Advisor compensation: Which clients know and
how do they pay?

Somer G. Andersona, Martin C. Seayb,*, Kyoung Tae Kimc,
Derek R. Lawsond

aMaryville University, 650 Maryville University Drive, St. Louis, MO, 63141
bKansas State University, 318 Justin Hall, Manhattan KS, 66506
cUniversity of Alabama, 312 Adams Hall, Tuscaloosa AL, 35487
dKansas State University, 303 Justin Hall, Manhattan KS, 66506

Abstract

Using the 2015 National Financial Capability Study Investor Survey, this study uses agency theory
to inform an exploration of characteristics associated with knowing how one’s financial advisor/broker
is compensated. This study further examines how individuals who do know the compensation method
choose between financial advisors with different compensation models. Proposed changes in the
financial advising regulatory landscape, as well as the pending changes to the CFP Board Standards
of Professional Conduct, brings greater emphasis on understanding consumers’ advisor compensation
preferences. Primary results indicate that clients who place importance on fees, that are more
knowledgeable about diversification, and that perform background checks are more likely to know
compensation methods. A follow-up analysis reveals distinct differences between individuals that
used each model, but also provide some mixed results, indicating that clients may not fully understand
the compensation paid to their advisors. Discussion and implications related to these results are
provided. © 2018 Academy of Financial Services. All rights reserved.

JEL classification: D10; G20; G41

Keywords: Agency theory; Compensation models; Financial advisor; Investor Survey; National Financial
Capability Study

* Corresponding author. Tel.: �1-785-532-1486; fax: �1-785-532-5505.
E-mail address: mseay@ksu.edu

Financial Services Review 27 (2018) 231-255

1057-0810/18/$ – see front matter © 2018 Academy of Financial Services. All rights reserved.



1. Introduction

Given the shift away from defined benefit (DB) plans to defined contribution (DC) plans,
U. S. workers have been tasked with an increased responsibility to plan and prepare for their
retirement. This task requires individuals to undergo complex financial decisions that involve
significant uncertainty, yet individuals often have difficulty making complex decisions under
uncertainly (Tversky and Kahneman, 1974). As such, the shift in retirement funding respon-
sibility has helped to ignite the growth of the financial planning profession (Seay, Kim, and
Heckman, 2016). A significant body of research exists investigating the relationship between
financial literacy, help-seeking behavior (Calcagno and Monticone, 2015; Collins, 2012;
Molton, Loibl, Samak, and Collins, 2013), and seeking professional advice (Finke, Huston,
and Winchester, 2011; Hanna, 2011; Lachance and Tang, 2012; Seay et al., 2016). However,
little research has examined how consumers choose between financial advisors, and more
specifically, how consumers choose between advisors with different compensation methods.

The Certified Financial Planner Board of Standards (CFP Board) provides three categories
of compensation models under which CFP certificants can operate: (a) commission only, (b)
fee-only, and (c) combined commission-and-fee. While advisors compensated by commis-
sion are paid based on specific product recommendations, fees are typically charged based
on assets under management (AUM), through an annual retainer, or through hourly or project
based planning charges. Further, commission-based advisors have historically been held to
a suitability standard, simply requiring that a product be suitable for a client given their
financial situation. On the other hand, fee advisors, operating in the Registered Investment
Advisor (RIA) space, have typically been held to a fiduciary standard that requires recom-
mendations be in the client’s best interest. The differences in how compensation has been
derived, and more specifically different standards of care, have often been used to differen-
tiate these advisors. Notably, commission-and-fee advisors operate in the hybrid space,
operating as a fiduciary when providing advice through an RIA and under the suitability
standard when operating through a broker-dealer.

Pending changes in the financial planning regulatory landscape brings greater emphasis on
understanding consumers’ advisor compensation preferences. While the Department of
Labor’s (DoL) Conflict of Interest rule (Federal Register, 2017) was vacated by the courts in
June of 2018, many financial services firms made structural changes in how their financial
professionals engage with clients. Further, much of the spirit of the DOL rule is now found
in the Security and Exchange Commission’s proposed rule Regulation Best Interest (Reg BI).
The current Reg BI proposal would require financial advisors previously held to the
suitability standard to act in the best interest of their customers at the time a recommendation
is being made, without placing the advisors, or their firms, interests ahead of the client.

Within the nonregulatory space, CFP professionals will be required to act as fiduciaries to
their clients when providing financial advice beginning October 1, 2019. This is a clear
expansion of the previous standard of conduct, which only required CFP professionals to act
in a fiduciary capacity when providing financial planning services. Financial advice is
defined very broadly, stated as any client communication that “would reasonably be viewed
as a suggestion that the Client take or refrain from taking a particular course of action with
respect to” an array of financial planning areas including, but not limited to, anytime the CFP

232 M. Seay / Financial Services Review 27 (2018) 231-255



professional has discretion of client assets, development and/or implementation of a financial
plan, and investment advice and strategies (CFP Board, 2017, p. 16–17).

In light of these pending changes, most, if not all, CFP professionals will engage in
fiduciary work with their clients. Similarly, under the proposed Reg BI the standards of care
between all financial professionals, although significantly different, will continue to converge
around putting the client’s interest first. This change would, in many ways, level the financial
planning client services market from a client care perspective, making it imperative to gain
an understanding of what factors clients consider when they differentiate between advisors
that use differing financial planning compensation models.

Given this backdrop, the current study primarily explores two central research questions.
First, what are the characteristics of clients that know their advisors’ compensation method?
Secondly, of those individuals who did know how their financial advisor was compensated,
what are the differences in clients that use commission-only, fee-only, and combined
commission-and-fee advisors? For empirical analyses, this study used two linked datasets,
the 2015 National Financial Capability Study (NFCS) and its supplementary dataset, the
2015 NFCS Investor Survey. This study provides insights into the client characteristics
that are associated with knowledge of financial advisor compensation and the use of a
certain type of financial advisor based on their compensation method. Furthermore, it
provides financial advisors insights into their clients’ characteristics and why they
choose to utilize their services.

2. Literature review

The use of professional financial advisors in the United States has been growing steadily
over the last several decades. The percentage of U.S. households that reported using a
financial advisor grew from 21% in 1998 (Hanna, 2011) to 29% in 2012 (Seay et al., 2016).
Previous research has established an understanding of who seeks financial advice and why
they do so, and the value that households receive from it. Agency theory provides a lens for
informing hypotheses and directional expectations related to an individual knowing or not
knowing their financial advisor’s compensation method.

2.1. Who seeks financial advice?

The characteristics of financial professionals that consumers are looking for has been
examined in previous studies. Bae and Sandager (1997) found that consumers were predom-
inantly interested in using a financial advisor for services related to retirement planning, tax
planning, and investment management. Furthermore, Bae and Sandager (1997) found that
consumers with both the desire for the personal assurance and confidence that a trusted
financial professional can provide and with lower levels of financial knowledge sought a
financial professional. The authors also reviewed what was important to consumers in
selecting their financial professional, showing that confidentiality, objectivity, competence,
honesty, and the ability to communicate were very important factors to prospective clients.
Prospective clients found access to the financial professional and compensation transparency

233M. Seay / Financial Services Review 27 (2018) 231-255



to be less important. Finally, education and certification were found as critical components
towards selection of a financial professional, with 92% of clients reporting that a financial
professional with the CFP designation is important (Bae and Sandager, 1997).

Much research has also examined the characteristics of consumers that sought financial
advice. Individuals that sought financial help from a financial professional typically had
higher levels of financial risk tolerance (Grable and Joo, 2001; Hanna, 2011; Joo and Grable,
2001; Marsden, Zick, and Mayer, 2011; Robb, Babiarz, and Woodyard, 2012), financial
satisfaction (Grable and Joo, 2001), wealth, educational attainment (Finke et al., 2011;
Hanna, 2011; Lachance and Tang, 2012), and income (Hanna, 2011). These individuals were
also older (Finke et al., 2011; Grable and Joo, 2001; Marsden et al., 2011) than their
counterparts and displayed more positive financial behaviors (e.g., regularly saved, elimi-
nated credit card debt, or planned for big purchases; Grable and Joo, 2001; Joo and Grable,
2001). Another important positive predictor in financial help-seeking is an individual’s
willingness to trust (Lachance and Tang, 2012; Martin, Finke, and Gibson, 2014). Finally,
individuals with more proactive attitudes and women were more likely to seek financial help
from a professional (Joo and Grable, 2001).

2.2. The value of financial advice

Some of the best evidence of the value of financial advice based on real client data were
provided by Marsden et al. (2011). They found that meeting with a financial advisor
encouraged more prudent retirement planning activities. Specifically, individuals were more
likely to have established long-term financial goals, calculated financial needs in retirement,
higher levels of more supplemental savings, diversified retirement portfolios, saved more
regularly, and spent more time learning about financial topics. Measurable long-term impacts
of meeting with an advisor were increased confidence in the client’s ability to retire and
larger emergency fund savings.

Montmarquette and Viennot-Briot (2015) found that having a financial advisor for a
minimum of four years had a positive and significant impact on client wellbeing. After
controlling for many variables (socio-economic, demographic, and attitudinal variables),
individuals using a financial advisor had significantly higher levels of financial assets. This
increase in asset levels was not purely because of investment related advice, as evidence
suggested that clients receiving financial advice created better savings disciplines.

