OpenAI's growth forecast is statistically impossible

The company’s published trajectory implies about $284B in revenue by 2030, versus roughly $13.1B in 2025. That is roughly an 85% compound annual growth rate over five years.

Michael Mauboussin at Morgan Stanley benchmarked that against every US listed company from 1950 onward, roughly 19,000 firm-period observations. No company at that starting scale has matched that pace. Not the dot-com era. Not the smartphone wave.

Under a normal approximation to the historical distribution, such an outcome sits far out in the tail (they describe it as on the order of ~10 standard deviations). For perspective, they note a 7 sigma positive outcome would be expected roughly once in ~780 billion trials, again assuming normality.

Growth rates are not perfectly normal, but the point stands: this would be a genuinely unprecedented outcome for a business already at this revenue scale.