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.
5,922 documents
The article explains robust portfolio optimization as a way to reduce the effect of errors in expected-return forecasts. Rather than optimize only for a single set of estimates, the methods consider adverse plausible cases and seek a portfolio that performs…
This research describes a bond-fund selection method built around return attribution. It expands the Campisi framework—which separates income, government-rate, credit-spread, and security-selection effects—with convertible-bond and monetary-policy effects.…
This 2018 weekly report reviews a sharp post-holiday decline in Chinese equities, noting that large-cap leaders held up better than smaller companies. It interprets price structure, valuation, and long-term support as signs that the market was in a potential…
This example organizes daily decisions across leveraged sector and broad-market ETFs using separate AI agents for technology, financials, healthcare, energy, and consumer-related groups. Each sector pod is instructed to consult recent news and macroeconomic…
This research summary argues that banks should be modeled separately in equity selection because their asset-heavy business structure and distinctive price behavior can make factors selected across the full market less effective within the sector. It…
This article explains risk parity as an allocation approach that assigns comparable risk contributions across assets or risk factors, unlike capital-weighted mixes such as a conventional stock and bond portfolio. It lays out assumptions behind the method,…
The document describes an indicator that uses principal component analysis to choose coefficients for instruments in a pseudo-stationary portfolio intended to return toward zero. It frames each instrument as a dimension in a multivariate dataset and uses PCA…
The document explains how multi-factor models describe asset returns through factor exposures, factor returns, and asset-specific residuals. It presents Barra as a framework for estimating portfolio risk by combining factor covariance with specific risk, and…
The document describes how to build daily return data for level-two industries and use it in stock selection. It proposes joining stock industry classifications with daily returns and float market capitalizations, then grouping by industry and date. Each…
This research summary describes using machine learning to predict equity returns from alpha factors. It compares LASSO, support vector machines, boosted decision trees, and random forests, selecting random forests for their relatively simple structure,…
This research summary proposes combining price-to-book ratio (PB) with return on equity (ROE) to find companies with stronger fundamentals and lower valuations in China’s A-share market. It treats ROE and other operating measures as indicators of value…
The document surveys a Chinese securities research team’s work on applying artificial intelligence to quantitative investing. It organizes that research around model evaluation, factor discovery, overfitting controls, synthetic data, and methods intended to…
This report tests the ratio of research and development spending to revenue as an equity-selection factor across industries. Single-factor tests find some effectiveness in technology-oriented sectors, including pharmaceuticals, electronics, communications,…
The document describes a market neutral stock factor strategy that estimates each stock’s beta against a broad US equity index using roughly one year of daily prices. At monthly formation, stocks are ranked by beta; the lowest beta group is held long and the…
This report challenges mean-variance optimization assumptions that returns are normally distributed, volatility captures risk symmetrically, and portfolios should maximize return per unit of risk. It instead frames investor concerns as preserving principal…
This 2017 review compares commodity trading adviser factors and explores ways to combine them. It covers time-series and return-signal momentum, roll yield, basis momentum, and changes in warehouse receipts and inventories. The report says standalone…
This research roundup summarizes three studies. The first examines whether unexpected US monetary policy announcements affected hedge fund alpha after the financial crisis, using event studies, structural-break tests, and Markov-switching models. It reports…
This report examines shortcomings in the Henriksson–Merton (HM) and Treynor–Mazuy (TM) models for assessing fund managers’ market and style timing. TM represents beta adjustment as a gradual quadratic pattern, while HM assumes a two-state exposure shift;…
The document outlines the five factors used to explain differences in stock returns: market excess return, company size, book-to-market value, profitability, and investment. It describes each as a comparison between groups of stocks, such as small versus…
This BigQuant example builds a daily Chinese-stock portfolio by ranking eligible shares on 30-day turnover variability relative to their industry group. It filters out risk-warning stocks and applies price and listing-age conditions, then selects five names…
This historical account explains how Bridgewater developed the All Weather approach from a broader effort to understand recurring economic relationships. Its core framework separates returns into cash, market beta, and manager alpha, then considers how…
This article presents a Chinese equity selection rule focused on stocks associated with the metaverse concept. Candidates must have a 30-day moving average that is rising and must not be listed on the STAR Market. The article gives screening logic in terms…
This market note reviews a modest rebound in Chinese equities and discusses the forces behind it. It attributes the recovery partly to expectations of improved second-quarter corporate earnings and reduced global risk aversion. It also cautions that…
This guide lays out a futures data workflow for a trading system. It starts with instrument settings, spread costs, and roll parameters, then gathers individual contract histories, builds roll calendars, creates multiple-price series, derives back-adjusted…