Blanchett and Kaplan (2013) explored the value of financial advice through the concept
of gamma. They evaluated the advantages of using a financial advisor as evidenced through:
(1) appropriate asset allocation strategies; (2) creating dynamic withdrawal strategies; (3)
understanding the appropriate use of guaranteed income strategies (e.g., annuities); (4)
choosing between tax-efficient strategies; and (5) building portfolios that account for risks
faced by retirees. Although theoretical and not driven by actual client data, improved
financial decision making associated with financial advisors generated increased returns
of approximately 1.6% annually within the context of retirement planning for retirees.
Grable and Chatterjee (2014) expanded upon this research to introduce Zeta, a function
of gamma and alpha (portfolio return beyond expected return within a portfolio), which

234 M. Seay / Financial Services Review 27 (2018) 231-255



focuses on the value of advice in reducing wealth volatility during turbulent markets.
They found that clients who met with a financial advisor before the Great Recession
experienced significantly less negative returns and wealth volatility, adjusted for the risk
taken within the portfolio.

2.3. Understanding compensation method

Previous literature provides little information in regard to which clients would be expected
to know how they are paying their financial advisors. In response to this gap, one study, the
2014 Advisor Impact Economics of Loyalty Study conducted by Nielsen, collected infor-
mation about knowledge of financial professional compensation methods from a sample of
1,229 respondents who worked with a financial professional and possessed investable assets
of at least $50,000 (Raskie, Herbison, and Martin, 2017). Descriptive results from that study
suggested that 68% of respondents knew how their financial professional was compensated.
However, not all respondents that knew the method being used to pay their financial
professional indicated that they knew the amount which the professional was being com-
pensated (Raskie et al., 2017). In fact, the results suggested that the more commission that
was involved with the compensation method, the less likely it was that the respondent knew
how much they were paying their financial professional. Specifically, 82% of respondents
who were paying fee-only, 79% who were working with a commission-and-fee professional,
and 67% of those who were paying commission-only indicated that they knew how much
their professionals were being compensated for their services.

Raskie et al. (2017) used the Nielsen study to investigate the relationship between
knowledge of compensation method and knowledge of the financial professional’s invest-
ment process and found evidence of a link between the two. Compared with those who did
not know the compensation method, those who used a fee-only professional had the highest
odds of knowing the investment process; with those using fee-and-commission coming in
next highest, and those using commission-only having the lowest odds of knowing the
investment process. They also investigated the association between compensation method
and having an understanding of the investment process and found that client understanding
of the investment process used by the financial service professional was affected by the
method of compensation. Consumers who knew how their advisor was compensated had a
better understanding of the professional’s investment process than those who did not. Use of
the commission-only method was not found to be a significant predictor of understanding the
investment process; however, using the fee-only and commission-and-fee methods were
associated with higher levels of understanding as compared with not knowing the compen-
sation method at all.

Although sparse, previous literature (Raskie et al., 2017) suggests that there is a relation-
ship between some individual client characteristics and the likelihood of knowing the
compensation method used to pay a financial professional. Agency Theory can provide a lens
that can be used to explore this relationship. Clients and financial advisors have an agency
relationship where the consumer (client) compensates the agent (financial advisor) to per-
form services on the consumer’s behalf (Jensen and Meckling, 1976; Raskie et al., 2017).
Because the financial advisor is presumably more informed about the services that are

235M. Seay / Financial Services Review 27 (2018) 231-255



delegated by the client, information asymmetry exists between the financial advisor and the
client (Jensen and Meckling, 1976). Information asymmetry creates agency costs that can
be incurred by both the client and the advisor in the form of monitoring and bonding costs.
The client can incur monitoring costs when he or she feels they must observe the behavior
of the advisor to make certain the advisor is protecting the client’s best interests. The advisor
may incur bonding costs, such as obtaining professional designations and licenses, to signal
to the client that they have the knowledge and expertise to perform the services in the client’s
best interests. There should be an inverse relationship between monitoring and bonding costs
where the client can reduce monitoring costs when the advisor has signaled an investment in
bonding costs. In this vein, clients who are sensitive to agency costs would be expected to
understand the method they were using to compensate their financial advisors.

Given the gap in research related to knowledge of how financial professionals are
compensated, the primary purpose of the current study is to explore the relationship between
client characteristics and knowledge of financial advisor compensation method at the mul-
tivariate level. Specifically, the associations between the following client characteristics: (1)
client socio-demographic profiles, (2) financial characteristics, (3) financial knowledge, (4)
attitudes toward advisor preference, and (5) compensation; and the dependent variable,
knowing or not knowing the advisor compensation method, were investigated. Informed by
agency theory, the following hypotheses were explored in the primary analysis.

H1: Financial knowledge will be positively associated with knowing the advisor com-
pensation method.

The financial knowledge block of characteristics explored in this study includes objective
knowledge about compound interest, inflation, bonds, mortgages, and diversification. Addi-
tionally, this block includes indicators of both subjective financial knowledge and confidence
in applying that knowledge. It is expected that there would be less information asymmetry
in the relationship between clients with high financial knowledge and their advisors. There-
fore, those with higher financial knowledge may better understand the advisor compensation
methods that are available and should choose the method that best fits their desired level of
agency costs.

H2: Attitudes about advisor preferences will be positively associated with knowing the
advisor compensation method.

Agreement with the importance of using an advisor to free up time, prevent losses, and
improve investment performance were the attitudes investigated in the advisor preference
block of characteristics. Additionally, this block includes whether or not the client performed
a background or registration check, and the degree of importance of an advisor’s professional
designation. It is expected that clients who had strong advisor preferences would be sensitive
to agency costs and would seek to understand the compensation method used by the advisor
so they could gauge the level of bonding costs incurred by the advisor and adjust monitoring
costs based on those preferences. For instance, agency costs would appear to be very
important to a client who went to the trouble to perform a background check on a financial
advisor. Therefore, it would be expected that this client would also seek to understand the
compensation method they were using to pay their advisor.

236 M. Seay / Financial Services Review 27 (2018) 231-255



H3: Compensation beliefs will be positively associated with knowing the advisor com-
pensation method.

The compensation beliefs investigated in this study related to agreement with the impact
on services received from an advisor who received commission on trading activity or
received selling incentives. Additionally, the degree of importance that a respondent placed
on advisor fees was also included in this block of characteristics. It is expected that clients
who had strong beliefs in this area would be more likely to know how their financial advisors
were paid.

2.4. Client preferences in compensation models

Beyond understanding how an advisor is paid, previous literature has provided evidence
that there is a relationship between some client characteristics and the compensation method
those clients used to pay their advisors (Raskie et al., 2017). However, more research is
required to understand how clients choose between compensation models. A first step in
understanding how client preferences influence advisor selection was provided by Seay et al.
(2017), who analyzed bivariate statistics of clients that used each compensation model.
Results indicated that commission-only clients were mostly older (65 and over), relatively
financially knowledgeable and confident in their knowledge about finances, and were
comfortable with their decision-making as it related to investments. These clients reported
placing less importance on professional designations and using a financial advisor to free
up time. Similarly, they were not as concerned about fraud, were less prone to believe
that commissions influenced advice received, and believed that the fees they paid were
reasonable.

Commission-and-fee clients were relatively younger (specifically, age 18 –34), more
racially diverse and were full-time workers (Seay et al., 2017). They reported the highest
levels of concern about fraud and were the most likely group of clients to perform a
background check on their advisor, yet they believed the fees they paid were reasonable.
These clients had the lowest financial knowledge but reported the highest subjective knowl-
edge of all client groups. They were confident in their investment decision-making, were the
most optimistic about the future prospects of the financial markets, and used an advisor to
free up time.

Lastly, fee-only clients were the largest group of clients, had high incomes, were finan-
cially knowledgeable but were not confident in their investment decision-making and also
reported the lowest subjective financial knowledge of all client groups (Seay et al., 2017).
Fee-only clients were interested in working with an advisor with a professional designation,
wanted to delegate decisions to their advisor, and utilized an advisor to free up their time.
Finally, fee-only clients had strong beliefs about commissions influencing advice received.

Given the basis for compensation model preference provided by Seay et al. (2017), a
secondary analysis within the current study explored the association between client charac-
teristics and the use of either a commission-only, fee-only, or combined commission-and-fee
financial advisor for those individuals who did report knowing the compensation model used
to pay their financial advisor. Given the heterogeneity of services provided within each

237M. Seay / Financial Services Review 27 (2018) 231-255



compensation method, it is difficult to make consistent a priori hypotheses related to
characteristics associated with the use of each method. Consequently, this secondary analysis
is purely exploratory in nature, and there are no directional expectations for which clients
might be using which method.

3. Method

3.1. Data and sample selection

This study utilized data from the 2015 NFCS, a nationally representative survey of
American adults administered between June and October 2015, and the follow-up 2015
NFCS Investor Survey that was administered in July 2015 to a subset of respondents who had
indicated ownership of nonretirement investments. The larger 2015 NFCS state-by-state data
were collected from an online survey of 27,564 adults in the United States, with approxi-
mately 500 observations per state plus the District of Columbia. The 2015 study included an
oversampling for large states, resulting in 1,000 respondents each from California, Illinois,
New York, and Texas. Respondents were recruited from established online panels and
selected using a nonprobability quota sampling methodology with criteria based on income,
gender, ethnicity, age, and education level.

