Enhancing Stock Investment Returns with Learning Aggressiveness and Trust Metrics
In addition , since interest rates impact agents ’ cumulative return overtime in the model , the model could be used to inform policy makers on the setting of the interest rate in the context of anticipated investor reactions . This is due to the fact that the government ’ s open-market operations induce liquidity effects that leads to interest rate behaviors ( Lucas , 1990 ), and market liquidity also affects the trading activities ( Chordia , Roll , & Subrahmanyam , 2001 ).
2 . Background
Investment strategies are usually classified as either passive or active portfolio
management strategies ( Barnes , 2003 ). Passive portfolio management only involves limited buying and selling . Passive investors anticipate the appreciation of their capital in the long term , and there is only limited engagement involved in their portfolio maintenance . In contrast , active portfolio management strategies can bring investors extra profits because they can benefit from short runs , but it requires their active engagement with the stock market . Thus , active engagement covers a wider range and frequency of stock price movement . Active equity portfolio management requires periodic forecasting of economic conditions , as well as portfolio rebalancing based on the forecasted conditions ( Grinold & Kahn , 1999 ). A degree of trust in the rationale for market movements and trends impacts the stock-trading activities and investors ’ risk control strategies ( Asgharian , Liu , & Lundtofte , 2014 ). Risk control methods make it possible for a dynamic portfolio management strategy to outperform the market ( Browne , 2000 ). A simple momentum and relative strength strategy has been outperforming the buy-andhold strategy by 70 % since the 1920s ( Faber , 2010 ). Performance can be improved by considering simple trends before taking positions . Overall , learning from past performance and from actions of others makes investors become better through their accumulated expertise ( Seru , Shumway , & Stoffman , 2010 ). This is , in turn , reflected in their actions in the market place . Clearly , this constant dynamic interplay between investors ’ changing strategies and hard-to-predict market movements makes the problem of understanding the market place and anticipating its behavior extremely challenging .
However , the automated methods described above simply provide a retroactive , aggregated model of the stock market . They don ’ t take into consideration interactions between / among agents , as they learn from each other and change their behavior and strategies accordingly . These interactions , in fact , can better inform investors about changing market conditions due to changed investment strategies . This clearly affects the performance of anyone ’ s investment portfolio .
The CAS methodology offers a natural framework for augmenting portfolio management strategies with simulation of individual agent interactions in the market place and their financial surroundings . A further advantage of CAS stems
51