Could FTX Have Been Avoided? The Better Polygraph Says Yes

How a deeper look at Sam Bankman-Fried’s voice could have spared the public from FTX disaster.

The fallout of recent FTX, Alameda Research, and Sam Bankman-Fried (SBF) events suggest to us that a world of responsible corporate governance might never be attainable. A world where investors can trust their leaders to make ethical decisions in their best interest. If there is one lesson to be learned from FTX, it is the importance of transparency. Transparency is required to ensure all those involved can see the full human picture and can make informed decisions based on the truth.

How do we create an environment of transparency and a greater sense of accountability to stop those who would intentionally mislead investors, customers and the public while making claims worthy of support and money?

The answer lies right in front of us: the vocal nuances behind tone of voice. At Helios we believe the truth is in the tone of voice, and it’s not just us. Decades of bio-signal and neuroscience research confirm the same. Tone of voice accounts for nearly 40% of human communication while words account for only 7%. The way one speaks is a better indicator than what they speak. In fact, tone is the number one passive indicator of what someone is thinking.

As indicators like CoinDesk’s whistle piece and twitter activity from Binance CEO CZ and Alameda CEO Caroline Ellison accelerated, culminating in the November 11, 2022 Chapter 11 announcement, we began to wonder: Could a deeper look into the voice of  “The Most Generous Billionaire” have saved investors from the cataclysmic effects of the FTX collapse?  

The answer we came upon was a resounding yes. The Helios team saw radical changes between SBF’s voice in pre- and post-collapse interviews. These changes were detected by our newest voice AI product - Earnings Polygraph™ - which is a mechanism for investors to pull back the curtain on hidden truths.

Tapping into Tone of Voice: How We Do It

The Helios platform leverages deep-sensing artificial intelligence (AI) and machine learning (ML) to interpret nuances in the audio behind an individual’s voice. It is trained on over 350,000 hours and 20 years of historical corporate events. The output generated is a set of valuable insights – including level of certainty, impulsiveness, volatility and many more.

By systematically analyzing voice tone, the Helios AI cuts out noise and picks up on meaning that matters.  While text and speeches can be scripted to wrap positive messaging around negative meaning, the subtle nuances found in tone cannot be so easily obscured. Tonal insights create a new realm of understanding; this is not technology rooted in NLP or text sentiment analysis. Tonal AI allows us to hear new things and derive greater meaning. Is the speaker certain of what they are saying? Does the word choice match the tone of voice, or are there incongruencies between the two? These are just some of the questions that can be answered with tonal analysis.

In the case of corporations, what the CEO controls - the company’s biggest moves account for up to  45% of a company’s performance. Understanding their tone of voice is crucial to increase transparency and to make more informed business decisions. A better understanding of a company’s C-suite is a better understanding of a company’s overall performance. This new understanding creates a sharper image of human conversational intelligence and greater market transparency.

Tonal Intelligence Could Have Saved Us from the FTX Burn

After the news broke, we decided to dig deeper and put SBF to the test with the help of our bionic ear.  The Helios team leveraged our AI engine to analyze pre- and post-collapse SBF interviews. From this, we derived voice signatures from our proprietary measurements which provide powerful insights into the future outlook of, in this case, FTX.

All interviews analyzed were focused on SBF alone and were published by outlets such as CNBC, Forbes, and The Wall Street Journal, among several others. Pre-collapse audio interviews (15+ minutes) were narrowed down to only those in 2022 through September. Post-collapse interviews analyzed (30+ minutes) are from mid-November 2022 and onwards.

Like human fingerprints, tone is unique to each individual and changes occur with emotions - doubt, excitement, fear, etc. It is useful to have a way to detect the changes, like Helios’ Earnings Polygraph™ which monitors voice tone for public securities on eight dimensions. If there is a significant change in the tonal fingerprint after an event, such as a public interview or earnings call, it is worth investigating.    

The team found SBF’s voice radically changed between pre- and post-collapse activities. In fact, when analyzing pre-collapse interviews, our models forecasted a clear negative outcome from as early as May 2022 for FTX; perhaps a precursor to the impending collapse. While post-collapse interviews showed a completely opposite tonal sentiment. Such drastic changes show SBF has been singing two different tunes [See figure below].

Had tonal analytics been leveraged to analyze privately-held FTX sooner, a warning would have been generated based on this study alone. Our results in this brief FTX analysis and many others demonstrate the importance tonal analysis plays in detecting uncertainty and in fostering transparency to protect investors from being blindsided.

Tonal Intelligence: The Revolution Has Only Just Begun

The public has witnessed fraud after fraud - Enron, Volkswagen and Wells Fargo to name a few.  Without transparency and accountability, these kinds of cases will persist.  Yesterday it was Elizabeth Holmes of Theranos; today it is Sam Bankman Fried of FTX; and tomorrow it will be another master manipulator who tries to blindside the public to the tune of billions of dollars. We live in a volatile world full of risk and uncertainty. However, insights derived from tone can increase our confidence to operate and manage risk.

Organizational psychologist and science author Adam Grant hit the mark when he said, “We pay too much attention to the most confident voices - and too little attention to the most thoughtful ones. It’s better to learn from complex thinkers than smooth talkers.”

In a world riddled with noise, a deeper look into tone of voice can help differentiate the complex and honest thinkers from the smooth talkers. Helios can help do just that with our deep-listening AI ears to provide quantifiable insights that empower users to make more informed decisions.

The use of voice technology through audio and video has experienced exponential growth in recent years. Despite the fact that tone exists in every aspect of human communication, it is still largely untapped and in its early stages in this context. All components of human communication, including word choice, tone of voice, body language, and facial expressions, are crucial for interpreting human intelligence. When one of these pieces is missing, the picture remains incomplete, particularly with regards to voice tone, which accounts for 38% of a message.

In the present day, text is optimized for communication, but in the future, tone will be as well. The potential applications for this technology are limitless and extend beyond the worlds of investors and corporate fraud. Large and small investors alike cannot afford to operate without understanding the nuances of tone. Similarly, marketers seeking greater engagement and personalization, call centers attempting to manage disgruntled customers, voters surveying political candidates, advertisers looking for production cost savings and audience understanding, and doctors diagnosing patients with illnesses such as Parkinson's disease and depression, all require a deeper understanding of the significance of tone in their respective fields. The use cases for this technology span across industries, each critical in its own right.

When we change the way we listen, what we hear changes too. By stepping towards the vocal nuances found in voice tone, we step into a rich multi-dimensional human experience that promotes transparency and truth above all else.

Want to Learn More About the Helios Earnings Polygraph™?

Click here to book a time with our team and learn how you can gain access to tonal analysis for thousands of company profiles.

About Helios Life Enterprises

Helios Life Enterprises (HLE) is a pioneer in voice-based tonal analysis. HLE is the first and currently only company to conceive and devise a widely available data platform that delivers systematic analytics of an executive’s voice during critical corporate events. These analytics provide novel information that is useful for predicting future earnings surprises, company performance, and cumulative abnormal returns in the context of M&A events. Core product offerings include the Comprehend product suite for buy-side investors of all sizes and in limited pilot Mercury; a scaled research and audio analytics platform that allows a firm to extract full tonal value from their own proprietary audio.

For more information or to request a demonstration, please visit https://www.helioslife.enterprises/.

###

BUSINESS CONTACT

Zeynep Yenisey

Director of Marketing

zeynep@helioslife.enterprises

518-212-7163