Blogs

Posted: October 31, 2018

The VIX is Fixed?! A Complaint is Remixed

We closed out our first VIX-is-Fixed post with a promise to deliver real-time updates on the state of the VIX complaint. While we’ve failed a bit in that regard, we’ve added a timely discovery update to this post to make up for it – keep scrolling to find out just what tricks the Plaintiffs have up their sleeves.

For an overview of the main players in the VIX action and an overview of the index itself, I’d suggest you review our Preview of the Tricks post. If you’re up to speed already, then buckle up for a deep dive into the zeros, banging, and a small ocean’s worth of data. The Consolidated Amended Complaint alleges negligence and violations of the Securities Exchange Act, Commodity Exchange Act, and Sherman Act, but gives over the majority of its eighty-odd pages to detailed explanations of the factual review and research carried out by Plaintiff’s attorneys that breaks down the vulnerabilities present in the VIX.

Bang Bang Bid

Plaintiffs devote some time in outset of their complaint to break down the two processes that Defendants used to manipulate the settlement process. These include “banging the close” and abuse of the “two-zero bid rule,” both of which I’ll take a minute to discuss here.

While “banging the close,” would, in other markets, refer to a form of manipulation in which a trader bought or sold large numbers of futures contracts during the closing period (clever naming!), in order to benefit futures positions purchased earlier in the day. Given what we already know about the way the VIX settlement process works (recap: highly dependent on thinly traded and illiquid financial instruments – namely, out-of-the-money SPX Options – and a short window trading window closing at 8:20 AM), the term isn’t quite on point for this particular form of manipulation, if only for the timing. “Banging the once-monthly, early-morning opening auction” didn’t have quite the same ring to it, I suppose. The complaint takes care to point out that a longer settlement window that occurred during normal market hours, or a more frequent measurement, would have made the manipulation more difficult—and that the CBOE declined to take such preventative steps. As a result, Plaintiffs claim, by placing higher mid premiums on puts at particular strike prices, the John Doe Defendants were able to “bang up” the level of the bid premium of that strike, and thus increase the settlement value for corresponding VIX Options and Futures. The process worked in reverse, too—Defendants could “bang down” the mid premium by placing lower ask premiums on puts at particular strike prices, and so decrease settlement values for the corresponding bids.

The second method of manipulation has to do with the process of the SOQ calculation. The calculation starts in the center of the pricing circle, and works its way outwards through increasingly out-of-the-money strike prices. The “two-zero bid rule” is so named because the calculation is supposed to stop at the point in which two zeros are found in a row—an indication, first, that the SPX Option is so far out of the money that pricing is no longer reliable for the settlement calculation, and, second, that traders are not particularly creative when it comes to assigning monikers. To circumvent the two-zero bid rule, Plaintiffs posit that the John Doe Defendants were spreading bids out across strike prices to ensure that there were never more than two consecutive zero bids ahead of the strike prices that Defendants wanted the SOQ process to take into account. By preventing two zero bids in a row from appearing naturally, Defendants forced the calculation to consider out-of-the-money strike prices appearing much deeper in the range than they otherwise would, thus skewing the settlement values for the expiring VIX Options and VIX Futures.

The formula used by the CBOE to calculate the VIX weighs the difference between strike prices on either side of a given strike price—a number that will nearly always be larger for more out of the money options—and that weighing factor will increase when the given strike price is smaller. That means that out-of-the-money put options (as opposed to call options) will have a greater impact on the ultimate VIX settlement price. This is the case because the strike prices for out-of-the-money put options will always be less than the prevailing at-the-money strike; out-of-the-money call options, in contrast, are always greater than the prevailing at-the-money strike price. While the complaint offers a much more in-depth look at the VIX calculation, let it suffice to say that these two factors lead to the result that put options that are the most out of the money have a disproportionate impact on the ultimate VIX settlement price. By circumventing the two-zero bid rule, defendants push the formula to consider bids deeper and deeper out of the money, thus amplifying the impact of their manipulations.

In God We Trust; All Others Must Provide Data

In addition to the academic analysis performed by Professor Griffin and Mr. Shams—whose results we discussed in our first installment—Plaintiffs also undertook their own extensive economic analysis to further demonstrate their the VIX process had been routinely exploited. By tracking trends in trading volume over time—and putting together into neat and colorful graphs for the more visual learners out there—Plaintiffs very effectively make their case for intentional manipulation.

