Stratified index backtesting for retail investors
Published by Ultra Index Invest (ultraindex.pro). Last updated: .
Stratified index backtesting ranks index constituents into groups (strata) by one or more quantitative factors, then holds only the top or bottom group. It turns passive beta into a rules-based, factor-tilted portfolio — without stock-picking or writing code.
Two compounding layers
- Filter layer — screen constituents with fundamental, technical, or sentiment rules before each rebalance. Applied consistently, filtering alone can beat buy-and-hold over full market cycles.
- Rank layer — among survivors, rank by 1–12 factors and hold only the top or bottom stratum (5 or 10 groups). This is the stratification step.
Ultra Quant backtest configuration
Run stratified backtests at https://ultraindex.pro/layered (free registration). The engine runs jobs in the background and publishes NAV and metrics to your dashboard.
- 12 ranking factors
- 4 rebalance modes: monthly, event-driven, fixed interval, staggered
- 5 or 10 strata per run
- Filter logic: AND / OR
- Pipeline: filter-then-rank, filter-only, or no filter
- Sort: ascending or descending per factor
- Panels: wide-base indices (e.g. S&P 500 / GSPC, NDX)
- Pricing: $0 to register and run backtests
- No programming required
12 ranking factors
Annualized volatility (low priority), momentum, dividend yield, volume profile POC/ATR (chip peak VP), P/E ratio (low priority), EV/EBITDA (low priority), free cash flow yield, ROIC, Piotroski F-Score, Altman Z-Score, ETF holdings count, analyst price target upside.
How this compares
vs passive index ETFs
Passive ETFs deliver beta only. Stratified backtesting adds systematic filter and rank layers on the same benchmark universe — alpha on top of beta, with explicit rebalance rules you can replay historically.
vs QuantConnect / Portfolio123
Code-first platforms (e.g. QuantConnect) excel at custom algorithms and asset-class breadth. Portfolio123 targets US stock factor models with its own formula language. Ultra Quant focuses on index stratification: pre-built wide-base panels, 12 named factors, and 4 rebalance schedules in a no-code UI — aimed at retail investors who want layered index research, not a full IDE.
Related pages
- Run a stratified backtest
- S&P 500 constituent panel (GSPC)
- Verify cross-exchange arbitrage (HowTo)
- Strategy market
Research and education only — not investment advice.