AI Case Study
SCT Capital Management $350 Adaptive Quant Trading hedge fund is fully automated
SCT Capital Management $350 Adaptive Quant Trading hedge fund employs a mix of short term strategies based on 40+ of the most liquid futures contracts globally. The advanced machine learning models claim to see patterns that are not easily observable to human traders. The results are unclear but Bloomberg reports mediocre returns with AI trading beating the broader hedge fund but not the stock market.
Fund And Asset Management
"AQT (Adaptive Quant Trading) Company: SCT Capital Management Investment Program Description: The AQT (Adaptive Quant Trading) program employs a mix of short term strategies based on volatility, momentum, and counter-trend patterns on 40+ of the most liquid futures contracts globally. The program is conservative in risk exposure with holding periods of 5 days or less, limited capital allocation to any one strategy, and tightly controlled overall portfolio leverage. Since inception, AQT has had a worst drawdown of 4.9% with a realized volatility of 7.7% based on monthly data. The program is 100% systematic, including the process of asset allocation where risk exposure in monitored and adjusted on a constant basis. The strategies and trading systems in the program have matured over twelve years of proprietary trading at Morgan Stanley and Deutsche Bank, and five years as an independent sub-advisor to Deutsche Bank. The newer strategies in AQT are adaptive, based on a periodic automated statistical learning theory procedure that takes as input our database of market indicators and produces as output many simple uncorrelated agents that are assembled into a well calibrated trading strategy. The database of indicators are based on a blend of 17 years of trading experience and market knowledge and advanced quantitative analysis. The philosophy underlying the newer strategies, which we intend to continue adding, is that markets are complex and computers are able to extract hidden emerging patterns based on powerful statistical methods that are not easily recognized by humans."
According to Bloomberg "today’s small group of fully automated AI strategies are off to a middling start. Their performance beats the broader hedge fund industry but not the stock market. Thirteen AI funds gained an average of 10.6 percent annually in six years through 2016, and rose 8.5 percent through October."
20 years ago founded one of the first machine-learning hedge funds, the $350 million Adaptive Quant Trading program at SCT Capital Management. “They can generate hypotheses, test them, and then tell humans, ‘This is interesting, go dig deeper.’ As machines add more value, it changes the nature of work humans do.”
"The database of indicators are based on a blend of 17 years of trading experience and market knowledge and advanced quantitative analysis."