DEMA Strategy
Dynamic European Macro Allocation
Multi-asset conviction-weighted strategy driven by proprietary European macroeconomic signals with continuous position sizing and monthly rebalancing
The Strategy
DEMA (Dynamic European Macro Allocation) is a quantitative multi-asset strategy that uses a proprietary set of European macroeconomic indicators to determine exposure across eight major asset classes: U.S. equities, international equities, emerging markets, treasuries, gold, and investment-grade bonds.
Unlike binary long/flat models, DEMA uses conviction-based continuous position sizing. Each ETF receives a weight between 0% and 50% based on the strength of macro signals, allowing the portfolio to express varying degrees of confidence across asset classes. Total portfolio exposure can range from 0% to 200%.
The strategy was developed using a strict walk-forward methodology. Signal selection was performed exclusively on pre-2010 data. Signal combination uses an expanding-window approach from 2010 onward with multiple layers of statistical validation. All reported results include transaction costs.
Walk-Forward Test Active Starting April 1, 2026, this strategy is being tested with market data in a walk-forward framework.
How It Works
Each month, the model evaluates hundreds of validated European macroeconomic indicators per asset class. These indicators are processed through a proprietary statistical combination framework that produces a conviction score for each ETF.
Positive conviction scores result in proportional allocation (up to 50% per ETF). Asset classes with negative conviction receive zero allocation. A drawdown control mechanism reduces exposure when portfolio losses exceed predefined thresholds.
The dashboard shows both backtested and walk-forward performance. The shaded area on the chart marks the walk-forward period starting April 2026. Everything before that date represents out-of-sample backtested results (from 2010). Everything after represents hypothetical out-of-sample performance.
The exact indicator selection, combination method, and model parameters are proprietary and not disclosed.
Simulation Settings
Set your start date and initial capital to configure your portfolio simulation.
Model Performance
Drawdown
Peak-to-trough declineTotal Exposure
Sum of all ETF weights over timeWalk-Forward Performance
Since Apr 1, 2026Current ETF Allocations
Model Asset Class Allocation
Model ETF Allocation
Model Allocation
Hypothetical model allocation. Does not represent actual holdings or investment advice.
| Symbol | Weight | Price | Shares | Value |
|---|---|---|---|---|
| Loading holdings... | ||||
Historical Asset Weights
Performance Metrics
Walk-Forward Test Status
Walk-forward testing started April 1, 2026
Monthly signal updates at end of month
Backtested Performance (2010-2026)
Performance Comparison
DEMA vs SPY Buy & Hold, out-of-sample walk-forward backtest from January 2010:
| Metric | DEMA | SPY B&H |
|---|---|---|
| Total Return | -- | -- |
| CAGR | -- | -- |
| Volatility | -- | -- |
| Sharpe Ratio | -- | -- |
| Max Drawdown | -- | -- |
| Avg Exposure | -- | 100% |
Walk-forward test begins April 1, 2026
Key Properties
DEMA uses European macroeconomic data to allocate across eight global asset classes. Each asset receives a conviction-weighted allocation based on the model's proprietary signal framework, allowing nuanced exposure scaling.
The strategy was validated through multiple statistical robustness tests including permutation testing, stress analysis, and regime-specific evaluation. All reported metrics include transaction costs.
The model rebalances monthly with a turnover dampener that blends target and previous weights to reduce unnecessary trading. Average portfolio exposure varies between 40% and 150% depending on macro conditions.
The signal generation methodology, specific indicators, and model parameters are proprietary.
Walk-Forward Test Disclaimer: This strategy (DEMA) is currently in a walk-forward testing phase starting April 1, 2026. All performance data before this date represents backtested results using a strict out-of-sample methodology. Past performance, whether backtested or live, does not guarantee future results. Transaction costs are included. This implementation is for research and educational purposes only. The exact methodology is proprietary and not disclosed.