SSRN Id3834335

UNSW
THE UNIVCRSITY Of NCW SOUTH WAL[S

SYDNEY · CANBERRA · AUSTRM IA
Law
University of New South Wales Law Research Series
From ‘Best Friends’ to ‘Best

Execution’: Rethinking Cross-
Border Payments
Dirk A. Zetzsche, Ross P. Buckley, Douglas W.

Arner and Maria Lucia Passador
[2021] UNSWLRS 45
Forthcoming Journal of Business Law
University of Luxembourg Law WPS 2021-002
Bocconi Legal Studies Research Paper No.
3834335
UNSW Law
UNSW Sydney NSW 2052 Australia
E: unswlrs@unsw.edu.au
W: http://www.law.unsw.edu.au/research/faculty-publications
AustLII: http://www.austlii.edu.au/au/journals/UNSWLRS/
SSRN: http://www.ssrn.com/link/UNSW-LEG.html
Electronic copy available at: https://ssrn.com/abstract=3834335

FROM ‘BEST FRIENDS’ TO ‘BEST EXECUTION’:
RETHINKING CROSS-BORDER PAYMENTS
DIRK A ZETZSCHE*, ROSS P BUCKLEY**, DOUGLAS W ARNER***

& MARIA LUCIA PASSADOR****
Cross-border payments suffer from slowness, poor transparency and limited

access, and resulting high costs. This article identifies the root of the problems as
being that relationships among payment institutions determine how orders are
routed. In this “best friends” environment we find little incentive for payment
institutions to put their clients’ interests first. To remedy this, we propose to apply
the best execution principle developed in the context of securities and derivatives
law to payments: financial institutions would be legally required to provide the
most advantageous order execution in terms of speed, risks and costs for their
customers. The best execution principle in payments regulation works best if paired
with (1) a fiduciary status for the service provider, (2) disclosure rules vis-à-vis
clients and regulators on the execution terms and priorities of individual
transactions, (3) market intelligence for regulators on the pre- and post-execution
environment, based on large scale data transfer from financial institutions engaged
in transactions, and (4) robust public and private enforcement. Putting the onus of
identifying and pursuing the cheapest payment route upon the financial institution
will likely promote innovations such as routing algorithms and distributed ledger
technologies for payments. Further, in light of the difficulties of implementing
multi-lateral, multi-national frameworks, this best execution principle could be
*
Professor of Law, ADA Chair in Financial Law (Inclusive Finance), and Coordinator, House of Sustainable
Governance & Markets, Faculty of Law, Economics and Finance, University of Luxembourg.
**
KPMG Law –King & Wood Mallesons Chair of Disruptive Innovation, Australian Research Council Laureate
Fellow, and Scientia Professor, UNSW Sydney.
***
Kerry Holdings Professor in Law, RGC Senior Fellow in Digital Finance and Sustainable Development and
Associate Director, HKU-Standard Chartered Foundation FinTech Academy, University of Hong Kong.
****
LL.B., LL.M., Ph.D., Post-doctoral researcher Faculty of Law, Economics and Finance, University of
Luxembourg.
This article benefitted from comments by, and discussions with, Benjamin Geva, Tres Wehrli, and Thomas
Lammers, as well as comments from participants in the Bank of International Settlements/Committee on
Payments and Market Infrastructures (CPMI) conference on “Pushing the frontiers of payments: towards faster,
cheaper, more transparent and more inclusive cross-border payments”, 18-19 March 2021.
The authors gratefully acknowledge the financial support of the Australian Research Council Laureate
Fellowship (FL200100007); the Hong Kong Research Grants Council Research Impact Fund; and the Qatar
National Strategic Research Foundation. The views herein are of the authors and not necessarily of the
Australian, Hong Kong or Qatar governments or the Australian Research Council.

[Best Execution in Cross-border Payments]
imposed unilaterally by any jurisdiction’s regulator interested in reducing costs of

cross-border payments.
Keywords: Best execution, international payments, correspondent bank.

JEL Codes: G21, G23; K22, K23.

