I saw a viewpoint today where the Coinbase CEO said that prediction markets need insider trading, as inside information can improve accuracy when pursuing truth signals. The essence of a prediction market is a truth aggregator. The purpose of the stock market is to help companies raise funds, providing capital liquidity and exit mechanisms. The reason the stock market doesn't allow insider trading is to protect ordinary investors from unfair trading. If insider trading were widespread, ordinary investors would perceive it as unfair, reducing trading and capital investment, leading to decreased market liquidity. Similarly, if we consider prediction markets to be truth aggregators rather than investment or trading markets, and therefore allow insider trading, then ordinary investors will naturally reduce their investment, leading to decreased liquidity and a loss of predictive value, which is detrimental to the long-term development of prediction markets. I believe prediction markets have three distinct roles: trading markets, truth aggregators, and bounty platforms. If most participants consider prediction markets to be trading markets, then protecting fairness and thus market liquidity is necessary. Regulation of insider trading is also necessary. If most participants consider prediction markets to be truth aggregators, and they are willing to invest in uncovering the truth and filtering out noise, then insider trading is permissible. If most people perceive prediction markets as reward platforms, then artificially altering the outcome of events is permissible. For example, predicting whether a company CEO or a president will step down by the end of the year, if the odds are favorable, could significantly influence the outcome. Even match-fixing in sports is permitted (though it would be illegal under sports regulatory bodies). From a perspective of positive social development, I would argue that prediction markets are a real-time mechanism for aggregating collective wisdom, capable of more accurately assessing political, economic, and sporting indicators than traditional polls or expert predictions, thus providing earlier signals for decision-making by relevant administrators. However, this hinges on participants recognizing that prediction markets are not fair trading markets, but rather mechanisms where participants bear the costs of information discovery.