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 brief support discussion addresses a backtest failure in a modular trading strategy. The response suggests that module m22 may not be receiving data from its upstream input, and recommends checking that connection and whether the input actually contains…
The guide explains how impermanent loss arises when the relative prices of two tokens change while they are held in an automated market maker (AMM) liquidity pool. It contrasts a liquidity position with simply holding the deposited assets, describes how…
This tutorial describes how to create a forward-looking label for whether a Chinese stock will close limit-down, outside a price limit, or limit-up on the next trading day. It uses a precomputed stock factor table and a price-limit status field whose values…
The author describes using an AI assistant to rebuild a customized trading application from VeighNa 3.9 to 4.3. The work covered a changed modular architecture, a new database backend, a redesigned strategy data model, and a rewritten desktop interface. The…
This Q&A examines a Chinese A-share small-cap strategy that selects the market’s smallest stocks by capitalization. The user expected the backtest to buy at its August 24 start date, but the first trade appeared on September 23. The explanation is that the…
This document describes a MetaTrader 5 panel for manual trade management, with automatic entries enabled only in Strategy Tester mode. Traders choose direction and adjust entry, stop, and target levels; the panel can suggest levels from hidden pivot-based…
This document describes an MT4 Expert Advisor that uses the Stochastic Oscillator to trade potential overbought and oversold reversals or momentum shifts. It lists default upper and lower thresholds of 80 and 20, with configurable %K, %D, and slowing inputs.…
This Chinese-language post describes a stock screen that selects shares with turnover between 3% and 12%, an initial listing year of 2021, and a candle formation labeled a Kute Intelligence Morning Star. The pattern is framed as a possible reversal after a…
This discussion addresses a backtest in which purchases across two mornings totaled more than the stated initial capital. The explanation attributes the apparent overspending to the simulation’s timing convention: the backtest accesses that day’s completed…
This report studies whether boosted tree models can extract stock return signals from analyst expectation factors that linear models may miss. The authors note that these factors can be complex, have substantial missing data, and correlate with earnings and…
This release note describes changes to Zipline, a Python framework for running algorithmic trading systems. It adds command-line and IPython notebook ways to execute algorithms, plus a history function that supplies rolling market data to a strategy. The…
This Chinese stock-screening post describes selecting shares priced below 12 yuan, with turnover between 3% and 12%, and a pattern it calls a reversal or “wraparound” candle. Its indicator conditions include a low price threshold and an RC value no greater…
The post proposes screening Chinese stocks for MACD above zero, upward separation of short moving averages, and a prior-day auction turnover condition. It presents the combination as a way to select for positive momentum while using trading activity as a…
This stock-selection note describes a short-term screen for Chinese equities based on three conditions: turnover between 3% and 12%, a ratio of outside to inside trading volume above 1.3, and a closing price above the previous day’s low. The combination is…
This research note examines shifts between large-cap and small-cap leadership in Chinese equities, using the CSI 300 and ChiNext indices as proxies. It argues that rotation can reflect macroeconomic conditions, industry trends, and market sentiment, and that…
This research summary studies iron ore returns using factors from the commodity supply chain as well as macroeconomic variables. It groups fundamental drivers around supply, demand, inventories, and costs, and reports findings from a vector autoregression…
This Expert Advisor pairs trend-change signals from an Iin_MA_Signal indicator with position sizing that responds to recent trade outcomes. Signals are acted on after a bar closes when the indicator shows a trend change. Buy and sell histories are evaluated…
The post describes a technical stock screen combining daily price amplitude above a threshold, a positive weekly MACD reading, and detection of a bullish morning star candlestick pattern. It presents the morning star as a three-candle reversal signal and…
The post outlines a stock screen focused on shares whose codes begin with 60, using data from 2021. It presents this universe as relatively large-cap and liquid, then proposes adding financial filters such as a current ratio threshold and a price-to-earnings…
The document summarizes research on whether actively managed equity funds earn stronger future returns when their trading increases the technological similarity of their holdings. It describes Active Technological Similarity (ATS), built from the overlap in…
This entry describes an Expert Advisor that uses the Center of Gravity indicator. It generates a signal when a bar closes with the indicator crossing its signal line, making the bar close the stated confirmation point. The entry specifies that the compiled…
This overview of a Chinese securities firm’s research describes a shared AI platform for financial data, model development, GPU computing, backtesting, performance analysis, and deployment. It also discusses deep reinforcement learning for stock selection…
This document describes an Expert Advisor that uses a Slow Stochastic oscillator for entry decisions. It offers alternative signal modes based on the oscillator changing direction, crossing the 50 level, or a change in the signal-line cloud color. The…
This guide explains how FreqAI trains and deploys adaptive machine-learning models in live or dry trading and in historical backtests. Live operation can retrain models as capacity permits, use the latest trained model for predictions, and apply limits on…