To explore the associations between client attributes and financial advisor compensation
models, the analytic sample was drawn from the 2,000 individuals who indicated owning
nonretirement investments and completed the 2015 NFCS Investor Survey. From this subset,
the sample was then limited to 955 respondents who indicated having a specific broker or
advisor. Observations with missing data for any variables in the regression model were
listwise deleted, leaving an analytic sample of 750 for the primary analysis. Of this 750, 142
respondents indicated that they did not know the compensation method used to pay their
advisor; while 608 did know the compensation method. The secondary analysis used only the
sample of 608 individuals who knew their compensation method. To obtain further details on
this dataset, see Mottola and Kieffer (2017).

3.2. Measurement of variables

The dependent variable, compensation method, was based on the following question from
the 2015 NFCS Investor Survey: “Which of the following types of fees do you pay for your
nonretirement investment accounts” with the response choices of: (a) a commission on
trades, (b) a percentage of the total value of assets managed, and (c) a fixed monthly or
annual fee. No information related to hourly planning fees was collected. Responses were
organized into four mutually exclusive categories: commission-only (n � 143) if the respon-
dent selected only option a; fee-only (n � 291) if the respondent selected option b and/or c
but not option a, commission-and-fee (n � 174) if the respondent selected option a along with
another option, and do not know (n � 142) if the respondent indicated that they did not know
the compensation method. For the primary analysis, we created a binary indicator of knowing
compensation method and coded it as 1 if the compensation method was located in the

238 M. Seay / Financial Services Review 27 (2018) 231-255



commission-only, fee-only, or commission-and-fee categories and 0 if the compensation
method was located in the do not know category. The secondary analysis, “compensation
method” was coded as 0 for commission-only, 1 for fee-only, and 2 for commission-and-fee.

Independent variables were organized into five blocks of client characteristics: (a) socio-
demographic, (2) financial, (3) basic financial knowledge, (4) advisor preference, and (5)
compensation beliefs. These variables and their measurements are discussed below. Addi-
tional measurement information is provided in the Appendix.

3.2.1. Socio-demographic characteristics
The NFCS includes a variety of demographic and socio-economic variables. However,

because of sample size limitations, some groups were condensed to ensure appropriate cell
sizes for the multinomial logit analysis. Socio-demographic variables included in the analysis
include age, gender, race, education, marital status, and employment status.

3.2.2. Financial characteristics
The 2015 NFCS and 2015 NFCS Investor Survey include both objective and subjective

assessments of an individual’s financial situation. Objective assessments include income and
approximate value of all investments in nonretirement accounts. Subjective measures include
respondents’ self-reported risk tolerance measured on a 10-point Likert-type scale, with
higher scores indicating greater willingness to take risks.

3.2.3. Financial knowledge
The 2015 NFCS asked five questions about the fundamental concepts of personal finance

drawn from a 5-question scale created and used by Lusardi and Mitchell (2009, 2011). For
the purposes of this analysis, binary variables were created to signify whether the respondent
correctly answered each question related to compound interest, inflation, bonds, mortgages,
and diversification. Additionally, respondents were asked to assess their subjective financial
knowledge and confidence in financial ability according to a 7-point Likert-type scale, with
higher scores indicating greater self-assessed financial knowledge and confidence in financial
ability.

3.2.4. Advisor preference characteristics
The NFCS 2015 Investor Survey provided several measures related to financial advisor

use and preference. Respondents were asked to assess the importance of each of the
following reasons for using a financial advisor: to free up my time, to help avoid losses, and
to improve investment performance. For this analysis, responses were coded as binary
variables with very important � 1, otherwise coded as 0. Respondents were also asked about
the importance of financial advisor designations (very important � 1, otherwise � 0) and
performance of background checks (yes � 1, otherwise � 0).

3.2.5. Advisor compensation characteristics
The 2015 NFCS Investor Survey asked a question that measured the respondents’ attitudes

toward sales-based compensation structures. Responses were coded as categorical variables

239M. Seay / Financial Services Review 27 (2018) 231-255



with the following levels: would not affect at all, would affect somewhat, and would affect
a great deal. Additionally, the survey contained a question that measured respondents’
agreement with fee importance when opening nonretirement investment accounts, assessed
on a 10-point Likert-type scale.

3.3. Empirical models

A binary logit was used to analyze the effect of the selected predictor variables on the
likelihood of knowing advisor compensation method. The model for the probabilities of
clients knowing the compensation method is:

Prob �Yi � j�Xi� � Pij �
exp� x�i�j�

1 � �
k�1

J exp� x�i�k�
, j � 0, 1.

We estimate probabilities for being in the J�1 type for clients with characteristics, Xi for the
group indicator J, 0 � do not know compensation method and 1 � know compensation method.

To analyze the effect of the selected predictor variables on the likelihood of utilizing the
three compensation methods, a multinomial logit model was used. The model for the
probabilities of clients using each compensation is:

Prob �Yi � j�Xi� � Pij �
exp� x�i�j�

1 � �
k�1

J exp� x�i�k�
, j � 0, 1, 2.

We estimate probabilities for being in the J�1 type for clients with characteristics, Xi. For
the group indicator J, 0 � commission-only, 1 � fee-only, and 2 � commission-and-fee.

Multicollinearity among the predictor variables was tested using variance inflation factor
(VIF) scores. Although the NFCS provides weighting information for the full sample,
weights are not provided for the subset represented by the NFCS investor survey. Conse-
quently, results were not weighted for the purpose of this research.

4. Results

4.1. Descriptive results

Tables 1 through 5 display the descriptive statistics of financial planning clients overall
and by compensation model.

4.1.1. Socio-demographic characteristics
Table 1 contains the descriptive statistics for socio-demographic characteristics. The

majority of the overall sample respondents were age 55 or older (60%); however, the
ages of the respondents using the commission-and-fee compensation method were more
evenly split between under 55 (52%) and over 55 (48%). Overall, the sample respondents

240 M. Seay / Financial Services Review 27 (2018) 231-255



were male (54%), White (83%), had at least some college education (92%), and married
(71%). There were some demographic differences noted between the respondents who
did not know their compensation and the rest of the sample groups. There were more
women (62%), there was more diversity (22% non-White), and more of these respon-
dents indicated that they had no college education (15%). Most of the overall sample
respondents were either employed full-time (39%) or retired (34%); however, the
majority of the respondents using the commission-and-fee compensation method were
employed full-time (59%). Respondents that knew their compensation method also had
a higher rate of being employed full time (41%) than respondents who did not know their
compensation method (29%).

4.1.2. Financial characteristics
Table 2 displays the descriptive statistics for financial characteristics. Most participants

reported a household income of $50,000 or more (85%) and a total nonretirement account
value of less than $500,000 (75%). Participants in the sample reported relatively high risk
tolerance (M � 6.19 out of 10); however, respondents using the commission-and-fee method
had the highest risk tolerance (M � 6.78 out of 10) compared with respondents using the

Table 1 Socio-demographic characteristics, 2015 National Financial Capability Study (NFCS) investor
survey

Compensation model

Full
sample
(N � 750)

Do not know
compensation
(N � 142)

Know
compensation
(N � 608)

Commission
only
(N � 143)

Fee only
(N � 291)

Commission-
and-fee
(N � 174)

Age
18 to 34 11.20% 11.97% 11.02% 6.29% 9.97% 16.67%
35 to 44 12.00% 9.15% 12.66% 12.59% 11.68% 14.37%
45 to 54 16.80% 14.79% 17.27% 17.48% 15.12% 20.69%
55 to 64 25.07% 25.35% 25.00% 23.08% 25.09% 26.44%
65 and older 34.93% 38.73% 34.05% 40.56% 38.14% 21.84%

Gender
Male 54.13% 38.03% 57.89% 61.54% 52.58% 63.79%
Female 45.87% 61.97% 42.11% 38.46% 47.42% 36.21%

Race
White 82.80% 78.17% 83.88% 89.51% 83.85% 79.31%
Non-White 17.20% 21.83% 16.12% 10.49% 16.15% 20.69%

Education
High school or less 8.00% 14.79% 6.41% 5.59% 6.87% 6.32%
Some college 24.40% 26.76% 23.85% 20.28% 24.05% 26.44%
College 67.60% 58.45% 69.74% 74.13% 69.07% 67.24%

Marital status
Married 70.67% 68.31% 71.22% 70.63% 72.85% 68.97%
Not married 29.33% 31.69% 28.78% 29.37% 27.15% 31.03%

Employment status
Full-time 39.07% 28.87% 41.45% 34.27% 34.71% 58.62%
Self-employed 10.00% 6.34% 10.86% 13.29% 10.31% 9.77%
Retired 33.87% 36.62% 33.22% 36.36% 38.83% 21.26%
Other 17.07% 28.17% 14.47% 16.08% 16.15% 10.34%

241M. Seay / Financial Services Review 27 (2018) 231-255



commission-only or fee-only methods (M � 6.24 and 6.12, respectively). Moreover, respon-
dents who did not know their compensation had the lowest risk tolerance (M � 5.56).

4.1.3. Financial knowledge characteristics
Descriptive statistics for basic financial knowledge characteristics are located in Table 3.

