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 proposed China A-share screen selects stocks associated with the metaverse theme, requires prior-day turnover above 8%, and targets year-over-year growth in net profit attributable to parent-company shareholders between 20% and 100%. The post explains…
This short Chinese-language community thread asks how win rate is defined for a trading strategy. The replies suggest one possible interpretation: the number of days with profitable total holdings divided by the number of backtest days. The participants…
This analysis uses on-chain valuation, miner, and network activity measures to assess Bitcoin’s 2025 market regime and frame questions for 2026. It explains MVRV, its standardized Z-Score, NUPL, and realized price as measures of holder cost basis and…
This utility module prepares trading results for performance analysis. It extracts returns, positions, and transactions from a backtest, normalizes dates, and converts positions into a format suitable for reporting. It also includes display helpers,…
This notebook walks through an Alphalens workflow for assessing alpha factors, which assign a value to each asset at each date and are judged by how those relative values relate to subsequent returns. It demonstrates loading daily stock prices, organizing…
This tutorial explains how to prepare data and use Alphalens to evaluate cross-sectional alpha factors. It distinguishes factor research, which tests whether relative signal values predict later returns, from trading strategy evaluation, which also needs to…
This short indicator description compares the upper and lower shadows of each candlestick and marks a candle when one shadow is sufficiently larger than the other. The user sets a multiplier threshold; for example, a threshold of two marks a candle when…
This indicator guide explains volume profile as a histogram of trading activity across price levels within a chosen time range. Depending on the data available, it can count estimated trades or aggregate real volume, and it can display distribution modes,…
This tutorial describes using a recurrent LSTM network to forecast stock closing prices from recent price history. It forms rolling sequences from historical data, reshapes them into the three-dimensional input expected by the network, and separates each…
The Delta indicator plots two series: the bar’s average price, calculated from its open, high, low, and close, and a second series derived from the decimal logarithm of the ratio between the previous average price and the current one. Its stated formula adds…
This Chinese A-share screening proposal combines the 9:25 opening move, technical-indicator crossovers, and relative trading activity. It selects stocks with an opening gain below 6% versus the prior close, ranks volume ratio and takes the top 100, and seeks…
The document introduces pandas as a Python library built on NumPy that simplifies data handling. It identifies the DataFrame as the chapter’s main subject and notes that pandas also includes Series and Panel structures. The available text provides only a…
This post outlines a study of autoregressive (AR) and mean-reverting (MR) models applied to CSI 300 prices and returns. It proposes using the models to fit and forecast the index series, and says the article will explain the modeling workflow and compare…
This overview introduces machine learning as a way to learn patterns from data and use them to make predictions. It outlines a typical workflow: divide data into training, validation, and test sets, fit a model, evaluate it, then tune it before applying it…
The article uses simulations to argue that a strategy with a modest standalone Sharpe ratio may still add value when its returns have low correlation with an existing portfolio. It models twenty daily signals with an annualized Sharpe ratio of 0.6, varies…
The essay presents technical analysis as a way to measure price behavior and respond to market change, rather than predict the future. It advocates a structural approach that classifies price movement, seeks entry and exit points, and treats conclusions as…
The document summarizes research on how macroeconomic conditions relate to cyclically adjusted price earnings ratios and stock returns. CAPE smooths earnings using a long inflation adjusted history, which can make it informative about long horizon returns…
The document describes an advance-decline line indicator intended to show trend strength and possible changes in direction. Its calculation uses the positive and negative components of candlesticks over a selected period and takes their difference. The…
The document explains how to find where two straight lines cross on a MetaTrader 4 chart. It maps chart time to the horizontal coordinate and price to the vertical coordinate, then represents each line using either two known points or one point and its…
This note discusses the relative roles of predictive factors and model design in quantitative strategies. It argues that factor quality usually sets the ceiling for potential performance, while model quality determines how effectively a strategy can use the…
This Chinese stock-screening post combines a measure of trading activity with historical revenue growth. Its activity measure multiplies the prior day’s turnover rate by the current day’s auction volume divided by the prior day’s total volume, and it…
This forum post addresses how to retrieve a time-series calculation for one specific date using SQL. The reported problem is that filtering the input directly to the target date can leave a rolling calculation empty, because the calculation lacks earlier…
The article describes an unsupervised approach to identifying market regimes from the returns of 17 US factors beginning in 1970. A Gaussian mixture model clusters observations by their return distributions; cross-validation is used to choose four clusters.…
This analysis studies minute-level order-book snapshots for Binance’s BTC/FDUSD market over a six-week period in 2025, measuring displayed depth at distances from 5 to 100 basis points around the midpoint. It finds that depth accumulates non-linearly:…