Is Proprietary Trading Detrimental to Retail Investors?

• Author: FALKO FECHT,ANDREAS HACKETHAL,YIGITCAN KARABULUT
• DOI: http://doi.org/10.1111/jofi.12609
• Published date: 01 June 2018
• Date: 01 June 2018

THE JOURNAL OF FINANCE •VOL. LXXIII, NO. 3 •JUNE 2018

Is Proprietary Trading Detrimental to Retail

Investors?

FALKO FECHT, ANDREAS HACKETHAL, and YIGITCAN KARABULUT∗

ABSTRACT

We study the conflict of interest that arises when a universal bank conducts propri-

etary trading alongside its retail banking services. Our data set contains the stock

holdings of every German bank and those of their corresponding retail clients. We in-

vestigate (i) whether banks sell stocks from their proprietary portfolios to their retail

customers, (ii) whether those stocks subsequently underperform, and (iii) whether re-

tail customers of banks engaging in proprietary trading earn lower portfolio returns

than their peers. We present affirmative evidence for all three questions and conclude

that proprietary trading can, in fact, be detrimental to retail investors.

INCREASING LIFE EXPECTANCY AND DEMOGRAPHIC changes have forced households

in many developed countries to substantially extend their own provisions for

retirement financing. These changes imply that households must become more

actively involved in managing their personal finances. However, given their

limited financial literacy (Lusardi and Mitchell (2007), van Rooij, Lusardi, and

Alessie (2011)), retail investors often depend on professional financial advice

when making investment decisions.1

∗Falko Fecht is at the Department of Finance, Frankfurt School of Finance and Management.

Andreas Hackethal is at Goethe University Frankfurt. Yigitcan Karabulut is at Rotterdam School

of Management, Erasmus University, and CEPR. The usual disclaimer applies. We are indebted

to Kenneth Singleton (the Editor), an anonymous Associate Editor, and two anonymous refer-

ees for their helpful suggestions and comments. We would like to thank Arnoud Boot; Joerg

Breitung; Martin Brown; Dell’Arricia Giovanni; Degryse Hans; Egemen Genc; Reint Gropp;

Luigi Guiso; Michalis Haliassos; Harry Huizinga; Roman Inderst; Stephan Jank; Janet Mitchell;

Morkoetter Stefan; Madureira Leonarda; Jonathan Reuter; Jean-Charles Rochet; Klaus Schaeck;

and Gregory Udell; as well as seminar participants at the 2011 NBER Summer Institute on House-

hold Finance, the European Conference on Household Finance, the Banking Models and Banking

Structure Conference at Banque de France, the 4 Swiss Winter Conference on Financial Interme-

diation, the Household Heterogeneity and Household Finance Conference at the Federal Reserve

Bank of Cleveland, the 47 Annual Conference on Bank Structure and Competition at the Fed-

eral Reserve Bank of Chicago, the 12 Symposium on Finance, Banking and Insurance at KIT, the

2011 Annual Meeting of the German Finance Association, the 2011 Annual Meeting of the Ger-

man Economic Association, Deutsche Bundesbank, Frankfurt School of Finance and Management,

University of Leipzig, University of Kiel, Tilburg University,and Bangor Business School for their

helpful comments and suggestions. The authors declare that they have no relevant or material

financial interests that relate to the research described in this paper.

1Recent survey evidence suggests that the use of financial advice is pervasive. For instance, a

survey conducted in the European Union shows that 80% of respondents report seeking professional

DOI: 10.1111/jofi.12609

1323

1324 The Journal of Finance R

Banks—particularly universal banks—should be well suited to provide such

guidance to their retail customers. Playing a key role in many financial mar-

kets, banks dispose of and process information that is crucial to the provision

of relevant financial advice that supports retail investors in their investment

decisions. For example, compared with retail investors, banks are likely to

have superior stock-selection and market-timing abilities, as they have more

and better resources (e.g., technology, human capital) to collect and process

information than do individual investors. Furthermore, banks can collect and

process information at lower cost because they can exploit economies of scale

in portfolio management and information acquisition. In addition, banks can

obtain superior information about firms through their close lending and other

direct business relationships (Acharya and Johnson (2007), Ivashina and Sun

(2011)), thus, economies of scope should exist, particularly between banks’

proprietary trading and portfolio management units and their retail banking

services.

However, financial advice provided to retail investors resembles a “credence

good”: because of their limited financial literacy, households typically cannot

assess the quality of financial products or services neither ex ante nor ex post.

