Which forecast of the past twenty years aged worst, and what did it miss?

Looking back at old predictions is humbling and useful. Some forecasts missed because the technology changed, others because people behaved differently from the model, others because something unexpected happened. Nassim Taleb's black swan idea describes rare, high-impact events that nobody foresaw but that seem predictable in hindsight.

We are asking readers to pick a specific forecast made in the past twenty years, from an expert, an institution or a company, and examine why it was wrong. A good answer would quote or link the original forecast, say when and by whom it was made, explain what happened instead, and identify the cause of the error: bad data, a bad model, a hidden assumption, overconfidence, or a surprise.

The most helpful answers also say what a better forecaster could have done at the time, and avoid the trap of hindsight bias, assuming the right answer was obvious.

Predictions that were right for the wrong reasons are welcome too.