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 note describes an attempt to forecast the S&P 500’s next daily closing price with a Temporal Convolutional Network. The model uses daily closes, minimizes squared error, and is regularized with methods such as early stopping and dropout. Adding…
The document is a software-focused question about implementing a multivariate Markov regime-switching GARCH model for a dependent variable with several regressors. The author reports difficulty finding an implementation in common statistical environments and…
The document distinguishes an option-price surface, organized by strike and expiry, from a surface describing the probability distribution of future underlying prices. It asks whether one can be derived from the other and whether option-pricing models…
The document describes a workflow for generating option implied volatilities from the Heston model: price options with a Lipton integral formula, then invert prices with a Black–Scholes–Merton implied volatility routine. The author asks whether parameter…
The document presents an attempted Monte Carlo implementation of Longstaff–Schwartz for an American put, then describes an effort to replace polynomial continuation-value regression with a neural network. The accompanying response questions whether the…
The document addresses fitting a Student t-copula to multiple time series when some series have strongly changing variance and inverse-CDF resampling produces convergence warnings and implausible values. The response recommends a copula-GARCH construction as…
The document asks how to model feature–return correlations that fluctuate around zero over time while remaining within the valid correlation range. One proposed construction starts with an unbounded diffusion, such as a mean-reverting Ornstein–Uhlenbeck…
The document compares sampling financial prices at fixed time intervals with constructing volume or dollar bars, where each observation is triggered by traded activity. It notes that returns from volume or dollar bars can appear closer to normally…
The document shows how to generate samples from a discrete distribution using a uniform random draw. Given outcomes with probabilities, first form cumulative probability thresholds. Compare each uniform draw with those thresholds and return the outcome whose…
The document explains that Black–Litterman views can come from any researched source of expected excess return. Examples include published forecasts, broker recommendations, machine-learning models, technical analysis that produces target prices, and…
The document asks how to calculate the error gradient for individual weights in a feed-forward neural network trained with resilient propagation. The answers explain that a weight’s gradient depends on the neuron’s activation function and its derivative, as…
The document collects references and software suggestions for estimating the drift parameter of geometric Brownian motion and related stochastic differential equation models. It points readers to a financial econometrics textbook, a study comparing…
The document suggests finance applications for natural language processing projects, emphasizing tasks beyond generic news or social-media sentiment analysis. Proposed research directions include extracting risk-factor exposures from company reports,…
The document discusses practical use of logistic regression to estimate probability of default (PD), especially for retail credit such as consumer loans and credit cards. It contrasts those settings with low-default portfolios, including government bonds,…
The document asks how to forecast the next day’s closing prices for S&P stocks using information from prior days. The author first considers fitting a lognormal distribution from a price bar and sampling possible closes with Monte Carlo simulation. The…
The question asks whether estimating an AR(1) mean model first and then fitting a stochastic volatility model could introduce bias, especially into the autoregressive estimate. The response points to a joint Bayesian approach that includes an AR(1) structure…
The document describes estimating separate currency movements from a network of forex pairs. A single pair such as GBP/USD identifies only the relative change between its currencies, so the two component changes cannot be uniquely recovered from that quote…
The discussion surveys ways machine learning can be applied to quantitative finance and points readers toward introductory resources. It names supervised prediction of returns and alpha-factor selection, classification-based strategy research, and…
The document discusses why a covariance matrix of changes in forward rates may fail to be positive definite when used for principal component analysis. It distinguishes a matrix with small negative eigenvalues caused by floating-point round-off from a matrix…
The document discusses a neural network trained on exogenous variables to forecast Bitcoin prices ahead and examines why its predictions appear delayed despite having no past-price inputs. The answer suggests that the model may effectively predict the latest…
The discussion surveys ways Fourier analysis can be applied in quantitative finance while cautioning against treating observed stock prices as periodic signals that can be extrapolated for forecasts. Its clearest established use is option valuation: when a…
The document considers using recurrent neural networks and macroeconomic inputs to forecast bank non-maturity deposits over a multi-year horizon, including whether core deposits must be modeled separately. Its main caution is that financial time series are…
The document surveys ways topological methods have been applied to finance and points readers toward tools and research. It introduces Mapper as a way to turn data into simplicial complexes for visualization, and persistent homology as a method used to study…
The answer advises choosing quantitative finance training according to the intended path: industry, teaching, or personal study. For industry roles, it argues that demonstrated skills matter more in day-to-day work than credentials alone, while a strong,…