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.
511 documents
This project applies a Random Forest classifier to intraday BTC/USD data to produce directional signals from technical features. It uses two years of one-minute OHLC observations and inputs including returns, percentage changes, RSI, ADX, moving-average…
This article turns machine-learning predictions into a rule-based EUR/USD strategy and compares its historical performance with buy and hold. Its indicators are Parabolic SAR, which trails price and reverses after a price break, and the MACD histogram,…
This report describes an introductory talk on algorithmic trading, including its growth in India over the preceding three to four years. It outlines how the speaker introduced basic trading strategies and built toward a more complex example by adding order…
The article presents paper trading as a way to practise buying and selling with virtual funds, evaluate a strategy on live market data, and learn trading platforms before committing capital. It recommends matching simulated account size and positions to…
The article introduces quantitative trading as the use of mathematical and statistical analysis, commonly applied to price and volume data. It describes using tools such as moving averages, ARIMA, exponential smoothing, and neural networks to investigate…
The document introduces convolutional neural networks (CNNs), explaining how convolutional filters create feature maps, pooling reduces dimensionality, and fully connected layers support classification or regression. It surveys several well-known CNN…
The document introduces core Python concepts, including syntax, indentation, variables, operators, conditions, loops, functions, modules, and libraries. It frames these basics in the context of algorithmic trading, where Python can be used to acquire and…
This project describes an intraday strategy for Indian equities built around the first five-minute candle. It classifies opening candles into gap-up or gap-down patterns, reversal setups using Bollinger Bands and candle shadows, engulfing patterns, and…
The document explains RippleNet’s role as a payments network for financial institutions and distinguishes it from XRP, the digital asset used as a possible bridge currency. It describes the XRP Ledger, validator consensus, trusted Unique Node Lists,…
The article compares two unsupervised clustering methods using daily RSI and ADX observations as an example for grouping stock behavior into possible bullish, bearish, and sideways regimes. K-means assigns observations to the nearest of a chosen number of…
The article presents data cleaning as a necessary stage between acquiring raw data and analyzing it or training machine learning models. It explains tidy data structure, variable types, and the importance of preserving the original source data alongside a…
This introduction explains portfolio management as selecting and combining assets to pursue a return objective while controlling risk. It contrasts passive, active, and aggressive management, and describes bottom-up security selection alongside top-down…
The article examines market effects associated with the early COVID-19 outbreak and the Russia–Saudi Arabia oil price dispute. It describes calculating average forward returns after historical drawdowns: compute cumulative returns and running peaks, identify…
The article presents a simple cross-venue arbitrage example and uses it to show how algorithmic strategies can be organized around events. A strategy quotes one instrument using prices from another, aiming to capture a specified spread, then places a hedge…
The article introduces Monte Carlo as a way to estimate expectations by simulating random variables and averaging their outcomes. It contrasts this approach with deterministic models, sketches the method’s history through Buffon’s needle and early…
The article introduces supervised and unsupervised learning, then focuses on supervised classification, where models learn from labeled examples to assign observations to categories. It distinguishes binary, multiclass, and imbalanced classification and…
The article examines how fixed and trailing stop-loss rules affect a strategy’s return distribution. Its central point is that stopped trades remain part of the results: a stop can cut large losses while also closing positions that might have recovered or…
This tutorial develops a simple S&P 500 trading signal using a support vector classifier. It derives two predictors from historical open, close, high, and low prices, labels the next day according to whether the index rises, and splits observations into…
This guide introduces algorithmic trading for retail traders, explaining how software applies predefined rules to market data and places orders. It names moving-average crossovers, momentum, and mean reversion as beginner strategy examples. It also describes…
This tutorial explains a Python workflow for retrieving cryptocurrency market data from CryptoCompare. It describes authenticating with an API key, listing available coin tickers, and requesting historical prices at daily, hourly, or minute intervals. The…
This tutorial presents a basic classification workflow using scikit-learn and the Iris dataset. It explains how features and labels are represented, why data should be split into training and test sets, and how a k-nearest neighbors classifier is created,…
The article describes collecting cryptocurrency price and volume observations at minute intervals, storing them for analysis, and accounting for delays caused by fetching data across many coins. It then presents a simple trend-following strategy that uses…
The article introduces ARFIMA models, which extend ARIMA by allowing the integration parameter to be fractional. This lets the model represent persistent dependence, or long memory, that may be diminished when prices are converted to returns through ordinary…
This project outlines a mean-reversion strategy for liquid, shortable stocks organized across five sectors. It first screens candidate pairs for correlation, then tests their spread for stationarity with the Augmented Dickey-Fuller test. When a qualifying…