Journal on Policy & Complex Systems Vol. 2, Issue 2, Fall 2015 | Page 57

Journal on Policy and Complex Systems
4.2 Trading Rules

The following formulas describe agents ’ basic decision rules in detail ( Su &

Hadzikadic , 2014 , 2015 ).
For example , if the values of buy threshold , buy period , self-confidence , and market momentum for a particular agent are 0.2 , 40 , 0.9 , and 0.7 , respectively , then the buying rule for this agent is :
If the stock price goes up 0.2 * ( 1 – ( 1 – 0.9 ) * 0.7 ) * 100 % = 18.6 % in the past 40 trading days , then , the agent takes a long position .
Similarly , if the values of sell threshold , sell period , self-confidence , and market momentum are 0.1 , 10 , 0.2 , and , 0.3 respectively , then the selling rule for this agent is :
If the stock price goes up less than 0.1 * ( 1 – ( 1 – 0.2 ) * 0.3 ) * 100 % = 7.6 % in the last 10 trading days , then , the agent takes a short position . Agents have two sets of trading rules . One is for the bull market , and the other is for the bear market . Agents identify current market status based on the recession indicator provided by the National Bureau of Economic Research ( NBER ). Agents switch trading rule sets once they have the latest knowledge of the market status change . Since the financial markets are not perfectly correlated with recessions , it usually takes about 250 trading days for a recession to start impacting the stock market . The recession signals from NBER are delayed for a year .
In addition , agents are permitted to sell short the stock at any time . This decision makes it possible for agents to capture profits in market downturns , as well as to hedge the potential risk with making erroneous decisions . An agent can sell short any amount of stock up to their available cash amount . In financial terms , a long position indicates the purchase of a security such as stocks , commodity , or currency , with the expectation that the value of these assets will rise in value over
54