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 whether a market maker can estimate fair value by modeling the limit order book (LOB) and its evolution without separately accounting for recent trades. It explains that trades help drive changes in the book, while resting orders also…
The document explains how technical indicators such as moving averages and MACD can be used in neural network trading models. Indicators can be calculated from market data and supplied as predictor variables to the network, alongside raw price data. The…
The document considers whether a strategy with fixed take-profit and stop-loss levels can be evaluated by treating wins as binomial outcomes. It cautions that a win count alone may not represent the strategy’s payoff process, and that a false-positive rate…
The exchange discusses whether least-squares Monte Carlo (LSM) for Bermudan options has numerical support and theoretical guarantees beyond a limited proof based on two exercise dates and a Markovian state on positive values. Mark Joshi replies that…
The document asks how reinforcement-learning environments such as FinRL simulate markets from historical data, and whether they use geometric Brownian motion. It outlines a GBM approach: calculate historical log returns, estimate their mean and standard…
The document surveys ways to obtain historical corporate earnings release dates for equity research, including a terminal data function, financial data providers, and company filing records. Some sources return structured quarterly data that can be used in…
The document asks how to reconstruct two related time series, unleaded and diesel prices, using principal component analysis. It provides eigenvalues and eigenvectors for the two price columns and wants to compare the original series with reconstructions…
The document collects recommendations for quantitative finance papers and research areas, spanning rough volatility, option-implied recovery of future distributions, asset pricing factors, rates options, behavioral finance, and reinforcement learning for…
This exchange addresses whether a stock prediction model should use adjusted closing prices or raw closes when its inputs include technical indicators and lagged prices. The answer distinguishes the target being modeled: adjusted prices are appropriate when…
The document asks how to assess feature autocorrelation in a trend strategy using XGBoost and SHAP when mean return and volatility appear to dominate feature importance. The author describes centering mean returns by the portfolio mean and scaling volatility…
The document describes an attempt to initialize nonlinear optimization for the Nelson–Siegel–Svensson yield curve model using ordinary least squares. The author derives spot rates from par bond coupons, constructs a coefficient matrix from maturities and…
The discussion introduces quantization as a way to approximate a continuous space with a finite set of representative points. In the response, this idea is associated with centroidal Voronoi methods: each generator represents a neighborhood, allowing a…
The document explains how to judge out-of-sample volatility forecasts beyond ranking models by measures such as MAPE or Mincer–Zarnowitz R-squared. It proposes comparing models through their time-by-time loss differences, using a loss function such as mean…
The document asks how to score a trading strategy in a supervised machine learning framework while accounting for exposure to the market. It compares raw strategy returns and a conventional Sharpe ratio with excess returns and a Sharpe ratio based on returns…
The document describes an attempted replication of a published method for functional principal component analysis (FPCA) of financial time series. The author uses crude oil futures contract data, represents log prices with a curve built from exponential…
The note compares traditional approaches to estimating the probability of default on commercial bank loans with machine learning methods. It names Merton and KMV as structural credit-risk models and random forests as a possible machine learning alternative.…
The document collects book recommendations for readers beginning quantitative finance, especially those drawn to mathematics. Its suggestions span derivatives pricing, financial statistics and data analysis, time-series analysis, numerical methods,…
The document examines whether CAPM and the Fama–French three-factor model should be viewed as cross-sectional models, time-series models, or tools for forecasting an individual asset’s future return. It distinguishes CAPM’s theoretical relation among…
The document asks whether researchers can create synthetic datasets through explicit, context-aware choices rather than relying on a generative AI model or a uniform library. Its proposed workflow begins with a problem context and a small set of candidate…
The document presents a Python function for estimating daily volatility from a close-price series. It searches the time index for the most recent observation at least one calendar day before each timestamp, computes the corresponding price returns, then…
The document addresses why the smoothing coefficient in simple exponential smoothing is normally restricted to values between zero and one. In the stated forecasting recurrence, the next forecast combines the latest observation with the prior forecast,…
The document frames an investment decision as a binary choice: invest in a stock index or stay out, using macroeconomic variables such as volatility measures, currency exchange rates, and Treasury yields. It proposes a classification approach, with…
The document compares a discrete Gaussian random walk with increments whose standard deviation is proportional to the square root of the time step against one using a different power of the step. Independent mean-zero Gaussian increments with variance equal…
The document presents an R example that fits univariate ARMA-GARCH models to three return series and combines them with a time-varying normal copula. It uses semiparametric distribution transformation with generalized Pareto tails, then repeatedly filters…