Remember me

Register  |   Lost password?

The Trading Mesh

Using Corporate Event Data to Navigate Low-Latency in Equity Options: Strategies for Institutional Traders and Market Makers

Tue, 09 Oct 2018 15:38:58 GMT           

Introduction

 

Options remain the closest community of market participants, and the murkiest. Even to its veterans, the options market can seem uncanny; its trading has always been full of mystery. Rumors abound of big-wins and equally huge losses turning on a dime, and there is common sleight of hand. It evolves at its own pace. But wild swings and tectonic shifts can also be born of simple mistakes. Such was the case in 2015, when a pair of microprocessor manufacturers was first reported to be in merger talks[1]. In only a few seconds, the news quietly generated a sizable option contract on one of the companies’ stock prices. Initially costing around $110,000 for more than 300,000 shares, the option went from completely out of the money to worth $2.4 million in less than half an hour.

 

The reporting subsequently proved wrong, adding intrigue and furrowing eyebrows. But in the end it wasn’t the windfall or the scant information that turned out to be significant; rather it was the speed. The option, most observers agreed, could only have been generated by an algorithm—an incredibly fast one, at that. While neither the first nor certainly the last, this was some of the most dramatic public evidence yet that low-latency and high-frequency trading (HFT) techniques had arrived—and for that matter, could work—in a space where they had only been casually embraced before.

 

Below we explore new trends in the development of event-based trading signals in the equity options market. Ultimately, we conclude that investors must construct a dual strategy—combining measurable post-trade execution quality analysis with pre-trade contextualized indicators of price movements—to effectively mitigate downside risk and exploit market inefficiencies with options.

 

The Challenge: Reading Corporate Tea Leaves

The growing cohort of investors represents a kind of ‘Goldilocks’ group of options participants. They are compelled to be able to execute in a few seconds, but without the need or wherewithal to move in microseconds. In short, timing matters. Equity options provide an effective way to play on ideas about a listed company’s price—be they changes in market sentiment one way or other, or mitigating short-term volatility. There are myriad ways to construct that option position, from simple binaries up to multi-strike butterflies and other wingspread strategies. Either way, doing so inevitably starts with the target strike (or strikes) and tenor for the option.[2] And they often focus in upon a corporate[3] event.

 

This exercise requires context—placing a premium on availability and the proper application of data.[4] Key characteristics of an option are shaped by isolating phenomena and information that drive changes in the underlying stock’s price. Indeed, some of the data points surrounding an event can be highly diverse, and at times hard to find. Their impact and relevance will be open to debate and, like the microprocessor merger mentioned above, crucial signals about the direction can sometimes come out of nowhere, or prove inaccurate.

 

Still, many scheduled corporate events can be planned for and actively monitored in advance.[5] Empirical research has shown this corporate “body language”—changes and modifications to the calendar and even the structure of the news, itself—can even provide accurate, predictive signals as to the future state of the company’s health.[6] These data points, whether public or primary sourced, can be absolutely crucial if timely and properly structured, verified for accuracy and populated into the strategy. They are essential tea leaves about corporate behavior and future performance of an equity; yet until recently, many investors have been slow to read them.[7]

 

Applications

There are at least two applications where this data should be applied -  corporate earnings calls and disclosure of dividends distribution have significant consequences for pricing, and their timing is fairly predictable. Studies using corporate event data have proven that the nature of the news—positive or negative relative to expectations—can be accurately surmised from changes in calendaring of these events.[8]

 

Deploying faster electronic execution enables firms more time to precisely game out these events, take heed of ongoing price movements[9] and optimize construction of the option without crossing the spread. Of course, it also allows HFT, low-latency firms and market makers to do the same, swooping in when orders increase around an event to pick off slower flow. Some of the challenges that result from this asymmetry are very complex. For example, correctly structuring sensitivities in multi-strike spread strategies, and aligning the cost of the option against the investor’s level of tolerable risk, are all the more difficult to do in a market where faster players might be lurking.

