daily factors and a five-day return label, and compares LightGBM and deep…