Apr 30, 2018

13 min read

The VIX & The Unwinding Of Short-Volatility Trades

The first in a series of deep dives on economic events centered around one or a few financial products. This one? An exploration of the VIX, what it is, how it works, volatility trading, short-vol strats, and mostly what happened on February 5th, 2018 when volatility skyrocketed and blew up one of the most profitable trading strategies of 2017.

Introduction

One thousand, one hundred and seventy-five points. That was how much the Dow Jones Industrial Average (DJX) dropped on Monday, February 5th, 2018 — the largest nominal drop in its 122-year history. This was accompanied by the S&P 500 (SPX) having its worst day since 2011, and new ten year lows for Asian & European indexes. All in all, over four trillion dollars were lost, on a day that shook markets around the world.

What caused this massive drop? The truth is no one really knows. Since financial markets are collections of millions of individuals, institutions, and governments trying to predict each other’s actions and act on available information, the exact reason behind any fluctuation is hard to pinpoint. However, the February 5th drop was not completely without warning.

The first signal was a report from the Bureau of Labor Statistics on wage growth. It showed growth in January hitting a post-recession high of 2.9%. Wage growth is a leading indicator of inflation, and inflation was something that had been largely absent in the nine years following the financial crisis. For half a decade, the United States had been experiencing a “Goldilocks” economy — a period of economic stability that features little inflation, low risk of recession, and close to full employment. In other words, an almost optimal state.

The second indicator was turmoil around the Federal Reserve and interest rates. The Fed had just appointed a new chair, Jerome Powell, a member of its Board of Governors who had largely voted in line with (ex-chair) Janet Yellen. He had, however, previously expressed grievances with the Fed’s latest measures of quantitative easing. Combined with the potential for impending inflation, concerns that interest rate increases were on the horizon were high. Interest rate hikes, while keeping inflation in check, make it more expensive for corporations to finance debt. This makes it harder for companies to expand and leads to less corporate investment.

Wage growth, potential inflation, and a fear of increased interest rates combined to create a perfect storm that began a mass selloff in the equities market. As The Atlantic’s Derek Thompson writes, “since the stock market is fundamentally a collective bet on the future profitability of publicly traded companies, these factors would altogether predict that investors would be willing to pay less to own a share in a typical company than they were a week ago.”

The Setup

The selloff began during the latter half of the trading day, really catching steam towards the end of market hours. Between 3 and 4 PM, the Dow slipped from an 800 point decline into a 1600 point plummet. The markets, in a span of hours, went from unnaturally tranquil to utter chaos.

Most affected by this chaos were volatility indicators. The DOW, S&P 500 and NASDAQ tumble(s) sent the CBOE’s VIX volatility index skyrocketing. Referred to as the markets “fear index”, the VIX volatility index tracks the implied volatility on thirty-day S&P 500 options. Put simply, the VIX is a measure of expected stock market volatility for the next 30 day period. When the markets move with unnatural volume or at an unexpected pace, the VIX tends to soar — and soar it did, rising 116% to its highest level since August 2015.

A Brief History Of The VIX

Though discussions surrounding a market volatility indicator have existed for decades, the first official paper describing such an idea is attributed to Prof. Menachem Brenner (NYU) and Prof. Dan Galai (Hebrew University of Jerusalem). Published in 1989, they called their system “Sigma”. “Sigma” was a tool that attempted to predict market volatility. It worked by calculating and aggregating a series of synthetic option prices showing thirty-day implied volatility, whose time to expiration stayed constant. This acted as a gauge of market expectations of volatility.

Volatility is a measure of an assets price movement, as opposed to the absolute price of the asset itself. Volatility trading comes down to placing (educated) bets on how much and how frequently an assets price will move. It is less concerned with the direction of the change, versus a prediction of how far it may shift and how often.

