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.
5,018 documents
The document discusses how to present a quantitative trading interview assignment that forecasts volatility for several stocks from a short high-frequency dataset. The candidate used a Bayesian multivariate stochastic-volatility model and wonders whether the…
The document considers how to encode credit ratings as categorical features when the distance between rating notches may not be linear. It notes ordinal encoding as a common approach and mentions probit, logistic, and exponential transformations as…
The document considers how to rank live bonds using predicted default probabilities from a classifier trained with matured bond outcomes. It questions an individual-bond score that takes the logarithm of a return-to-default-probability ratio. The response…
The document discusses ways to estimate corporate probability of default from company financial statements or internal credit ratings. It mentions structural approaches such as the Merton model, accounting-based scores such as Altman’s Z and Ohlson’s O, and…
The document considers how to track changing dependence between two time series when a changing regression slope may reflect shifting variances rather than a stronger relationship. It raises similar interpretive concerns about time-varying quantile…
The document considers how to normalize daily trading volume and value for stock time-series models, including recurrent neural networks. It contrasts whole-series statistics such as global minimum, maximum, mean, and standard deviation with normalization…
A hidden semi-Markov model allows the probability of leaving a hidden state to depend on time spent in that state, unlike a standard hidden Markov model’s memoryless transition structure. The document asks whether a trading model could also use cumulative…
The response suggests comparing a volatility forecast from a non-normal GARCH model with the volatility implied by option prices. If the options appear inexpensive relative to the forecast, the proposed position is to buy calls and puts; if the implied…
This exchange addresses how to retrieve the long-run variance component from a component GARCH model fitted with the rugarch package in R. The questioner can obtain conditional volatility with the package’s standard accessor but needs the time-varying…
The document asks whether a Gaussian Kalman filter applied to a regression with a changing slope can be used to estimate the changing correlation between two series. It sets out a linear observation model, first with an intercept and then without one, and…
This collection presents examples of incorrect AI answers to quantitative finance questions and briefly contrasts them with corrections. Topics include converting yield quotations between compounding frequencies, whether mean-variance portfolio theory…
The question seeks accessible research methods for identifying which of many candidate factors influence broad equity index returns, with interest in machine learning beyond ordinary linear regression. The accepted response frames this as a return…
The document considers using principal component analysis on industry peers’ fundamental metrics to inform equity analysis. It provides variance shares and loadings for thirteen components, then asks what the first two components represent. The response…
The document considers a stochastic differential equation whose drift switches between positive and negative values according to the sign of the state. The drift is bounded but discontinuous, so it fails the usual Lipschitz condition associated with standard…
The document explains how dollar bars can be used in a machine-learning workflow despite their irregular time spacing. A bar retains the familiar structure of a time bar, such as open, high, low, and close, so these values can be supplied as model features.…
The document discusses ways to investigate whether returns lead volatility or volatility leads returns. One suggested exploratory approach is to compare correlations between time-shifted returns and volatility with correlations between time-shifted…
The document surveys ways to turn trading signals into mathematical strategy rules. One suggestion is to combine implemented signals in a factor model and estimate coefficients, which can produce a fitted formula and measures of statistical fit. The response…
The document identifies a suspected comparison error in the Optimal Number of Clusters procedure associated with de Prado and Lewis's work on detecting false investment strategies. The algorithm recursively reclusters groups whose silhouette-based…
The document considers a monthly return factor model in which one factor explains a substantial share of returns in some periods, but its relationship with returns reverses in others. Pooling the periods yields a lower overall R-squared. This pattern alone…
The document asks where to find algorithmic trading strategies that were once profitable but later stopped working. Its answer points to public compilations of strategies, while cautioning that detailed disclosure is unlikely. It gives several reasons:…
The question asks whether quantitative finance has a counterpart to complex economics, focusing on markets as dynamic systems in which participants’ forecasts and trading can influence future prices. It contrasts the usual emphasis on option pricing and…
The discussion considers a trading research process that clusters historical intraday observations, such as daily vectors of hourly exchange rates, and then measures the next day's average return for each cluster. A researcher might use the return associated…
The discussion distinguishes the two-stage Fama–MacBeth procedure from a linear factor model implementation. In the procedure described, asset betas are first estimated, potentially with a rolling window, then a separate cross-sectional regression is run for…
The document surveys several approaches to modeling limit order books and market microstructure: agent-based models, point-process models, propagator models, and Markov-chain approaches such as the Queue Reactive model. It describes the strengths associated…