SellInMayBlog_Hero

Sell in May & Go Away?

Q4 Activism Alarm: New algorithm empowers corporate and institutional investors to predict activist trading

27 March 2017

By Cory Todd

0 comments

ActivismAlarm_Hero (1)

Share

We’ve got some big news. Today we’ve released new research on a predictive analytics model that helps to forecast and recognize activist trading.

For IR teams, this means having unprecedented access to real-time data that empowers them to be proactive. It means that real-time trade monitoring is now available to scan for potential activist movements in their stock, so they can form a plan of action. Getting ahead of activist trading has just gotten easier — and taking important steps to securing defensive measures at the management and board level can happen sooner.

For institutions, it means predicting potential activist accumulations before they are made public. It means making more profitable decisions based on complete knowledge of activism. Institutions can now hedge existing positions, enhance current strategies, or build novel strategies of their own with a complete picture of the activist landscape.

 

So how does it work?

Top activist trading is a repeatable event. As a result, it has the ability to be predicted. Q4’s Activist Alarm (AA) is the first of its kind to combine long-term outlook characteristics with short-term indicators such as daily options trading and stock markets. This combination of datasets provides a precise and complete view of activist trading.

While all activist campaigns are noteworthy, we view campaigns initiated by the largest 50 activists as the most significant events to predict. These activists account for nearly half of all dissident campaigns and, again, because their buying happens in a pattern, it can be predicted. For the purposes of our research, we focused our attention on determining and creating predictive indicators based on this subset of activists.

 

InsetImage1 (7)

Top 50 activist campaigns by year

 

Q4’s scoring system uses sophisticated models and historical analysis to create a baseline, or assumptive “normalcy”, for each stock. Through machine learning, algorithms closely monitor equity and options order flow, including liquidity analysis, volume velocity, and trade pattern recognition, to identify when current trading activity veers from preset norms.

When these norms are broken in such a way as to match the historical “footprints” left by activist investors, the alarm score begins to spike. These spikes are an indication that an activist investor could be building a significant position in the security. Q4’s AA is a differentiated model for detecting real-time activist movements, allowing one to stay ahead of tomorrow’s trading activity.

The AA is measured on a zero to 10 scale, with 10 being most “at risk” for potential activist buying. Through using the same datasets that comprise the AA, we are further able to predict the probability that a top-50 activist will increase their position in the next 90 days. Again through binning the results one to ten, we can predict close to a 42 percent likelihood of an activist increasing their position over the next 90 days.

 

InsetImage3 (2)

Probability top 50 activist increases their position in the next 90 days (%)

 

Q4’s AA is rooted in over two decades’ worth of experience from a team of veteran stock surveillance analysts, experts in real-time monitoring, and the tracking of activist shareholders such as Pershing Square, Carl Icahn, and JANA Partners. Q4’s unique expertise in these highly specialized niche markets has given rise to a new area of focus in predictive analytics: activist trading recognition.


Find out more about our research on activist trading:
download our latest whitepaper to learn more about why activists target a stock and how predictive models have come a long way in predicting activist trading.

 

Cory Todd is the vice-president, quantitative analytics at Q4.

0 comments

Leave a Reply

Interesting Articles

Q4 DESKTOP

Next Gen IR Platform

Learn more