Most respondents in the sample correctly answered objective financial knowledge questions
related to compound interest (91%), inflation (83%), bonds (55%), mortgages (91%), and
diversification (79%); however, respondents using the commission-and-fee method had a

Table 2 Financial characteristics, 2015 National Financial Capability Study (NFCS) investor survey

Compensation model

Full
sample
(N � 750)

Do not know
compensation
(N � 142)

Know
compensation
(N � 608)

Commission
only
(N � 143)

Fee only
(N � 291)

Commission-
and-fee
(N � 174)

Household income
Less than $35,000 7.60% 11.27% 6.74% 11.19% 6.87% 2.87%
$35,000 to $49,999 7.07% 9.86% 6.41% 6.29% 6.53% 6.32%
$50,000 to $74,999 24.27% 28.87% 23.19% 24.48% 22.34% 23.56%
$75,000 to $99,999 21.07% 20.42% 21.22% 20.28% 19.93% 24.14%
$100,000 or more 40.00% 29.58% 42.43% 37.76% 44.33% 43.10%

Total value of non-retirement accounts
$0 to $49,999 20.93% 27.46% 19.41% 24.48% 19.24% 15.52%
$50,000 to $99,999 13.07% 12.17% 12.17% 10.49% 11.68% 14.37%
$100,000 to $249,999 20.53% 19.90% 19.90% 16.08% 20.62% 21.84%
$250,000 to $499,999 20.13% 21.22% 21.22% 19.58% 21.65% 21.84%
$500,000 to $999,999 13.73% 9.15% 14.80% 13.99% 15.46% 14.37%
$1,000,000 or more 11.60% 7.75% 12.50% 15.38% 11.34% 12.07%

Risk tolerance 6.19 5.56 6.34 6.24 6.12 6.78
Mean (SD) (2.05) (2.14) (2.01) (2.01) (0.20) (2.00)

Table 3 Basic Financial Knowledge, 2015 National Financial Capability Study (NFCS) investor survey

Compensation model

Full
sample
(N � 750)

Do not know
compensation
(N � 142)

Know
compensation
(N � 608)

Commission
only
(N � 143)

Fee only
(N � 291)

Commission-
and-fee
(N � 174)

Financial knowledge (correct %)
Compound interest 91.07% 85.21% 92.43% 90.91% 93.81% 91.38%
Inflation 83.33% 82.39% 83.55% 90.91% 84.19% 76.44%
Bond 55.20% 41.55% 58.39% 66.43% 55.67% 56.32%
Mortgage 91.07% 86.62% 92.11% 95.10% 91.41% 90.80%
Diversification 79.33% 64.79% 82.73% 84.62% 85.22% 77.01%
Subjective financial

knowledge, mean (SD)
5.83 5.51 6.00 5.96 5.87 5.90

(0.86) (0.99) (0.81) (0.79) (0.83) (0.82)
Confidence in financial

ability, mean (SD)
6.45 6.25 6.50 6.47 6.59 6.39

(0.92) (1.19) (0.84) (0.92) (0.71) (0.94)

242 M. Seay / Financial Services Review 27 (2018) 231-255



lower percentage of correct answers on each question compared with respondents using the
other two compensation methods. Further, respondents who did not know their compensation
also had a lower percentage of correct answers on each question compared with those who
did know their compensation. In terms of self-assessed financial knowledge measures,
respondents rated themselves at the higher-end of the subjective financial knowledge and
confidence in financial ability scales (M � 5.83 and M � 6.45 out of 7, respectively).
However, those paying commission-and-fee reported lower confidence in financial ability
(M � 6.39) than those using the other two methods. Those who did not know their
compensation reported the lowest subjective financial knowledge (M � 5.51) and confidence
of all groups (M � 6.25).

4.1.4. Advisor preference characteristics
Table 4 contains descriptive statistics related to advisor preference characteristics. Most

respondents found it very important to use their advisor to help them prevent losses (79%)
and improve investment performance (84%). Only 30% of respondents found it very
important to use an advisor to help them free up time, while 35% of those who did not know
their compensation and only 20% of respondents using the commission-only method found
this factor to be very important. Concerning advisor selection, 59% of the respondents felt
that it was very important for their financial advisor to have possession of a professional
designation; but 63% of those who did not know their compensation method and only 50%
of respondents using the commission-only method found professional designations to be very
important. Only 25% of the overall sample respondents reported that they had performed a
background, registration, or license check on a financial advisor; however, 39% of respon-
dents using the commission-and-fee method and only 8% of those who did not know their
compensation had done so.

Table 4 Advisor preference, 2015 National Financial Capability Study (NFCS) investor survey

Compensation model

Full
sample
(N � 750)

Do not know
compensation
(N � 142)

Know
compensation
(N � 608)

Commission
only
(N � 143)

Fee only
(N � 291)

Commission-
and-fee
(N � 174)

Use advisor to free up time
Very important 29.60% 34.51% 28.45% 19.58% 30.93% 31.61%
Not very important 70.40% 65.49% 71.55% 80.42% 69.07% 68.39%

Use advisor to prevent losses
Very important 79.47% 81.69% 78.95% 80.42% 81.44% 73.56%
Not very important 20.53% 18.31% 21.05% 19.58% 18.56% 26.44%

Use advisor to improve investment performance
Very important 84.40% 84.51% 84.38% 81.12% 86.94% 82.76%
Not very important 15.60% 15.49% 15.63% 18.88% 13.06% 17.24%

Performed advisor background/
registration check

25.07% 7.75% 29.11% 23.08% 26.46% 38.51%

Importance of professional designation
Very important 59.20% 62.68% 58.39% 49.65% 62.20% 59.20%
Not very important 40.80% 37.32% 41.61% 50.35% 37.80% 40.80%

243M. Seay / Financial Services Review 27 (2018) 231-255



4.1.5. Advisor compensation characteristics
In regard to the dependent variables, the majority of the sample (81%) indicated that they

knew the compensation method. Of those that did know their compensation method, almost
half of the respondents reported paying compensation on a fee-only basis (48%), with the rest
of the sample reporting compensation based on commission only (23%) or both commission
and fees (29%). Table 5 displays descriptive statistics related to advisor compensation
characteristics. Respondents overwhelmingly felt that advisor compensation based on sales
factors such as commissions on trading activity and selling incentives would at least
somewhat affect advice received. In terms of fee importance, respondents felt that fees and
pricing structure related to nonretirement accounts were important factors when opening an
account (M � 7.9 out of 10), but those who did not know their compensation method found
it to be less important (M � 7.2).

4.2 Regression results—primary analysis

To analyze the effect of the selected predictor variables on the likelihood of knowing the
compensation method, a binary logit model was used. Results from the binary logistic
regression analysis are presented in Table 6. In regard to model fit, Pseudo R2 was 0.195.
Variance inflation factors ranged from 1.1 to 4.3, indicating that multicollinearity among the
predictor variables is not a concern in this analysis. Holding all else equal, variables from
each block of characteristics except for the financial characteristics block were found to be
significant predictors of knowing the compensation method (p � 0.05).

Hypothesis one expected a positive relationship between financial knowledge and know-
ing the advisor compensation method; however, knowledge of diversification was the only
variable in the financial knowledge block that was significantly associated with knowing the
compensation method. Specifically, those answering the question correctly had 2.5 times
higher odds of knowing the compensation method than those who answered the question

Table 5 Compensation beliefs, 2015 National Financial Capability Study (NFCS) investor survey

Compensation model

Full
sample
(N � 750)

Do not know
compensation
(N � 142)

Know
compensation
(N � 608)

Commission
only
(N � 143)

Fee only
(N � 291)

Commission-
and-fee
(N � 174)

Commission on trading activity impacts advice received
Would not affect at all 21.60% 16.90% 22.70% 32.17% 19.24% 20.69%
Would affect somewhat 42.40% 47.89% 41.12% 41.26% 39.18% 44.25%
Would affect a great deal 36.00% 35.21% 36.18% 26.57% 41.58% 35.06%

Selling incentive impacts advice received
Would not affect at all 14.13% 11.97% 14.64% 16.08% 15.12% 12.64%
Would affect somewhat 38.53% 46.48% 36.68% 36.36% 34.71% 40.23%
Would affect a great deal 47.33% 41.55% 48.68% 47.55% 50.17% 47.13%

Fee importance, mean (SD) 7.89 7.15 8.07 8.06 7.99 8.20
(1.93) (2.33) (1.78) (1.62) (1.88) (1.72)

244 M. Seay / Financial Services Review 27 (2018) 231-255



Table 6 Binary logistic regression results estimating probability of knowing compensation method (N � 750)

Characteristic B SE B Odds ratio p value

Intercept �.3281 1.210 0.007**
Socio-demographic characteristics

Age (ref group: 65 and older)
18 to 34 0.114 0.472 1.121 0.809
35 to 44 0.450 0.470 1.569 0.338
45 to 54 0.412 0.401 1.510 0.304
55 to 64 0.175 0.300 1.192 0.559

Gender (ref group: Female)
Male 0.570 0.227 1.767 0.012*

Race (ref group: Non-White)
White 0.421 0.289 1.524 0.145

Education (ref group: College)
High school or less �0.780 0.351 0.459 0.027*
Some college �0.002 0.261 0.999 0.996

Marital status (ref group: Not married)
Married �0.182 0.254 0.834 0.474

Employment status (ref group: Full-time)
Self-employed 0.280 0.445 1.322 0.531
Retired �0.196 0.345 0.822 0.576
Other �0.623 0.312 0.534 0.044*

Financial characteristics
Household income (ref group: $100,000 or more)

