AI Case Study

The Financial Times reports that "SAE now employs 80 portfolio managers and researchers, who include more than 30 PhDs in computer science, physics and engineering. The goal is not to achieve dramatic wins but consistent market-beating results. SAE’s research process borrows from academia. An analyst proposes an investment idea, and is then assigned a “referee” who spends a week trying to demolish the thesis. It then gets presented to an approval board, which decides whether it should be included in investment portfolios and what kind of weighting it should be given. These days, the competitive advantage from such signals can quickly be lost, so SAE often gives a heavy weighting to new signals, and gradually reduces the money allocated as their effectiveness “decays”. A good example is Glassdoor, where employees can anonymously review their companies. Two decades ago investors might have looked at Fortune magazine’s ranking of the happiest workplaces, but today SAE can systematically monitor for signs of a company’s employees getting happier or restive. That analysis has turned out to be a good predictor of stock returns."

"Although BlackRock executives stress that SAE stands on its own two feet, the hope is that resources ploughed into the quant arm will be a catalyst for the entire company. One example is the real-time economic gauges SAE has created for the BlackRock Investment Institute, based on a host of alternative data sets like internet searches, online invoices and even traffic patterns. But the cross-unit pollination will need to be much broader and bigger to be judged successful."