Volatility Forecast Chart User Guide
Visualize volatility forecasts across multiple horizons with risk metrics
Contents
Tool Overview
Essential Concepts
How to Use
Data Interpretation
Practical Applications
Understanding Data
Troubleshooting
Tips & Best Practices
How to Use
Tool Layout
The Forecast Chart tool has three main components:
- Forecast Curve
The main chart showing annualized volatility (y-axis) across forecast horizons in days (x-axis). The curve starts at current volatility and converges toward long-run levels. You can click anywhere on the curve to select that specific day and view its risk metrics.
- Horizon Buttons
Quick-access buttons to jump to standard horizons: 1d, 1w, 1m, 6m, 1y. Click to update the risk metrics panel for that horizon.
- Risk Metrics Panel
Shows VaR and ES at 95% and 99% confidence for the selected horizon. Updates when you click horizon buttons or adjust the tau slider.
- Summary Metrics
Located below the chart, displays three key reference values: 1-Year Forecast (where volatility is headed), Long-run Vol (the model's equilibrium level, shown in green), and Persistence (how slowly shocks decay).
- Horizon Details
In the tool panel, shows details for the selected day: the day number, forecast volatility (color-coded by stress level), difference from long-run, and percentage convergence toward long-run.
Using the Horizon Explorer
Click the horizon buttons to focus on specific forecast periods:
- 1 Day (1d)
Tomorrow's forecast. Most influenced by current conditions. Best for daily trading decisions and short-term risk limits.
- 1 Week (1w)
5-day average forecast. Smooths daily noise while reflecting current regime. Useful for weekly position sizing.
- 1 Month (1m)
22-day average forecast. Shows medium-term outlook. Compare to 1-week to see expected trend direction.
- 6 Months (6m)
126-day average forecast. Begins converging toward long-run volatility. Useful for strategic planning.
- 1 Year (1y)
252-day average forecast. Essentially the model's long-run volatility estimate. Compare to current levels for regime context.
Understanding Risk Metrics
The tool calculates Value-at-Risk and Expected Shortfall for the selected horizon:
- VaR 95%
The loss level exceeded only 5% of the time. For a $100 position with -3.2% VaR, expect to lose more than $3.20 on roughly 1 in 20 days.
- VaR 99%
The loss level exceeded only 1% of the time. More conservative than 95% VaR. Used for stress testing and regulatory capital calculations.
- ES 95% (Expected Shortfall)
Average loss when VaR 95% is exceeded. Also called CVaR. More sensitive to tail risk than VaR alone.
- ES 99%
Average loss in the worst 1% of scenarios. The most conservative standard measure. Important for stress testing.
Using the Tau Multiplier
For models with a tau parameter (Spline-GARCH, MF2-GARCH), the tau slider enables scenario analysis:
- Adjusting the Slider
Move from 0.5× to 2× to see how different long-run volatility levels affect forecasts. Higher multipliers simulate higher-volatility regimes. The slider background is color-coded to preview the resulting 1-year forecast at each setting.
- Scenario Analysis
Use tau adjustment to stress-test. What if long-run volatility increases 50%? How does that change your 6-month VaR? This helps plan for regime changes.
Hover and Click Interactions
The chart responds to mouse movements and clicks:
- Hovering
Move your mouse over the forecast curve to see a vertical line and tooltip showing: the day number, forecast volatility, difference from current volatility, and difference from long-run. Positive differences are shown in red/orange; negative in green.
- Clicking
Click anywhere on the chart to select that specific day. The tool panel will update to show the forecast details and risk metrics for your selected day.
Data Interpretation
The shape of the forecast curve reveals important information about current market conditions and expected volatility evolution.
Interpreting Curve Patterns
- Elevated and Decaying
Current volatility above long-run average, forecast declining over time. Normal pattern after volatility spikes. Risk is elevated short-term but expected to normalize.
- Depressed and Rising
Current volatility below long-run average, forecast rising toward normal. Calm conditions expected to eventually give way to typical volatility levels.
- Flat at Long-Run Level
Current volatility near long-run average, forecast relatively flat. Stable regime with no significant changes expected.
Practical Applications
Risk Management
Use forecast volatility to set position sizes and risk limits. Higher predicted volatility warrants smaller positions or wider stop-losses. VaR estimates help quantify potential losses.
Trading Decisions
Check the forecast before large trades. If volatility is expected to decline, consider waiting for calmer conditions. Use VaR to size positions appropriately for your risk tolerance.
Options Analysis
Compare statistical volatility forecasts to implied volatility from options markets. The forecast term structure shows what your model expects. Divergences may signal trading opportunities.
Understanding Data
The Forecast Chart computes multi-step ahead predictions using the estimated GARCH model parameters.
How Forecasts are Calculated
For each horizon, the tool iterates the GARCH variance equation forward, starting from current conditional variance. Multi-day forecasts average the variance path. Results are converted to annualized volatility.
Troubleshooting
Common Questions
Why is my forecast curve nearly flat?
A flat curve means current volatility is close to long-run volatility and the model expects little change. This is typical during stable market periods. The curve becomes steeper after volatility spikes.
Why is my VaR so high at longer horizons?
VaR scales with the square root of time for multi-day horizons. A 1-day VaR of 2% becomes roughly 4.5% over 5 days (2% × √5). Longer horizons have higher VaR even if volatility is mean-reverting.
Why don't I see the tau slider?
The tau slider only appears for models with a tau parameter (Spline-GARCH, MF2-GARCH). Standard GARCH, GJR-GARCH, and EGARCH don't have this parameter and won't show the slider.
Technical Questions
How is multi-day VaR calculated?
The tool computes the average variance forecast over the horizon, converts to volatility, then applies normal distribution quantiles. This assumes returns are approximately normal. Actual tail risks may be larger.
How do forecasts work for liquidity (ILLIQ) models?
For liquidity models, the forecast shows projected illiquidity rather than volatility. The interpretation is similar: elevated illiquidity forecasts suggest higher trading costs. VaR and ES are not computed for liquidity forecasts.
Tips & Best Practices
Get the most out of the Forecast Chart tool with these recommendations.
Recommended Workflow
- 1. Check Current vs Long-Run
Compare the left edge (current) to the right edge (long-run) of the curve. This tells you whether volatility is stressed or calm relative to normal.
- 2. Compare Horizons
Click through 1d, 1w, 1m to see forecast evolution. Big differences between horizons indicate strong mean reversion expectations.
- 3. Assess Risk Metrics
Review VaR and ES for your relevant horizon. For daily trading, use 1d metrics. For position holding periods, match the horizon.
- 4. Stress Test (if available)
For models with tau, use the slider to explore regime scenarios. What if long-run volatility increases significantly?
Was this page helpful?