Page 137 - 49A Field Guide to Genetic Programming

12.5 Financial Trading, Time Series, and Economic Modelling 123

12.5 Financial Trading, Time Series

Prediction and Economic Modelling

GP is very widely used in the areas of financial trading, time series prediction
and economic modelling and it is impossible to describe all its applications.
It this section we will hint at just a few areas.
Chen has written more than 60 papers on using GP in finance and eco-
nomics. Recent papers have looked at the modelling of agents in stock
markets (Chen and Liao, 2005), game theory (Chen, Duffy, and Yeh, 2002),
evolving trading rules for the S&P 500 (Yu and Chen, 2004) and forecasting
the Heng-Sheng index (Chen, Wang, and Zhang, 1999).
The efficient markets hypothesis is a tenet of economics. It is founded
on the idea that everyone in a market has “perfect information” and acts
“rationally”. If the efficient markets hypothesis held, then everyone would
see the same value for items in the market and so agree the same price.
Without price differentials, there would be no money to be made from the
market itself. Whether it is trading potatoes in northern France or dollars
for yen, it is clear that traders are not all equal and considerable doubt has
been cast on the efficient markets hypothesis. So, people continue to play the
stock market. Game theory has been a standard tool used by economists to
try to understand markets but is increasingly supplemented by simulations
with both human and computerised agents. GP is increasingly being used
as part of these simulations of social systems.
Neely, Weller, and Dittmar (1997), Neely and Weller (1999, 2001) and
Neely (2003) of the US Federal Reserve Bank used GP to study intra-day
technical trading on the foreign exchange markets to suggest the market is
“efficient” and found no evidence of excess returns. This negative result
was criticised by Marney, Miller, Fyfe, and Tarbert (2001). Later work by
Neely, Weller, and Ulrich (2006) suggested that data after 1995 are consis-
tent with Lo’s adaptive markets hypothesis rather than the efficient markets
hypothesis. Note that here GP and computer tools are being used in a
novel data-driven approach to try and resolve issues which were previously
a matter of dogma.
From a more pragmatic viewpoint, Kaboudan shows GP can forecast in-
ternational currency exchange rates (Kaboudan, 2005), stocks (Kaboudan,
2000) and stock returns (Kaboudan, 1999). Tsang and his co-workers con-
tinue to apply GP to a variety of financial arenas, including: betting (Tsang,
Li, and Butler, 1998), forecasting stock prices (Li and Tsang, 1999; Tsang
and Li, 2002; Tsang, Yung, and Li, 2004), studying markets (Martinez-
Jaramillo and Tsang, 2007), approximating Nash equilibrium in game the-
ory (Jin, 2005; Jin and Tsang, 2006; Tsang and Jin, 2006) and arbitrage
(Tsang, Markose, and Er, 2005). Dempster and HSBC also use GP in for-
eign exchange trading (Austin, Bates, Dempster, Leemans, and Williams,