SVPulse · Cross-asset risk
Volatility spillover
Also called: vol spillover, volatility transmission
Volatility spillover is the transmission of volatility shocks from one asset or market to others, measured with VAR, GARCH or information-flow models.
Definition
Volatility spillover describes how turbulence in one market raises volatility in others, for example from equities to credit, from one region to another, or from energy to inflation-sensitive assets.
It is measured as the share of one asset's future volatility that can be attributed to shocks in another, or as the strength of lagged links between their volatility series.
Intuition
Markets are linked through shared investors, funding, hedging and fundamentals. When one market gets stressed, leveraged holders reduce risk elsewhere, hedgers move to correlated instruments, and information about the global outlook spreads. Volatility therefore travels, often faster than price direction does.
Knowing which assets transmit and which receive shocks helps frame where stress may show up next and how diversified a portfolio really is in turbulent periods.
A familiar pattern: a sharp rise in volatility in a major equity index often shows up within days as higher volatility in credit, emerging-market currencies and commodities held by the same investors. In a spillover table, the index would appear as a net transmitter for that period, and the other markets would show large shares of their variance coming from it.
How it's calculated
Common approaches:
- VAR variance decomposition (Diebold–Yilmaz): fit a vector autoregression to realized or implied volatility series and compute the share of each series' forecast-error variance due to shocks in the others.
- Multivariate GARCH (for example BEKK): estimate how lagged squared shocks in one return series enter another's conditional variance.
- Lagged causality tests on volatility series: Granger causality or transfer entropy from one asset's volatility to another's.
Each method needs a volatility input. Realized volatility from daily ranges or intraday data is common; implied volatility captures the market's forward-looking view.
How to read it
High spillover from one asset to many others marks it as a transmitter; assets mostly on the receiving end are receivers. Spillovers usually rise during crises, so a sustained increase in total spillover is itself a sign of a more connected, fragile market.
Results depend on window length, VAR lag order and forecast horizon. Check that a link persists across reasonable settings before relying on it.
Common mistakes
- Confusing correlation of returns with transmission of volatility.
- Reading statistical lead–lag as proof of a causal channel.
- Using overlapping assets in different time zones without aligning closing times, which creates artificial leads.
- Ignoring that orderings matter in non-generalized VAR decompositions.
- Treating a spillover estimate from one window as a permanent property of the market; transmitters and receivers change with the regime.
In Saqrid
On the SVPulse map, lines between assets show measured volatility spillover, strongest links first, with the top transmitter and receiver.
SVNexus can build its asset network from implied-vol spillover links, alongside return correlation, DCC correlation, transfer entropy and tail dependence.
Go deeper
Saqrid Institute lesson: Volatility Spillover (Diebold–Yilmaz). Lesson text is open to read without an account.