I am always happy when I wake up early on a cold morning in Riyadh and meet my old friends at our favorite cafe in the Business Gate complex. There, we discuss the news, as well as the performance of major corporations and brands. We agree and often disagree, as it is natural for people to differ, particularly in how we express emotions, opinions, and sentiments.
One of my friends mockingly remarked, as he held his smartphone to tweet his dissatisfaction on the X platform: “Would you believe I'd rather pay extra to purchase what I need from Amazon instead of Temu? After my first purchase from Temu, I began receiving annoying emails and WhatsApp messages full of silly emojis on an almost daily basis.”
We all read his tweet, which he wrote in local dialect and in which he blended humor and frustration. Today, we all use social media and various online platforms to exchange opinions, discuss the pros and cons of products, choose restaurants while traveling, and even make stock market projections. We also see tweets evaluating the services provided by hospitals and private schools.
Investors, CEOs, hedge fund managers, marketing executives, consulting firms, and customer service analysts obtain crucial data from the positive or negative sentiments and views people convey online. They use sentiment analysis (also known as opinion mining), which integrates linguistics, psychology, data science, and artificial intelligence, to do so. In other words, my friend’s tweet is linguistic data containing personal emotions and feelings that are analyzed by advanced machine learning and deep learning algorithms. This enables decision-makers to make informed choices and improve their financial performance.
Companies and brands are not the only ones using opinion mining. Media institutions also use software such as Chartbeat, Google Analytics, and Social Mention to attain crucial insights into their audience. This software helps media companies better understand their audience’s interests, engagement levels, and the number of people visiting their websites. These tools also help them categorize readers and viewers based on their geographical location and gender. Simply put, AI-supported analysis is driven by human analysts who oversee the process of gathering signals and comments from people across the globe into a single file to allow decision-makers to take appropriate steps.
It literally involves mining data from social media, as well as all of the relevant written and visual content shared on it, to reach a broad conclusion evaluating the sentiments as negative or positive, with a precision that sometimes approaches 90 percent. Social media content is shared spontaneously and is not subject to the scrutiny of professional editing of traditional media.
However, the way in which social media data is used to predict the performance of stock and hedge funds is truly astonishing. This analysis is conducted in parallel with the fundamental and technical analyses used to monitor the volatile financial market environment, which constantly fluctuates. Stock markets are influenced by the news cycle and real-time discussions on social media.
In an Indiana University Bloomington study, researchers conducted a review of many academic research projects and training modules on the use of artificial intelligence in opinion mining on social media. The study concluded that predictions based on analyses of a large number of tweets over a specific period, of the performance of the Dow Jones Industrial Average the next day, are 87.6 percent precise.
This can be illustrated by another example, that of a large company owned by a man who has held the title of the richest man in the world several times. At the end of 2023, he was thought to be the third richest, with an estimated net worth of $195 billion.
Elon Musk presents a prime example of how conveying emotions and opinions - and, of course, sharing rumors and jokes - can be either a blessing or a curse for global stock-market investors. Musk's tweets have had a significant impact on the stock market performance of Tesla, a company that made electric cars. When Tesla went public in 2010, its stock was initially listed at $17.
According to the data available at the time of writing (May 31, 2024, 14:40 Riyadh time) Tesla's shares are now listed at $178.79 on global indices, including the Nasdaq 100 and the S&P 500. It is worth noting that the company's stock has both risen and fallen over the years, exceeding $700 at certain periods.
On April 1, 2018, Elon Musk shared a sudden and shocking Tweet announcing that Tesla had gone bankrupt. We later learned that the tweet was just an April Fools' joke. Immediately, investors and hedge fund managers analyzed the data and sentiments on X (formerly Twitter) and reacted to the flood of responses from Musk's tens of millions of followers.
The result was disastrous, with Tesla’s shares plummeting by 7% when markets opened. However, the stock quickly recovered, improving dramatically over the next two days, rising over 13% after Musk had reassured investors, shareholders, and hedge fund managers that he was joking.
What does this mean for investors? Overall, it is alarming news, as mass panic selling of stocks can be stirred at any moment. No sentiment drives stock prices down like fear, and this is true across the board, from oil markets to pharmaceutical companies. Therefore, sentiment analysis is crucial for helping investors determine what actions are appropriate at any given moment. Emotions and feelings push people to constantly react to news, comments, and opinions of celebrities, either in fear, ignorance, or hope.