Less than $35,000 �0.452 0.482 0.637 0.349
$35,000 to $49,999 �0.561 0.447 0.571 0.210
$50,000 to $74,999 �0.234 0.305 0.790 0.439
$75,000 to $99,999 �0.210 0.312 0.811 0.503

Total value of non-retirement accounts
(ref group: $1,000,000 or more)

$0 to $49,999 0.145 0.504 1.156 0.774
$50,000 to $99,999 �0.378 0.494 0.685 0.444
$100,000 to $249,999 �0.029 0.456 0.971 0.949
$250,000 to $499,999 0.334 0.463 1.399 0.467
$500,000 to $999,999 0.192 0.486 1.211 0.694

Risk tolerance 0.033 0.058 1.034 0.563
Basic financial knowledge

Compound interest 0.603 0.352 1.827 0.087
Inflation �0.158 0.326 0.854 0.628
Bond 0.274 0.233 1.316 0.239
Mortgage 0.147 0.374 1.158 0.695
Diversification 0.934 0.273 2.545 0.001**
Subjective financial knowledge 0.210 0.139 1.234 0.131
Confidence in financial ability 0.058 0.118 1.060 0.622

Advisor preference
Very important to use advisor to free up time �0.279 0.244 0.756 0.251
Very important to use advisor to prevent losses �0.050 0.313 0.951 0.874
Very important to use advisor to improve

investment performance
0.106 0.333 1.111 0.751

Performed advisor background/registration check 1.392 0.372 4.021 0.000**
Professional designation very important �0.286 0.231 0.751 0.214

Compensation beliefs
Commission on trading activity impacts advice received

(ref group: Would not affect at all)
Would affect somewhat �0.232 0.376 0.793 0.537
Would affect a great deal �0.385 0.416 0.680 0.355

Selling incentive impacts advice received
(ref group: Would not affect at all)

Would affect somewhat �0.486 0.443 0.615 0.273
Would affect a great deal �0.231 0.469 0.793 0.621

Fee importance 0.243 0.056 1.275 �.0001***
Pseudo R2 0.195

†p � .10, *p � .05, **p � .01, ***p � .001.

245M. Seay / Financial Services Review 27 (2018) 231-255



incorrectly. Answering questions correctly that were related to compound interest, inflation,
bonds, and mortgages was not associated with knowledge of the compensation method.
Regarding the subjective financial knowledge variables, self-assessed financial knowledge
and confidence in financial ability, neither client characteristic was associated with knowing
the compensation method.

Attitudes about advisor preferences were expected to be positively related to knowing the
compensation method in hypothesis two. Only one variable in this block of characteristics
was significantly related to knowing the compensation method. Specifically, those who had
performed an advisor background or registration check on their financial advisor had four
times higher odds of knowing the compensation method compared with those who did not
perform these checks. Clients who indicated that it was very important to use a financial
advisor to free up time, prevent losses, and improve investment performance and those who
found professional designations to be very important did not have significantly different odds
of knowing the compensation method than those who did not find these variables to be very
important.

In Hypothesis 3, it was expected that compensation beliefs would be positively associated
with knowing the advisor compensation method. One item in the compensation beliefs block
was associated with knowing the compensation method. Specifically, an increase in the belief
that fees were important when opening nonretirement investment accounts was associated
with 28% higher odds of knowing the compensation method. Believing that commission on
trading activity or selling incentives would impact advice received was not significantly
associated with knowledge of the compensation method.

Several variables in the socio-demographic block of characteristics were associated with
knowing the compensation method. Men had 1.8 times higher odds of knowing the com-
pensation method in comparison to women. Education was also a significant predictor of
knowing the compensation method. Compared with having a college education, those with
an education of high school or less had lower odds of knowing the compensation method;
however, there was no significant difference between completing a college education and
only having some college. In terms of employment status, there were no significant differ-
ences between clients who classified themselves as full-time versus those who considered
themselves self-employed or retired. However, clients with an employment classification of
“other” had significantly lower odds of knowing the compensation method. Age, race,
marital status, household income, the total value of nonretirement accounts, and risk toler-
ance were not found to be significant predictors of knowing the compensation method.

4.3. Regression results—secondary analysis

As a follow-up analysis, a multinomial logit model was used to analyze the effect of the
selected predictor variables on the likelihood of utilizing the three compensation methods
using only the sample of respondents who reported knowing the compensation method used
to pay their advisor. Results from the multivariate logistic regression analysis are presented
in Table 7. In regard to model fit, Pseudo R2 was 0.12. Holding all else equal, variables from
all five blocks of client characteristics were found to be significant predictors of compen-

246 M. Seay / Financial Services Review 27 (2018) 231-255



T
ab

le
7

M
ul

tin
om

ia
l

lo
gi

st
ic

re
gr

es
si

on
re

su
lts

es
tim

at
in

g
pr

ob
ab

ili
ty

of
us

in
g

co
m

pe
ns

at
io

n
m

od
el

(N
�

60
8)

C
om

pe
ns

at
io

n
m

od
el

Fe
e-

on
ly

vs
.

co
m

m
is

si
on

-o
nl

y
C

om
m

is
si

on
-a

nd
-f

ee
vs

.
co

m
m

is
si

on
-o

nl
y

C
om

m
is

si
on

-a
nd

-f
ee

vs
.

fe
e-

on
ly

B
SE

B
O

dd
s

ra
tio

p
va

lu
e

B
SE

B
O

dd
s

ra
tio

p
va

lu
e

B
SE

B
O

dd
s

ra
tio

p
va

lu
e

In
te

rc
ep

t
1.

50
4

1.
49

7
0.

31
5

2.
40

0
1.

63
1

0.
14

1
0.

89
6

1.
36

4
0.

51
1

So
ci

o-
de

m
og

ra
ph

ic
ch

ar
ac

te
ri

st
ic

s
A

ge
(r

ef
gr

ou
p:

65
an

d
ol

de
r)

18
to

34
0.

77
5

0.
54

2
2.

17
0

0.
15

3
1.

22
8

0.
58

8
3.

41
3

0.
03

7*
0.

45
3

0.
45

9
1.

57
3

0.
32

4
35

to
44

(0
.1

20
)

0.
45

4
0.

88
7

0.
79

1
0.

01
5

0.
50

4
1.

01
5

0.
97

6
0.

13
5

0.
43

6
1.

14
5

0.
75

6
45

to
54

(0
.0

78
)

0.
40

3
0.

92
5

0.
84

6
0.

08
8

0.
44

9
1.

09
2

0.
84

4
0.

16
7

0.
39

0
1.

18
1

0.
66

9
55

to
64

0.
15

1
0.

31
4

1.
16

3
0.

63
0

0.
39

7
0.

36
4

1.
48

7
0.

27
6

0.
24

6
0.

31
2

1.
27

8
0.

43
1

G
en

de
r

(r
ef

gr
ou

p:
Fe

m
al

e)
M

al
e

(0
.4

00
)

0.
23

9
0.

67
0

0.
09

3†
0.

03
4

0.
26

8
1.

03
5

0.
89

9
0.

43
4

0.
22

0
1.

54
4

0.
04

9*
R

ac
e

(r
ef

gr
ou

p:
N

on
-W

hi
te

)
W

hi
te

(0
.5

59
)

0.
36

2
0.

57
2

0.
12

3
(0

.7
02

)
0.

38
2

0.
49

6
0.

06
6†

(0
.1

43
)

0.
28

1
0.

86
7

0.
61

0
E

du
ca

tio
n

(r
ef

gr
ou

p:
C

ol
le

ge
)

H
ig

h
sc

ho
ol

or
le

ss
0.

26
8

0.
49

0
1.

30
7

0.
58

5
0.

41
6

0.
54

6
1.

51
6

0.
44

6
0.

14
9

0.
44

7
1.

16
0

0.
73

9
So

m
e

co
lle

ge
0.

31
8

0.
29

3
1.

37
4

0.
27

7
0.

64
9

0.
32

4
1.

91
3

0.
04

5*
0.

33
1

0.
26

0
1.

39
2

0.
20

3
M

ar
ita

l
st

at
us

(r
ef

gr
ou

p:
N

ot
m

ar
ri

ed
)

M
ar

ri
ed

(0
.1

92
)

0.
27

7
0.

82
5

0.
48

7
(0

.3
98

)
0.

30
3

0.
67

2
0.

18
9

(0
.2

06
)

0.
24

9
0.

81
4

0.
40

9
E

m
pl

oy
m

en
t

st
at

us
(r

ef
gr

ou
p:

Fu
ll-

tim
e)

Se
lf

-e
m

pl
oy

ed
(0

.2
67

)
0.

39
4

0.
76

6
0.

49
8

(0
.7

61
)

0.
42

9
0.

46
7

0.
07

6†
(0

.4
94

)
0.

37
0

0.
61

0
0.

18
2

R
et

ir
ed

0.
40

4
0.

36
6

1.
49

7
0.

27
0

(0
.5

75
)

0.
41

3
0.

56
3

0.
16

4
(0

.9
78

)
0.

35
1

0.
37

6
0.

00
5*

*
O

th
er

0.
00

3
0.

35
8

1.
00

3
0.

99
3

(0
.7

60
)

0.
41

3
0.

46
8

0.
06

6†
(0

.7
63

)
0.

34
5

0.
46

7
0.

