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 presents an attempted Kalman filter implementation for estimating a changing hedge ratio between two cointegrated stock log-price series. The proposed state is a single beta that evolves over time, with the second stock’s log price treated as…
The discussion compares Bayesian and frequentist approaches in finance and argues that method choice should depend on the data, model, and question. Frequentist tools include least-squares estimation and hypothesis testing. Bayesian methods add prior…
The document raises a practical concern about combinatorial purged cross-validation (CPCV) in financial backtesting. By combining separated time blocks into a training sample, the method can make observations on either side of a gap appear adjacent. That can…
The document raises a preprocessing question: how to transform a time series with changing volatility before using it as a predictor in a regression model. The author reports trying several scaling methods without obtaining satisfactory stationarity, and…
The document discusses how financial pattern discovery can extend beyond traditional technical analysis. It distinguishes loosely defined chart patterns and indicator rules from approaches that state hypotheses mathematically and evaluate them with…
The document suggests three input features for training a binary classifier to predict bid or ask price movement from high-frequency order book and trade data: bid-ask spread, bid-ask volume imbalance, and signed transaction volume. Signed volume is positive…
The discussion surveys proposed links between chaos theory, nonlinear dynamics, fractals, and financial markets. It points to research on nonlinear dynamics, self-similarity, the Hurst exponent, and the use of fractal ideas to model asset returns. One…
The document poses an optimal stopping problem for a firm's profit modeled as a Brownian process with drift and volatility. The value is expressed as accumulated discounted profit up to a stopping time, then rearranged into a term involving an expected…
The document examines whether a small subset of original variables can represent the first principal component while contributing little to the others. It explains that principal components form a coordinate basis for the original feature space. In a…
The document advises aspiring practitioners to learn statistics and machine learning before applying models to market data, then study financial modeling and backtesting, including transaction costs. It explains model fitting in terms of prediction error and…
The document raises two questions about using combinatorial purged cross-validation (CPCV) for a binary trading model: whether to select hyperparameters by averaging metrics across splits or across reconstructed paths, and whether evaluating those parameters…
This discussion considers whether commercial derivatives-pricing software will reduce the need for pricing quants. The responses describe a changing role: less development of entirely new pricing formulas in some settings, and more adaptation of existing…
The document asks how to validate a simple credit scoring system when a portfolio has no historical defaults or losses. The score is a weighted sum of borrower variables, with both variables and weights chosen through expert judgment. Since outcome-based…
The document discusses what mathematical preparation high-frequency trading may require. One answer emphasizes applied tools such as linear algebra, statistics, time-series analysis, optimization, and sometimes machine learning, while noting that practical…
The document asks how to identify a useful subset of roughly one hundred market related variables for explaining or predicting global stock returns. It highlights two data settings: fewer observations than predictors and at least as many observations as…
The document considers a sequential decision game: three coins have known head probabilities, but their assignments are hidden. The player can pay for either a single identified flip or a paired flip whose outcome reports only the total number of heads, then…
The document describes a proposed volatility forecasting exercise using ARCH, GARCH, and EGARCH models on daily returns for two stocks. The author divides the observations into a training portion and a later testing portion, estimates model parameters on the…
The document asks for examples of spectral clustering in finance and why it might be selected over other clustering methods. One response describes a time-series decomposition approach: apply singular value decomposition to a univariate series, cluster the…
The document describes an effort to calibrate a simple stochastic volatility model with the Efficient Method of Moments (EMM), as a step toward calibrating Heston parameters from historical equity returns. The author reproduces a SAS example in Python,…
The discussion addresses how to decide whether a predictive model has too many explanatory variables, especially when adding variables keeps improving in-sample fit. The answers emphasize that there is no universal cutoff. One warning sign is…
The document asks how a physics undergraduate with some Python, machine-learning coursework, and numerical computing experience can prepare for quant roles. The answer proposes using QuantLib as a practical test of programming and financial modeling ability,…
The question concerns regressions linking Bitcoin prices with daily Reddit post counts and sentiment. The author reports that price appears integrated while the social measures appear stationary, yet the series show strong correlation, high regression fit,…
The document describes an attempt to forecast a differenced FTSE 100 series with an ARMA mean model and GARCH disturbances using the rugarch package. The author encounters missing volatility forecasts when requesting rolling forecasts and a plotting error…
The document collects approaches to generating synthetic price series that remain within a chosen range while exhibiting a specified, but imperfect, relationship to historical prices. One answer proposes modeling returns with an Ornstein–Uhlenbeck process…