Several potential agency problems, such as misselling, can thus arise between

banks and their retail customers.2Given their diverse lines of business, such

as proprietary trading and retail banking, under the same roof, these agency

problems might be particularly severe for universal banks. For example, when

actively trading on their own account, to avoid a price impact, banks may face

incentives to direct their retail customers to those stocks that they sell from

their portfolios, which may not necessarily suit the needs of their customers.

Thus, a question arises as to whether retail customers benefit from their banks’

greater ability to provide guidance based on market knowledge from propri-

etary trading, or whether banks face incentives to exploit their uninformed

retail customers, as banks trade stocks on their own account.3

To address this question, we study a unique data set provided by Deutsche

Bundesbank that comprises the individual stock investments of each German

bank and those of its retail customers. Using a series of panel regressions,

we first examine the relationship between the stock investments of banks and

those of their retail customers at the individual security level and find that

when a bank sells a given stock from its proprietary trading portfolio, its retail

customers tend to buy the same stock in that period. The direction of stock

advice before purchasing investment products (Chater,Huck, and Inderst (2010)). In a U.S. survey,

Hung et al. (2008) document that 73% of all investors consider financial advice when making

investment decisions.

2Misselling is generally understood as the practice of misdirecting customers to buy a product

that does not suit their specific needs (Inderst and Ottaviani (2009)). This practice also includes

selling customers financial instruments with an inferior risk-return profile.

3Recent studies that address this issue, such as Hackethal, Haliassos, and Jappelli (2012),

Karabulut (2013), and Foerster et al. (2017), find evidence that the involvement of bank advisors

negatively affects individual portfolio performance, despite having some benefits. However, these

studies use data from only one advisory firm.

Is Proprietary Trading Detrimental to Retail Investors? 1325

flows between bank and customer portfolios is negative only when banks sell

stocks, not when they buy stocks. Further tests confirm that this finding is not

a mere artifact of banks’ market-making activities or retail investors’ herding

behavior and show that it is robust to using alternative empirical specifications,

variable definitions, sampling restrictions, and econometric methods.

Tostrengthen the causal interpretation of our results, we exploit a reasonably

exogenous shock that affected the stock investments of a subset of the sample

banks over a certain period: Banks that obtained state assistance from the

German Special Fund for Financial Market Stabilization were required to shut

down most or all own-account trading and/or to substantially reduce their risky

asset holdings during the assistance period. We exploit this exogenous sales

pressure to isolate the causal link between bank and customer portfolios and

obtain qualitatively and quantitatively similar results. We further explore the

potential motives underlying this effect and provide several pieces of empirical

evidence suggesting that banks do so to avoid a price impact. For example,

the tendency of banks to sell stocks to their retail customers is significantly

stronger for illiquid stocks (as measured by a higher Amihud ratio) in periods

when financial markets are relatively more illiquid (i.e., after the collapse

of Lehman Brothers in September 2008) and when the stock positions that

banks liquidate are larger, all of which are associated with larger adverse price

impacts.

We next analyze whether the observed behavior of banks has any nega-

tive implications for the portfolio performance of their retail customers. We

find that retail customers experience trading losses from those transactions in

which they purchase stocks that their banks sell from their proprietary trading

portfolio and that these losses increase in the trading profits of the bank in the

same stock. We further find that stocks sold by banks directly to their cus-

tomers not only significantly underperform stocks that are held or purchased

by banks but also underperform retail customers’ other stock investments in

their portfolios. These results suggest a potential conflict of interest between

the proprietary trading activities of banks and their retail banking divisions. It

is possible, however,that retailcustomers could still benefit from banks’ greater

ability to provide guidance, given banks’ ability to draw on market knowledge

from proprietary trading, which could compensate for the observed losses of

retail customers. To address this possibility, we compare the stock portfolio

performance of customers of banks with proprietary trading units with that

of customers of banks without proprietary trading units. Using several perfor-

mance indicators, we find that the stock portfolios of customers of banks with

proprietary trading desks significantly underperform those of retail customers

of banks without proprietary trading desks. Overall, our results reveal a poten-

tial conflict of interest arising from combining proprietary trading and retail

banking under one roof, which seems to negatively affect the stock portfolio

performance of retail investors.

These findings have important implications for the ongoing discussion about

splitting up universal banks and separating their investment banking activity

(in particular, proprietary trading) from their commercial and retail banking