VIX history: the seven major volatility spikes 1990-2026 and their macro contexts

Since 1990, the VIX has produced twenty-six distinct episodes of closes above 30 (separated by at least two months), three above 50 and two above 80 (Eco3min calculation on Cboe closes, January 1990 to September 2026). Each spike documents a specific market dislocation; none repeats identically.

The empirical catalog that follows covers the seven major episodes from 1990 to 2026, from LTCM 1998 to the yen carry trade unwind 2024, with macro context, spike magnitude, resolution duration, and associated S&P 500 drawdown.

Selection methodology

A major VIX spike in this inventory combines a peak above 45 on Cboe’s daily-high series and an identifiable macro-financial dislocation whose mechanism the catalog can document. The criterion is not purely mechanical: it leaves aside summer 2002 (45.08 at the close on 5 August, absorbed into the dot-com phase), the May 2010 Flash Crash (45.79 at the close on 20 May, covered in section 5) and 24 August 2015 (the yuan devaluation, a single close above 40, at 40.74). The April 2025 tariff shock (52.33 at the close on 8 April, three consecutive sessions above 40) meets both conditions and is covered in a postscript at the end of section 7.

For each episode, the documented variables are: the date of initial 30 breach, the closing peak, the spike duration (days above 30), the associated S&P 500 drawdown, and the macro context of the trigger. Data comes from the VIXCLS series published by FRED based on CBOE quotes, and, for the 2008 intraday record and the selection threshold, from CBOE’s daily-high history.

LTCM, Russia, and the VIX’s installation as a barometer (1998)

The first major modern VIX spike comes in September-October 1998. Russia’s surprise ruble devaluation on 17 August 1998 and the Long-Term Capital Management (LTCM) hedge fund collapse in mid-September triggered a major cross-asset dislocation: credit spreads, swap spreads, OTM option premiums, all widened simultaneously. The VIX first closed above 30 on 4 August, closed above it without a break from 21 August, at 44.28 on 31 August (the day of the S&P 500 trough), peaked at 45.74 at the close on 8 October 1998, and stayed above 30 until 28 October, 48 consecutive sessions.

The S&P 500 drawdown, from the 17 July peak to the 31 August closing trough, stopped just short of the 20% bear-market threshold; the trough was retested on 8 October. The coordinated intervention of Wall Street dealers to take over LTCM’s positions, under the New York Fed’s auspices, stabilised the market from mid-October. The VIX dropped back below 30 on 29 October and held there from early November. The episode durably installed the VIX in risk desks’ radar: it is the first episode where the index functioned as a recognised systemic stress indicator and not merely as a statistical curiosity.

Dot-com and 11 September 2001 (2000-2002)

The dot-com bubble bursting from March 2000 produced a progressive compression without a single characteristic spike. Setting aside the 11 September shock, the VIX oscillated between 16.5 and 34.7 at the close from March 2000 to March 2002, below 25 nearly two sessions out of three, with closing peaks of 33.49 on 14 April 2000 and 34.72 on 3 April 2001; it was at the depth of the bear market, in summer 2002, that it reached 45.08 at the close (5 August). The S&P 500 drawdown from the March 2000 peak to the October 2002 trough was comparable in depth to the GFC’s, but spread over a far longer period.

11 September 2001 produced an isolated spike within this phase. The VIX, closed during the 11 to 17 September market suspension, closed at 41.76 on reopening day, 17 September, and peaked at 43.74 on close on 20 September. From 10 to 21 September the S&P 500 fell into correction territory, followed by a quick rebound. The episode is atypical for its exogenous nature and the rebound speed; it has served as reference for acute geopolitical shocks ever since.

The GFC, absolute reference of the series (2008-2009)

The global financial crisis produced the largest episode of the 1990-2026 VIX series. The VIX crossed 30 on 15 September 2008 (Lehman Brothers bankruptcy day), reached 89.53 intraday on 24 October 2008, and closed at 80.86 on 20 November 2008. The first remains the all-time intraday high, and the second has been topped at the close only by 16 March 2020 (82.69). The VIX closed above 40 from 29 September 2008 to 7 April 2009 (bar a few sessions between 38 and 40), a little over six months, and above 30 for 170 consecutive sessions, from 15 September 2008 to 18 May 2009: itself an unprecedented duration.

