Vix spike in August not driven by actual volatility – BIS

The spike in the Cboe volatility index (Vix) during August’s market turmoil was not due to a fundamental rise in volatility, a Bank for International Settlements (BIS) bulletin has argued.

The bulletin, published on October 29, says the asymmetric widening of bid-ask spreads played a key role in exacerbating the Vix spike in August.

Shortly before markets opened on August 5, the Vix reached a peak comparable to the levels in the 2008 financial crisis and Covid-19. Whereas the two past Vix peaks indicated actual crises, the August market turmoil was much more benign, raising questions on the index’s reliability.

The authors – Karamfil Todorov and Grigory Vilkov – note that the Vix value is based on mid-values of option quotes rather than actual trades. This, they say, means the Vix is sensitive to the bid-ask spreads, which widened before markets opened on August 5 as market-makers increased their ask prices.

Crucially, the authors say, the widening bid-ask spreads were not accompanied by an outsize increase in actual trades during the pre-market period. This, they add, is presumably because traders were deterred by the extreme bid-ask spreads. Additionally, most of the increase in pre-market trading on August 5 was for at-the-money options, which have lower weight in the Vix formula.

The authors also note that Vix exchange-traded funds and dispersion trades were unlikely to be the main drivers behind the spike.

The authors argue the pre-market Vix can be less effective in indicating volatility because it can be significantly affected by illiquidity with wide bid-ask spreads. They add that the Vix during regular trading sessions, which is calculated in a more liquid market, could be seen as a more informative indicator of volatility.

“Despite these differences, Vix is treated very similarly in intraday versus pre-market periods by market participants and the media,” they write.