The easiest way to trade volatility is using options. From the original Black-Scholes option pricing model (and most subsequent variations/improvements) we see that the price of an option is in part determined by the implied volatility of the underlying asset price. Traders are able to take positions on implied volatility by hedging all other price factors, exposing themselves to volatility shifts alone. The relationship between an asset’s price and volatility is called its Vega.

When markets are weak, investors tend to hedge their trades in the options markets. As the need to hedge increases, the premium paid for options contracts increases as well. Since markets tend to move faster in a downturn than during growth phases, an indicator tracking the volatility of options prices on a market basket could function as an indicator of investor confidence. This was the idea behind “Sigma”.

In the mid-1980’s, Brenner and Galai shopped “Sigma” to a few stock exchanges and the Chicago Board Options Exchange (CBOE) but could not make a sale. They published their research as well as descriptions of a market built around the index in 1989. In 1993, the CBOE eventually rolled out their own volatility index (motivated by the Black Monday crash of 1987). This was the birth of the VIX.

Created by Dr. Robert Whaley (Vanderbilt University), who cited Brenner and Galai’s research, the VIX was a marketable implementation of “Sigma”. It was designed to measure the market expectation of thirty-day volatility implied by at-the-money S&P 100 index (OEX) option prices. Later in 2003, the VIX was updated to use S&P 500 index (SPX) option prices. It soon rose to prominence as the benchmark indicator of stock market volatility and investor sentiment. This gave rise to the VIX’s colloquial name — the “fear index”.

Financial Instruments Derived From The VIX

For the first decade of its existence, Wall Street treated the VIX as little more than an efficient market gauge. It was not directly tradeable and there was no official market around it because pricing any type of derivative with the original formulation was impossible.

This changed in 2002 with a phone call. Legend holds that billionaire Mark Cuban called Goldman Sachs looking for a way to hedge his market risk. He wanted to use the VIX as a form of insurance since it tended to rise when markets fell, but Goldman Sachs had no such product.

Two traders, Devesh Shah and Sandy Rattray, set out to satisfy his request. In 2002, the markets were in a post-deregulation frenzy, and almost every priceable asset was being packaged and securitized. They saw an opportunity to make volatility tradeable on a large scale by reworking and standardizing the internal formula behind the VIX.

Shah and Rattray reformulated the VIX such that each dollar of the indicator corresponded to an expected S&P 500 daily move of less than 0.1% per day, for the next 30 days. For example, a VIX value of $10 indicated market expectation that the S&P 500 would move (on average) less than 1% in either direction, per day, for thirty days. Shah and Rattray turned their work over to the CBOE, who lost no time in trying to create this market for volatility.

As previously mentioned, the VIX is not directly tradeable, mainly due to the way its price is calculated. Most indexes use a static basket of underlying assets and periodically rebalance their allocations. The VIX, on the other hand, is priced using a dynamic basket of at-the-money, 23 to 37-day S&P 500 options. The basket is impossible to recreate for a variety of reasons, but mainly because the strike price of options chosen is a function of time, the SPX spot price, and its volatility. Thus the VIX has a theoretical underlying basket as opposed to a real one, making it impossible to replicate, “crack-open” or otherwise perform activities needed for a normal Exchange Traded Product (ETP).

CBOE’s solution to the VIX’s “tradability” problem was to create a derivative market centered around the index. VIX futures were launched in 2004, followed by VIX options in 2006. The CBOE marketed these derivatives as risk-management tools, making the same pitch to investors that Cuban made to Goldman Sachs.

The Thesis

The fundamental way to make money using the VIX is to engage in either long-vol or short-vol trading. Long-vol traders believe that market volatility is likely to increase (and thus VIX future price), whereas short-vol traders are betting on relative calm, and the markets maintaining their status quo (i.e. sinking VIX futures). The reason a profit exists in this type of trade is that the VIX exhibits what is known as a “futures premium”. Like any other futures contract, VIX futures settle against their underlying asset at expiry. But because the VIX is untradeable, the settlement can only be done at the spot price of the VIX, and no cash-and-carry arbitrage is available.

