Abstract
Mutual fund is a popular investment vehicle for investors. Investors usually judge fund manager performance relative to target benchmarks. Fund managers, on the other hand, are interested in knowing how/why they perform well or poorly relative to their peers in different aspects of fund management as well. To acquire more insights about this issue and design a comprehensive performance measure, fund management function is conceptualised as a three-stage production process. To assess overall and stage-level performance, a network data envelopment analysis model is developed. The stage-level processes are deemed to operate under two different environmental conditions—levels of risk exposure. Operation under different levels of risk exposure is modelled through conditions imposed on the intermediate measures. A new index proposed to assess linkage performance is demonstrated empirically to improve discriminatory power of performance. Further applications of the proposed model are discussed.
1. Introduction
U.S. mutual fund (MF) industry is the largest in the world with nearly $16 trillion in assets at year-end 2015. Fifty-two per cent of these MF assets are equity funds with bond funds (21%), money market funds (18%) and hybrid funds (9%) making up the rest (Investment Company Institute, 2016). Equity, bond and hybrid MFs are typical long-term investments whereas money market funds provide shortterm yields. Interest in MFs is generally widespread across households, business and institutional investors. MF is an attractive financial instrument for households because MFs are managed by financial experts and owning shares in a MF is a cost effective way of diversification. In the U.S., approximately eighty-nine per cent of MF assets are held by households. However, as large number of companies offers a wide choice, MF selection is not an easy task. Rating agencies such as Morningstar give guidance to this end by providing star ratings based on risk-adjusted return. Nevertheless, it is in the best interest of investors, whether big or small, to have a general overview of fund performance in addition to risk-adjusted return. Fund managers, on the other hand, will be interested in knowing why/how they perform well or poorly overall as well as in different aspects of fund management compared to their peers. Our aim is to investigate this issue through a novel multistage data envelopment analysis model.
