AI Rally vs Historical Bubbles: 3 Years Duration, 131% Gains vs 244% Average

Bank of America’s bubble comparison table reveals the AI rally has already exceeded average bubble duration (3.0 years versus 2.55 average) while delivering only half the typical gains (131% versus 244% average). This positioning suggests either the rally has room to run or it represents a more moderate cycle than historical extremes.

Duration vs Magnitude

The AI rally is longer than average but less extreme in gains. This could indicate:

• Bull case: Room to run toward the 244% average before exhaustion
• Bear case: Already mature by duration, gains concentrated in few names
• Base case: A more moderate cycle reflecting learned caution from previous bubbles

What Makes This Cycle Different

Unlike previous bubbles, AI’s core beneficiary (NVIDIA) generates massive cash flow ($60B FCF). The economic foundation differs from profitless dotcom growth.

The Comparison Limits

Historical bubble comparisons assume similar dynamics. But AI’s infrastructure — as explored in the economics of AI compute infrastructure — concentration, cash generation profiles, and global adoption patterns may not map cleanly to historical precedents.

The table provides context, not prediction. Whether AI follows historical patterns or breaks them remains the central investment question.

For deeper analysis of market cycles, subscribe to The Business Engineer.