 

With the right data, other defensive challenges are more easily solved for. This can be as straightforward as missing a change in the timing of the earnings call; likewise, there may be mistaken reporting or even a misdated announcement from the company, itself. Changing dividend amounts and calculated payouts—or the suspension or resumption of a dividend—frequently present another, similar issue of timing. A revised dividend will likely modify an equity’s Sharpe ratio (risk-free versus risk-adjusted returns). Depending on the portfolio’s risk management tolerances, that may artificially skew—and typically, limit—the types of options that can be used.

 

Rearguard Action: Data Feed Requirements

Neither of these problems—misplacement of the contract’s tenor, or an incorrect strike—is unique to a HFT environment. However, they are far more likely to be punished. Likewise, firms that can prove out their defensive techniques are more likely to be able to use that same logic, venue knowledge, and technology in reverse. Even if they cannot move quite as quickly as the top players, concerted investment in this area might just provide two new tools for the options arsenal.

 

How to do it? Faster execution capabilities are unavoidably important; yet a more complete rearguard action requires building competencies around event data—sourcing it, qualifying its accuracy, digesting it, and putting it into action in a low-latency setting.

 

The other component—predictive analytics that compare patterns of corporate behavior leading up to events against a stock’s “point in time” pricing—is more qualitative in nature  and trickier to develop internally. An earnings date that seems askew or payout that appears completely unachievable cannot be evaluated for its veracity without context—which requires filtering tools and curation. Vetting these signals requires not only machine-readable news and natural-language processing elements to discover them (often but not limited to press releases), but extensive warehousing of data to reference and intuitive sentiment analysis guiding the output.

 

Concluding Thoughts

Options trading was once a technologically sluggish, space. Expectations were low because, besides occasional murmurs of a big play or a big flop, it was as much a testing ground for theory as it was a good way to hedge or short. For better and worse, the arrival of higher speeds has changed all that.

 

Today, amongst this louder din of high-frequency and lower-latency trading, there is a natural and understandable twitch to try and catch up—to join the herd mentality. For the many participants in the middle, though, to whom that seem a losing proposition, we suggest an alternative approach: a focus on the calendar—not the microseconds—and a mastery of reading for mistakes. Mistakes can include inaccurate dates as well as content and both need to be considered.  Then as you increase the speed of the transaction and combine it with accurate data, you achieve an overall better deployment of the data.  With a prior back-testing of your strategy, and an ongoing evaluation of your execution quality, your analysis and subsequent transactions should enable you to mitigate downside risk and exploit market inefficiencies with options.  

 

 

 

 

[1]How one trader made $2.4 million in 28 minutes” reported by Stephen Gandel in Fortune, April 1, 2015.

[2] Investors have increasingly turned to weekly contracts to more closely hone in on events. See “For Traders, a Weekly Gamble” by Kaitlyn Kiernan, Wall Street Journal, January 2014.

[3] Corporate event data can be defined as any single-stock event that can cause volatility in the marketplace such as (but not limited to) the following categories:  analyst/investor events, calendar events, company events, company metrics, investor conference related events, corporate actions, dividends, earnings, government and public offerings.

[4] Wall Street Horizon recently explored the development of research in this area in an earlier whitepaper, “Exploring Corporate Event Data and Volatility: Considerations for Academic and Financial Industry Research”

[5] Johnson and So’s 2016 Paper “Time Will Tell: Information in the Timing of Scheduled Earnings News” documents the managerial aspects and institutional self-perception implicit in earnings calendar revisions, proving their predictive power of future earnings.

[6] For further discussion, see “How Reading ‘Corporate Body Language’ Can Boost Stock Returns by Evie Liu, Barron’s , July 2018 and “Reading the Signals: “What Does a Company’s ‘Body Language’ Tell Investors” by Max Bowie, Inside Market Data,  May 2018.

 

 

 

[7] “The Unspoken Signals in Earnings Releases” by Rachel Emma Silverman, Wall Street Journal, December 2014.

 

[8] See Deltix Quantitative Research Team, “An Automated Trading Strategy Using Earning Data Movements from Wall Street Horizon”, 2015. The study used a time-series analysis of ten years of data, proving genuine alpha can be realized by tracking forward advances (positive) and delays (negative) of earnings announcements.

 

[9] See Shrihari Santosh’s study, “The Speed of Price Discovery: Trade Time vs. Clock Time”, 2016. Santosh finds that asset prices primarily incorporate information, both public and private, through the process of trading.