Automated investment management: Comparing the design and performance of international robo‐managers

• Published date: 01 September 2022
• Author: Nils Helms,Reinhold Hölscher,Matthias Nelde
• Date: 01 September 2022
• DOI: http://doi.org/10.1111/eufm.12333

Eur Financ Manag. 2022;28:1028–1078.1028

|

wileyonlinelibrary.com/journal/eufm

DOI: 10.1111/eufm.12333

ORIGINAL ARTICLE

Automated investment management:

Comparing the design and performance of

international robo‐managers

Nils Helms |Reinhold Hölscher |Matthias Nelde

Department of Business Studies and

Economics, Technical University

Kaiserslautern, Kaiserslautern, Germany

Correspondence

Nils Helms, Erwin‐Schrödinger‐Straße

52, 67663 Kaiserslautern, Germany.

Email: helms@wiwi.uni-kl.de

Abstract

Robo‐managers offer automated asset management;

however, their overall performance is highly

debated. We analyze 15 robo‐managers from

Germany, the United States and the United King-

dom by conducting a comprehensive qualitative and

quantitative study. The qualitative comparison

shows considerable differences between the various

robo‐managers, not only across but also within

countries. The quantitative evaluation utilizes dif-

ferent measures to evaluate the performance of the

robo‐manager sample. Our results indicate that

each country has one particularly favourable robo‐

manager. Furthermore, we find that the costs and

characteristics of rebalancing measures have only a

small effect on performance.

KEYWORDS

automated investment management, digitalization, performance,

robo‐advisor, robo‐manager, qualitative and quantitative study

JEL CLASSIFICATION

G2, G11, G29

EUROPEAN

FINANCIAL MANAGEMENT

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and

reproduction in any medium, provided the original work is properly cited.

© 2021 The Authors. European Financial Management published by John Wiley & Sons Ltd.

We thank John A. Doukas, the editor, and an anonymous referee of European Financial Management as the paper has

enormously benefited from their comments.

1|INTRODUCTION AND LITERATURE REVIEW

The question of how to successfully invest in financial markets is a recurring and highly

debated topic for consumers, firms and academics alike. In light of the digitalization taking

place in the financial industry, new technologies to support and improve investment decisions

have emerged (Aziz et al., 2021). Robo‐advisors are at the forefront of those trends, promising

successful investment decisions by machines rather than human advisors. These companies

rely on algorithmic trading strategies that can be made automatically, without hardly any

human decisions in the process. Although each robo‐advisor has its own platform and stra-

tegies, a few commonalities concerning their procedures are observable. First and foremost,

robo‐advisors ask potential investors several standardized questions regarding their investment

history, experience with specific instruments and asset classes, investment horizon and risk

preferences. The use of algorithms allows the robo‐advisor to generate an optimized portfolio

structure based on the answers to the questions. The portfolio offered includes different pro-

portions of equities, bonds, real estate and commodities depending on the investor's risk pre-

ferences. Concerning the investment horizon, robo‐advisors usually adjust the portfolio weights

on an ongoing basis, because of the natural drift of the asset structure from the originally

assigned weights over time. The underlying idea is that this standardized and automated advice

can be offered at a reasonable price while delivering good financial performance.

However, the usefulness of robo‐advisors is discussed controversially. On the one hand, the

suppliers and some industry experts welcome the idea of a simple, transparent and inexpensive

way to gain financial profits, especially in times of low interest rates. On the other hand,

investor demand remains below supplier expectations (Hock, 2020). The objective of this paper

is to examine robo‐advisors from different countries from a qualitative and quantitative per-

spective based on distinct, well‐defined criteria.

The current state of research on robo‐advisory can be summarized as follows: even though

robo‐advisors have been around for over a decade, academic research is still scarce. Apart from

a large number of rather popular scientific and journalistic articles, few significant publications

address the legal (Fein, 2015; Ji, 2017), regulatory (Baker & Dellaert, 2018; Grischuk, 2017) and

financial issues. Henceforth, we will analyze the current state of research, focusing on the

financial sector. The majority of the articles in the financial area either focus on the optimal

design patterns of robo‐advisor services or aim to study changes in customer behaviour,

portfolio composition or financial performance through robo‐advisor use.