02
7*

Fi
na

nc
ia

l
ch

ar
ac

te
ri

st
ic

s
H

ou
se

ho
ld

in
co

m
e

(r
ef

gr
ou

p:
$1

00
,0

00
or

m
or

e)
L

es
s

th
an

$3
5,

00
0

(1
.2

41
)

0.
51

9
0.

28
9

0.
01

7*
(1

.9
26

)
0.

69
3

0.
14

6
0.

00
5*

*
(0

.6
85

)
0.

62
9

0.
50

4
0.

27
6

$3
5,

00
0

to
$4

9,
99

9
(0

.8
78

)
0.

53
1

0.
41

6
0.

09
8†

(0
.6

44
)

0.
59

5
0.

52
5

0.
27

9
0.

23
4

0.
49

1
1.

26
3

0.
63

4
$5

0,
00

0
to

$7
4,

99
9

(0
.8

94
)

0.
32

8
0.

40
9

0.
00

6*
*

(0
.6

07
)

0.
36

5
0.

54
5

0.
09

6†
0.

28
7

0.
30

0
1.

33
3

0.
33

8
$7

5,
00

0
to

$9
9,

99
9

(0
.6

51
)

0.
32

5
0.

52
2

0.
04

5*
(0

.3
36

)
0.

35
5

0.
71

5
0.

34
4

0.
31

5
0.

28
8

1.
37

1
0.

27
4

T
ot

al
va

lu
e

of
no

n-
re

tir
em

en
t

ac
co

un
ts

(r
ef

gr
ou

p:
$1

,0
00

,0
00

or
m

or
e)

$0
to

$4
9,

99
9

0.
28

2
0.

46
9

1.
32

6
0.

54
8

(0
.7

02
)

0.
54

3
0.

49
6

0.
19

6
(0

.9
84

)
0.

47
7

0.
37

4
0.

03
9*

$5
0,

00
0

to
$9

9,
99

9
0.

46
5

0.
49

1
1.

59
2

0.
34

3
0.

05
2

0.
54

4
1.

05
3

0.
92

4
(0

.4
13

)
0.

46
0

0.
66

1
0.

36
9

$1
00

,0
00

to
$2

49
,9

99
0.

78
7

0.
44

6
2.

19
6

0.
07

8†
0.

52
9

0.
49

6
1.

69
7

0.
28

6
(0

.2
58

)
0.

41
6

0.
77

3
0.

53
5

$2
50

,0
00

to
$4

99
,9

99
0.

58
6

0.
41

3
1.

79
7

0.
15

6
0.

39
4

0.
46

1
1.

48
3

0.
39

3
(0

.1
92

)
0.

39
7

0.
82

5
0.

62
8

$5
00

,0
00

to
$9

99
,9

99
0.

70
3

0.
42

9
2.

01
9

0.
10

1
0.

44
5

0.
47

7
1.

56
0

0.
35

1
(0

.2
58

)
0.

41
3

0.
77

3
0.

53
2

R
is

k
to

le
ra

nc
e

0.
01

0
0.

06
1

1.
01

0
0.

87
1

0.
07

6
0.

07
1

1.
07

9
0.

28
1

0.
06

6
0.

05
9

1.
06

8
0.

26
6

(c
on

ti
nu

ed
on

ne
xt

pa
ge

)

247M. Seay / Financial Services Review 27 (2018) 231-255



T
ab

le
7

(C
on

tin
ue

d)

C
om

pe
ns

at
io

n
m

od
el

Fe
e-

on
ly

vs
.

co
m

m
is

si
on

-o
nl

y
C

om
m

is
si

on
-a

nd
-f

ee
vs

.
co

m
m

is
si

on
-o

nl
y

C
om

m
is

si
on

-a
nd

-f
ee

vs
.

fe
e-

on
ly

B
SE

B
O

dd
s

ra
tio

p
va

lu
e

B
SE

B
O

dd
s

ra
tio

p
va

lu
e

B
SE

B
O

dd
s

ra
tio

p
va

lu
e

B
as

ic
fin

an
ci

al
kn

ow
le

dg
e

C
om

po
un

d
in

te
re

st
1.

00
2

0.
45

2
2.

72
3

0.
02

7*
0.

84
7

0.
49

8
2.

33
2

0.
08

9†
(0

.1
55

)
0.

42
5

0.
85

7
0.

71
6

In
fla

tio
n

(1
.0

34
)

0.
40

1
0.

35
6

0.
01

0*
(1

.1
51

)
0.

42
1

0.
31

6
0.

00
6*

*
(0

.1
17

)
0.

30
3

0.
89

0
0.

69
9

B
on

d
(0

.3
92

)
0.

24
7

0.
67

6
0.

11
3

(0
.1

77
)

0.
27

4
0.

83
8

0.
51

8
0.

21
5

0.
22

3
1.

23
9

0.
33

5
M

or
tg

ag
e

(0
.3

47
)

0.
49

5
0.

70
7

0.
48

4
0.

01
5

0.
54

2
1.

01
5

0.
97

8
0.

36
2

0.
39

0
1.

43
6

0.
35

3
D

iv
er

si
fic

at
io

n
0.

30
7

0.
34

7
1.

35
9

0.
37

7
(0

.1
45

)
0.

36
9

0.
86

5
0.

69
2

(0
.4

52
)

0.
29

5
0.

63
6

0.
12

6
Su

bj
ec

tiv
e

fin
an

ci
al

kn
ow

le
dg

e
(0

.2
27

)
0.

16
0

0.
75

8
0.

08
3†

(0
.2

10
)

0.
17

6
0.

81
0

0.
23

3
0.

06
7

0.
14

8
1.

06
9

0.
65

2
C

on
fid

en
ce

in
fin

an
ci

al
ab

ili
ty

0.
30

7
0.

15
3

1.
35

9
0.

04
5*

0.
01

8
0.

15
2

1.
01

8
0.

90
4

(0
.2

86
)

0.
14

0
0.

74
9

0.
03

9*
A

dv
is

or
pr

ef
er

en
ce

V
er

y
im

po
rt

an
t

to
us

e
ad

vi
so

r
to

fr
ee

up
tim

e
0.

61
8

0.
28

2
1.

85
5

0.
02

9*
0.

65
4

0.
31

0
1.

92
3

0.
03

5*
0.

03
6

0.
24

1
1.

03
7

0.
88

1

V
er

y
im

po
rt

an
t

to
us

e
ad

vi
so

r
to

pr
ev

en
t

lo
ss

es
(0

.3
61

)
0.

32
5

0.
69

7
0.

26
6

(0
.8

11
)

0.
34

5
0.

44
4

0.
01

9*
(0

.4
50

)
0.

28
2

0.
63

8
0.

11
1

V
er

y
im

po
rt

an
t

to
us

e
ad

vi
so

r
to

im
pr

ov
e

in
ve

st
m

en
t

pe
rf

or
m

an
ce

0.
25

3
0.

34
6

1.
28

7
0.

46
6

(0
.0

23
)

0.
37

3
0.

97
8

0.
95

2
(0

.2
75

)
0.

32
4

0.
75

9
0.

39
6

Pe
rf

or
m

ed
ad

vi
so

r
ba

ck
gr

ou
nd

/
re

gi
st

ra
tio

n
ch

ec
k

0.
18

1
0.

28
3

1.
19

9
0.

52
1

0.
30

6
0.

30
3

1.
35

8
0.

31
3

0.
12

5
0.

24
6

1.
13

3
0.

61
2

Pr
of

es
si

on
al

de
si

gn
at

io
n

ve
ry

im
po

rt
an

t
0.

48
5

0.
24

7
1.

62
5

0.
05

0*
0.

35
6

0.
27

9
1.

42
8

0.
20

1
(0

.1
29

)
0.

23
4

0.
87

9
0.

58
1

C
om

pe
ns

at
io

n
be

lie
fs

C
om

m
is

si
on

on
tr

ad
in

g
ac

tiv
ity

im
pa

ct
s

ad
vi

ce
re

ce
iv

ed
(r

ef
gr

ou
p:

W
ou

ld
no

t
af

fe
ct

at
al

l)
W

ou
ld

af
fe

ct
so

m
ew

ha
t

1.
01

4
0.

36
1

2.
75

6
0.

00
5*

*
0.

73
2

0.
39

1
2.

08
0

0.
06

1†
(0

.2
82

)
0.

37
1

0.
75

5
0.

44
8

W
ou

ld
af

fe
ct

a
gr

ea
t

de
al

1.
85

0
0.

40
4

6.
35

9
�

.0
00

1*
**

1.
17

8
0.

43
8

3.
24

8
0.

00
7*

*
(0

.6
72

)
0.

40
5

0.
51

1
0.

09
7†

Se
lli

ng
in

ce
nt

iv
e

im
pa

ct
s

ad
vi

ce
re

ce
iv

ed
(r

ef
gr

ou
p:

W
ou

ld
no

t
af

fe
ct

at
al

l)
W

ou
ld

af
fe

ct
so

m
ew

ha
t

(0
.8

06
)

0.
42

8
0.

44
7

0.
05

9†
(0

.1
53

)
0.

48
0

0.
85

8
0.

75
1

0.
65

4
0.

43
1

1.
92

3
0.

13
0

W
ou

ld
af

fe
ct

a
gr

ea
t

de
al

(1
.1

55
)

0.
44

7
0.

31
5

0.
01

0*
(0

.4
74

)
0.

50
0

0.
62

2
0.

34
3

0.
68

0
0.