Cash-and-carry arbitrage is what causes the futures prices of items such as commodities to closely track their underlying asset prices. For most futures, one can calculate the dollar cost involved with owning the asset for a specified amount of time, and can then compare this to the price of a futures contract for that same amount of time. If either one is out of sync with the other there is room to profit via arbitrage. With the VIX, the purchasing and holding of the underlying assets are impossible, thus the futures price remains considerably independent of the VIX spot price. This leads to a futures risk premium that investors can take advantage of.

To step back a moment, the VIX is priced based on options — which are derivatives (of the SPX, in this case). Thus the VIX can itself be considered a derivative of a derivative. VIX futures are “derived” even further, a third-degree derivative. Understanding the risks involved in this type of asset are difficult for all but the savviest investors. Thus VIX derivative trading remained a niche activity even as late as 2008.

Then along came the financial crisis. The CBOE could not have asked for a better marketing opportunity. The VIX doubled and tripled, seemingly overnight, hitting all-time highs of $80 and more. In a sinking market, VIX and its derivatives seemed to be the only thing going up.

As previously mentioned, however, trading VIX futures and options remained an activity restricted to seasoned professionals. These investors had deep pockets and could afford losses, or deeply understood the functionality of the VIX. This left a large percentage of the market unable or unwilling to capitalize on the VIX gains. Simultaneously, ETPs were becoming more and more popular.

Barclays PLC noticed this problem and in January of 2009 launched The Barclays iPath S&P 500 VIX Short-Term Futures ETN — referred to by its ticker symbol VXX. Amidst a turbulent economic climate, with the S&P 500 hitting a 12-year low, trading of the Barclays ETN exploded. Investors were reeling from losses as markets took the largest beating since the Great Depression. When an opportunity arose to trade volatility — and make a killing doing it, investors of all kinds jumped at the opportunity.

What had previously been the domain of specialized hedge funds and institutional investors was now open to retail traders. A variety of VIX based ETPs cropped up in the years following Barclay’s introduction. Bill Speth, VP of research at CBOE remarked that it was “the great democratization of volatility. Investors who never would have opened a futures account or traded an option could and did trade these exchange-traded products”. An $8 billion industry emerged seemingly overnight, made entirely of products tied to a single index.

Some of these offerings took long positions on the VIX like the Barclay’s VXX, ProShares UVXY, and VelocityShares TVIX. Others took a negative view and synthetically shorted the VIX like VelocityShares (aptly named) XIV, ProShares SVXY or VelocityShares ZIV. What they all shared was volatility (intraday price movements of 15% were a common occurrence) and to a lesser extent, leverage. Some inverse VIX products offered up to three or four hundred percent exposure to the VIX and VIX futures.

Convergence or Explosion?

As market conditions improved and the United States exited one of the worst downturns in recent history, volatility began to drop. Interest flowed out of products that tracked the VIX and into those that “shorted” it (by selling futures), wagering that volatility would stay low. In what would become an incredibly profitable strategy, both money managers and retail investors bet heavily on the short-vol trade. With the recent Goldilocks market conditions — low-interest rates, inflation, and volatility, short-vol trading turned into an almost $2 trillion trading strategy. Employed by investors of all types, it became so prolific that Bloomberg coined VIX based trades the volatility-financial complex.

“[It was] the great democratization of volatility. Investors who never would have opened a futures account or traded an option could and did trade these exchange-traded products”
— Bill Speth (VP Research, CBOE)

In January, ETrade, a major retail stockbroker, saw one of the heaviest months of trading in their 20-year history. That same month, the two largest exchange-traded products betting on low volatility took in an unprecedented $1.7 billion in invested capital. There was a good reason for the heavy interest. Leading up to the February volatility spike, the XIV (inverse-VIX ETN) boasted an incredible 10-year total return of 1,518%, and the SVXY (inverse-VIX ETF) claimed a return of 70% on the year. It was estimated that for every dollar of VIX futures movement, $512 million of investor capital was at stake.

