Limitation

The agent-based financial markets literature re­veals some of the objective and design limitations of such models. On the objective limitations side, one of the major limitations is that the majority of agent-based markets focus on replicating some of the stylized facts of real markets, while they pay much less attention to identifying the key aspects that are responsible for the emergence of real financial market behaviour.

With regard to the design limitations, the complexity of some of the agent-based financial markets, which prevent from identifying the key aspects those are responsible for the emergence of financial market behaviour. In addition, the timing issue represents a crucial limitation on agent-based markets. In the majority of agent­based markets, trading is synchronized and, most importantly, there are no representations of physical time in terms of the market. In addition, most of these models use low-frequency timing such as daily timing, in order to represent time, which does not reflect the real market. This is a clear indication that these two assumptions are not realistic and there is a vital need to address this issue. Besides, one of the main agent-based market constraints is their validation. The lack of market data, especially high frequency data, in order to validate the accurate of agent-based market, appears to be an obvious critical issue in many agent-based markets.

In this work, we have presented an agent-based FX market in which we attempt to cope with some of the existing limitations associated with previ­ous work in agent-based markets, in an attempt to precisely replicate FX market trading activity.

4.