The S&P 500 drawdown, from the October 2007 peak to the 9 March 2009 trough, is the deepest of the period. The nature of the crisis (forced deleveraging, interbank dislocation, massive US Treasury intervention) explains the spike duration. This spike takes its place among the others in the dated sequence of volatility crises read by regime. Contrary to the usual pattern where a spike resolves in weeks, the VIX did not durably return below 30 until May 2009, eight months after Lehman, and the S&P 500 kept falling until 9 March 2009, three and a half months after the VIX’s closing peak.

Flash Crash, eurozone and fiscal cliff (2010-2012)

The post-GFC period produced three spike episodes without reaching 2008 levels. The Flash Crash of 6 May 2010, triggered by a poorly calibrated algorithmic sell order on e-mini S&P 500 futures, saw the VIX close at 32.80, then at 40.95 the next day, before returning below 30 on 10 May; the episode’s peak, 45.79 at the close on 20 May, came two weeks later. The S&P 500 plunged intraday that day before rebounding, but the index kept falling for two months, into correction territory below its April 2010 peak.

The European sovereign debt crisis of August-October 2011 produced a more durable spike. The VIX crossed 30 on 4 August 2011, the day before S&P’s downgrade of US sovereign debt to AA+ (announced on Friday 5 August after the close), peaked at 48.0 on 8 August, and stayed above 30 for 50 consecutive sessions, until 13 October. The S&P 500 drawdown between the 29 April peak and 3 October 2011 stopped just short of the bear-market threshold. Final resolution coincided with Mario Draghi’s whatever it takes declaration in July 2012, but the VIX did not close below 20 again until 19 January 2012.

The late-2012 fiscal cliff produced a third, more moderate spike (peak at 22.7), which resolved within a few sessions after the US Congress agreement on 2 January 2013.

Volmageddon and post-2017 stress (2018-2019)

5 February 2018 produced an atypical episode in VIX history. After months of record compression (annual average VIX 11.1 in 2017, closing low 2017 at 9.14), the VIX leapt from 17.3 to 37.3 in a single session. The immediate cause was the implosion of several short-volatility ETNs (notably Credit Suisse’s XIV, which lost 96% of its value in one session) which forced massive unwinding of short-vol positions by market-makers hedging these products.

The S&P 500 drawdown remained relatively contained (a correction between 26 January and 8 February), and the VIX dropped back below 20 by end-February. The episode is documented in detail by Augustin, Cheng and Van den Bergen (Financial Analysts Journal, 2021), who reconstruct the loop of VIX futures buying by short-vol ETPs into the close and point to the procyclicality of short-vol structured products and the opacity of their systemic impact. It has served as reference for configurations where VIX structural compression precedes a sudden repricing.

Late 2018 produced another episode (closing peak at 36 in December 2018, S&P 500 drawdown just short of a bear market), linked to the Fed’s tightening cycle and US-China trade tensions.

COVID and yen carry trade unwind (2020-2024)

The COVID crash of March 2020 produced the second-largest spike of the series. The VIX crossed 30 on 27 February 2020 (39.16 at the close, after 25.03 on the 24th), peaked at 82.69 on close on 16 March 2020 (the all-time closing high of the series), and stayed above 30 for 50 consecutive sessions, until 7 May, with brief returns above that level in June and around the November election. The S&P 500 drawdown reached -34% between 19 February and 23 March, the fastest drawdown in modern history. The resolution itself was atypically rapid thanks to coordinated Fed (0% rates, unlimited QE, liquidity programmes) and Treasury (CARES Act) intervention. The Eco3min catalogue of historical market crises places them on a common timeline.