45
4

1.
97

4
0.

13
4

Fe
e

im
po

rt
an

ce
(0

.1
14

)
0.

06
9

0.
89

2
0.

09
8†

(0
.0

35
)

0.
08

0
0.

96
5

0.
65

9
0.

07
9

0.
06

7
1.

08
2

0.
23

5
Ps

eu
do

R
2

0.
12

0

†p
�

.1
0,

*p
�

.0
5,

**
p

�
.0

1,
**

*p
�

.0
01

.

248 M. Seay / Financial Services Review 27 (2018) 231-255



sation method (p � 0.05). Significant results from each multinomial logit model are detailed
below.

4.3.1. Fee-only compared with commission-only compensation method
Elements of financial knowledge were found to be associated with using the fee-only

versus the commission-only compensation method, although there were confounding results.
First, those who correctly answered the compound interest question had 2.7 times higher
odds of using the fee-only than the commission-only method, and a one-unit increase in
self-assessed confidence in financial ability was associated with a 36% increase in the
odds of using the fee-only method. However, those who correctly answered the inflation
question had lower odds of using the fee-only than the commission-only method. These
divergent results were surprising, and remained present in a variety of different model
specifications.

Advisor preference and compensation beliefs were also associated with advisor compen-
sation method. As related to advisor preferences, those who found it very important to use
an advisor to free up time and those who found a professional designation to be very
important had 86% and 63% higher odds, respectively, of using the fee-only method than the
commission-only method. However, confounding results were found related to compensation
beliefs. While respondents who believed commissions affect advice received were more
likely to use a fee-only advisor, respondents who believed selling incentives impacted advice
received were less likely to use a fee-only advisor as compared with a commission-only
advisor.

Lastly, household income was found to be associated with advisor choice, as higher income
households were more likely to use a fee-only advisor. When compared with households with
income of $100,000 or more, those making less than $35,000, between $50,000 and $74,999, and
between $75,000 and $99,999 had lower odds of using the fee-only compensation method. None
of the variables in the socio-demographics block of characteristics were significant predictors of
compensation method in this model (p � 0.05).

4.3.2. Commission-and-fee compared with commission-only compensation method
One element of financial knowledge was associated with compensation method in this

model. Respondents who correctly answered the inflation question were less likely to use the
commission-and-fee method than the commission-only method. Two characteristics related
to advisor preference were significant in this model. Respondents who found it very
important to use an advisor to free up time had 92% higher odds of using the commission-
and-fee method; and those who found it very important to use an advisor to prevent losses
had lower odds of using the commission-and-fee method. Related to compensation beliefs,
compared with respondents who felt that advisors receipt of commission on trading activity
would not impact advice received, those who thought advice would be a great deal impacted
had higher odds of using the commission-and-fee method compared with the commission-
only method.

In terms of socio-demographic and financial characteristics, age, education, and household
income were associated with compensation method in this model. Compared with those aged
65 and older, respondents in the youngest age group of 18 to 34 had 3.4 times higher odds

249M. Seay / Financial Services Review 27 (2018) 231-255



of using the commission-and-fee method compared with the commission-only method.
Compared to those with a college degree, respondents with only some college had 91%
higher odds of using the commission-and-fee than the commission-only method. Households
making less than $35,000 annually had lower odds of using the commission-and-fee than the
commission-only method.

4.3.3. Commission-and-fee compared with fee-only compensation method
One characteristic of financial knowledge was associated with compensation method in

this model. A one-unit increase in self-assessed confidence in financial ability was associated
with a decrease in the odds of using the commission-and-fee method compared with the
fee-only method. In terms of financial characteristics, the total value of nonretirement
accounts was also a significant predictor of compensation method in this model. Compared
with respondents who reported having $1,000,000 or more in nonretirement accounts, those
who reported having between $0 and $49,999 were less likely to use the commission-and-fee
method than the fee-only method. Two elements of socio-demographic characteristics,
employment status and gender, were found to be associated with compensation method in
this model. Compared with respondents who were working full-time, those who identified as
being retired or other had lower odds of using the commission-and-fee compensation method
than the fee-only method. Compared to females, male respondents had higher odds of using
the commission-and-fee compensation method than the fee-only method.

5. Discussion

The primary purpose of this paper was to investigate the characteristics of individuals that
are associated with knowing the compensation method used to pay their financial advisor. Of
those respondents who did know how their financial advisors were compensated, a secondary
analysis explored the characteristics between clients who used the commission-only, fee-
only, and combined commission-and-fee methods to pay their advisors. Using data from the
2015 NFCS Investor Study, this study is one of the first attempts to explore whether clients
knew how they paid their advisors, and how clients make choices between compensation
models at the multivariate level. Although significant variation was found between clients
who knew and did not know how their advisors were compensated, and between compen-
sation models for those who did know how their advisor was paid, some results were
somewhat contradictory in nature.

Given that this study represents the first use of the NFCS Investor survey to evaluate
advisor compensation methods at the multivariate level, an evaluation of the sample quality
in relation to previous research is important. Sample descriptive results were largely con-
sistent with previous research that suggested clients who seek financial help from financial
professionals are older, have higher levels of income, are better educated, and have higher
risk tolerance (Finke et al., 2011; Grable and Joo, 2001; Hanna, 2011; Marsden et al., 2011).
However, most of the clients in this sample were male, despite previous research indicating
women were more likely to seek financial help from a professional (Joo and Grable, 2001).
Similarly, incongruences were found with Bae and Sandager’s (1997) suggestion that

250 M. Seay / Financial Services Review 27 (2018) 231-255



consumers sought a financial professional because they had lower levels of financial knowl-
edge, as respondents in this sample were very knowledgeable about compound interest,
inflation, mortgages, and diversification. However, this pattern is consistent with more recent
research indicating that financial knowledge, specifically understanding diversification and
mortgages, is positively related with the use of a financial advisor (Calcagno and Monticone,
2015; Seay et al., 2016).

Regarding subjective financial knowledge, previous research found that clients were most
likely to pay for financial advice if they believed themselves to be less financially knowl-
edgeable (Finke et al., 2011). However, clients in the sample assessed their own financial
knowledge very highly on average. In support of Bae and Sandager’s (1997) suggestion that
certification was a critical component when clients were selecting a financial professional,
most of the financial planning clients in this sample found certification to be very important.
Regarding advisor compensation, previous research that found compensation transparency to be
less important to prospective clients (Bae and Sandager, 1997) was not supported as the clients
in this sample found fees and pricing structure to be important factors when opening nonretire-
ment accounts. Overall, it appears that the NFCS Investor Survey provides a reasonable sample
of individuals that use financial advisors, but it likely is not perfectly generalizable.

Moving to the primary analysis, some evidence was generated to support all three
hypotheses related to financial knowledge, advisor preference, and compensation beliefs
blocks. Hypothesis 1 proposed that higher financial knowledge would be associated with
knowing the compensation method. This hypothesis was supported for those who correctly
answered the diversification question, but it was not supported for the other objective
financial knowledge questions nor the two subjective questions related to confidence and
ability. This limited result indicates that one’s level of financial knowledge may not inform
the level of agency costs that a client is willing to incur in the client/financial advisor
relationship, or that it may be a domain specific relationship. Specifically, diversification, as
compared with the other knowledge questions, may be measuring a higher level of invest-
ment sophistication that is more closely tied to scrutiny of investment professionals.

Hypothesis 2 proposed that clients with strong advisor preferences would be more likely
to know the compensation method used to pay their advisors. This hypothesis was supported
for those who had performed a background or registration check on an advisor. This finding
suggests that clients who are sensitive to agency costs will seek to understand how their
advisors are being paid. However, there was no significant relationship found between the
level of importance placed on professional designations and knowing the compensation
method; which indicates that either these designations are not functioning as a signal of an
investment in bonding costs or that professional designations are not a strong enough signal
of knowledge and expertise to motivate clients to adjust their monitoring costs. Preferences
about advisor use (to free up time, to prevent losses, and to improve investment performance)
were not associated with knowing the compensation method, indicating that clients who
outsourced financial decisions for these specific reasons were not taking agency costs into
account when choosing their financial advisor.

Finally, Hypothesis 3 posited that strong beliefs about compensation would be associated
with knowing the compensation method used to pay the advisor. Results did provide
evidence of a positive relationship between finding fees to be important and knowing how

251M. Seay / Financial Services Review 27 (2018) 231-255



the advisor was compensated, suggesting that clients who were sensitive to agency costs
would be more likely to gain an understanding of how their advisors were being compen-
sated. However, there were no relationships revealed between beliefs about commissions and
sales incentives and knowing compensation method.

Lastly, a follow-up exploratory analysis was conducted to determine characteristics
associated with each advisor compensation model among those that were aware. Instead of
a clear picture, results related to financial knowledge were confounding. Clients who
correctly answered the compound interest question were less likely to use a commission-only
advisor as compared with the other methods; but those who answered the inflation question
correctly preferred the commission-only method above commission-and-fee. In terms of
self-assessed financial knowledge, the more knowledgeable that a client felt about finances,
the less likely they were to use the fee-only method as compared with the commission only
method. However, this pattern did not align with confidence in financial ability, as higher
confidence was associated with a preference for the fee-only method as compared with the
commission-and-fee method. Taken together, it is hard to get a clear understanding of how
knowledge informs compensation model choice. It should be noted that these results may be
due to the wide variation in services that are provided by different advisors operating within
each compensation model. This could be an indication that there is something going on
beyond compensation model, potentially quality of the financial advisor, which is driving the
advisor choice and confounding the analysis. However, these results could also indicate that
there is confusion in the market about what each of these fee-models entails.

