Skip to main content
V-Lab
V-Lab

Volatility Analysis Page — User Guide

View historical volatility charts, forecasts, and model statistics for a specific security

Contents

  • Page Overview

  • Essential Concepts

  • Navigation Guide

  • Data Interpretation

  • Practical Applications

  • Understanding Data

  • Troubleshooting

  • Tips & Best Practices

Page Overview

The Volatility Analysis page displays a time series chart of historical volatility, a summary table with forecasts, and model estimation details. Use it to understand a specific asset's risk characteristics and see where volatility is headed.

V-Lab estimates volatility using GARCH-family models that capture how volatility clusters over time and responds asymmetrically to positive and negative returns. V-Lab offers 12 different volatility models, from standard GARCH to specialized variants for different market conditions.

Volatility Chart

The main chart shows historical volatility as a time series. Use the date range buttons (6M, 1Y, 2Y, 5Y, 10Y, All) or custom date pickers to examine specific periods. Add comparison series to see how volatility relates to VIX or other assets.

GARCH Model Selection

V-Lab offers 12 volatility models. GJR-GARCH is the default for all assets. Use the Related Models panel to switch between available specifications.

Statistical Output

View model parameters, estimation details, and summary statistics. The Estimation Info panel shows the mathematical specification and coefficient values driving the forecasts.

Comparative Analysis

Add VIX, other securities, or different models to the chart. Compare volatility levels and patterns across assets to understand relative risk and correlation.

Was this section helpful?

These terms appear in the charts and tables on this page:

Volatility Prediction

The model's forecasts of volatility at multiple horizons (1 Day, 1 Week, 1 Month), shown prominently in the header as metric tablets. Displayed as annualized percentages with change indicators.

Why it matters: An up arrow in red means volatility is forecast to increase. A down arrow in green means it's forecast to decrease. This gives you a quick read on the current risk trend.

Annualized Volatility

Volatility scaled to a yearly basis. A 25% annualized volatility means the asset typically moves within plus or minus 25% over a year. V-Lab displays all volatility values as annualized percentages for easy comparison across assets and time periods.

Why it matters: The chart y-axis shows annualized volatility percentages. Higher values mean larger expected price swings. During crises (2008, 2020), you'll see values spike well above normal levels.

GJR-GARCH Model

The default volatility model at V-Lab. Extends standard GARCH by capturing the asymmetric response to returns: negative returns increase volatility more than positive returns of the same magnitude. Named after Glosten, Jagannathan, and Runkle (1993).

Why it matters: Markets exhibit this asymmetry empirically: bad news spikes volatility more than good news calms it. GJR-GARCH captures this pattern, making forecasts more accurate during downturns.

Volatility Persistence

The tendency for current volatility levels to influence future volatility. High persistence means today's volatility strongly predicts tomorrow's. In GARCH models, persistence is measured by alpha + beta, where values near 1 indicate very slow decay of volatility shocks.

Why it matters: After a volatile day, expect more volatility. After calm days, expect continued calm. This clustering effect means volatility forecasts are most accurate in the short term.

Mean Reversion

The tendency for volatility to return toward its long-run average over time. Extreme volatility eventually subsides; unusually low volatility eventually rises. The speed of mean reversion depends on persistence.

Why it matters: The 1-year forecast converges toward the long-run average while short-term forecasts reflect current conditions. Compare them to see whether volatility is expected to normalize.

Was this section helpful?
Understanding the Interface

The Volatility Analysis page is organized into several key areas that work together to give you a complete picture of the asset's risk characteristics.

Chart Controls
  • Model Selection

    The header shows which GARCH model is being used (e.g., GJR-GARCH, EGARCH). The Related Models panel on the right lists all available models for this asset. Click any model name to switch.

  • Adding Comparison Series

    Click the Compare button to overlay additional volatility series. You can add VIX, other ticker symbols, or different models to see relative volatility levels.

  • Time Period Controls

    Use the preset buttons (6M, 1Y, 2Y, 5Y, 10Y, All) to quickly adjust the chart's time range. The From/To date pickers allow custom date ranges.

  • Viewing the Chart

    The chart displays the selected date range. Use the preset buttons or custom date pickers above the chart to adjust the time period shown.

Volatility Prediction Display

The header displays multiple forecast horizons as metric tablets, showing annualized volatility predictions for different time frames:

  • 1 Day Prediction

    Tomorrow's predicted annualized volatility, with a colored chip showing the change from the previous forecast. Most responsive to recent market moves.

  • 1 Week Prediction

    Predicted average volatility over the next 5 trading days. Smooths out daily noise while still reflecting current conditions.

  • 1 Month Prediction

    Predicted average volatility over the next 22 trading days. Shows where volatility is expected to settle in the near term.

Summary Table

The table below the chart shows current statistics and forecasts at a glance:

  • Price

    The most recent closing price for the security.

  • Return

    The latest daily return (percentage change from the previous close).

  • Avg Week Vol

    Average volatility over the past 5 trading days. Shows recent volatility conditions.