The 5 August 2024 episode remains one of the most instructive of the decade. The VIX, at 16.36 at the close on 31 July, leapt on 5 August to one of the highest intraday peaks of the series, but closed at only 38.57. The trigger was the forced unwind of the yen carry trade, itself caused by the Bank of Japan’s surprise rate hike on 31 July and a weaker-than-expected US employment report on 2 August. The Nikkei lost 12.4% on 5 August, the largest single-day drop since 1987. The S&P 500 fell far less the same day, but the VIX reacted as to a major crash. The resolution was quick: the VIX dropped back below 25 within three sessions (23.79 on 8 August) and below 18 within seven (16.19 on 14 August). The episode illustrates how a cross-asset shock (FX, futures, equity) can produce a VIX spike without a corresponding equity crash. Read alongside: VIX contrarian thresholds and historical equity rebounds.

Dated postscript, 22 September 2026: the April 2025 tariff shock meets this inventory’s criterion and is, in effect, an eighth episode, visible on the chart above. After the reciprocal tariff announcement of 2 April, the S&P 500 posted two heavy daily losses on 3 and 4 April; the VIX closed at 45.31 on 4 April, at 46.98 on 7 April and at 52.33 on 8 April, its highest close since April 2020, before the 90-day pause announced on 9 April (a sharp S&P 500 rebound that day, VIX back to 33.62). The S&P 500 drawdown from the 19 February peak to 8 April stopped just short of a bear market; the VIX stayed above 30 for ten consecutive sessions and dropped back below 20 on 12 May. Since then no episode has exceeded 31.05 at the close (27 March 2026), as of 21 September 2026 (Cboe daily closes; Eco3min calculation).

Recurring patterns and historical-analysis limits

Three recurring patterns appear in the catalog. First: post-spike resolution speed has accelerated since 2018. Measured in consecutive sessions closing above 30, resolution took 48 sessions in 1998, 170 in 2008-2009, 50 in 2011 and 50 again in 2020; it fell to isolated sessions in February 2018, 1 in 2024 and 10 in April 2025. The acceleration holds for positioning shocks (2018, 2024, 2025), not for the real shock of 2020, which lasted as long as 2011. The SPX options market maturation, deeper liquidity, and faster central bank intervention explain this acceleration.

Second: the VIX intraday / VIX close ratio has stretched since 2010. Recent spikes capture more intraday than at close, making comparisons on daily series less informative than on intraday series. The full series of threshold breaches and post-spike returns documents this stretch.

Third: none of the seven episodes repeats identically. The GFC is unique in its duration, 9/11 in its pure exogenous nature, Volmageddon in its structural trigger, the yen carry trade unwind in its cross-asset concentration. Mechanical extrapolation from a past episode to a future one is the most frequent reading error. The general analytical framework for reading the VIX documents this epistemological limit.

What the index construction does to the comparison

The catalog compares readings produced under one methodology, not always observed in real time. The current VIX, designed by the Cboe with Goldman Sachs and published in 2003, prices a 30-day implied variance swap on a basket of SPX options weighted by 1/K². The VIXCLS series starts on 2 January 1990, and its 1990-2002 values were backfilled by the Cboe under the 2003 methodology for historical comparison, a span that covers the LTCM and 11 September episodes. VXO, the original Whaley 1993 index, was discontinued on 31 August 2021 (FRED: VXOCLS). Its 1986-2021 history helps gauge the effect of that methodological break on historical levels.

In October 2014, the methodology was revised to add weekly SPXW options. Before that, the two monthly expirations bracketing the 30-day horizon could sit far apart, sometimes 25 and 55 days, which made the 30-day interpolation less precise.

Closing values are themselves a convention. VIXCLS is the value fixed by the Cboe at the close of the SPX options session, 4:15 PM ET, from the latest bid-ask quotes. It usually differs from the last intraday print by a few hundredths of a point and does not reflect the extended session.

Truncation is the most delicate step. The calculation stops on each side as soon as two consecutive strikes carry a zero bid, so when liquidity retracts on deep out-of-the-money puts, the cut-off moves closer to spot and the index can understate the volatility that would have been paid on distant strikes. The VIX is also blind beyond 30 days, and it can sit low while crash protection is expensive. A low VIX with an elevated SKEW index, which measures the relative pricing of deep out-of-the-money puts, is a configuration documented in 2017, before Volmageddon.

Last updated — 29 October 2026