First, Plaintiffs note that the data shows a “uniquely disproportionate” number of puts placed, rather than calls, during the settlement period. When considered across the whole of expiry and non-expiry days, the ratio of puts to calls placed was similar (ranging from 1.74 to 1.82). During the settlement window, however, that ratio ballooned to 5.93—a 224% difference. It is more than a coincidence, Plaintiffs insist, “that market actors just so happened to be preferring the type of order that would maximize a manipulative effect during exactly the time when such manipulative effect was possible.”

Second, Plaintiffs track increases in trading volumes on settlement days, particularly for the SPX Options that would have a more significant impact on the SOQ process. Notably, prior to February 13, 2018, 92% of settlement days saw higher trading volume than the Tuesday that preceded them. This pattern of higher trading volume held true for SPX put options that were out of the money—an unusual thing to see, because, in a manipulation-free market, one would expect to see lower trading volumes simply because, as the option is less and less likely to be exercised, there will be less and reason to pay for that option. What Plaintiff’s data shows, however, is that the more out of the money an SPX put option was, the more it was being traded. Weird. What’s more, this abnormality was particularly present for options that had a wider gap between strike prices—as discussed above, these are the trades that would have the biggest impact on the VIX SOQ settlement process. And, again, this is a pattern true only for trades within the settlement window. Weirder? Finally, the data also shows that, at exactly 30 days to maturity, out-of-the-money options (and only out of the money options) saw a spike in trading volume. As a reminder: SPX Options were only included in the VIX SOQ calculation if they were 30 days to maturity, and out of the money. Therefore, the increased trading volume on settlement days is being driven by trading only in those instruments that could have an impact on the SOQ process. Weirdest!

Third, Plaintiffs argue that the data shows “routine” exploitation of the two-zero bid rule. Such exploitation must be occurring, they posit, because the total number of actively quoted SPX Options (that is, options with a non-zero ask quote) did not change very much between 8:30 AM and 8:40 AM on settlement days. What did change, and to a statistically significant degree, was the number of those SPX Options in that time that were VIX-eligible: two zero bids in a row, the circumstance or “gap” that would render subsequent bids ineligible, occurred far less frequently during settlement windows than it did during other time periods.

Plaintiffs next compared the VIX benchmark to the VIX itself. While careful to stress that the question is not whether the VIX moved, but whether it move differently when a settlement was involved, Plaintiffs make several arguments that the data does show the VIX acting differently around the settlement window. For example, they point to the fact that there was a much larger gap between the start and end of the day for settlement Wednesdays, and that, on settlement Wednesdays, there is a much larger gap between the VIX at the start of the day as compared to fifteen minutes later. All of the differences identified, Plaintiffs note, were statistically significant.

Finally, Plaintiffs point to one final, telling oddity in the data. All of these patterns suggesting market manipulation were consistent across the time frame studied, up until February 2018. After this date, many of the volume anomalies surrounding settlement Wednesdays abated. Why the change, one might ask? Well, Plaintiffs have a pretty good answer to that question: on February 13, 2018, it was for the first time publicly reported that FINRA was investigating the manipulation of VIX pricing. What better motivation to desist manipulations than the menace of unmasking?

They don’t know that we know that they know . . .

The next section of Plaintiff’s complaint aimed to show that the CBOE knew or was reckless in its disregard of the fact that the VIX settlement process was being manipulated. Unlike all other participants, the CBOE had a front-row seat to all of the settlements, and access to all of the data needed to determine the identity of the manipulators responsible for the rigging.

Rather than disclosing the manipulations that they could easily have observed, the CBOE instead, according to Plaintiffs, made misleading statements about the integrity of the VIX Options and VIX Futures. It further failed to take advantage of the many viable alternatives to its flawed settlement processes—such as calculating prices by using the average of prices across a three-hour window during normal market trading time, as other volatility-related products typically do.

Plaintiffs further point out that it is explicit in the CBOE’s own rules and those that they were subject to obligated them to police for and prevent manipulation. As a “board of trade,” for example, the CEA requires that the CBOE “have the capacity and responsibility to prevent manipulation, price distortion, and disruptions of the delivery or cash-settlement process through market surveillance, compliance, and enforcement practices and procedures.” Similarly, the CBOE rules hold that traders may not “engage or attempt to engage in any fraudulent act or engage in any scheme to defraud, deceive, or trick, in connection with or related to any trade or other activity related to the Exchange.”

Plaintiffs also point to the CBOE’s glaring motivation to maintain the VIX as a “premier” product. The CBOE has an exclusive licensing agreement with Standard & Poors, which permits only CBOE to list SPX Options. This gives the CBOE a lock on the SPX and VIX markets, and those proprietary products generate far higher revenue for the CBOE than any of their multiply-listed options. Plaintiffs go further, describing the three products—SPX Options, VIX Options, and VIX Futures—as “cash cows” for the CBOE, consistently representing about half of that entity’s total revenues. At risk of mixing metaphors, it would be detrimental to the CBOE, or so Plaintiffs argue, for the company to bite the cow that feeds it.