  • Avg Month Vol

    Average volatility over the past 22 trading days. A slightly longer-term view of recent conditions.

  • Min Vol

    The minimum volatility observed over the full estimation period. Shows the calmest conditions this asset has seen.

  • Max Vol

    The maximum volatility observed over the full estimation period. Usually corresponds to crisis events like 2008 or 2020.

  • Avg Vol

    The average volatility over the full estimation period. This is the long-run baseline for this asset.

  • Vol of Vol

    How much volatility itself fluctuates. Higher values mean more unpredictable risk levels.

  • 1 Week Pred

    Model forecast for average volatility over the next week.

  • 1 Month Pred

    Model forecast for average volatility over the next month.

  • 6 Month Pred

    Model forecast for average volatility over the next six months. Starts converging toward the long-run average.

  • 1 Year Pred

    Model forecast for average volatility over the next year. Essentially the model's estimate of long-run volatility.

Plot Options

The menu bar above the chart provides several customization options:

  • Date Range Presets

    Click 6M, 1Y, 2Y, 5Y, 10Y, or All to quickly change the time span. Each preset adjusts both the chart and the summary statistics.

  • Custom Date Range

    Use the From and To date pickers to specify exact dates. Useful for examining specific events or matching analysis periods.

  • Compare

    Opens a dialog to add comparison series. Enter ticker symbols to add other assets, select VIX for market volatility context, or add different models for the same asset.

  • Sub Plot

    Add a subplot below the main chart showing daily returns, prices, or both. Useful for correlating volatility spikes with specific price movements.

  • Line Style

    Toggle between plain lines and lines with data point markers. Adding points helps identify exact values at specific dates.

  • Legend Position

    When multiple series are displayed, adjust where the legend appears on the chart (top-left, top-right, etc.) to avoid obscuring data.

Was this section helpful?
Understanding Statistical Output

The page provides detailed statistical information to help you assess the model's quality and understand the volatility dynamics:

  • Volatility Prediction Summary

    The header tablets show forecasts at 1 Day, 1 Week, and 1 Month horizons. Compare these to see whether volatility is expected to increase, decrease, or remain stable.

  • Model Parameters (Estimation Info)

    Click to expand the Estimation Info panel. It shows the GARCH model coefficients (omega, alpha, beta, gamma for asymmetric models), estimation period, and number of observations used. High persistence (alpha + beta near 1) means volatility shocks decay slowly.

  • Summary Statistics

    The table shows historical min/max/average volatility plus forward-looking predictions. Compare current volatility to the historical range to assess whether conditions are normal or extreme.

Reading Volatility Patterns

The volatility chart reveals several important patterns that help you understand risk dynamics:

  • Volatility Clustering

    High volatility days tend to follow high volatility days, and calm periods persist. This clustering is captured by the GARCH model and is why short-term forecasts are most accurate.

  • Asymmetric Response (Leverage Effect)

    Negative returns increase volatility more than positive returns of the same size. This is why GJR-GARCH often outperforms standard GARCH: it explicitly models this asymmetry through the gamma parameter.

  • Mean Reversion

    Extreme volatility eventually returns toward the long-run average. After spikes (like 2008 or 2020), watch the longer-term forecasts to see how quickly the model expects normalization.

Was this section helpful?
Forecast Interpretation

Use the forecasts to inform your risk decisions:

  • Short-Term Forecasts (1 Day, 1 Week)

    Most reliable. Based on current conditions and recent volatility. Use for daily position sizing, stop-loss placement, and short-term option pricing decisions.

  • Medium-Term Forecasts (1 Month, 6 Month)

    Useful for strategic planning. Shows how volatility is expected to evolve. If currently elevated, these forecasts show the expected path back to normal.

  • Long-Term Forecasts (1 Year)

    Converges to the unconditional (long-run) volatility. Use as a baseline for what's 'normal' for this asset. Current conditions have little influence on this forecast.

Practical Applications
  • Risk Management

    Use volatility forecasts to set position sizes and risk limits. Higher predicted volatility warrants smaller positions or wider stop-losses.

  • Portfolio Construction

    Compare volatility across assets in your portfolio. Assets with lower or more stable volatility may deserve larger allocations during uncertain periods.

  • Options Analysis

    Compare V-Lab's statistical volatility forecasts to implied volatility from options markets. Discrepancies may indicate trading opportunities.

Was this section helpful?
GARCH Model Framework

V-Lab uses Generalized Autoregressive Conditional Heteroskedasticity (GARCH) models to estimate and forecast volatility. These models capture two key empirical facts: volatility clusters over time, and it responds asymmetrically to returns.

Available Volatility Models
  • GARCH (1986)

    The baseline model capturing volatility persistence and clustering. Today's conditional variance depends on yesterday's conditional variance and yesterday's squared return. Simple, robust, and effective for many assets.

  • GJR-GARCH (1993)

    Extends GARCH with an asymmetry term (gamma) that allows negative returns to have a larger impact on future volatility than positive returns. The default choice for all assets at V-Lab because it captures the empirical leverage effect.

