Realistic Expectations and Limitations to Consumer-Facing Robo-tic Advisors

Abstract

We present a benchmark life-cycle simulation model (RFSM) that incorporates the financial, demographic, and mortality positions of retired households. By adjusting several features we nest a specific type of consumer-facing (generic) robo model that attempts to minimize the user's workload by imputing key inputs. We calibrate both models under robo policies using the Health and Retirement data for our benchmark model and robo-imputed data for the Generic-Robo model. Our findings indicate that retirees using Robo advisors tend to have large variances in consumption; overspending in a few years with sizeable shortfalls in a number of years. Our RFSM model, however, promotes smoother consumption profiles for retirees. Lastly, retirees using the Generic-Robo model experience more underfunded years and scenarios than those using our RFSM approach.