The complaint further reviews the CBOE’s actions with respect to the VIX, and to products with similar settlement process. These show that the CBOE is, contrary to appearances, able to identify manipulative acts, and to address them—albeit only publicly years after the manipulation originally occurred.

Yup, that hurt.

What is a complaint without some damages? Plaintiffs make the general assertion that, based on assurances by the CBOE as to the accuracy and fairness of the settlement process, investors—plaintiffs and members of the class—were harmed because they “poured billions of dollars” into transactions in products that were “not the result of regular forces of supply and demand.” Instead, Plaintiffs and class members were “tricked” into trading SPX Options, VIX Options, and VIX Futures at prices that were made inaccurate as a result of misconduct on the part of the CBOE and the Doe Defendants. That manipulation mean that Plaintiffs and class members were forced to pay more, or accept less, for those products than they would otherwise have done, had the market been a truly free on.

The complaint goes on to explain in detail the different ways that such harm occurred, depending on what options and futures the various plaintiffs and class members held, and what they did with them. They argue reliance on the fairness of the VIX SOQ process: such transactions would not have occurred had they known of the manipulation. In the alternative, Plaintiffs also put argue for a presumption of reliance under Affiliated Ute, because their claims are partially predicated upon material omissions of fact by the Defendants, and, in the alternative, a presumption of reliance pursuant to the fraud-on-the-market doctrine.

No Time Like the Present

The final section of Plaintiff’s brief argues that Defendants, being the overachievers that they were, took their inherently self-concealing manipulation, and worked to affirmatively conceal it. Any applicable statute of limitations has been tolled, Plaintiffs argue, because Plaintiffs and class members did not—and could not have, due to Defendant’s hidden misconduct—discover that Defendants were manipulating the VIX or VIX-linked instruments.

The very nature of the SOQ process, with its anonymized trading, made it impossible for Plaintiffs to discover the facts comprising their claim until very recently. Moreover, Plaintiffs claim that Defendants not only knew of the practices detailed in their brief, but “knowingly, actively, and affirmatively concealed th[ose] facts,” and “actively misled Plaintiffs as to the true nature of VIX Options and VIX Futures, as well as the SOQ Process,” through public statements—such as those denouncing the whistleblower letter sent to the SEC and CFTC.

In making these arguments, Plaintiffs invoke the discovery rule, the doctrine of equitable tolling, and fraudulent concealment, and further claim that Defendants are estopped from relying on any statute of limitations defense in this action.

A Formal Request to Spill the Tea

Plaintiffs are in the unusual position, of course, of having no insight into exactly which trades were affected by this manipulation, because they cannot identify the traders responsible for this manipulation. To address this issue, Plaintiffs have filed a request for advanced discovery of the non-anonymized trading date currently in the possession of the CBOE. While Judge Shah previously expressed a “preference to ‘test’ plaintiffs’ claims before allowing discovery,” as Plaintiffs acknowledged in their discovery motion, they argue that their comprehensive complaint clearly demonstrates that no such testing is needed:

The specificity and robustness of plaintiffs’ complaint allegations show that the court need not wait for completion of the motion to dismiss process to know this is not a blind fishing expedition. […] Granting this motion is also appropriate because plaintiffs’ proposed discovery requests are narrowly targeted to their need to identify the Doe defendants.

This request is being made now largely out of fear that the statute of limitations will run before the motion to dismiss process is complete and formal discovery begins (likely not before summer 2019, assuming the claims survive the motion to dismiss). The Doe Defendants, once identified and named, will surely contend that the two-year statute of limitations clock began to run when the Griffin & Shams paper was published in May of 2017.

While Judge Shah considers this request, the Defendants—the identified ones, at least—will be hard at work on their motion to dismiss this complaint. Considering the level of detail Plaintiffs’ analysis provided, it may be an uphill battle, but the SSD Manipulation Monitors look forward to their efforts, and to summarizing the same for you. For that update—barring any adjustments to the currently-posted briefing schedule—watch this space sometime after the 19th of November.

This post was written by Alexandra M.C. Douglas.

We welcome your feedback. If you have questions or comments about this post, please e-mail John M. Lundin, the Manipulation Monitor’s editor, at jlundin@schlamstone.com or Alexandra M.C. Douglas at adouglas@schlamstone.com or call John or Alexandra at (212) 344-5400.

Click here to subscribe to this or another of Schlam Stone & Dolan’s blogs.

View posts