4. Alternative data: the edge off the price chart

Level 1: Foundations of Algorithmic Trading

For decades, quantitative trading was built almost exclusively on price, volume and company fundamentals. Today, much of the competitive edge of the most sophisticated funds comes from somewhere else: alternative data, information that is not born in a financial market but that can anticipate what is about to happen in one.

Textual data is extracted from news, social media or earnings reports, and processed with natural language processing techniques to measure market sentiment or detect events that may move the price before they become headlines.

Satellite imagery makes it possible to monitor things as concrete as activity at a factory, the number of cars in a department store car park, or harvest forecasts — information that can anticipate a company's quarterly results weeks before they are published.

Web scraping consists of gathering data published on websites — product prices, reviews, job postings — to gauge a business's activity or its position against competitors. Google Trends, for its part, reflects search behavior globally and gives clues about genuine consumer interest in a product or service.

Data from Internet of Things sensors is also used, to measure economic activity or supply chain efficiency in real time, along with credit card transaction data, which offers a very direct snapshot of consumer spending and, with it, an early clue about the revenue a retail company is going to report.

What all these data sources have in common is that none of them sit on a traditional financial terminal. Whoever can obtain them, clean them and turn them into a useful signal before the rest of the market has a real edge — albeit a temporary one, because as soon as an alternative data source becomes popular, its advantage erodes.