Behavioral and structural opportunities will persist until both human nature changes and academic / industry practices are radically reformed. After several years of allocating to hedge fund managers who employed pattern recognition and other systematic trading strategies, we began to recognize a pattern of our own. Results in live trading rarely approach those developed in back-tests. It’s not that the math is wrong. It's that these simulations are consistently based on false assumptions. And unfortunately, most managers start with identical, false assumptions. Even more concerning, the nearly universal approach to developing these systems - “torturing data with false assumptions” - is incapable of adapting to continuous market changes. At Reformation Technologies, we look for sustainable and deep behavioral and structural opportunities. Pattern recognition, data mining and optimization are for the other guys.
Behavioral and structural opportunities will persist until both human nature changes and academic / industry practices are radically reformed. After several years of allocating to hedge fund managers who employed pattern recognition and other systematic trading strategies, we began to recognize a pattern of our own. Results in live trading rarely approach those developed in back-tests. It’s not that the math is wrong. It's that these simulations are consistently based on false assumptions. And unfortunately, most managers start with identical, false assumptions. Even more concerning, the nearly universal approach to developing these systems - “torturing data with false assumptions” - is incapable of adapting to continuous market changes. At Reformation Technologies, we look for sustainable and deep behavioral and structural opportunities. Pattern recognition, data mining and optimization are for the other guys.