The work of Jung et al. (2018) falls into the first category, investigating the optimal design of

a robo‐advisor for risk‐averse investors. Glaser et al. (2019) analyze how a robo‐advisor should

be designed to be optimally adapted to customer needs. Further, Rossi and Utkus (2020)

examine investors’needs and wants in terms of financial advice, clarifying why individual

investors choose to hire financial advisors in the first place and the implications for the design

of robo‐advisers.

In the second category, for example, is the research of Gulden (2019), who examines the

factors that drive the acceptance and use of robo‐advisors. Scheurle (2017) considers how an

investor's portfolio composition changes through the use of a robo‐advisor. Further, D'Acunto

et al. (2019) analyze how the characteristics and portfolios of robo‐advisor users differ from

those of nonusers. Additionally, Rossi and Utkus (2021) study how former self‐directed in-

vestors change their portfolios after using a hybrid robo‐advisor service and which investors

benefit the most. Belanche et al. (2019) investigate the characteristics that determine if a person

is willing to use a robo‐advisor. D'Hondt et al. (2019), in turn, investigate which users would

HELMS ET AL.EUROPEAN

FINANCIAL MANAGEMENT

|

1029

benefit the most from a robo‐advisor. Cheng et al. (2019) analyze the role of trust in the

acceptance of robo‐advisor services.

Only a very small number of studies conducted a comprehensive comparison of certain

features of robo‐advisor services. Tertilt and Scholz (2019) focus on the determination of an

investor's risk profile. They examine the depth and quality of the risk‐profiling tools used by

robo‐advisors. Nelde (2019), in turn, compares the investment behaviours of private investors,

the market portfolio and robo‐advisors in Germany. Nelde's (2019) research aims to assess the

scale of rationality in the overall decision‐making process based on a comprehensive evaluation

of the German robo‐advisor market. Puhle (2019) carries out a comparison of German robo‐

advisors with regard to realized performance. However, the study does not rely on observed

asset structures; instead, the asset allocation is estimated with regression analysis. Additionally,

the performance figures are not independently collected or calculated but, rather, copied from

the external finance platform Brokervergleich.de, whose reliability remains questionable. In a

recent study, D'Acunto and Rossi (2021) propose a taxonomy of robo‐advisors based on a

literature review, market observations and clearly defined dimensions, enabling the segmen-

tation of business models according to their degree of personalization, involvement, discretion

and human interaction.

The literature clearly shows that the robo‐advisor industry lacks global qualitative and

quantitative performance analysis. We are not aware of any scientific article that carries out a

comprehensive survey of international robo‐advisors. Furthermore, we are also elaborating on

the decision‐making principles applied by robo‐advisors in regard to portfolio construction and

composition, as well as the differences observed between the individual providers. These ob-

servations are linked to whether and how the investment performance of robo‐advisors can be

compared. If comparability is ensured, it should be possible to study and identify whose per-

formance in a real market setting is superior to that of competing robo‐advisors.

The remainder of this paper is structured as follows. In Section 2, we first develop a

catalogue of the criteria that serves as a basis for our qualitative analysis. We then compare

multiple robo‐advisors from three different countries. Substantial differences in robo‐advisor

characteristics are further analyzed. On the basis of the findings of this qualitative analysis,

Section 3presents the methodological foundation for the quantitative analysis, using econo-

metrics (bootstrapping) to establish quantitative comparability. The limitations of the analysis

are also discussed. On the basis of these considerations, we carry out a comprehensive per-

formance analysis of the robo‐advisors in Section 4. The effects of rebalancing and fees and

their relation to the overall results are explicitly addressed. Lastly, Section 5presents the key

results.

2|QUALITATIVE ANALYSIS OF THE ROBO‐MANAGER

PLATFORMS

The starting point of this analysis is the assessment of several robo‐advisor platforms, focusing

on the evaluation of qualitative characteristics. First, we will explain the difference between

robo‐advisors and robo‐managers, as well as the selection criteria of our study sample. Second,

we will develop the qualitative characteristics and, third, analyze them in the context of the

selected companies.

1030

|

EUROPEAN

FINANCIAL MANAGEMENT

HELMS ET AL.