Pulled from every Market Bubble episode. Ranked by how much the hosts genuinely dug in — charts, theses, price action — not just how often a name got dropped. Every row links to the exact moment.
Loading…
Tokens are extracted from the transcripts by a language model, then normalized (caption manglings merged, non-assets like platforms and categories filtered out). Low-confidence hits are dropped. Some error remains — treat the tail of the list with more suspicion than the top.