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 responds to a request for time-varying BBX copula implementations in Matlab or R. It points to R packages that provide BBX families and spatial or spatiotemporal bivariate copulas, while noting that temporal variation may require additional…
The document compares ordinary least-squares projection onto the leading principal components with an attempted machine-learning approach. It raises practical concerns about representing each component's weight vector as model inputs and about digitizing…
The document explains how to interpret latent states estimated in a standard stochastic-volatility model for log returns. In the stated model, the conditional standard deviation of returns at time t is beta multiplied by the exponential of half the latent…
The document outlines a least-squares Monte Carlo setup for valuing a Bermudan swaption. At the final exercise date, the payoff is the positive part of the underlying swap value. At earlier exercise dates, the method regresses discounted future swaption…
The document explains the role of the two random draws in acceptance-rejection sampling. A candidate is drawn from a proposal density, and a separate uniform draw determines whether that candidate is accepted according to the target-to-proposal density ratio…
The document poses a parameter estimation problem for a time series modeled by a process with a known-form characteristic function and unknown parameters. The author has considered recovering a probability density through Fourier inversion and then applying…
The response compares several evaluation methods by the data they use and the questions they address. Backtesting applies a defined strategy or model to real historical time-series data to assess how it would have performed. Historical simulation also uses…
The document explains that a classifier is not inherently a weak learner: its performance depends on the structure of the data and the task. Linear methods may perform poorly on nonlinear data, while their performance can differ when classes are linearly…
The document raises a preprocessing question for principal component analysis across assets with different volatilities. It compares scaling observations by volatility and centering them, taking logarithms of prices and centering, and clipping series to a…
The document asks how to simulate the conditional expectation of the time integral of the squared state of a Markovian stochastic differential equation. It decomposes the target into the integral already accumulated up to the conditioning time and a future…
The document proposes a practical framework for combining analyst return paths when their errors share information, change with market conditions, and may include useful contrarian views. It starts from forecast-error covariance: under quadratic loss,…
This discussion considers backtesting platforms for users who need to supply their own market data and use third-party libraries in strategy code. The replies suggest Python-based approaches, including the Python Algorithmic Trading Library, Zipline, and…
The document outlines a basic way to use a hidden Markov model (HMM) in portfolio allocation. The model estimates latent market states, asset drifts, and state-specific covariance matrices. For each state, the estimated return and risk inputs can then be…
The document surveys competing views on using genetic algorithms (GAs) to find financial trading rules or construct portfolios. The main concern is that searching many combinations can produce rules that fit historical data by chance, with data snooping and…
The document asks whether volatility models with statistically significant parameter estimates should be preferred over otherwise similar specifications whose estimates are insignificant. The examples include sGARCH, EGARCH, and TGARCH models fitted with…
The document discusses how to train a trading model intended to operate only in low-volatility conditions. The question contrasts supplying volatility as a feature with explicitly excluding unsuitable market states. The response recommends retaining…
The question describes difficulties estimating an ARIMA(1,1)–GARCH(1,1) model by maximum likelihood in R. Its custom likelihood recurses through residuals and conditional variances, and the optimizer encounters non-finite starting values, invalid square…
The document explains a proposed forecasting procedure from a paper on stock prediction with hidden Markov models. For a current observation, the trained model calculates its likelihood, then historical observations with matching or nearby likelihoods are…
The document surveys ways entropy concepts can be applied to financial time series. Suggested uses include studying market behavior and crash forecasting, connecting entropy with the Kelly approach to money management, and deriving option-pricing results…
The discussion considers why machine-learning stock-prediction studies appear less often in leading finance journals than in specialist or lower-profile outlets. It offers several explanations: short-horizon stock returns are noisy and hard to forecast;…
The question describes an option-pricing framework in which aggregate dividends and a state variable evolve with Gaussian shocks, while a stochastic discount factor determines risk-adjusted valuation. The underlying equity claim is represented as a sum of…
The document outlines an affine term structure model in which log bond prices are linear functions of latent factors. With as many observed yields as factors, the factor vector can be recovered from yields by inverting the model’s loading matrix, assuming…
The document outlines a self-study route for a mathematics graduate preparing for quantitative finance. It frames the field as a combination of mathematics, finance, and numerical programming, with the needed balance depending on the role. Derivatives work…
The document examines a practical difficulty in implementing tick imbalance bars from a financial machine-learning text. The proposed threshold depends on the expected number of ticks per bar and the expected imbalance in tick signs. The questioner reports…