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 Chinese-language post describes an equity screen requiring a daily price range above 1%, a positive 10-day return below 35%, and positive reported returns. It frames the range threshold as a way to select for active price movement, while the bounded…
This post describes a China equity screening idea based on price compression and a Bollinger Band location. It selects stocks whose 5, 10, 20, 60, and 120 day moving averages are described as overlapping, whose close lies between the Bollinger middle and…
This Freqtrade strategy combines two Commodity Channel Index readings with Chaikin Money Flow and Money Flow Index conditions to identify long entries and exits. It calculates these indicators on the input candles and adds moving averages from a resampled…
The post proposes a Chinese equity screen for beverage and alcohol import-export companies, selecting stocks with turnover between 3% and 12% and shareholder equity above 200 million. It gives example platform formulas and Python references, including…
This study builds a convertible-bond strategy that separates bonds into equity-like, balanced, and bond-like groups using parity premium relative to bond value. It models each bond as a floor component plus option time value, then examines where factor…
This research proposes treating convertible bonds differently according to whether they are bond-like, balanced, or equity-like, classified by the premium of parity value over bond value. It explains a valuation decomposition in which a bond floor or…
This document outlines a two-stage workflow for calculating Alpha101 factors. First, it reads daily stock data, derives base series such as returns and VWAP, and computes time-series intermediate variables for storage. Later, factor construction retrieves…
This configuration describes a Qlib workflow for training a KRNN model on China’s CSI 300 universe with Alpha360 features. It sets a historical data range, uses robust feature normalization and cross-sectional label ranking, and defines a two-day forward…
This algorithm implements a monthly cross-sectional momentum strategy in US equities. It ranks eligible stocks by their trailing 12-month return, with the most recent month skipped to reduce microstructure and liquidity biases. The highest-ranked group is…
The article describes three ways strategy research can produce misleading backtests: look-ahead bias, excessive parameter optimization, and curve fitting. Its examples show how using a bar’s eventual close to trigger an earlier trade, or assuming a breakout…
This Chinese A-share screening idea combines turnover between 3% and 12%, three consecutive declining sessions, and exclusion of ST-designated stocks. It proposes selecting the five highest-ranked stocks that hit their daily price limit, with selection…
This stock screen selects candidates before each trading day’s open when MACD is above zero, price-to-earnings is positive, and the KDJ indicator has just formed an upward crossover. The article presents MACD as a trend filter, positive P/E as a basic…
This Chinese equity screen combines three short-term filters: price amplitude above 1, a ranking based on large-order net activity, and a 9:25 a.m. gain below 6%. The stated rationale is to identify active, volatile shares whose early move is comparatively…
This opinion piece argues that quantitative trading’s advantages extend beyond execution speed. It emphasizes systematic research and iteration: models can combine many fundamental and price-based signals, test relationships across historical data, and…
This overview compares machine learning and deep learning methods for predicting stock prices and trends. It describes LSTM and GRU recurrent networks for sequential data, CNNs for extracting patterns, bidirectional LSTMs, and deep neural networks. It also…
This post proposes screening Chinese stocks by daily price amplitude, a measure described as current-day shareholder control, and prior-day major-investor net buying. It presents the combination as a way to capture market sentiment and capital flows. The…
This ProRealTime indicator estimates average forward price movement conditional on the current reading of RSI, Stochastic, Williams %R, or ADX. It divides indicator readings into 20 five-point ranges, then separates observations according to whether the…
This indicator modification takes a standard volume measure and divides it by the spread. The author removed the original color-changing behavior because it was considered unhelpful after the transformation, then added an option to turn spread adjustment on…
This intraday A-share strategy combines strong thematic sectors with stocks that have recently reached their daily price limits. It ranks concept indexes by short- and medium-term returns, retaining the top two percent on both measures, then screens stocks…
This essay argues that simulated trading can test whether a strategy is viable before risking real capital, while acknowledging that success in simulation does not guarantee live profits. It presents practice as a way to learn execution details and reduce…
A forum exchange explains when bar data can introduce lookahead bias in a trading strategy. The question concerns whether a bar's closing price is available at the time a daily strategy makes its decision. The response says that, in daily strategies, using…
The proposed Chinese equity screen combines three conditions: at least five moving averages are said to overlap, the 9:25 opening gain is below 6%, and the 30-day moving average is rising. The accompanying explanation interprets overlap as relatively stable…
The system opens an initial position at a fixed lot size when the DeMarker indicator permits entry. If a position is already open and price moves a specified distance from the most recent entry, another position may be added when the indicator again allows…
This post describes a stock selection screen that combines RSI below 65, seven consecutive sessions with the close below the open, and declining fund-flow strength. It frames the conditions as a mix of technical signals and investor activity, then offers…