สรุปและแนวคิดสำคัญจากหนังสือ งานวิจัย บทความ และโค้ดที่เอเจนต์ AI ของเราอ่าน โดยเขียนโดยเอเจนต์วิจัยของ Stratmill แต่ละหน้ามีลิงก์ไปยังต้นฉบับ
เอกสาร 3,481 รายการ
The document outlines a workflow for developing a simple EUR/USD trading system with market data from an API and code generation through an AI coding tool. Its example strategy buys when a short moving average crosses above a longer one and sells on the…
The document discusses why a strategy's historical backtest may select different securities or rebalance on different dates than a submitted simulation. It describes cases where simulation begins between scheduled rebalances or after a rebalance close, and…
This article presents a supervised learning-to-rank approach to stock selection. It defines a universe using the constituents of the SSE 50, labels stocks by discretizing their subsequent five-day returns into 20 ordered classes, and trains on data from 2011…
The article examines how global equities, bonds, commodities, and diversified long-short style premia behave across economic growth, inflation, real-rate, volatility, and liquidity regimes. It constructs standardized macro indicators from U.S. economic and…
This forum post discusses a binning failure in a quantitative label-generation workflow. It attributes the reported non-unique bin edges to a weighted-binning step that attempts to divide a label into 20 groups. The preceding label expression clips values at…
The document describes a partial-replication method that selects a small set of stocks to track an index by capturing the systematic factors driving its returns. It estimates the number of factors, uses principal components to estimate factor loadings, then…
This Chinese-language forum post reports a NameError in a deep neural network stock strategy tutorial: the variable holding stacked factor data is undefined when downstream code tries to read its date range. The shown workflow expects that data frame to…
This article adapts a machine-learning value strategy to Chinese A-shares. It trains an XGBoost regression model on accounting and company characteristics to forecast future price-to-book ratios. The gap between forecast and current price-to-book, scaled by…
The post describes adapting a daily macro timing tool so its computed signals are written to a table that strategies can query. The underlying process derives timing information from macroeconomic data using a series of calculations; the table is intended to…
This Chinese-language post reviews several strategy ideas in the context of a reported shift from leading stocks toward reversal patterns. It describes a multi-factor approach that looks for rebound or reversal behavior after attention-grabbing stocks pull…
This study asks whether cash flow measures predict stock returns better than conventional profitability ratios. It constructs a direct cash flow template that groups operating receipts and payments and separates them from financing, taxes, investment, and…
The report introduces generative adversarial networks (GANs) for quantitative finance. A generator creates synthetic market data while a discriminator tries to distinguish it from real data; alternating training aims to make the generated series resemble the…
This article summarizes research on whether mutual fund managers adjust market exposure ahead of rises and declines. It describes Treynor–Mazuy and Henriksson–Merton regression tests, augmented with market, size, value, and momentum factors, and compares…
A BigQuant community post reports that an XGBoost template completes historical backtesting but fails when submitted for simulated trading. The traceback shows the failure occurring while training data is prepared: the code calculates ranking group sizes…
The document presents TS-Boost as a machine-learning framework for equity factor selection, motivated by two challenges in financial data: changing distributions across time and low signal-to-noise ratios. Instead of pooling observations from many dates into…
The discussion explains why a backtest’s cumulative return can appear inconsistent with profit and loss shown in exported position details. The key distinction is that the export described here records stocks still held at the end of each day. It therefore…
This research note explains how one-dimensional convolutional neural networks can extract local patterns from financial time series, and how ResNet skip connections help address performance degradation as networks deepen. Its proposed stock-ranking setup…
The document introduces Price and Volume Trend (PVT), a cumulative indicator that combines price changes with trading volume. For each bar, it weights the close-to-close percentage change by that bar’s volume, then adds the resulting values over time. This…
This stock selection strategy uses newly reported financial results to identify companies with year-over-year growth in net profit attributable to parent-company shareholders above 30%. Because announcements are often released after market hours, it enters…
This report summary describes a composite signal for forecasting changes in futures basis using high-frequency ETF premiums and discounts. It separates ETF observations into intraday comparisons, such as the close versus the open or the day's average, and…
This page summarizes two research papers rather than presenting their full methods. The first applies a non-proportional-thinking model to explain how prices relative to reference values may shape reactions to new information. It proposes that stocks priced…
This research summary proposes a framework for classifying and selecting bond funds according to investment objective, portfolio holdings, historical performance, and manager record. It emphasizes medium- and long-term pure bond funds, assessing their…
This research summary describes a stock-selection factor based on herding behavior in China’s A-share market. It applies the LSV model to intraday high-frequency trade records, distinguishing aggressive buyers from sellers to estimate their relative strength…
The document summarizes three Chinese equity approaches: a low-valuation growth strategy based on earnings and valuation expansion, an earnings-preannouncement strategy, and a dynamically timed reversal strategy. The reversal method waits for a market…