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 considers a method for comparing calendar months across historical data: calculate correlations across eight daily variables, then use Euclidean distance to identify similar months. The replies point toward established time-series methods for…
The question concerns fractal dimension, rescaled-range analysis, the Hurst exponent, and possible short-term prediction for a stock index. One response recommends applying rescaled-range analysis to log price differences. It distinguishes a local Hurst…
The document asks how to compare growth across products with different starting sales levels. It considers whether an indicator can assign the same value to products that double their sales, even if their absolute increases differ. One response proposes…
The document asks whether psychometric or behavioral assessments can help evaluate loan applicants who lack conventional credit histories, particularly in developing countries. It sketches an approach in which questionnaire responses or psychological factors…
The document asks how to value a compound option when its underlying instrument is itself option-like and has additional conditions. A preferred share with conversion rights and dividends is given as an example. The author is considering Monte Carlo methods…
The document outlines ways to study whether investor sentiment influences stock returns or returns influence sentiment. It distinguishes lexicon-based sentiment measures from methods that label text polarity using stock returns. The latter can create…
The post describes a proposed way to adapt an Avellaneda–Stoikov crypto market-making strategy to changing conditions. The author runs backtests with several risk-aversion and quoting parameters, records market features and short-interval equity changes,…
The document presents a learner’s request for a scalar example of the extended Kalman filter, motivated by difficulty implementing a nonlinear power growth state-space model. It first gives a local-level observation equation, then asks how to estimate a…
The document poses a modeling question: whether GARCH can be applied to time-series models that include exogenous variables, such as an ARIMAX model. It gives cash-flow forecasting as an example and asks whether the model’s residuals could be described with…
The document collects examples of quantitative forecasts for football's World Cup, including reports from financial institutions and a research paper using bookmaker and betting-exchange odds. The described method adjusts quoted odds for bookmakers' profit…
The document asks whether an agent trained to play Atari games from screen images and scores could learn to trade by observing market data and choosing actions. It introduces deep Q-networks as an example of reinforcement learning, where the agent improves…
The document considers whether changing performance across symbols and weeks might reflect shifts between stationary and nonstationary behavior. Its response recommends testing returns rather than price levels under a weak-stationarity framework, since price…
The document asks how a return model with low R² could still support profitable decisions. It considers using predicted returns as expected returns for directional trades and Kelly sizing, then raises two concerns: predictions may cluster near zero, and…
The document discusses how to label future price movements for a trading prediction model. One approach checks whether the price is higher or lower at the end of a fixed interval; another labels a sample positive if price reaches a favorable level at any…
The document asks whether European options under GARCH volatility require simulating the full sequence of stock prices and variances, or whether forecast average volatility can be inserted into a geometric Brownian motion solution. The response explains that…
The document raises a question about constructing buy, sell, and hold labels for a machine-learning trading classifier. In the example it describes, hourly observations are assigned classes according to whether log returns cross positive or negative…
The document is a short reading list for exploring deep reinforcement learning and related machine-learning methods in quantitative finance. It points readers toward work on optimal execution, foreign-exchange trading, price formation, options pricing, and…
The document outlines a proposed reinforcement-learning setup for trading a single asset pair. The suggested observation includes time, holdings in both assets, and market information such as quotes, order-book data, or automated-market-maker reserves.…
The document discusses how to process stock returns for a momentum prediction task. The accepted explanation describes building short-term and medium-term signals from cumulative returns, using daily data for the shorter horizon and monthly data for the…
The discussion surveys reported uses of F# in banks, trading firms, and quantitative finance teams. Examples include valuation and derivative pricing, energy trading software, sports trading models, data cleaning, and analytics. Some teams combine F# with C#…
The document describes a challenge in training a model to classify stock news as a long, short, or no-bet opportunity while producing usable probabilities. Expected-value optimization can favor taking bets even when the estimated edge is tiny, because…
The document surveys learning resources and possible uses of quantum computing in quantitative finance. It highlights optimization and simulation as areas of interest, including quantum annealing for constrained graph problems that may help represent…
The document clarifies that an option pricing model and a method for solving or evaluating it are separate concepts. Black–Scholes is a model with a closed-form solution under its assumptions, but the same model can also be evaluated using numerical…
The document considers how to represent clients’ non-optimal exercise of Bermudan swaptions, with the observation that decisions may respond to past rate movements as well as current rates. One proposed route is supervised learning: build observations…