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
This post proposes a rule-based screen for Chinese A-shares that combines three characteristics: classification in the metaverse industry, circulating share capital no greater than 5.5 billion shares, and a principal business related to beverage and alcohol…
This research summary explains why correlated stock-selection factors can make portfolio exposures difficult to control. Factor correlations can vary over time, so even an equally weighted combination may carry an unintended concentration in a factor such as…
This document describes a daily stock allocation strategy that uses recent SEC insider reports. It filters for executives’ open-market purchases and sales, excluding awards, gifts, option exercises, and planned sales. The trading process begins with equal…
The document summarizes a report comparing the historical performance and future prospects of index funds and actively managed funds. Its stated takeaway is that their past performance showed no clear overall winner. A slight historical advantage for active…
The document outlines a framework that combines monetary and credit conditions to time Chinese equities and bonds, then adds momentum, oil, or inflation signals for commodities, equities, and bonds. It argues that monetary policy can lead markets because of…
This example describes an AI trading workflow for leveraged long and inverse ETFs. A researcher ranks the available funds, separate bull and bear agents present competing cases, and an interpreter turns their views into target account weights. A trading…
This configuration describes a Qlib workflow for training a LightGBM model on Alpha360 features and evaluating its predictions on CSI 300 constituents. The label compares closing prices across the next two reference periods. The data handler drops missing…
This configuration specifies a Chinese equities prediction and portfolio backtest using Qlib, LightGBM, and the Alpha158 feature handler. It pairs daily labels with one-minute features, resampling the minute data at 14:56. The listed data span begins in 2008…
The document examines why ESG ratings from different providers can disagree, citing differences in factor definitions, data sources, weighting, and treatment of missing disclosures. It warns that ratings may reward companies with greater reporting resources…
This report surveys gold’s supply and demand, major trading venues, and common investment instruments. It notes that investment demand and central bank purchases are especially relevant to prices, and describes the roles of London spot trading, U.S. futures,…
This announcement outlines a multi-factor strategy for OKX that ranks a universe of coins by a composite factor value and divides them into three roughly equal groups. It proposes going long the lowest-ranked group and short the highest-ranked group. If…
The report discusses using the Black–Litterman (B-L) framework to incorporate an investment manager’s judgments into a quantitative factor portfolio. It frames factor information coefficients as the return inputs and their covariance matrix as a…
This brief summary outlines several approaches to rotating among equity industries. It discusses adapting the investment clock framework from asset allocation to sector selection, while noting that macroeconomic data may be delayed or unreliable and that…
This Chinese language literature digest summarizes research on return predictability and portfolio covariance estimation. One study examines stock prices and dividends in the Netherlands, the United Kingdom, and the United States over nearly four centuries.…
This research summary examines how broad equity factors behaved across Chinese A-share industries from July 2009 through November 2018. It links differences in factor effectiveness to industry characteristics: small-cap effects were stronger in industries…
This BigQuant example assembles a China A-share workflow from stock selection through backtesting. It selects eligible main board, STAR Market, and ChiNext stocks associated with the CSI 1000, then calculates current and lagged volume relative to a five-day…
This meetup note explains factor exposure as an asset’s sensitivity to factor returns. It sketches a way to estimate that sensitivity: form a low-versus-high valuation long-short portfolio, treat its return series as the factor return, and regress an asset’s…
This article discusses practical choices when tuning a short-term stock-selection strategy: starting capital, number of holdings, capital allocated per holding, holding period, and factor selection. Its author argues that the number of names and per-name…
The document examines whether rolling Fama-French five-factor (FF5) alphas can guide U.S. sector rotation. It compares FF5 with the three-factor model (FF3), using monthly returns for ten U.S. industry portfolios from 1964 to 2014. The rotation signal is…
This Qlib configuration defines a temporal convolutional network workflow for ranking CSI 300 stocks. It uses Alpha360 features, robust feature normalization, missing-value filling, and cross-sectional rank normalization of labels. The prediction target is…
This survey describes how more than 800 U.S. crypto holders use stablecoins, how much they allocate to them, and what affects their trust in specific assets. It reports that 88% hold stablecoins and gives allocation ranges, common uses, and differences by…
This configuration describes a multi-asset systematic trading framework that combines rules for breakouts, relative and absolute momentum, moving-average trends, carry, acceleration, and skew-related factors. The rules use multiple horizons and include…
This meetup Q&A surveys several quantitative trading topics: using Level 2 order-book data, reproducing a legacy strategy, allocating idle capital to a gold ETF, factor decomposition, building CNN strategies, and portfolio optimization. Most answers point…
This strategy combines a restricted China A-share universe, model-based ranking, technical timing, and portfolio controls. It starts from a manually selected group of roughly 100 to 300 large-cap constituents associated with the CSI 150 universe. A…