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.
21,023 documents
This brief strategy proposal starts with the constituents of the CSI 300 and looks for stocks whose latest price remains within an 8% pullback from the highest price of the past 30 days. It also compares average trading volume over the latest seven sessions…
This stock-screening approach combines two technical indicators with basic company filters. It selects Chinese A-share companies with market capitalization below 10 billion yuan, positive profitability, RSI below 65, and MACD above its zero line. The post…
This brief description presents a modified Chaikin volatility indicator with a stochastic view, displayed as a signal cloud. It identifies the original author and notes that the indicator was first published in MetaTrader 4 before being implemented for…
This overview introduces ten machine learning methods and sketches how they work, with examples of possible financial uses. It covers linear and logistic regression, K-nearest neighbors, support vector machines, decision trees, random forests, neural…
The document discusses converting one-minute stock data into daily observations for high-frequency price and volume factors. The example query constructs features such as volume shares across intraday intervals, price and volume correlations, correlations…
This document describes a software class for a probabilistic neural network used for classification. A constructor takes the input-vector size and number of target classes; class labels are zero-based and must be consecutive. Training data are supplied as…
The document describes the Volume Price Confirmation (VPCI) indicator, which is intended to express how price and volume relate. It identifies fast and slow moving-average periods as inputs and defines VPCI as the product of three components: a price…
This article proposes a qualitative framework for finding Chinese stocks with potential for medium- to long-term growth during annual and quarterly reporting periods. It combines three checks: whether recognized active investment firms appear among…
This indicator compares price pivots with ten momentum and volume measures, including RSI, MACD, momentum, OBV, stochastic, and Chaikin Money Flow. It counts how many enabled oscillators show each of four divergence types at the same pivot: regular bullish…
This platform discussion addresses how to understand a precomputed factor, using a MACD signal-line factor as its example. The response says that the platform does not provide the calculation process for such precomputed data. Users can inspect the factor's…
This stock-selection screen filters for shares with turnover between 3% and 12%, an external-to-internal volume ratio above 1.3, current volume above 10,000 lots, and an opening price equal to the day’s high. The accompanying rationale treats turnover, the…
This document introduces an oscillator-style indicator that presents total volume accumulated over a selected calculation period. It describes two independent display choices: volume may be shown in absolute units or relative values, and it may appear as a…
The document summarizes an equity factor study on whether analyst attention predicts future company fundamentals and stock returns. It proposes adjusting a conventional analyst-coverage measure with a linear regression that removes effects associated with…
The document describes a MetaTrader script that collects broker-specific trading costs for one symbol and exports them for a companion model used in triple-barrier labeling. It samples spreads from ticks and reports tick-weighted, dwell-time-weighted, and…
This Chinese-language post proposes a stock-selection screen based on a daily price movement threshold, a listing history longer than one year, and a ratio between today’s auction volume and the prior day’s volume that falls between 0.5 and 2. The rationale…
This indicator description isolates the trend component of Holt’s double exponential smoothing rather than displaying the smoothed on-chart value used in a typical presentation. It proposes reading that component as an oscillator for trend seeking, with…
The document introduces a financial time-series study using returns on the CSI 300 index. It says the study applies ARCH and related models, compares them, and attempts to fit and forecast both prices and volatility. This places the work in the area of…
This Chinese-language forum post raises a backtesting problem: gaps in stored Bitcoin market data can leave an open position without later bars on which the strategy could close it. The example describes a long period of missing exchange candles and asks how…
This indicator applies Symbolic Aggregate approXimation (SAX) to search for historical price windows resembling the latest closed-bar window. It z-normalizes each window to emphasize shape, compresses it into segment averages, and encodes those values as a…
This installment of an event-driven backtesting series explains how to summarize portfolio performance from an equity curve. It describes three measures: annualized Sharpe ratio, maximum peak-to-trough drawdown, and drawdown duration, defined as the number…
The document explains a threshold-selection method for pairs trading when the modeled log price follows an Ornstein–Uhlenbeck process. It builds on earlier work by deriving expressions for the expected first-passage time with two-sided boundaries, then uses…
This research summary examines four MSCI smart beta equity indexes—value weighted, equal weighted, momentum, and minimum volatility—against a capitalization-weighted benchmark. Using monthly observations from 1999 through 2014, it applies principal component…
This research summary describes stock-selection factors built from tick-level aggressive buy and sell amounts. It applies an informed-trading probability model to filter noisy order flow, then forms informed sell share, informed buy share, and net informed…
This technical report describes a dtype-sensitive error in VeighNa’s linear-decay time-series calculation using Polars rolling_map. In the reported example, a sparse binary series with one nonzero value produces zeros when stored as Int64, while the Float64…