A new [academic] year, a new manipulation to monitor. Today’s subject is none other than the fixing of the VIX.
It’s time for a trip back in time. Not very far back, though – just to the 23rd of May, 2017, the date on which John Griffin and Amin Shams published Manipulation in the VIX?, a 37-page (or 57-with-references), extensively researched, paper on, manipulation of the Chicago Board Options Exchange’s (CBOE) Volatility Index (VIX). VIX is a widely tracked index that gauges the 30-day implied volatility of the market, and, while Griffin and Shams couldn’t muster up a particularly exciting title for their own paper, the VIX itself is often referred to as the market’s “fear index” or “fear-gauge.”
As the litigation-minded among you might expect, this publication set off more than a few alarm bells – and had in house counsel calling their lawyers. Between March and July of 2018, more than twenty-eight VIX manipulation actions were filed; they now have been MDL’d to Judge Manish S. Shah of the Northern District of Illinois as In re: Chicago Board Options Exchange Volatility Index Manipulation Antitrust Litigation., No. 1:18-cv-04171, MDL No. 2842 (N.D. Ill.). The plaintiffs in these cases vary, but are largely investors and other trading firms, many of which are Chicago-based. Named defendants include the various CBOE entities, but several cases are also captioned as “[Plaintiff] v. Does” — without insight into the CBOE data bases, Plaintiffs aren’t yet able to identify the parties responsible for the alleged manipulation.
The consolidated master amended complaint (“CMAC”) is now due on September 19, 2018, while a joint status report proposing a briefing schedule for motions to dismiss is to be filed by October 5, 2018. The Manipulation Monitor will be following this one closely, so watch this space for details. But while we wait for the official MDL complaint, the time seems ripe—given that we have an academic paper and nearly thirty complaints already in existence—to set up our blogging on this line of cases with some background on the VIX itself, and exactly how the Doe defendants achieved “mischief managed” status.
The Building Bricks of the VIX
So, let’s start with the major players. The CBOE is the largest options exchange in the U.S., and offers options contracts on hundreds of stocks and exchange-traded funds, as well several popular stock indexes. Trading on the CBOE happens through a clearing mechanism provided by the Options Clearing Corporation (“OCC”), which brings together buyers and sellers and handles all subsequent transfers of securities or cash between option buyers and sellers at expiration. The CBOE Volatility Index—or “the VIX,” as we’ve been referring to it, whilst wondering if Drake named this one as well—is based on real-time prices of near-the-money options on the S&P 500 Index (“SPX”), and is designed to reflect the investors’ consensus view of the future (at least, the 30-day future) expected stock market volatility. The “fear index” received its nickname from its tendency to spike significantly when investors believe the market is behaving bearishly.
Allegations in the various complaints involve futures and options contracts linked to the VIX Index, which are traded on the CBOE Futures Exchange (“CFE”). While the first exchange-traded VIX futures contract was only introduced in 2004—and VIX options contracts in 2006—combined trading activity has grown to over 800,000 contracts per day, and 2017 saw record-setting trading volumes for both products. It is the unusual settlement procedures for these VIX contracts that, according to Griffin and Shams, render the market susceptible to manipulation.
When a futures or options contract expires, the settlement price—the reference price against which the contract is measured—is a critical variable. If the settlement price can be manipulated, then so too can the value of the derivatives at settlement. Usually, a settlement price would be based on the spot price of the underlying assets. In contrast, the settlement price for a VIX contract is determined by a monthly auction that yields a price quoted using the ticker symbol VRO. This auction is conducted using a Hybrid Opening System, or “HOSS,” in which traders submit bids and offers for SPX options listed by the CBOE; this yields an opening price for such options from which VRO is calculated. VRO in turn determines the final settlement value of VIX derivatives. This auction starts at 7:30am Chicago time, and ends an hour later, when trading on non-expiring VIX futures and options begin.
Verily Verily, VIX is but a Dream
According to Griffin & Shams, there are three features about the VIX that “might leave it open to manipulation: multiple connected markets with different price-order elasticities, cash settlement, and a finite window to manipulate.” The first factor has to do with the fact, explained above, that there are actually two different markets at play here. The “lower level” SPX options market—to which the “upper level” VIX derivatives market is tied—is fairly illiquid, with large bid-ask spreads, while the VIX derivates are large and liquid. This means that a comparatively small number of trades in the SPX options market can move the price of those options significantly; these price movements will have a corresponding, and disproportionately large, effect on the VIX. The second factor, cash settlement, ties into these price movements. Unlike an asset with a physical settlement—where an anomalous price at settlement would just result in an investor taking possession of a physical asset at an inflated price—if the VIX is artificially inflated at settlement, the would-be manipulator can cash out their position at the deviated price. The third factor is the short period of time concerned. The one-hour period of the auction means that a would-be manipulator need only intervene for a very discrete period of time. Manipulation for the low-energy types, if you will.
Manipulation in the VIX goes on to explain exactly how the theoretical VIX tricksters have carried out their scheme. Keeping in the theme of threes, the triad of steps required for such manipulation include: (1) opening long positions in the VIX derivatives prior to settlement; (2) submitting aggressive buy orders in the SPX options during the settlement auction, causing the auction-clearing of prices of SPX options to rise, and, with them, the VIX settlement prices; and, finally, (3) obtaining the higher price for the upper-level VIX futures or options when they settle.
Griffin and Shams point out that such trading will leave reviewable patterns in the data, and their paper, quite naturally, goes on to review just that. For starters, they highlight spikes in average daily trade values that are not related to any S&P 500 market-related event, but solely to the date of VIX settlement. These volume increases occur primarily on options with VIX sensitivity, and researchers found that “the relationship is highly statistically significant.” Because the pattern of trading spikes is not observable in other, nearly identical, OEX and SPY options, or even in ITM options excluded from VIX settlement calculations, they conclude that this “evidence consistent with attempted manipulative activity.”
Being the thorough academics that they are, Griffin and Shams also considered whether any alternative phenomena could be driving SPX option trading volume spikes at settlement. For example, they investigate whether the VIX settlement option provides an opportunity for those with “pent up demand”—stemming from the normally illiquid SPX market—to trade deep OTM options. This theory was scuppered by their observation that comparable OTM options actually traded at lower volumes at the time of settlement. The pair further considered two possible hedging explanations, but were able to show both that traders were not attempting to hedge positions in VIX derivatives through the underlying SPX options, nor were rolling their hedging positions into SPX options in a way that mimicked the VIX weighting formula.
Observing all of these patterns, Griffin and Shams concluded that the most likely explanation for the market behavior that they could see was, in short, manipulation.
What they could not do, however, was identify the manipulators. Electronic trading on the CFE and CBOE is conducted anonymously, and only the CBOE—and the manipulators themselves—know the identities of the parties entering the manipulative orders that the researchers have observed. Luckily for Plaintiffs, the CBOE and CFE are central repositories of a range of information related to the trades. The only now question is how many associates will it take to parse the results of the (eventual) discovery demands . . . .
Details on the specific legal allegations and potential damages to be discussed in our next installment, so watch this space for real-time* updates.
*Okay, maybe not quite real time. I’m not NBC’s Sprinting Intern . . .
This post was written by Alexandra M.C. Douglas.