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
The document summarizes research on using corporate earnings-call transcripts to predict subsequent stock-price movements. Earnings calls contain management discussion of financial performance and information that investors and analysts may use when making…
This study examines why publicly listed Asian companies disclose environmental, social, and governance information. Using panel data and regression models, it tests whether economic sustainability performance, profitability, leverage, and company size are…
This indicator method applies a linear regression channel to RSI values and treats excursions beyond the channel as potential mean-reversion signals. When RSI rises above the upper band, it marks a possible sell; when RSI falls below the lower band, it marks…
The document proposes adding a simple API to retrieve selected strategy performance metrics over a specified date range. Examples include the Sharpe ratio, annualized return, benchmark return, and alpha. The request responds to a gap between an existing…
This essay draws on Nassim Taleb’s The Fool’s Random Walk to argue that decisions should account for chance, uncertainty, and the role of luck in apparent success. For traders, its central distinction is between counting probabilities and gambling on…
This notebook demonstrates how to prepare synthetic prices and sparse event signals for Alphalens. It creates a small panel of prices for six securities, then marks selected date-security pairs in an event factor while leaving other entries missing. The…
The discussion explains that a supervised model learns from factors paired with labels, where a label defines the target. For stock selection, the target might be future returns over a chosen horizon. The model and label together determine what predictions…
The Aroon oscillator is defined as the difference between Aroon Up and Aroon Down. Each component measures how recently the market reached a high or low over a chosen lookback period; subtracting the two provides a single indicator intended to express which…
This proposed equity screen selects stocks with turnover between 3% and 12%, a listing year of 2021, and a dividend ratio above 25% for 2019. The article presents these as activity, company-age, and historical payout filters. It cautions that relying on a…
The document answers whether a factor can be neutralized for both market capitalization and industry exposure. It recommends including market capitalization as a continuous explanatory variable and industry membership as dummy variables in the same…
This event study examines how Chinese A-share prices behave around formal earnings reports, especially when companies previously issued earnings forecasts. It compares reported results with forecast ranges and groups companies by forecast category or by the…
This report compares groups of Chinese equity factors across industries, focusing on cash flow, valuation, size, momentum, liquidity, volatility, beta, and alpha. Its stated findings come from portfolio net value behavior, industry-level results, ranking…
This report summary describes a stock-duration factor inspired by bond duration and constructed using company financial data and analyst consensus estimates. It reports that the factor’s cross-sectional behavior varies with market conditions and across…
The post raises a data-timing concern when using iWencai in quantitative research. Its author says that get_iwencai() does not introduce future data, while some uses of query_iwencai() may do so. The practical takeaway is to check whether a query relies only…
This research summary asks whether financial data vendors act as information intermediaries in equity markets. It focuses on FirstCall, a Thomson Reuters service, and whether the timing of its distribution of earnings information affects how investors…
This 2020 research summary argues that the Shenzhen 100 could suit long-term allocation, describing it as a liquid large- and mid-cap index with both growth and blue-chip characteristics. It cites fundamentals reported at the time, including return on equity…
This Chinese-language post proposes screening mainland Chinese stocks using an RSI below 65, a daily gain above 1%, and a main-board listing. It adds a liquidity condition: yesterday’s turnover rate multiplied by today’s opening-auction volume divided by…
The document introduces network value to transactions (NVT) as a crypto-market analogue to the price-to-earnings ratio. It defines NVT as market capitalization divided by transaction volume, treating on-chain transfer activity as a rough measure of network…
The document describes a utility for saving the historical OHLCV data available on a MetaTrader 5 chart to a CSV file in the platform’s common data directory. It is presented as an alternative when the standard Symbols request tool fails to return all…
This note outlines a nonlinear Kalman-style filtering method attributed to John Ehlers. It first smooths price with an exponential moving average or a three-pole filter, calculates the difference between price and that smoothed baseline, and smooths the…
The analyzer evaluates closed-trade results from a CSV containing a date and profit for each trade. It reports how much gross profit came from the largest winners at several cutoffs, and calculates the share of net profit attributable to the top N trades. A…
The report compares on-chain behavior around Bitcoin volatility shocks in bull and bear periods. It defines a shock as a day when absolute daily return exceeds two standard deviations above its historical average, then studies 21 shock days in each market…
These release notes describe additions to Zipline’s Pipeline API in version 0.9.0. New datasets expose buyback authorizations and dividend information organized by ex-date, payment date, or announcement date. Related built-in factors measure business days…
The Adaptive Price Zone is a price channel built from two moving-average calculations. One moving average is applied to a configurable price series, and a second is applied to the bar range, defined as the high minus the low. A width setting scales the range…