Help & Documentation
A complete reference guide for system configuration, mathematical models, inputs, and output metrics.
System Inputs
Standard Formatting
For all manual ticker and weight fields:
- Tickers: Space separated (e.g.,
SPY QQQ). - Weights: Space separated (e.g.,
60 40). Must sum to 100% if not using optimization. - Order: First ticker = first weight, etc.
Backtest Start Date
The starting point for the engine. Safely truncates to the earliest shared date if assets lack earlier data.
Portfolio Tickers
Standard ticker symbols. Apply leverage with custom-ticker notation (e.g., SPY?L=2 for 2x leverage).
Optimization Model
Includes 26 mathematical optimization objectives across hierarchical allocation, risk budgeting, volatility, downside risk, tail risk, drawdown risk, Gini risk, diversification, robust CVaR, and utility. Examples include Hierarchical Risk Parity, Minimum Tail Loss (CVaR), Distributionally Robust Tail Loss (CVaR), Maximum Gini Ratio, Maximum Sharpe Ratio, and Maximum Log Utility.
Lookback Window
"Expanding" uses all history. Rolling windows force the model to adapt to recent market regimes.
Min/Max Weight Bounds
Prevent over-allocation to a single asset. Enter space-separated percentages matching your tickers.
Starting Value vs. Current Value
Starting Value = backtest baseline. Current Value seeds the Forecast Engine and Live tab.
Cash Flows
Simulate recurring deposits (positive) or withdrawals (negative) to model accumulation or decumulation phases.
Rebalance Frequency
How often the portfolio returns to target weights. Value Averaging (SIG) is an advanced dynamic strategy.
SIG System (Value Averaging)
Set a Quarterly Growth Target. Outperformance skims into a Safety Sleeve; underperformance deploys from it.
Outputs & Metrics
Compound Annual Growth Rate.
Annualized standard deviation of returns.
Deepest peak-to-trough decline.
Return per unit of total risk.
Like Sharpe, penalizes only downside volatility.
CAGR divided by Max Drawdown.
Excess return vs. benchmark.
Volatility relative to the benchmark.
Merton EWMA-SBB
Simfolio uses one fixed portfolio-level forecast engine: positive-part historical return drift, an absolute-return EWMA volatility overlay, empirical standardized residuals, filtered empirical tails, and stationary residual blocks.
Portfolio-Level Target
Cross-asset behavior, tactical rules, leverage, cash allocation, and rebalancing are already embedded in the realized portfolio return stream before simulation.
Volatility & Tail Modeling
Forward shocks come from standardized residual blocks with a data-fit volatility overlay and filtered empirical tails, preserving the portfolio's realized non-Gaussian behavior without fitting assets separately.
Cash flows in path
Scheduled deposits or withdrawals are injected on their actual dates into every simulated wealth path.
Forecast Distribution Metrics
Forecast performance metrics are summarized across the full simulation ensemble rather than a single trajectory.
Probability of Loss
Likelihood the portfolio ends the forecast horizon below the invested capital implied by the configured starting value and cash-flow schedule.
Value at Risk (VaR 5%)
The 5th-percentile terminal threshold across the simulated scenario set.
CVaR 1% / CVaR 5%
Average outcome inside the selected worst tail of the simulated distribution once the VaR threshold has already been breached.