最新报道:A Cornell University study analyzing over 2 million Reddit comments found that sentiment-based trading strategies outperformed buy-and-hold by up to 84%. Quantitative strategist Maksim Baradziuk explains how algorithms can integrate behavioral biases like overconfidence and loss aversion to improve trading performance, especially during high-volatility periods. His research shows that behavioral-aware algorithms can identify emotional market extremes and enhance risk management, as demonstrated during the COVID-19 market shock.