Methodology
Competitors describe their methodology. We publish the source code.
Every Blueprint is built by an open-source engine you can read, run, or fork. The named mechanisms below aren’t marketing language — each is a thin wrapper around an auditable algorithm. For an audience that inspects methodology before it trusts it, that’s the whole point.
The open-core strategy
The engine is open source. The curation is the product.
ETFWealthIQ runs an open-core model. The backtesting engine — Boutquin.Trading — is MIT-licensed and public. So are the construction models, the covariance estimators, the factor-regression and attribution tools, the Monte Carlo and walk-forward validators, and the transaction-cost and slippage models. These are academic algorithms published in textbooks; hiding them would prove nothing. Open-sourcing them proves they were implemented correctly.
What stays proprietary is the curation: which ETFs qualify, in what proportions, under which constraint bands and rebalancing rules — the researched, validated answer to “which allocation, and why.” The engine can build anything; the Blueprints are the editorial judgment layered on top. The math is auditable; the answers are the work.
Open source (MIT)
- Backtesting engine and event pipeline
- 10+ portfolio-construction models
- 7 covariance estimators (incl. DenoisedLedoitWolf)
- Factor regression, attribution, correlation
- Monte Carlo + walk-forward validation
- Transaction-cost and slippage models
Proprietary (the product)
- Blueprint configurations — the curated ETF mixes
- US / Canadian regional implementations
- The Decumulator bucket profiles and bands
- Pre-computed backtests and rendered reports
- The curated ETF universe and inclusion criteria
- The web application, dashboards, and rebalancing service
The named mechanisms
Each one maps to a class in the open-source engine.
Subscribers — and skeptics — can audit the math. The mechanism is a thin wrapper; the algorithm underneath is public.
IQ Diversification Score
The one number that separates a portfolio that looks diversified from one that is. A correlation-based ratio of how much genuine diversification a Blueprint actually buys — because holdings that fall together in a crash aren’t diversification, however different they look on paper. Rendered as a gauge.
Drawdown Catalogue
Every discrete drawdown period in a Blueprint’s history — depth, duration, and recovery — as a sortable table and area chart.
IQ Factor Fingerprint
A multi-factor regression showing loadings on market, value, size, momentum, and quality — what really drives the returns, with an R-squared badge.
IQ Stress Test
A 1,000-run Monte Carlo simulation rendered as a fan chart of 5th / 50th / 95th-percentile paths — because a single backtest path can mislead.
Attribution Report
A Brinson-Fachler decomposition of returns versus benchmark — how much came from allocation, selection, and their interaction.
Rebalancing Signal
A threshold-based trigger that compares your positions against the Blueprint’s targets and produces a specific, cohort-timed trade list.
The full roster also includes the Walk-Forward Report, the Correlation Explorer, the Asset-Location Illustrator, and the Decumulator Bucket Allocator — each backed by a named class in the engine or the pipeline layer.
Staggered rebalancing
The 13-cohort model: one cohort per week, not everyone on the same day.
Each quarter has 13 weeks. ETFWealthIQ runs 13 independent portfolio cohorts — one per week — instead of rebalancing every subscriber on a single day. When you join, you lock in a rebalancing week. On the Tuesday of that week, the pipeline computes fresh weights using market data through Monday’s close, and you execute at your own pace across the remaining trading days.
Why Tuesday
Zero US market holidays fall on a Tuesday — the most predictable trading day — and it sidesteps Monday overnight volatility.
Anti-front-running
If everyone rebalanced the same day with known target weights, the trades could be front-run — especially on less-liquid ETFs. Staggering across 13 weeks, each with independently computed weights, destroys the signal.
Market-impact mitigation
Quarterly turnover spread across 13 weeks means each cohort moves a fraction of the total — a small, invisible share rather than one concentrated spike.
Natural scarcity
Each week has finite cohort capacity. “Lock in your rebalancing slot” is a real operational constraint, not manufactured urgency.
Validation
Walk-forward, out-of-sample — so a Blueprint isn’t curve-fit to one history.
We run the test that tries to break our own models. A backtest that only fits the past it was tuned on proves nothing. Walk-forward validation re-derives parameters on an in-sample window, then tests them on the next out-of-sample window it has never seen — rolling forward through history. Every backtest is also paired with Monte Carlo uncertainty ranges, because a single historical path can flatter or mislead. The validators that do this are part of the open-source engine, so skeptics can verify the claims themselves.
- Backtested on actual ETF price data across multiple market regimes — not synthetic reconstructions.
- Out-of-sample walk-forward testing, so Blueprints aren’t fit to one slice of history.
- Monte Carlo uncertainty ranges on every backtest — ranges, not single-point promises.
Why open source wins
“Publish the source code” beats “describe the methodology.”
Anyone can describe a methodology in prose. Prose can’t be run, can’t be forked, and can’t be checked line by line. Published source can. For a research-oriented, skeptical DIY investor, an algorithm you can read and execute yourself is a higher form of proof than any claim — and it’s the form of proof a black-box robo-advisor or a fixed-weight blog model can never offer. The engine is public on GitHub. Inspect the math, then decide what to trust.
See the four Blueprints the engine builds →See the methodology applied to your general profile.
Eight questions, no dollar amounts, no signup wall. You’ll get your IQ Score, the Blueprint archetype that best fits your general profile, and the full backtest the engine computed behind it.
Take the free IQ ScoreImportant disclaimers
Educational content only. ETFWealthIQ provides financial education only. Model portfolios are illustrative examples presented for educational purposes and are not personalized investment recommendations. ETFWealthIQ is not a registered investment adviser, portfolio manager, or financial planner. Past performance of model portfolios does not guarantee future results. Before making any investment decision, consult with a qualified financial professional who understands your individual circumstances, goals, and risk tolerance.
Backtested results. All historical performance data represents backtested results computed by open-source software (Boutquin.Trading) using actual historical index and ETF return data. Backtested performance is hypothetical and does not represent actual trading. The source code used to generate these results is publicly available for audit. Actual investment results may differ materially. Past performance is not indicative of future results.
Canadian investors. This educational content does not constitute investment advice under Canadian securities law. The provision of investment advice in Canada requires registration with provincial securities regulators, which ETFWealthIQ does not hold.
Open-source software. The Boutquin.Trading backtesting engine is open-source software provided under the MIT License. It is a general-purpose research tool. Use of this software to make investment decisions is at your own risk. The software authors and ETFWealthIQ are not responsible for any losses incurred.