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.
4,194 documents
The document explains principal component analysis of changes in yields or forward rates across maturities. The covariance matrix captures how rate changes move together; its eigenvectors are factor loadings, while the corresponding eigenvalues describe the…
The document considers how to distinguish a manager’s deliberate stock selections from positions that simply differ from a benchmark. That distinction is difficult when the benchmark is unknown or poorly matched to the fund’s style, and when a fund holds…
The document explains how to estimate factor loadings for a hedge fund performance regression. It recommends an ordinary least squares linear model, with the fund’s excess return as the dependent variable and the seven proposed factors as predictors. The…
The document explains a test for whether a factor such as the size factor in a three-factor equity model adds pricing power beyond other, correlated factors. Simply checking whether the factor’s average return is small can mislead: removing it may change the…
The document discusses whether a low-volatility factor can be added to CAPM or the Fama-French three-factor model without invalidating the model. Its answer explains that CAPM attributes expected returns to market-related systematic risk, while return…
The document explains a way to remove a portfolio’s estimated exposure to a Fama–French factor. First, regress the portfolio’s excess returns on the chosen factor to estimate its beta. Then combine the portfolio with a short position in the factor…
The discussion compares a price-based pairs trading framework with the statistical arbitrage approach described by Avellaneda and Lee. In the former, log prices are related through a fitted spread, cointegration tests such as Johansen's can identify…
The document raises a problem in estimating short-term statistical factor models from equities trading in markets with different hours. When one market closes before another, their daily returns cover different time intervals. A Japanese stock may respond to…
The document asks how to construct a factor from stocks ranked by idiosyncratic volatility after accounting for the Fama–French market, size, and value factors. The proposed long–short return subtracts the returns of low-volatility stocks from those of…
The document describes a factor-sorted portfolio study that tests individual CAPM intercepts using time-series regressions. It sorts stocks into ten portfolios by a characteristic, estimates each portfolio's excess return against the market excess return,…
The document describes why illiquidity can affect asset prices and returns. Transaction costs such as taxes and brokerage fees act as market frictions; investors may require compensation for bearing them, which can influence the prices they are willing to…
The document explains the distinction between cash flow news and discount rate news in asset pricing. A Campbell–Shiller decomposition separates unexpected returns into revisions to expectations of future dividends and revisions to expected future returns.…
The note explains why indicators that forecast changes in risk premia need not contradict the Efficient Markets Hypothesis. The key distinction is between predictable differences in expected returns associated with risk exposure and predictable risk-adjusted…
The discussion considers how to interpret portfolio exposures to the market, size, and value factors in the Fama-French three-factor model. One response characterizes portfolios through their individual loadings: exposure to size and value distinguishes a…
The document asks how square root market capitalization weights in cross-sectional factor regressions interact with a size factor. It describes a two-stage factor construction process: estimate each stock’s factor exposure with time-series regressions, then…
The document considers how to express twelve-month stock returns and return on assets when using them in a cross-sectional regression alongside log market value. It compares fractional and percentage units for the return and ROA, and asks whether ROA should…
The document reviews PepsiCo as a large food and beverage company, emphasizing its long dividend record, diversified brands, and efforts to reshape its product mix. It describes acquisitions aimed at wellness and clean-label consumers, a divestiture in…
The document describes Maple Finance as a DeFi lending platform aimed at institutional participation, combining KYC checks and credit assessments with blockchain-based lending pools. It highlights real-world asset tokenization, including trade finance,…
The document compares Cardano and Ethereum using GitHub commits, active developer counts, project totals, and repository counts. It reports that Cardano recorded 21,439 commits over the past year, slightly more than Ethereum’s 20,752, while Ethereum had a…
The research examines whether selecting crypto assets by market capitalization relative to fees produced better returns than holding Bitcoin during the 2023–2025 period it analyzes. Its fee metric annualizes 30-day fees, treats lower ratios as stronger…
This market-neutral strategy ranks USDT perpetual contracts using a weighted composite of cross-sectional price momentum and funding-rate information. It goes long the highest-scoring coins and short the lowest-scoring ones. Funding intervals are normalized…
This UMD strategy ranks stocks by returns over a long lookback while omitting the most recent month, then buys the strongest group and shorts the weakest. It updates the selections monthly, uses equal weighting, and delays positions by one period before…
The code describes three candidate factors for ranking crypto assets: an ATR-based volatility measure, a volume-distribution measure, and an inverse-price measure. The volatility function calculates ATR over 14 periods, ranks the observations, and maps the…
The document outlines an automated process for turning a natural-language crypto factor idea into a calculated signal and evaluation report. A language model identifies the factor’s direction and data needs, generates a JavaScript function, and the workflow…