As a young investor in the late 1990's the conventional wisdom was that "sin stocks" (gambling, alcohol, and weapons stocks) outperformed others. Investors could choose to either be greedy and profit or be principled and underperform. The narrative fit the transactional "greed is good" nature of the times. But it turns out the story isn't so simple.

Inflation spotting, meme stocks (see our Jan newsletter), cryptocurrencies, and the green transformation – lots happening in markets. Today we focus on the latter.

With a our new ESG dataset, we see that the performance of ESG investments is more often the opposite of the conventional wisdom - invesors can both do good AND outperform.

BALI

In 2019 MarketPsych had an all-company meeting in Bali. We made good memories but also experienced the dark side of paradise – pre-pandemic Bali was overrun with tourists (yes, us too!) and the infrastructure didn’t hold up. Internet fails frequently, roads are clogged, and the beaches are covered in plastic garbage after rainstorms. A vlogger captured this drone shot at Canggu beach, Bali after a storm.

At our Bali meeting we started repurposing our natural language processing (NLP) engines to focus on ESG, so we (and others) could study how environmental management, social values, and government weakness hurt development. A picture of a team discussion in Bali, below.

For example, we saw Danone single-use cups littering ravines and floating in ocean plastic shoals. The Indonesian government was not effectively managing waste, and Danone was making it worse. As investors, we wondered if there would be any blowback for Danone or Bali longer-term.

TEXT ANALYSIS WITH AI

At the Bali meeting we outlined a new software architecture. We already used natural language processing (NLP) to scan thousands of news and social media outlets in real-time in a dozen languages (including Indonesian!). And we adapted this software for ESG applications.

Below is a simple depiction of our three-stage process digitizing key themes in media text – in this case trust in the management team of Luckin Coffee – converted into a ManagementTrust score.

We create hundreds of scores like PollutionControversy for 30,000 companies and GovernmentTrust for 252 countries and territories. As we test the data, we see many of the findings align with recent behavioral economics research.

QUANTITATIVE RESULTS

For example, recent Oxford University research found that happier employees are more productive – that is, happy call center employees sell 13% more than their unhappy peers. Happy workers are more productive (weforum.org) & "Does Employee Happiness have an Impact on Productivity? (SSRN)"

Our Workplace Sentiment score quantifies news and social media mentions of happy (vs unhappy) working environments. We ranked S&P 500 companies on their monthly score, and we plotted the next month stock price returns of the top and bottom 5% of companies. The return os a portfolio of companies with positive office sentiment is plotted in the green line, and the returns of the most unhappy workplaces is depicted in the red line.

Month-after-month, companies with happier employees outperform while companies with unhappy employees underperform.

SIN STOCKS OUTPERFORM? NOT LATELY

Not all of our results are so consistent – we decided to investigate “climate sin stocks”. We tracked a basket of the top 10% of companies in the S&P 500 associated with Sustainability Controversy in the media since 2006. We used a monthly rotation model.

Note that these unsustainable companies outperformed the remaining 90% of the S&P 500 from 2006 until the oil price collapse in 2014. Since then Unsustainable companies have been significantly underperforming while Sustainable companies have been outperforming.

ELON MUSK

The media perceives Tesla as a sustainable company, but portrayals of Musk's management style are often negative. Below we can see how the views of Musk plummeted when news of him smoking marijuana and drinking whiskey on the Joe Rogan podcast was released. The red dotted line is the minutely share price of Tesla (TSLA) and the blue line is a 60-minute moving average of the media sentiment about Tesla’s management team (including CEO Musk).

That said, one (or a few) management mistakes does not a bad stock make. While Tesla is an exception, when we tested the S&P 500 stocks on their management sentiment scores, we saw that over time companies associated with higher management sentiment generally outperform those with lower.

HOUSEKEEPING AND CLOSING

There are many factors affecting share prices, and the three we showed above - workplace happiness, sustainability, & management sentiment - are only three.

We also see some anti-ESG factors suggesting positive returns, and we'll report on those as well if they appear robust. For example, we initially saw that our score "DeceptiveAdvertising" was associated with higher stock returns for the S&P 500, but on further examination we found top tech companies were the primary inclusions, and there is no broadly stable predictive relationship there.

We’ve been looking through our data, and in coming months we’ll share new insights around the evolution of the economy. Most of our positive ESG results are replicable at yearly horizons, across countries, and with larger stock universes.

We're part of a free webinar "Building Quant Strategies with ESG Data" on Wednesday March 17th at 9:30am ET / 14:30 CET by NorQuant in case you’d like to learn more!

Semoga Sukses! (Best wishes!)
The MarketPsych Team

REFERENCES

Bellet, Clement and De Neve, Jan-Emmanuel and Ward, George, Does Employee Happiness have an Impact on Productivity? (October 14, 2019). Saïd Business School WP 2019-13, Available at SSRN: https://ssrn.com/abstract=3470734

LinkedIn Twitter