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.
62 documents
This article explains Hierarchical Risk Parity (HRP) as an alternative to covariance-inversion methods such as the Critical Line Algorithm. It identifies estimation errors, unstable matrix inversion, computational burden, and the loss of meaningful asset…
The article introduces Black-Litterman as a Bayesian approach that combines CAPM equilibrium returns with investor views to produce portfolio allocations. It motivates the method by describing common mean-variance optimization problems: sensitivity to…
This announcement describes the early contents and development plans for MLFinLab, a Python package based on methods from a financial machine learning text. Its covered techniques include financial data structures built from raw tick data, such as imbalance…
This review explains how climate change can affect financial institutions through physical hazards such as floods and droughts, and transition pressures such as new climate policy, technology shifts, litigation, and changing customer demand. It maps these…
This article applies the Ornstein–Uhlenbeck (OU) process to mean-reverting spreads, including those used in pairs trading. It contrasts Euler–Maruyama simulation, which introduces discretization error, with Doob’s exact simulation method, which uses the…
This article explains how minimum spanning trees (MSTs) represent relationships among assets using a connected graph with minimal total edge weight. It describes visualizing trees with industry colors and market-cap node sizes, and reviews measures such as…
This document reviews research practices for applying machine learning and quantitative methods to investing. It outlines common barriers to financial machine learning, including the interdisciplinary nature of the work, limited data, and markets shaped by…
The article explains Theory-Implied Correlation (TIC), a method for estimating portfolio correlations by combining observed correlations with an externally specified hierarchy of assets. It describes three stages: fit a hierarchical tree to empirical…
The article describes an experiment applying meta-labeling to S&P 500 E-mini futures data. It combines event-based sampling, the triple-barrier method, and meta-labeling with two example strategies: trend following and mean reversion using Bollinger Bands.…
The document explains online portfolio strategies that find historical market windows resembling current conditions. CORN measures similarity with Pearson correlation rather than Euclidean distance and uses the resulting matches to guide portfolio weights.…
This article outlines a research workflow for quantitative finance teams, from reviewing prior work to framing a research question, planning a study, conducting analysis, preparing a paper, and organizing group learning. It recommends assessing the quality…
This article develops a way to choose entry thresholds for a spread used in mean-reversion trading. A position is opened when the spread crosses an upper or lower boundary and closed when it returns to its mean. Tight boundaries create more trades with…
This overview compares online portfolio momentum approaches across six equity and market-index datasets. Exponential Gradient updates portfolio weights using recent relative performance, with a learning rate and regularization intended to limit abrupt…
This introduction compares four portfolio selection benchmarks using a collection of 23 ETFs with closing prices from 2008 to 2016. Buy and Hold starts with fixed allocations and lets weights drift with asset prices; Best Stock selects the strongest asset…
The article introduces cointegration as a way to find a stationary spread from non-stationary asset prices. If two price series share common long-run trends, a weighted combination may remove those trends; the resulting spread can fluctuate around a stable…
The article describes the entry challenge in quantitative finance as learning both the financial ideas behind markets and the technical skills used to analyze them. It situates the field across mathematics, statistics, finance, and computing, with…
This article explains how a Planar Maximally Filtered Graph (PMFG) represents similarities among assets while preserving more network structure than a Minimum Spanning Tree. It ranks nodes by a combination of graph centrality measures, then compares…
Meta labeling adds a secondary classifier to a primary model that already proposes a trade direction or classification. The primary model is tuned for high recall, accepting some false positives; the secondary model then estimates whether those proposals are…
This overview unifies common copula-based pairs strategies around conditional probabilities, which estimate whether each asset appears relatively overvalued or undervalued given the other asset. Unlike spread-only signals, the two leg-specific estimates can…
This essay discusses how asset owners, asset managers, and companies can support sustainable investing by incorporating environmental, social, and governance considerations alongside financial analysis. It presents long-term ownership and broad market…
This article explains how to build vectorized equity curves while distinguishing long-only return calculations from long-short pair-trading P&L. For a single asset or a long-only portfolio with positive value, it recommends calculating portfolio returns and…
The article presents Model Fingerprints as a way to describe how machine learning features affect predictions. It estimates partial dependence by varying one feature while averaging predictions over other observations, then separates that dependence into…
This project update describes research notebooks for financial machine learning topics, including tick, volume, and dollar bars; CUSUM event filtering; vertical barriers; and triple-barrier labels. It outlines comparisons of bar sampling using weekly count…
This article applies optimal stopping theory to a mean-reverting spread formed from two co-moving assets. It models the spread with an Ornstein–Uhlenbeck process, estimates the process parameters and asset hedge ratio by maximizing average log-likelihood,…