The hot hand fallacyis a cognitive bias where people believe that a string of recent successes (a "hot streak") makes future successes more likely, even when outcomes are largely independent or random.
It originates from basketball lore: fans, players, and commentators assume a shooter who has made several shots in a row has a "hot hand" and is more likely to keep scoring. Classic research (e.g., Gilovich, Vallone, and Tversky in 1985) suggested this perception is mostly illusory—past shots don't meaningfully predict the next one in truly random or independent trials. (Note: later studies have nuanced this, finding small real effects in some skilled contexts, but the overbelief in streak continuation remains a strong bias.)
In Investing (and Specifically AI Stocks)
In financial markets, the hot hand fallacy appears as extrapolation bias: investors over-weight recent strong performance and assume it will persist, often ignoring fundamentals, mean reversion, competition, valuation, or external risks.
Relation to AI stocks:
The AI boom (especially since 2022–2023) has featured massive gains in stocks like NVIDIA, Super Micro Computer, Broadcom, and others tied to GPUs, data centers, and AI infrastructure. Investors have piled in, chasing momentum because "AI is the future" and these names have been "hot."
This creates a hot hand mentality: "NVIDIA has crushed earnings and stock price for X quarters → it (and the sector) will keep doing so indefinitely." People extrapolate short-term hype and revenue surges into long-term inevitability, buying at elevated valuations without fully accounting for risks like:
Competition and commoditization (e.g., new AI chip entrants, open-source models reducing hardware dependency).
Capex bubbles (hyperscalers overbuilding data centers).
Regulatory, energy, or adoption slowdowns.
Mean reversion — exceptionally high growth rates are hard to sustain forever.
This bias fuels momentum trading, FOMO buying, and bubbles. When the streak eventually cools (as all do), late buyers suffer sharp drawdowns. It's the mirror image of the gambler's fallacy (expecting a reversal after a streak), but in bull markets, the hot hand version often dominates for longer runs.
Why It Persists
Representativeness heuristic: We expect small samples (recent performance) to resemble the broader "story" (AI transforming everything).
Recency bias: Recent winners feel more salient than historical data or base rates.
Narrative appeal: AI has a compelling long-term thesis, which amplifies the emotional pull of recent success.
Countering it: Focus on base rates, forward-looking fundamentals (e.g., P/E ratios, competitive moats, actual AI ROI for customers), diversification, and probabilistic thinking rather than streak-chasing. Past performance (even spectacular) does not guarantee future results—especially in fast-moving tech sectors.