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 article frames the execution of a large stock order as a tradeoff between market impact and exposure to price risk. Executing quickly may push the price against the trader, while slower execution leaves inventory exposed to uncertain price movements. It…
This short tutorial introduces Python control flow for writing programs. It names conditional logic with `if` statements and two forms of repetition: `while` loops, including use of `break`, and `for` loops. These constructs let a programmer choose which…
This stock screen combines a turnover-rate range of 3% to 12%, a pattern described as a reversal or engulfing move, and positive institutional fund flow. The accompanying discussion treats turnover as a way to filter trading activity and institutional flow…
This paper guide summarizes research on convolutional networks, model efficiency, and image segmentation. It describes channel attention through feature recalibration, depthwise separable convolutions for reducing computation, and encoder-decoder or…
This indicator fits a linear regression to closing prices from the most recent segmentation point through the current bar. It estimates residual standard deviation and uses a configurable multiplier to place upper and lower bands around the regression line.…
This Chinese-equity selection idea filters stocks associated with the metaverse theme, requiring prior-day actual turnover to fall between 3% and 28% and share capital to exceed the stated threshold of 200 million. The note describes turnover as a measure of…
This indicator aims to mark possible transitions into and out of range-bound conditions using the directional indicators DI+ and DI− alongside Parabolic SAR. It compares the SAR position with the close and checks which directional indicator is greater,…
This document reports a two-year daily Bitcoin–USDT backtest on Binance, covering January 2019 through January 2021. The strategy is identified only as “AwesomeStrategy”; no entry or exit rules, position sizing, or execution assumptions are supplied, so the…
The article explains idiosyncratic volatility for individual stocks and extends the idea to residual volatility for indices and commodity futures. It argues that elevated volatility is more useful for anticipating the size of future moves or divergence from…
This brief forum exchange concerns an SQL error encountered after adding a z-score standardization function to factor data. The reported issue is that a query joining tables cannot determine which table supplies a column named date. The response recommends…
This stock screen combines three technical conditions: at least five moving averages converge, the KDJ indicator has just formed a bullish crossover, and the stock’s 10-day return is positive but below 35%. The stated rationale is that moving average…
This reference page catalogs technical indicators and data utilities available in QTPyLib for use with bar data. The built-in list covers volatility and range measures, moving averages, channels, momentum oscillators, returns, volume-related measures, price…
The B-indicator estimates trend strength by comparing the absolute smoothed trend with a noise estimate and scaling the ratio to a 0–100 oscillator. It is designed to show whether price movement is more trend-like or noise-dominated, without indicating…
The article explains how a multi-asset daily market-data panel can be turned into model-ready factor and label tables. It separates the workflow into defining train, validation, and test date ranges; adding feature expressions and a future-return label;…
This indicator note explains an oscillator derived from Heikin-Ashi candles. It defines the Heikin-Ashi close as the average of the current bar’s open, high, low, and close, while the Heikin-Ashi open is calculated from the previous Heikin-Ashi open and…
This Chinese-language competition update reports diagnostics for the ten highest-ranked anonymous factors and tracks a CSI 1000 index-enhancement portfolio. It describes average style exposures estimated through daily multivariate regressions and…
The document asks how to express a proposed factor in a platform's formula language. The factor is defined as the negative correlation, over a rolling interval, between a security's VWAP series and its volume series. In principle, this transforms the…
This document introduces two high-frequency trading examples: handling a dataset for reinforcement learning and predicting price trends. It explains that the dataset is represented by a Qlib DatasetH object, which can be serialized to disk and reloaded.…
The report outlines a multi-asset allocation framework that links macroeconomic risks to broad asset classes, styles, and industries. It identifies economic growth, interest rates, exchange rates, inflation, and credit as five central risk factors. A…
This report compares hourly returns, trading volume, and volatility for BTC, ETH, DOGE, XRP, USDC, and USDT across several centralized exchanges. It uses standard deviation of hourly returns as its main volatility proxy and also examines Garman-Klass…
The research note studies tail dependence between individual stocks and the broader market. Using copulas, it estimates upper-tail and lower-tail correlation, which describe whether assets tend to move together during extreme outcomes. The reported…
The forum response explains one possible reason that strategy messages written through a logging method appear in the application but not in log files: file output may filter messages below a configured severity threshold. It says that strategy messages…
The discussion explains a rolling sequence window as a way to assemble recent observations into a single training example. For instance, a model can receive a window of past factor values spanning several days, with the window advancing through time to…
This indicator models each price bar as a disturbance whose effect fades over time. It treats bars with Open above Close as rising bubbles and bars with Open below Close as falling drops, then uses the Open–Close difference as the disturbance amplitude. A…