At first glance, circumstances appeared to be auspicious. For the third quarter, SentinelOne reported adjusted earnings of 7 cents per share, beating out the consensus view of 5 cents per share. On the top line, the company generated $258.91 million, exceeding analysts' consensus target of $257.7 million. Further, total revenue jumped by 23% against the year-ago level, with customers with annualized recurring revenue (ARR) of $100,000 or more growing 20% to 1,572 in the most recently concluded quarter.
Despite encouraging talking points from management, which emphasized growing demand for the company's AI-native security platform, a holistic solution that combines data, intelligence and defense, investors were fixated on guidance. Unfortunately, SentinelOne projected fourth-quarter sales to reach about $271 million, disappointing analysts who were expecting $273.09 million.
Not surprisingly, investors read between the lines. During yesterday’s afterhours session, S stock fell more than 7%. On Friday afternoon in the open-market session, the equity dipped about 13%.
Despite the pain of the earnings miss, there's a science lesson here. What we're witnessing is heteroskedasticity in action. With SentinelOne delivering disappointing guidance, the event trigged a volatility shock against S stock: price gets noisy, ranges expand and the distribution of outcomes widens. In other words, volatility isn't constant but clusters around impactful events.
Later, as the market digests the news, uncertainty bleeds out, with volatility eventually collapsing back toward its baseline. This eventual transition from high volatility to low volatility is also heteroskedasticity in action.
What we're going to do is to measure this phenomenon to potentially ...
