Blighty, Boris, & Brexit—Investing with Data Science

Updated: Sep 19, 2020

A month before the UK electorate gave Boris Johnson and the Conservative party the strongest mandate in recent history, Toggle’s AI noticed that monetary conditions in the UK were changing. More specifically, the system alerted users that the drop in the Bank of England’s Reserve Assets to such a low level had historically warranted a strengthening of GBP against the US Dollar.

GBPUSD history chart with respect to reserve asset growth in the UK

This highlight was but one data point in the sea of uncertainty surrounding the UK Election and the likelihood of Brexit. Though the system had no context for the pre-election anxiety by investors, businesses, and consumers, it easily identified a dislocation. Data points like these can spur thought and may lead investors to action. For market participants, it often comes back to a question of value...

Has an asset sold off enough? Can resolving political uncertainty unlock value? These questions might be easier to assess if investors can find supporting evidence to enhance their theses—factual evidence instead of opinion is the bedrock of data-science.

The highlight came after a week when GBPUSD had sold off ~1%, and suggested a rally of nearly 6% over the coming 6 months. An investor that had assessed the post-election outlook for the Pound might have concluded that the risk/return was strongly in her favor. Overnight on December 12th, GBPUSD broke through 1.35 on the election results, a 5.5% return on GBP linked to the drop in reserve assets.

Less than 6 weeks out, the asset had moved close to what had been predicted over a 6 month period. A disciplined investor would see the thesis come true and consider exiting the position...a 5.5% realized return vs a 5.9% historical benchmark is close enough where the risk-return probably isn’t worth it. Not just because you’re eking the last bit blood out of the stone, but also because you understand that recent move may mean revert.

Machines are somewhat ignorant of the broad context, but they notice things that we often miss. Investors can contextualize machine-generated discoveries, and AI can act as a tactical guide only if paired with a disciplined discretionary investor who can think through the consequences.

In this case, the investor would have been right to take money off the table when examining the risk-reward...as of 12/17 GBPUSD dropped to 1.31