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
The document compares Ethereum, Polygon, BNB Chain, Avalanche, Arbitrum, Optimism, and Base by their early network histories. It describes how each network differs in architecture, consensus, execution environment, token use, and launch context, then uses…
This note describes a two-stage ranking factor for equities. For each stock, it ranks the recent ten-day low-price observations through time, then ranks those values across stocks on the same date. The intended interpretation is that a larger final factor…
This post describes an equity screening recipe that ranks stocks by volume ratio, uses MACD to identify a purported rounded-bottom pattern, and limits the 10-day price gain to a stated range above zero and below 35. The intended logic is to find stocks with…
The document explains beta as an asset’s sensitivity to market returns and gives the covariance of asset and market returns divided by market-return variance as its calculation. It illustrates the calculation with a short set of monthly portfolio and…
The Oscillator Phase Change Index (PCI) is presented as an indicator for identifying the current market phase. Its configurable inputs are the lookback period, applied price, and overbought and oversold levels. The stated index scales the ratio of upward…
This brief forum exchange asks how to interpret a random forest model’s predicted label after returns have been divided into twenty classes. The question is whether the output represents a class identifier or a return value, and whether regression metrics…
This forum post raises questions about translating factor functions from BigQuant version 1.0 to version 3.0. It asks whether the RSI function retains its name or uses a prefixed variant, and whether ranking should use a cross-sectional or unprefixed…
This MT5 tool helps traders review planned or open exposure across several instruments as a portfolio. Given comma-separated symbols and lot sizes, it calculates a Pearson correlation matrix from historical bar-to-bar returns over a configurable timeframe…
This document describes a class for applying an exponential Ornstein–Uhlenbeck model to mean-reverting portfolio prices. It inherits fitting and portfolio construction from an OU model, then works in log-price space to estimate optimal liquidation levels,…
The document explains the geometric mean as a way to summarize values that compound or multiply over time. It contrasts this approach with the arithmetic mean, which averages values by summing them, and notes that the geometric mean is suited to proportional…
This post proposes an A-share stock screen using three conditions: daily amplitude above 1, a rising bottom pattern, and a daily position-increase share above 5%. The text frames amplitude and rising lows as technical filters and the position-increase…
The document describes a Time-Averaged Moving Average (TAMA), a smoothed price line built by applying a moving average to a Time-Averaged Price (TAP) calculation. Users can choose the calculation period, applied price, smoothing period, and smoothing method.…
This stock selection rule screens for shares with an intraday amplitude above the stated threshold, prior-day turnover above the stated amount, and no limit-up close on the prior day. The article presents the screen as a way to find active, liquid stocks…
The document reproduces a pandas performance warning raised during a backtest while data is being written to HDF storage through PyTables. The warning says that a block of columns contains mixed object types that PyTables cannot map directly to its native C…
This brief description presents SpearmanStack_X20, an indicator display that places twenty lines of the Spearman Rank Correlation indicator in a single chart window. It identifies the tool as an MQL4 implementation first published in 2010 and notes that the…
This Chinese equity screen combines three conditions: daily price amplitude above 1%, a large order net-volume ranking threshold, and revenue in 2021 exceeding revenue in 2018 by more than a stated ratio of 1.1. The article describes the revenue comparison…
This overview describes how to read simulated strategy performance based on historical data. It defines several commonly reported measures: total return and annualized return for gains over time, maximum drawdown for peak-to-trough loss, Sharpe ratio for…
This guide explains common measures in quantitative strategy backtests: total and annualized return, maximum drawdown, Sharpe ratio, win rate, and the ratio of average winning to losing trades. It frames them as complementary views of performance, risk, and…
This broad educational overview introduces mathematical tools used in stock analysis and algorithmic trading. It covers descriptive statistics, including measures of central tendency and dispersion, and explains how visualizations can help reveal…
The document introduces the Commodity Channel Index (CCI) as an oscillator that measures typical price relative to its moving average, scaled by mean absolute deviation. It notes that the conventional scaling factor was chosen so that many readings fall…
The Murray Math Oscillator is described as a market-direction indicator built from a closing price’s position relative to a rolling high-low range. It uses the highest and lowest prices over a chosen period to define the range, then adjusts the midpoint with…
This overview introduces seven sorting methods that may be useful when strategy code needs to order data: quicksort, merge sort, heapsort, selection sort, bubble sort, insertion sort, and Shell sort. It describes the central steps of most methods, including…
The document proposes describing price movement with discrete first and second derivatives: velocity represents the rate of price change, while acceleration represents changes in that rate. It uses a car slowing before reversing as an analogy for a market…
This research review presents the frog-in-the-pan hypothesis: investors may respond less promptly to a sequence of small, gradual price moves than to a similarly sized change concentrated in a few sharp moves. It describes an information-discreteness measure…