This post develops a seasonal ARIMA forecast for household electricity consumption using four years of minute-by-minute meter data from UCI, achieving 0.2774 kW mean absolute error. The analysis covers data cleaning, stationarity testing, multiple modeling approaches including GARCH and state space models, and backtesting procedures to predict daily power use.
Ledger is a planned command-line toolkit for retail traders that applies professional quantitative validation techniques—walk-forward testing, cost modeling, and look-ahead-bias detection—to identify common backtest failures before risking real capital. The developer is gauging interest before building it.
Forexfin is a free platform offering eight trading calculators, live market data, and a trade journal with tagging and analytics. The paid Trader tier adds tilt detection to flag revenge trades and size creep, plus automated weekly and monthly retrospectives scored against actual trading statistics.
A backtest of Connors and Alvarez's 2008 mean-reversion trading strategy on SPY from 1993–2026 found that no parameter variant beat buy-and-hold, and the strategy's low drawdown reflects low market exposure rather than skill. The analysis demonstrates that backtests can pose unanswerable questions within their own data and that adversarial review is essential to catch methodological errors that bias results.