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
The document describes deploying the small Kronos pretrained model on a quantitative research platform to forecast five-minute candlestick data for stocks. Kronos represents OHLCV sequences as hierarchical discrete tokens using a specialized tokenizer, then…
This research summary compares Naive Bayes, linear discriminant analysis (LDA), and quadratic discriminant analysis (QDA) as supervised methods for multi-factor stock selection. The described process extracts and preprocesses features and labels, retrains…
The document proposes separating sector allocation from industry selection because broad sector shifts are linked more closely to economic conditions, while industry performance depends on factors such as earnings, valuation, and capital flows. It argues…
The strategy screens Chinese equities associated with the metaverse theme. It requires the previous day’s actual turnover to be between 3% and 28%, and at least one limit-up event during the prior 25 days. The rationale is that the industry theme may offer…
This stock-selection idea combines three filters: at least five moving averages converge, turnover lies between 2% and 9%, and the stock’s gain at 9:25 is below 6%. The document interprets clustered averages as relatively stable price behavior, the turnover…
This event outline presents a learning series on using large language models and AI agents in quantitative research workflows, with examples centered on VeighNa strategy development. Topics include agent and tool interaction, model tool calls, MCP, task…
This article describes a Chinese stock selection screen combining three conditions: a daily high-low amplitude above one, a three-session limit-up sequence on the prior day, and a prior-day 9:15 indicative matching price at the limit-down level. It frames…
This Chinese-language guide outlines how to build and run automated strategies on FMZ. It describes the platform’s strategy editor, exchange account configuration, hosted execution, and support for several programming languages and crypto venues, as well as…
The document defines interfaces for a backtesting system that processes historical market events and order interactions. A local processor can submit, modify, and cancel orders, expose positions and state values, report market depth and recent trades, and…
The document presents a proposed screen for Chinese equities that combines a daily-range condition, a dividend payout ratio above 25% for 2019, and alignment among multiple moving averages. It explains the intended rationale as combining price volatility,…
This market-timing method combines moving-average crossovers, MACD, DMA, and TRIX signals to reduce reliance on any single trend indicator. Each indicator contributes a positive point when its faster or primary line crosses upward through its comparison…
This guide explains how to prepare tick-by-tick trades and full order-book updates for HftBacktest, noting that this level of historical data is not commonly available for free in the way daily bars are. For Binance Futures, it describes collecting raw feed…
This note argues that quantitative strategy results depend on the combination of input factors and predictive models. It says that different factors can produce substantially different backtests when paired with the same model, and that a given factor can…
This document describes an Expert Advisor that uses a universal Bezier curve to generate trading signals. It checks for a change in the curve’s direction and forms a signal when a bar closes. The approach is therefore based on curve direction changes rather…
This study examines whether U.S. equity mutual fund managers adjust market exposure when broad stock mispricing changes. It builds a market mispricing measure from a financing-based long-short factor, then orthogonalizes it against macroeconomic variables. A…
This factor study examines whether the share of a stock’s trading volume occurring in the opening call auction contains information useful for equity selection. It builds an opening call-auction volume proportion factor from relative volume and moving…
This indicator searches historical candles for examples whose wick-to-body structure resembles the latest candle, then projects the average price path that followed those analogues. Users can compare general wick length, the upper wick, the lower wick, or…
This document describes a MetaTrader expert advisor that combines a moving average with a Momentum indicator. It evaluates signals when a new bar forms, uses the previous bar's close and the current bar's open when opening positions, and refers to prior…
This brief post introduces a Python implementation of Markowitz portfolio theory, a framework for constructing portfolios by balancing expected return against risk. It says the article works through the theory and implementation, but the supplied document…
The document outlines a China equity industry-rotation strategy using the Shenwan Level 1 industry classification. It ranks industries with a composite momentum measure formed from recent 5-day, 20-day, and 60-day returns, weighted 0.4, 0.3, and 0.3. Each…
This document presents an automated strategy that builds pseudo-Renko bricks from ordinary candlestick closes, calculates a moving average on those brick closes, and trades when the brick close crosses the average. It also includes a minimum time interval…
The document explains how to use VeighNa’s DataRecorder module to save live market data to a database. Recorded ticks and one-minute bars can later be viewed in DataManager, used for historical backtests, or supplied to trading strategies during live…
The document describes an updated step stochastic indicator and lists changes to its implementation and configuration. It says the calculation has been rewritten as a reusable function, with alert support and multi-timeframe operation. Users can choose among…
The document raises a measurement question about Rank IC, a rank-based measure of association between model predictions and subsequent returns. The author reports that Rank IC computed on a custom training set differs substantially from the value produced by…