An agency theory scale for financial services

The purpose of this paper is to draw on agency theory (AT) to develop and validate a scale to investigate customers’ loyalty towards salespersons under agency problems. Especially with credence services, customer-salesperson relationships are beset by problems that stem from personal differences and conflicts of interests.

The context is financial investment services in Indonesia. Following Churchill’s (1979) procedure, initial qualitative surveys generated a pool of 44 items for the three AT asymmetries of risk, goal and information. Expert panel validation and exploratory factor analyses of a dataset (n = 429) confirmed a four-dimensional structure of 30 items. Finally, confirmatory factor analyses using a second dataset (n = 299) tested the effects of the four asymmetries on customer loyalty, and how the effects are moderated by customer trust and relationship duration.

Overall, agency problems negatively influence customer loyalty. Information asymmetry has two discriminantly distinct dimension, quality and timeliness. Low trust amplifies the effects of all asymmetries, whereas relationship duration reduces the effects of only risk asymmetry. Comparison with a three-factor model shows that the four-factor scale is superior.

Service firms desiring long-term customer relationships need to be cognisant of potential asymmetries created by salespersons, and develop initiatives to minimise potential fallouts from each asymmetry. These include proper alignment of compensation structures, accurate and timely communications of product information, and matching risk profiles between customers and salespersons. Regular customer feedback regarding perceived asymmetries would also help early problem detection. Finally, the findings would inform the development of policy matters and industry best practices.

Besides contributing to the small stream of research that applies AT to marketing, this study is the first to develop and validate an AT scale that incorporates all three asymmetries.