Risk and diversification of nonprofit revenue portfolios: Applying modern portfolio theory to nonprofit revenue management

• DOI: http://doi.org/10.1002/nml.21385
• Published date: 01 December 2019
• Author: Heng Qu
• Date: 01 December 2019

RESEARCH ARTICLE

Risk and diversification of nonprofit revenue

portfolios: Applying modern portfolio theory to

nonprofit revenue management

Heng Qu

The Bush School of Government and Public

Service, Texas A&M University, College

Station, Texas

Correspondence

Heng Qu, The Bush School of Government

and Public Service, Texas A&M University,

4220 TAMU College Station, TX

77843-4220.

Email: hqu@tamu.edu

Abstract

This study uses modern portfolio theory (MPT) to estimate

the risk of nonprofit revenue portfolios and examines to

what degree the revenue concentration measure based on

Herfindahl–Hirschman Index is associated with the portfo-

lio risk measure based on MPT. The findings suggest that

nonprofits with greater revenue concentration have lower

revenue portfolio risk in the whole sample analysis. How-

ever, it is plausible that this result is dominated by organi-

zations reliant on commercial income, which comprise

over half of the sample. In fact, when examined sepa-

rately, the relationship varies by an organization's primary

funding structure. While higher revenue concentration is

positively associated with portfolio risk for organizations

relying on donations or those without a consistent primary

funding source, it appears to associate with a lower portfo-

lio risk for commercial organizations and those relying on

government grants. This study reflects on the concept of

diversification derived from portfolio theory and calls

attention to a more nuanced approach to nonprofit revenue

strategy.

KEYWORDS

Herfindahl–Hirschman Index, modern portfolio theory, revenue

concentration, revenue diversification, risk, return

Received: 23 December 2018 Revised: 11 July 2019 Accepted: 12 July 2019

DOI: 10.1002/nml.21385

Nonprofit Management and Leadership. 2019;30:193–212. wileyonlinelibrary.com/journal/nml © 2019 Wiley Periodicals, Inc. 193

1|INTRODUCTION

Nonprofit organizations receive income from various sources: individual donations, corporate contri-

butions, foundation grants, government funding, service fees, investment income, among others.

With each revenue source comes uncertainty (Note: It is also important to consider uncertainty under

different time frames. For example, a 3-year government grant may reduce its overall uncertainty for

the grant period. However, it may not remove the shorter-term uncertainty. Unexpected events, such

as government shutdown, may affect an organization's cash flow.), thus having more diversified rev-

enue sources is often recommended as a strategy to increase the overall funding stability in case there

is a decline in any source. However, in pursuing an additional revenue stream, nonprofit organiza-

tions risk the resources that could have been spent on providing additional services or obtaining an

alternative promising revenue stream. Therefore, nonprofit organizations face the critical task of

determining a revenue mix that is sufficient and stable in supporting their missions given resource

constraints.

Scholars have long recognized this challenge in nonprofit financial management. Central to the

scholarly inquiry is whether it is more beneficial for nonprofits to have more diversified than concen-

trated revenue streams. Using the Herfindahl–Hirschman Index (HHI) as a measure of revenue diver-

sification, many empirical studies have provided evidence that relying on diversified revenue sources

is associated with decreased financial vulnerability (e.g., Chang & Tuckman, 1991; Greenlee &

Trussel, 2000; Hager, 2001) or lower revenue volatility (e.g., Carroll & Stater, 2009; Wicker,

Longley, & Breuer, 2015). In contrast, other studies suggest that revenue concentration can contribute

to organizational efficiency (Frumkin & Keating, 2011) and financial capacity growth (Chikoto &

Neely, 2014).

Different from this dichotomous view of nonprofit revenue composition, a few studies have

focused on what is and how to choose an optimal combination of revenue sources, borrowing

insights from modern portfolio theory (MPT; Kingma, 1993; Grasse, Whaley, & Ihrke, 2016; Qu,

2016). In his pioneering work, Kingma (1993) applied MPT to modeling nonprofit financial predict-

ability. According to his model, the goal of nonprofit managers is to choose the optimal combination

of revenue streams that minimizes unpredictable revenue changes for a level of expected revenue. To

achieve this goal, nonprofit managers must consider the return on each revenue stream, the variance

of the returns on each revenue stream, and the covariances between them. In particular, revenue

diversification works by carefully selecting revenue sources, the returns on which do not rise and fall

together all the time. Building on Kingma's model, subsequent research used Form 990 data to esti-

mate the return–risk-covariance properties of nonprofit revenue sources at the subsector level, and

use the information to identify theoretically optimal portfolios for arts nonprofits (Grasse et al.,

2016) and other subsectors (Qu, 2016). Both studies have shown that revenue diversification as mea-

sured by lower levels of HHI does not always correspond with theoretical efficiency.

This study examines the relationship between HHI-based revenue diversification index and MPT-

based revenue portfolio risk measure, controlling for basic organizational characteristics. Building on

the rich literature that examines the relationship between revenue diversification index and financial

health, this study explores organizational portfolio risk as an alternative financial health variable.

More specifically, it extends the limited research that applies portfolio theory to nonprofit revenue

composition. Different from Grasse et al. (2016) and Qu (2016), this study estimates the properties

of revenue sources at the organizational level (rather than subsector level) and calculates empirical

portfolio risk (rather than hypothetical optimal portfolios) using the actual information on shares of

revenue sources in total revenue for each organization in a given year. It allows us to explore

194 QU