The Fund That Made 1,000% and Died in Fifteen Minutes

For seven years there was a trade that worked almost every day. You sold volatility. Specifically, you bought a product that was short VIX futures, and because volatility spent most of the post-2009 decade falling or staying low, and because the VIX futures curve normally slopes upward so that a short position collects the roll, the line went up. From launch in November 2010 to its January 2018 peak, the flagship product returned more than 1,000%. In 2017 alone it gained 176%.

It was, in the phrase the Bank for International Settlements used afterwards, collecting pennies in front of a steamroller. On 5 February 2018 the steamroller arrived, and it took about fifteen minutes.

The equity market barely moved

This is the fact that makes the story genuinely instructive rather than merely dramatic. The S&P 500 fell 4.1% on 5 February 2018. That was only about the thirty-first largest daily decline since the VIX began in 1993. An unpleasant day. Not a crash, not 1987, not March 2020.

The VIX, on the same day, closed at 37.32 against 17.31 the previous Friday. That is +115.6%, the largest one-day rise in the index's history — nearly double the previous record of +64.2% set in February 2007. In after-hours trading it printed above 50.

An ordinary bad day in the underlying produced a record-breaking day in the derivative. That gap is the whole story, and it is a general property of leveraged and inverse products, not a quirk of 2018.

The mechanism: everyone had to buy, and everyone knew it

Products that promise a fixed daily exposure — minus one times an index, or two times an index — have to rebalance at the end of every session to restore that ratio. The arithmetic is unforgiving and completely public.

As the VIX rose through 5 February, the leveraged long volatility products needed to buy more VIX futures to maintain their target exposure. The inverse products needed to buy VIX futures to cover the losses on their short. The post-mortem by the Bank for International Settlements — the bank for central banks — is worth quoting exactly: “both long and short volatility ETPs had to buy VIX futures … Due to the mechanical nature of the rebalancing, a higher VIX futures price necessitated even greater VIX futures purchases by the ETPs, creating a feedback loop.”

Everyone in the market knew this. Traders began bidding up VIX futures from around 15:30 in anticipation of the flow. The peer-reviewed reconstruction (Augustin, Cheng and Van den Bergh, Financial Analysts Journal, 2021) estimates the two main inverse funds alone had to buy roughly $2.6 billion of VIX futures — about 93,000 contracts — into the close. Including the leveraged long funds, around 113,000. For context, the whole market's average daily volume in the preceding week was about 400,000 contracts, and total open interest was around 600,000.

Roughly a quarter of a normal day's entire volume, needed by a known set of buyers, at a known time, in a known direction. The BIS recorded that at 16:08 more than 115,000 contracts changed hands within a single minute at heavily inflated prices — “roughly one quarter of the entire market.”

The French market regulator’s report on the episode (AMF, Heightened volatility in early February 2018: the impact of VIX products, April 2018) contains the sentence to remember: the fund was destroyed “due to the change in the future price between 4pm and 4:15pm” — that is, entirely after the cash equity market had closed for the day. A holder who watched the S&P close down 4% and went to make dinner had already lost almost everything and did not know it.

The clause in the prospectus

The XIV product was an exchange-traded note — a debt instrument issued by a bank, not a fund holding assets. Its terms contained an Acceleration Event: if the indicative value fell to 20% or less of the prior day's closing value, the issuer could terminate the note early. The prospectus even spelled out the scenario: if the underlying futures rose more than 80% in a day, it was “extremely likely” the notes would be accelerated.

That clause had sat in the documents for seven years while the product returned a thousand percent. It was invoked once, and once was enough.

One footnote from that afternoon deserves attention, because it is about disclosure rather than volatility. The issuer stopped updating the note's published intraday indicative value at around 16:10, leaving a figure near $24.70 on the screen until 17:08 while the true value was around $4.22. In that window investors bought 28.8 million notes for $823.6 million at an average of $28.60 — a transfer of roughly $700 million to better-informed sellers, and the basis of the litigation that followed.

The fund that survived, and what it had to become

The sister product, SVXY, was an ETF rather than a note. It had no acceleration clause, so it lost roughly 90% and kept trading. Three days later it took in $300 million of new money in a single session — investors buying the dip in a product whose dip had just been ninety percent.

On 27 February 2018 its sponsor cut the target exposure from −1x to −0.5x, and cut the leveraged long product from 2x to 1.5x, on the same day. That is the industry's own verdict on whether the original design was survivable. On the sponsor's arithmetic, a repeat of 5 February would now cost the half-leveraged fund about 48% instead of 90%.

Why nobody saw it coming, which is the actual lesson

These products were not marketed to fools, and the risk had been modelled. One widely-circulated pre-event analysis — by Vance Harwood at Six Figure Investing — concluded that a drop into termination territory would require an equity event on the scale of October 1987 — a 20.5% one-day crash. It got there by assuming VIX futures would keep under-reacting to VIX spikes, as they had in 2010, 2011 and 2016.

There was no structural reason they had to. The assumption was an observed regularity, not a law, and it had been reinforced by exactly the seven years of calm that had also grown the products from marginal to systemically relevant. By late 2017 the whole VIX futures market had a notional size around $7 billion, while the leveraged and inverse volatility ETPs alone held around $5 billion. The trade had eaten enough of its own market that its own rebalancing had become the thing most likely to kill it.

There had also been a warning that everyone had lived through: both products fell about 60% in 2015. Traders who remembered it hedged. Newer holders had learned a different lesson from the same event — that dips in this product were opportunities.

The retail translation

Almost nobody reading this will short VIX futures. Everyone reading this will meet the underlying pattern.

A strategy that wins constantly and loses catastrophically feels like skill for years. Selling premium, averaging into strength, martingale sizing, holding through drawdowns because it always came back — all of these produce a beautiful equity curve and a hidden distribution of outcomes in which one tail eats the entire history. Seven years of daily confirmation is not evidence of safety; it is exactly what this shape looks like from inside.

Read what the instrument is allowed to do to you. The acceleration clause was public for seven years. Leveraged and inverse products, notes versus funds, daily-reset mechanics, early termination, forced rebalancing — these are all in writing and almost nobody reads them until the day they matter.

Know when your product's real risk is transacted. The damage happened after the cash close. Futures gap over weekends, options assign at expiry, and rollover distorts price. A stop-loss protects you during the hours the market lets it work.