#预测市场 Seeing this research report from Kalshi, I am reminded of the various prediction failures I have seen over the years. The 25-month data shows that the accuracy of prediction markets in forecasting inflation is 40% less than the consensus on Wall Street. This is not new to those of us in the field—it's just that this time, someone finally used data to put it on the table.

Do you remember the period from 2021 to 2022? Traditional economists confidently claimed that inflation was "temporary," only to be proven wrong by reality. Meanwhile, those trading in prediction markets, because they had real money on the line, became more sensitive to risks. This is the power of incentives—when your money is at stake at the betting table, you study the data more diligently than analysts who only talk without action.

Even more interestingly, when actual data deviates significantly from expectations, the advantage of prediction markets can be amplified to 67%. This indicates that during times of greatest uncertainty, the collective wisdom of market participants performs the best. This is not luck; it is the result of the combined forces of collective intelligence and economic incentives.

However, I must say that how long this advantage can be maintained is still uncertain. Historically, those seemingly invincible prediction models have failed over time. For prediction markets to truly replace the Wall Street consensus, they need to go through more cycle tests. But at least now, with solid data support, the credibility of this tool is indeed increasing.