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
This note describes a custom Williams Percent Range indicator for use with an expert advisor. Its purpose is to expose the plotted line color, line width, and two threshold levels as input settings, so the indicator display can match the EA’s trading rules…
This code example demonstrates a data workflow for a FinRL stock-trading project. It fetches a single stock’s price history through both yfinance and FinRL’s Yahoo downloader, then downloads a Dow 30 universe over configured training and trading periods. A…
This retrospective contrasts rule-based stock selection with machine-learning ranking and describes backtesting as a way to evaluate a strategy on historical market data. Its central caution is that a strong fit on a small sample can reflect an irrelevant…
The document describes an Expert Advisor that trades signals from the DigitMacd oscillator. It enters when a bar closes and the indicator’s cloud changes color, making the color transition the strategy’s central trigger. The EA depends on a separately…
This configuration defines a Qlib experiment that trains an IGMTF model on Alpha360 features for CSI 300 stocks. It uses historical data from 2008 through 2020, with training through 2014, validation in 2015–2016, and a held-out test period beginning in…
This research report examines market timing with breadth measures built from the constituents of the CSI 300. A diffusion indicator estimates the share of index stocks showing positive trends, offering a view of how broadly market direction is shared. The…
This assignment turns a discretionary idea—finding concentrated holdings in recent hot industries—into a proposed equity research process. It suggests first identifying strong sectors with a sector momentum factor, then ranking stocks within those sectors…
The document describes a framework for evaluating equity factors and combining selected factors into a portfolio. It estimates factor returns with periodic cross-sectional robust regressions, measures the relationship between factor exposures and subsequent…
This article explains industry rotation in China’s A-share market, describing how economic cycles, policy, fundamentals, and herding can cause leadership to shift across sectors. It outlines a momentum approach that ranks industry indexes by weighted returns…
This Chinese equities strategy looks for stocks with a bullish moving-average stack: the five-day average exceeds the ten-day, then the twenty-day, forty-day, and one-hundred-twenty-day averages. It adds a pullback condition: the day’s low must fall below…
This release note describes changes in OctoBot 1.0.2. ChatGPT strategy profiles can now be backtested for some settings using prompts precomputed from historical exchange-pair data downloaded by the service. The ChatGPT profile also shifts from daily trading…
This forum post reports a VeighNa CTA backtest that reaches the historical-data loading stage but loads zero records. The script then completes initialization and replay with no trades, before result calculation fails because the daily results table lacks a…
This A-share stock screen combines a technical condition, a fundamental filter, and a short-term activity measure. It selects stocks with RSI below 65, parent-company net profit growth above 20% and at most 100%, and a turnover-based measure between 0.5 and…
The document explains how to measure Expert Advisor runtime in MetaTrader 5 while excluding delays that can distort the tester’s reported duration. It distinguishes total run time, which can include history synchronization and tick preprocessing, from net…
This overview explains how deep learning models can be assembled from input, intermediate, and output layers in a visual strategy-building platform. It surveys layer families including convolution, pooling, recurrent networks, embeddings, noise and dropout,…
This guide explains how to turn a trading idea into a Freqtrade strategy, from generating a template to defining indicators, entry and exit signals, stop losses, and optional position adjustments. It describes how Freqtrade represents candle data in pandas…
The document contrasts selecting stocks by relative rank with selecting them by an absolute model score. A ranking method can always choose the highest-ranked names in a universe, even when their scores are weak. The proposed alternative sets a minimum score…
This proposed Chinese equity screen selects stocks whose daily high-low range exceeds 1%, that had at least one limit-up day in the preceding 25 trading days, and whose circulating market value exceeds 10 billion yuan. The article interprets the range…
This document explains MACD as a way to compare a short-term and a long-term exponential moving average of closing prices. Their difference, DIFF, is smoothed into a signal line called DEA; MACD is calculated from the gap between DIFF and DEA. The standard…
The post outlines a stock screen combining three conditions: intraday high-low range relative to the open above 1%, a stated 2019 dividend-yield measure above 25%, and daily change from the open between -5% and 2.6%. It presents the conditions as a way to…
The article explains robust portfolio optimization as a way to reduce the effect of errors in expected-return forecasts. Rather than optimize only for a single set of estimates, the methods consider adverse plausible cases and seek a portfolio that performs…
This comparison explains differences between TqSdk and vn.py that matter when adapting existing trading strategies. vn.py is presented as an integrated package with market data, trading connections, storage, and interface components. TqSdk instead uses…
The article discusses data integration challenges when developing strategies across US equities and forex. It highlights differences in update speed, price conventions, and data formats, arguing that timestamp misalignment and latency can create gaps between…
The article proposes reducing noise in stock prices with wavelet decomposition, fitting ARMA models to the resulting coefficient series, forecasting those coefficients, and reconstructing a one-step-ahead price estimate. Its denoising explanation relies on…