More consistent patterns were found around advisor use preferences. Respondents who
found it very important to use an advisor to free up time overwhelmingly preferred either
method over the commission-only method. Alternatively, clients who found it very important
to use an advisor to prevent losses preferred the commission-only method, but only when
compared with the commission-and-fee method. Regarding professional designations, clients
who thought it was very important that a financial advisor have a professional designation
preferred the fee-only to the commission-only method; but there was no significant prefer-
ence between the commission-and-fee and fee-only methods. These preferences seem to
reveal client beliefs that service level is related to compensation structure. The results echo
the patterns revealed in Seay et al. (2017), and may indicate that advisors using a fee-only
advisor are looking to outsource financial decision making, while those using commission
based advisors are looking more for a partner to facilitate their financial decisions, as
opposed to offload on to.

5.1. Limitations

Several limitations exist in this study. First, the primary analysis was limited to 750 clients who
reported owning nonretirement investments, with the follow-up analysis further limited to 608
individuals who knew the type of compensation model they were using to pay their financial
advisor. This sample restriction limits the type of client that is analyzed, and results may not be
generalizable to the entire population of financial planning clients in the United States. Addi-
tionally, the analyses were based on self-reported survey data and perhaps respondents were
confused about the compensation method they were using for their financial advisor. It is also

252 M. Seay / Financial Services Review 27 (2018) 231-255



possible that respondents were confused about the difference between commissions and incen-
tives, given the confounding results in that area. Finally, the NFCS 2015 Investor Survey did not
collect information about compensation based on hourly fees, so either clients who use this type
of compensation structure are missing from this analysis or those who use an hourly fee model
chose one of the three models available to them in the survey.

6. Conclusion

Using data from the 2015 NFCS Investor Survey, this study explored the associations
between client socio-demographic characteristics, financial and financial knowledge charac-
teristics, advisor preferences, and compensation beliefs and the likelihood of knowing the
compensation method they used to pay their financial advisor/broker. A secondary analysis
explored the relationship between these client characteristics and the use of either a
commission-only, fee-only, or commission-and-fee compensation model for their financial
advisor. Primary results revealed patterns between client characteristics and knowledge of
how their advisors were compensated. Specifically, clients with knowledge about diversifi-
cation; those who took the time and initiative to perform a background or registration check
on their advisor/broker, and those who found fees to be important had higher odds of
knowing how their financial advisors were being compensated. These results indicate that
clients with a higher level of investment sophistication and those who are sensitive to agency
costs will seek to understand how their financial advisors are compensated. In understanding
this relationship, it is important to point out that 19% of respondents did not know how their
advisor was paid. This is consistent with previous literature, and highlights the limited
information that many clients gather, or retain, in assessing the use of a financial advisor.

For clients who did know the compensation method, results revealed some relationships
between client characteristics and preferred compensation models. However, the results also
provided evidence of some misunderstanding related to advisor compensation. Overall, there
are clear differences in the services clients are expecting based on the type of compensation
method used, and given clear regulatory trends leveling the standard of care between
financial services professionals, it is important to continue to gain an understanding of what
services are important to which clients, and how clients will be able to differentiate between
financial advisors in the future.

253M. Seay / Financial Services Review 27 (2018) 231-255



References

Bae, S. C., & Sandager, J. P. (1997). What consumer look for in financial planners. Journal of Financial
Counseling and Planning, 8, 9–16.

Blanchett, D., & Kaplan, P. (2013). Alpha, beta, and now . . . gamma. The Journal of Retirement, 1, 29–45.
Calcagno, R., & Monticone, C. (2015). Financial literacy and the demand for financial advice. Journal of Banking

& Finance, 50, 363–380.
Certified Financial Planning Board of Standards (CFP Board). (2017). Code of ethics and Standards of Conduct.

(available at https://www.cfp.net/about-cfp-board/proposed-standards).
Collins, J. M. (2012). Financial advice: A substitute for financial literacy? Financial Services Review, 21, 307.
Federal Register. (2017). Rules and Regulations, 82(66), 16902–16918. (available at https://www.gpo.gov/fdsys/

pkg/FR-2017-04-07/pdf/2017-06914.pdf).
Finke, M. S., Huston, S. J., & Winchester, D. D. (2011). Financial advice: Who pays. Journal of Financial

Counseling and Planning, 22, 18–26.

Appendix 1 Description of key variables in the 2015 National Financial Capability Study (NFCS)

Variable Description

Compensation method “Which of the following types of fees do you pay for your non-retirement
investment account(s)?”

Non-retirement account value “Not including retirement accounts, have any investments in stocks, bonds,
mutual funds, or other securities?”

Objective financial knowledge
Compound Interest “Suppose you had $100 in a savings account and the interest rate was 2%

per year. After 5 years, how much do you think you would have in the
account if you left the money to grow?”

Inflation “Imagine that the interest rate on your savings account was 1% per year
and inflation was 2% per year. After 1 year, how much would you be
able to buy with the money in this account?”

Bonds “If interest rates rise, what will typically happen to bond prices?”
Mortgage “A 15-year mortgage typically requires higher monthly payments than a

30-year mortgage, but the total interest paid over the life of the loan
will be less.”

Diversification “Buying a single company’s stock usually provides a safer return than a
stock mutual fund.”

Subjective financial knowledge “On a scale from 1 to 7, where 1 means very low and 7
means very high, how would you assess your overall financial
knowledge?”

Confidence in financial ability “I am good at dealing with day-to-day financial matters, such as checking
accounts, credit and debit cards, and tracking expenses.”

Reason for advisor use “Below are some reasons that people might use a financial adviser. How
important is each of the following to you, personally?”

Importance of designation “If you were looking for a financial adviser, how important would the
person’s professional designations or certifications be in your decision
to work with that person?”

Performed advisor registration/
background check

“Have you ever checked with a state or federal regulator regarding the
background, registration, or license of a financial professional?”

Sales based compensation
impacts advice received

“How much do you think each of the following would affect the advice
that a financial adviser gives to you?”

Fee importance “How important to you were the fees and pricing structure when opening
your non-retirement investment accounts?”

254 M. Seay / Financial Services Review 27 (2018) 231-255



Grable, J. E., & Chatterjee, S. (2014). Reducing wealth volatility: The value of financial advice as measured by
zeta. Journal of Financial Planning, 27, 45–51.

Grable, J. E., & Joo. S. H. (2001). A further examination of financial help-seeking behavior. Journal of Financial
Counseling and Planning, 12, 55–73.

Hanna, S. D. (2011). The demand for financial planning services. Journal of Personal Finance, 10, 36–62.
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership

structure. Journal of financial economics, 3, 305–360.
Joo, S. H., & Grable, J. E. (2001). Factors associated with seeking and using professional retirement-planning help.

Family and Consumer Sciences Research Journal, 30, 37–63. https://doi.org/10.1177/1077727X01301002
Lachance, M. E., & Tang, N. (2012). Financial advice and trust. Financial Services Review, 21, 209–226.
Lusardi, A., & Mitchell, O. S. (2009). How Ordinary Consumers Make Complex Economic Decisions: Financial

Literacy and Retirement Readiness (No. w15350). National Bureau of Economic Research.
Lusardi, A., & Mitchell, O. S. (2011). Financial Literacy and Planning: Implications for Retirement Wellbeing

(No. w17078). National Bureau of Economic Research.
Marsden, M., Zick, C. D., & Mayer, R. N. (2011). The value of seeking financial advice. Journal of Family and

Economic Issues, 32, 625–643. http://dx.doi.org/10.1007/s10834-011-9258z
Martin, Jr., T. K., Finke, M., & Gibson, P. (2014). Race, trust and retirement decisions. Journal of Personal

Finance, 13, 62–71.
Montmarquette, C., & Viennot-Briot, N. (2015). The value of financial advice. Annals of Economics and Finance,

16, 69–94.
Mottola, G. R., & Kieffer, C. N. (2017). Understanding and using data from the National Financial Capability

Study. Family and Consumer Sciences Research Journal, 46, 31–39.
Moulton, S., Loibl, C., Samak, A., & Collins, J. (2013). Borrowing capacity and financial decisions of

low-to-moderate income first-time homebuyers. Journal of Consumer Affairs, 47, 375–403.
Raskie, S., Herbison, G., & Martin, J. (2017). Do designations attract a more knowledgeable consumer? Journal

of Financial Service Professionals, 71, 80–96.
Robb, C. A., Babiarz, P., & Woodyard, A. (2012). The demand for financial professionals’ advice: The role of

financial knowledge, satisfaction, and confidence. Financial Services Review, 21, 291–305.
Seay, M. C., Anderson, S. G., Lawson, D. R., & Kim, K. T. (2017). Identifying variation in client characteristics

between financial planning compensation models. Journal of Financial Planning, 30, 40–51.
Seay, M. C., Kim, K. T., & Heckman, S. J. (2016). Exploring the demand for retirement planning advice: The

role of financial literacy. Financial Services Review, 25, 1–20.
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty: Heuristics and biases. Science, 185, 1124–1131.

255M. Seay / Financial Services Review 27 (2018) 231-255