Evento

QWAFAFEW Denver - Rick Davis from BrandLoyalties, Inc.

"Big Data" Meets "Deep Learning"

"Big Data" and more specifically the web luminosity of consumer oriented brand names offers an opportunity to adapt Peter Lynch's "brand loyalty" based investment strategy for the 21st Century. While managing the Magellan Fund at Fidelity Investments from 1977 to 1990, Peter used what he called "the power of common knowledge" (i.e., consumer familiarity with and loyalty to brand names) to select candidate equities that warranted his rigorous in-depth fundamental analysis. During that time the average annual return experienced by Magellans shareholders was over 29%.

But when leveraging "Big Data" into history's largest and most timely "brand loyalty" focus group it quickly becomes obvious that the web is messy, turbulent and unstructured. Context is critical, and "rules based" single-variate analysis generally cannot deal with context sensitive and chaotic data. Fortunately, Machine Learning is inherently multi-variate and it can deal with rapidly evolving environments by finding relationships and dependencies that rule writing human experts simply cannot see within the chaos of "Big Data."

This presentation will address 5 questions:

1. What is our Big Data and how is it captured?
2. What is the nature of the signals that can be produced?
3. Can those signals produce Alpha in fundamental/discretionary trading strategies?
4. Can those same metrics enhance returns in traditional quantitative/systematic portfolios?
5. Why does ML/AI work so well with Big Data?