[2021]
CONTENTS
Contents .................................................................................................................................................. 3
I Introduction .......................................................................................................................................... 4
II Rents in the Cross-border Payments Chain ......................................................................................... 6
A Correspondent Banking................................................................................................................... 6
B Closed Loop Systems ...................................................................................................................... 9
C Regulatory Approaches to Enhancing Competition and Efficiency.............................................. 10
1 Standardisation ........................................................................................................................... 10
2 Infrastructure Provision and Operations: Reducing Infrastructure Costs .................................. 10
3 Fee Regulation ........................................................................................................................... 11
III Introducing Best Execution: Learning from Securities Regulation ................................................. 11
A Best Execution: An Emerging Principle of Securities Regulation ................................................ 12
B Legal Framework of Best Execution ............................................................................................. 12
1 Fiduciary Duty basis .................................................................. Error! Bookmark not defined.
2 EU: MiFID ................................................................................................................................. 16
3 US Securities Regulation ........................................................................................................... 13
C Lessons from the Securities Context ............................................................................................. 20
IV Best Execution for Cross-border Payments? ................................................................................... 23
A Fiduciary Concepts in Payments ................................................................................................... 23
B Contrasting Cross-border Payments and Securities Transactions ................................................. 24
1 Legal Differences ....................................................................................................................... 24
2 Type of Risk ............................................................................................................................... 28
3 Transactional Differences .......................................................................................................... 28
C Implications ................................................................................................................................... 29
1 Impact of Payment-specific Legislation .................................................................................... 29
2 DLT-based Payments ................................................................................................................. 30
3 Risks and Incentives................................................................................................................... 31
V Transforming Cross-border Payments through Best Execution? ...................................................... 32
A From One to Many Links? ............................................................................................................ 32
B Drive towards DLT?...................................................................................................................... 33
VI Policy Considerations ...................................................................................................................... 34
A Institutional Coverage? ................................................................................................................. 34
B Design of a Best Execution Rule................................................................................................... 35
C Allocating Volume-based Commissions? ..................................................................................... 35
D Openness to Innovation ................................................................................................................. 36
E Minimum Number of Order Routes To Be Considered ................................................................ 36

F Facilitating Enforcement: A RegTech Approach .......................................................................... 36

VII Conclusion ...................................................................................................................................... 36
I INTRODUCTION
The Financial Stability Board (FSB), the Bank of International Settlements (BIS) and the
Committee on Capital Market Infrastructures (CPMI) identify four impediments to efficient
cross-border payments: (1) costs, (2) slowness, (3) limited access and (4) lack of transparency.1
Costs comprise “transaction fees, account fees, compliance costs, applied FX conversion rates
and fees, fees across along the payment chain, and liquidity cost for prefunding”2, and charges
for cross-border payments can be “up to 20 times those for domestic transactions”.3 Slowness
is due to poor technical integration, manual processes, and diverging legal requirements.4
Slowness increases costs as long settlement times elevate counterparty and FX risk. Issues
increase with developing countries, where interfaces are non-standard or illiquid currencies are
involved.5 Limited access is due in particular to proprietary technical systems and high barriers
to becoming a partner of any financial institution involved in cross-border payments (due to
market integrity and anti-money laundering/terrorist financing requirements). Transparency is
currently severely limited since volume, fee data and names of institutions involved are rarely
published or disclosed to regulators.6
1
Financial Stability Board, Enhancing Cross-Border Payments: Stage 1 Report to the G20 (Report, 9 April
2020) 1 <https://www.fsb.org/wp-content/uploads/P090420-1.pdf> (‘Enhancing Cross-Border Payments’);
Committee on Payments and Market Infrastructures, Cross-border Retail Payments (CPMI Papers No 173, 16
February 2018) <https://www.bis.org/cpmi/publ/d173.htm>. This time-related issue is perhaps even worse
given that ‘the lack of common communication or messaging standards across systems often hinders seamless
interoperability’: European Central Bank and Bank of Japan, Synchronised Cross-Border Payments (Report,
June 2019) 1 <https://www.ecb.europa.eu/paym/intro/publications/pdf/ecb.miptopical190604.en.pdf>. For a
general overview, see Jon Cunliffe, ‘Cross-border Payment Systems Have Been Neglected for Too Long’,
Financial Times (online, 14 July 2020) <https://www.ft.com/content/a241d7e0-e1de-4812-b214-
b350cbb7d046>.
2
Enhancing Cross-Border Payments (n 1) 1.
3
Hal S Scott and Anna Gelpern, International Finance, Transactions, Policy, and Regulation (Foundation
Press, 23rd ed, 2020) 744. For potential cost reductions in cross-border correspondent banking see: Gene Neyer
and Benjamin Geva, ‘Blockchain and Payment Systems: What are the Benefits and Costs?’ (2017) 11 Journal of
Payments Strategy & Systems 215, 218-9.
4
Morten Linnemann Bech, Yuuki Shimizu and Paul Wong, ‘The Quest for Speed in Payments’ (March 2017)
Bank for International Settlements Quarterly Review 57 <https://www.bis.org/publ/qtrpdf/r_qt1703g.htm>.
5
Whereas mainstream countries are ‘moving towards one common global standard for financial messaging,
called ISO 20022. Global adoption of this standard is accelerating with a number of high-value payment market
infrastructures already live and more planned to go live by 2023’: Chris Hadorn and Courtney Trimble, ‘A New
Standard for Payments’, KPMG (Web Page) <https://home.kpmg/xx/en/home/insights/2020/02/payments-
standard.html>.
6
On the general lack of transparency, see KPMG, Cross-Border Interbank Payments and Settlements: Emerging
Opportunities for Digital Transformation (Report, November 2018) 13-4 <https://www.mas.gov.sg/-
/media/MAS/ProjectUbin/Cross-Border-Interbank-Payments-and-Settlements.pdf>.