Devesh Shah, one of the aforementioned traders at Goldman Sachs (and re-formulator of the VIX) was shocked. “Everybody knew this was a huge problem,” said Shah “everybody knows that inverse VIX is going to go to zero at some point, and all these inverse and leveraged products, not just in the VIX but elsewhere too, at the end of the day cost people a lot of money.”

The day came sooner than later. On February 5th, with the Dow plummeting and S&P 500 bleeding value, the market calm turned into panic. With this panic came waves of volatility. The VIX spiked 116%, rising from $17.31 and closing at $37.32, even breaking the $50 level at certain points during the day. Not quite as high as during the 2008 financial crisis, but with investors heavily betting on calm after close to a decade of market tranquility, this spike in volatility was catastrophic.

Prior to 4 PM, losses were taken but most traders would live to fight another day. At 4 PM, however, the rebalancing for ETPs takes place. Once the dust settled, the XIV, SVXY and all other short-volatility and inverse-VIX products were decimated. In particular, the XIV and SVXY (with a combined $3.4 billion in assets) fell by 92.6% and 83% respectively. The XIV from $99.00 to $7.32, SVXY from $71.82 to $12.20.

Consequences

Prior to the rebalancing, the top five inverse-VIX products (XIV, SVXY, ZIV, VMIN, XXV) had an aggregate asset total of over $3.7 billion. After the rebalancing, this amount dropped to $525 million. The fallout mounted quickly.

On the following day, over a dozen inverse-VIX products were halted to stabilize their value. Credit Suisse AG (owner of VelocityShares), the largest issuers (and owners) of the XIV moved to liquidate the product. Nicole Sharp, a spokesperson for Credit Suisse commented that “The XIV ETN activity is reflective of today’s market volatility. There is no material impact to Credit Suisse.”

Major institutional investors were some of the largest holders of the XIV. Apart from Credit Suisse, they include Deutsche Bank, Morgan Stanley, Citadel, Two Sigma and Jane Street. Jane Street was also one of the largest holders of SVXY, holding 16% of the fund’s value. While it’s easy to say the impact on these banks and funds will be negative, it is difficult to say how much.

“It’s kind of sad that these products exist in the first place, but it’s hard to stop it. If you stop this, something else will come up. Bitcoin will come up.”
— Devesh Shah (Ret. Partner, Goldman Sachs)

Investors in the XIV will have their notes redeemed at least a 90% discount to the previous day’s price. However many of these institutional investors were also hedged in the UVXY, a long-VIX ETF that soared 66% on the day and continued to rise after hours. Retail investors on the other hand, quickly found forums to voice their complaints. The online stock trading forum, Stocktwits, saw an increase in negative sentiment surrounding the XIV and SVXY by over 500%. Individuals also took to spaces such as Reddit to share the stories of their losses.

Unfortunately, while the tale of the short-vol trade illustrates some of Wall Street’s worst behavior — creating products that are extremely risky, difficult to understand, and hard to value, it is unlikely to have a significant impact on the status quo. A few investment products such as the XIV were liquidated, and a few funds such as the LPM Preservation & Growth fund will be redeemed early (LPM losing 82% of the fund’s value in a single day). But if there is profit to be made from such risky trades it is unlikely to stop. As Devesh Shah eloquently states, “It’s kind of sad that these products exist in the first place, but it’s hard to stop it. If you stop this, something else will come up. Bitcoin will come up.”

Sources

This paper is an amalgamation of incredible sources on financial news. I had an insane amount of fun researching the topic and all the information here is a synthesis from an aggregate of 20–30 sources. I have listed as many as I could keep track of below. For more information on sources used please contact abhishek.nayar[at]yale[dot]edu