Trust research is traditionally found in marketing, economics, psychology, sociology and organisational behaviour literatures. Complexity theory provides an alternative lens through which to conceptualise trust. Three complexity theories (adaptation, self-organisation and self-organised criticality) are integrated with marketing and psychological theory to provide an understanding of trust as an emergent phenomenon and to guide the design of an intelligent agent simulation which will be tested subsequently within the CRM data of a major financial institution.Both fuzzy logic and evolutionary strategies are employed within a multi-agent simulation of interactions between wealth management advisors (WMAs) and their clients. In the simulation, fuzzy logic represents the agents' behavioural rules which are derived from complexity, marketing and psychology theories. The results obtained using the RePast based simulation show the advantages of evolutionary learning in optimising WMA-customer relationships and how this learning in turn affects WMA strategies as they seek to reduce client churn.