  • EGARCH (1991)

    Models log-volatility rather than variance, which ensures forecasts are always positive without parameter constraints. Also captures asymmetry through a different specification. May perform better for assets with extreme volatility events.

  • Specialized Models

    V-Lab also offers APARCH, Spline-GARCH, MEM variants, GAS-GARCH, and MF2-GARCH for specific scenarios. Check the Related Models panel or Documentation for details on these advanced options.

Data Updates and Methodology

Volatility estimates update daily after market close (typically by 6 PM ET for US assets). V-Lab estimates models using the full available price history with maximum likelihood estimation. Price data comes from standard market data providers using adjusted closing prices.

Was this section helpful?
Common Questions
Why does the volatility forecast take time to respond to news?

GARCH models incorporate persistence: today's volatility depends on past volatility, not just today's return. A single large move increases forecasts, but it takes sustained volatility to push forecasts significantly higher. This prevents overreaction to one-day events and reflects how markets actually behave.

Which model should I use?

GJR-GARCH is the default and works well for most assets because it captures the leverage effect (negative returns increase volatility more than positive returns). For assets with extreme tail events, try EGARCH. For analyzing long-run volatility trends, consider Spline-GARCH. Check the Related Models panel to see what's available for your asset.

Why do model parameters change over time?

Parameters are re-estimated as new data arrives. Market dynamics evolve: volatility persistence may increase during crises and decrease during calm periods. Updated parameters ensure forecasts reflect current conditions rather than outdated relationships.

How well do the models handle extreme events?

GARCH models capture volatility clustering and increase forecasts during stress. However, extreme events by definition exceed model expectations. The models are best for typical market conditions and near-term forecasting. Use the forecasts as guidance, not guarantees, especially during unusual market conditions.

Data Limitations

Volatility estimates are based on daily closing prices and may not capture intraday dynamics or overnight gaps. The models assume returns follow a specific statistical process, but actual market behavior is more complex. Forecasts are most reliable in the short term and during normal market conditions.

Technical Questions

More detailed questions about the models and data:

Why did the model parameters change suddenly?

Models are re-estimated as new data arrives. If a significant market event reveals different volatility dynamics, parameters update to reflect this. Major market events like COVID-19 in 2020 caused noticeable parameter shifts across many assets.

What lag lengths does V-Lab use?

Most models use GARCH(1,1) or GJR-GARCH(1,1), meaning one lag of conditional variance and one lag of squared returns. This parsimonious specification works well empirically for most assets and is standard in academic and industry practice.

How did the models perform during crises?

During 2008 and 2020, volatility forecasts increased sharply as market stress unfolded. The models captured the clustering and persistence of high volatility. Short-term forecasts tracked realized volatility reasonably well, though extreme daily moves still exceeded predictions.

Can volatility forecasts be negative?

No. The GARCH model structure ensures variance forecasts are always positive. If you see unusual values, it may indicate a data issue or an asset with insufficient history for reliable estimation.

How do I download the volatility data?

Click the Download menu above the chart. You can download the chart as an image or download the data as a CSV file. Downloading the CSV requires you to be logged in. The CSV includes dates, volatility values, and model parameters with column headers explaining each field.

Was this section helpful?
Recommended Analysis Workflow
  • 1. Check the Header Predictions

    Start by looking at the 1 Day, 1 Week, and 1 Month predictions in the header. The change indicators tell you immediately whether volatility is rising or falling.

  • 2. Examine the Chart

    Look at the volatility time series. Is current volatility high or low compared to history? Are there recent spikes? Use the date range buttons to focus on specific periods.

  • 3. Review the Summary Table

    Compare current volatility (Avg Week Vol, Avg Month Vol) to the historical range (Min Vol, Max Vol, Avg Vol). This tells you whether conditions are normal or extreme.

  • 4. Check Estimation Info

    Expand the Estimation Info panel to see model parameters. High persistence (alpha + beta near 1) means volatility shocks decay slowly. The gamma coefficient shows asymmetry strength.

  • 5. Add Context with Compare

    Use the Compare button to add VIX or related assets. This shows whether elevated volatility is asset-specific or reflects broader market stress.

Best Practices
  • Combine with Fundamental Analysis

    Volatility forecasts show statistical risk. Combine with earnings calendars, news, and fundamental research for a complete picture.

  • Compare Models Periodically

    Use the Related Models panel to see how different GARCH variants estimate volatility. If they diverge significantly, investigate why. It may reveal important dynamics.

  • Validate Forecasts

    Track how forecasts compare to realized volatility over time. This helps you calibrate confidence in the predictions for specific assets.

  • Adjust Risk Budgets

    When volatility forecasts increase, consider reducing position sizes or tightening risk limits. The model is warning you that larger moves are expected.

  • Be Regime-Aware

    Volatility tends to be higher during certain periods (earnings season, Fed meetings, crises). Use historical patterns to anticipate when forecasts may understate risk.

Was this section helpful?

Was this page helpful?