Momentum Acceleration Ranker

What it is

A price-action ranker that scores a 231-ticker universe (S&P 100 plus an AI-datacenter supply-chain map) with one formula: weighted multi-horizon returns scaled by the percentile of normalized price acceleration. A React/KaTeX dashboard shows rankings, parameter sweeps, walk-forward backtests, regimes and an allocation order sheet. The Recursive Momentum Trader grew out of it.

The problem

A transparent, testable momentum signal beats eyeballing charts.

How it works

1. Pure-formula NumPy module with a synthetic-data test suite; yfinance fetcher with batching, backoff and a parquet cache
2. Build scripts emit rankings, sweep and walk-forward JSON so the Vercel deploy is a static site with no Python runtime
3. Dashboard tabs for rankings, backtests, sweeps, regimes, picks and allocation
4. Later added a bear/flat/bull regime classifier and lane models synced from the Recursive Momentum Trader

Stack

Disciplines

For your business

Quant signal dashboards that turn a research formula into a deployable, testable ranking product.