Everyone wants to work for Tesla, right? Who wouldn’t be happy at a firm with a charismatic, headline-grabbing CEO who isn’t afraid to be at the bleeding edge of technology? And as a clean energy company, they must do pretty well on environmental, social and governance metrics (ESG)—factors that more and more investment firms are incorporating to drive alpha.  

Well, when quantitative investment firm PanAgora Asset management analyzed sentiment among Tesla employees, the data told a more complicated story. It seems that Tesla’s recent Q1 losses are taking their toll on staff. And staff sentiment is a harbinger of performance.

If you see the sentiment of employees falling, it’s going to lead to lower stock prices in the future

George Mussalli

Using machine-learning technology, the group analyzed the social media “chatter” of employees, said George Mussalli, chief investment officer and head of research for equity, at PanAgora.  

“From this chatter, we see that despite the positive headlines about Tesla—and their self-driving car technology is leading the pack—internally the mood is getting more and more sour. Tesla is probably missing deadlines and the financial losses there have been weighing on them for weeks,” said Mussalli, who was speaking at a Bloomberg Invest event held in New York on June 4.

He added that only recently have firms been able to track something as opaque as employee sentiment. “Previously, if you had asked a CEO, ‘Are your employees happy?’, 100% of CEOs will say, ‘Yes.’ And there is no way to verify that,” Mussalli said. “But today we use web-scraping technology to monitor the chatter of employees on social media. We take all that and then we apply machine learning to capture sentiment.”

This case study was an example of how quant firms can use ESG data to improve performance. While Tesla superficially would appear to perform well on ESG metrics, employee satisfaction is an important marker of future stock performance.

“This precedes stock prices. If you see the sentiment of employees falling, it’s going to lead to lower stock prices in the future,” Mussalli said.

He said that PanAgora believes there is alpha in ESG and that belief is based on the premise that most assets in US companies are intangible, including employee sentiment.

Material World

The ESG space is wide and deep. As such, firms have found it challenging when it comes to sifting through all these datasets to find materiality—the relevance of information.

“There are dozens of ESG factors, but they are not all applicable across the universe,” Mussalli said.

While Disney and Netflix might both be classified as media companies, for example, Disney’s environmental footprint is plainly far deeper, he added. This is crucial as portfolio managers would prefer to make like-for-like comparisons, rather than extrapolate insights.

ESG performance is measured by materiality. For many metrics in ESG, firms have a lot of consistent, measurable data available. In the case of carbon, however, measuring impact is difficult—as WatersTechnology recently reported.

Asha Mehta, portfolio manager and director of responsible investing at Acadian Asset Management, who was speaking on the same panel, said that while new tools are becoming increasingly available to help firms find value in a sea of information, they also help to highlight the fact that there are a lot of unknowns.

“The carbon pricing issue is a classic example,” Mehta said. “We recognize that there is momentum towards a low-carbon future, so how do we price these into portfolios? It’s considerably challenging given the issues out there with a lack of data.”

While Acadian tracks 40,000 securities around the globe, only 2,500 of these companies have reported their carbon emissions.  

“So one of our challenges is to identify values for all that missing data, and we use some of the tools that we have along the lines of imputation and web scraping,” to accomplish that, said Mehta.