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.
1,124 documents
This notebook describes producing predictions for a US equities case study’s holdout window after selecting a configuration on validation folds. It resolves the selected configuration, preserves its checkpoint choice, and constructs a new training…
This notebook explains how hidden Markov models infer unobserved market states from returns and recent volatility. It first establishes simple comparison rules: a volatility index threshold for stress and price relative to a long moving average for trend. A…
The notebook assembles end-to-end pipelines for daily equities and hourly crypto perpetual futures, covering acquisition, validation, source labeling, and storage. Its equity example joins historical WikiPrices data with a more recent Yahoo feed. Since the…
This notebook presents a framework for checking whether a momentum signal’s apparent quality survives reasonable research choices. Using real ETF data before the designated holdout period, it varies one lookback parameter at a time and measures…
The notebook contrasts two ways to divide a large equity order into child orders over a trading session. TWAP sends roughly equal quantities at each interval and needs no market forecast. VWAP allocates shares according to an estimated intraday volume…
The notebook explains how to compare prediction models for S&P 500 options without allowing incomplete or mismatched evaluation samples to distort the results. It identifies each model by family, configuration, and checkpoint, checks that predictions meet…
This notebook builds uncertainty intervals for ETF forward returns using Ridge regression and three conformal methods: split conformal prediction, conformalized quantile regression, and adaptive conformal inference. It evaluates them with walk-forward folds…
This notebook uses double machine learning to estimate whether a perpetual-futures premium z-score is associated with the following eight-hour return after adjustment for six pre-treatment controls. It compares effects across high- and low-volatility regimes…
This notebook explains a conditional autoencoder that models stock returns as characteristic-dependent factor loadings multiplied by latent factors. It first estimates characteristic-managed portfolio returns with a joint cross-sectional least-squares fit,…
This dataset guide describes a collection of continuous CME futures contracts spanning equity indexes, rates, energy, metals, currencies, agriculture, and livestock. It explains the hourly source data and derived daily frequency, multiple contract tenors,…
This notebook studies how allocation rules affect ETF strategies after a model has already ranked assets. It holds the prediction sets and rebalance calendar fixed, selects the same top-k basket for each comparison, and varies the weights assigned within…
This notebook compares linear models on ETF features built from overlapping return windows and related momentum, volatility, and oscillator measures. Such predictors are collinear, making ordinary least squares coefficients unstable. The experiment fits…
This notebook tests whether portfolio construction can rescue a cost-defeated, every-bar NASDAQ-100 strategy. It selects leading baseline predictions using the declared selection universe, then runs the allocation comparisons on the full universe. The sweep…
This notebook takes leading validation prediction configurations from an equal-weight signal-stage baseline and tests alternative allocation methods on the same predictions. It compares score-based and conformal weighting with inverse volatility, risk…
This notebook evaluates how registered ETF prediction sets perform when converted into portfolios. It crosses each validation prediction set with declared entry schemes, such as different top-k holdings, while keeping the trading calendar, rebalance cadence,…
Fractional differencing applies a non-integer power of the lag difference operator to log prices. Its binomial weights decay across past observations, so the result combines price changes with some surviving level information. The fractional order controls…
This chapter describes a path from research to live execution built around strategy logic that can run in backtest, paper, and live settings. It covers broker and platform integrations, deployment loops, data and feature parity checks, order lifecycle…
This notebook builds features for a cross-sectional hypothesis: unusually costly long positioning in perpetual futures may precede relative weakness as crowded trades unwind. It transforms funding rates and premium-index data into levels, historical…
This chapter maps financial data across market, fundamental, and alternative sources, then compares how asset classes shape what prices, liquidity, and observations mean. It emphasizes that timestamps, adjustments, identifiers, and revisions are part of a…
This notebook describes an LSTM model that predicts returns from ordered windows of stock features in a US equities panel. Unlike a flat tabular model, the LSTM processes sessions sequentially and carries a hidden state through each window. Learned gates…
This notebook presents TimeGAN, a recurrent generative adversarial model for producing synthetic financial sequences. Its five modules learn an embedding and reconstruction, predict successive latent states, generate sequences from noise, and distinguish…
This document explains a shared method for generating walk-forward validation splits from case study evaluation settings. It delegates fold construction to a calendar-aware splitter, works from unique timestamps so panel rows stay aligned, and steps backward…
This notebook implements a long-only RSI mean-reversion rule for BTC/USDT perpetual bars using an event-driven backtesting engine. It aggregates intraday observations into UTC daily bars, calculates a simple rolling gain-and-loss RSI, enters when the…
This notebook applies two sequential detectors to real validation prediction errors from two US equities linear-model configurations over a calendar year. A two-window mean-shift detector looks for changes in error magnitude, while a bad-day frequency…