Simulation of trust in client-wealth Management Adviser relationships

This paper describes the components of a two-phase model for simulating trust amongst clients and their Wealth Management Advisers (WMAs). In phase one, an artificial life model was used to assess the dynamics of trust. In phase two, the model is extended to utilise real data from a corporate database of client information. The Alife model highlighted the need for information not captured directly, requiring sophisticated inference techniques. Fuzzy logic is used to describe client behaviour with rules found through evolutionary optimisation. Analysis of mutual information between time series of clients' investments is used to determine links between clients.