[2021]
The dysfunctional state of cross-border payments results in higher-than-necessary transactions

costs. In a 2021 report, McKinsey & Company put global payments revenue generated from
cross-border transactions at of $206.6 billion on an approximately $25 trillion transaction
volume.7 To put these numbers in dimension, the Bank of England noted that ‘a cross-border
payment can cost up to 10 times more than a domestic payment.’8
These costs are massive, and with contemporary technology, largely unnecessary. Such costs
retard cross-border trade and damage the efficiency of product and services markets, and thus
society. Enhancing efficiency of cross-border payments has thus become a general policy goal
of the Group of 20 and of regulatory bodies around the globe. This article contributes to that
goal by analysing to what extent the best execution principle developed in the context of
securities and derivatives brokerage law could address the shortcomings of cross-border
payments.
We argue that introducing the best execution principle could improve the current structure of
cross-border payments which predominantly rests on a system of large, globally connected
correspondent banks.9 The current system may be best characterised as one of “best friends”,
wherein the relationship among the payment institutions involved (i.e., correspondent banks or
group companies within a closed-loop system) determines which institutions the orders are
routed through. In turn, payment institutions charge their clients on a “per cost” basis. Some of
the payment institutions even benefit from volume-based rebates, kick-backs (a type of “soft
dollars” benefit), and reduced rates elsewhere in the payment chain. Overall, there is little
incentive for payment institutions to put the interests of their clients first by working to seek
the best compromise in terms of speed, costs and risks in this current “best friends”
environment.10
A “best execution” obligation would be potentially transformative: it would require payment
institutions to consider, exclusively, their clients’ interests in terms of speed, risk and costs
when choosing the route or the order to take. Based on experience with the best execution
principle in securities law, we would expect, following its adoption in payments law, the large-
scale introduction of digital routing systems to identify the offer that constitutes best execution.
Furthermore, we expect that more links between correspondent banks, new FinTech service
providers, and public payment networks (including regional integration systems) will be
established. Innovative technologies, including routing algorithms and distributed ledger
7
Philip Bruno, Olivier Denecker and Marc Niederkorn, Global Payments 2021: Transformation Amid Turbulent
Undercurrents (McKinsey & Company Report, October 2021) 6
<https://www.mckinsey.com/~/media/mckinsey/industries/financial%20services/our%20insights/the%202021%
20mckinsey%20global%20payments%20report/2021-mckinsey-global-payments-report.pdf>.
8
Bank of England, Cross-border Payments', Bank of England (Web Page, 15 June 2021)
<https://www.bankofengland.co.uk/payment-and-settlement/cross-border-payments>.
9
In general, Tara Rice, Goetz von Peter and Codruta Boar, ‘On the Global Retreat of Correspondent Banks’
(March 2020) Bank for International Settlements Quarterly Review 37, 39
<https://www.bis.org/publ/qtrpdf/r_qt2003g.pdf>; ‘New Correspondent Banking Data – the Decline Continues
at a Slower Pace’, Bank for International Settlements
<https://www.bis.org/cpmi/paysysinfo/corr_bank_data/corr_bank_data_commentary_2008.htm>.
10
Committee on Payments and Market Infrastructures, Payment Aspects of Financial Inclusion in the Fintech
Era (CPMI Paper No 191, April 2020) 46 <https://www.bis.org/cpmi/publ/d191.pdf>.

technology (DLT) systems, will most likely be implemented to identify excess liquidity in less
frequently traded currencies or currency pairs and enhance payments efficiency.
This article is structured as follows. In Part II we show how the current structure of cross-
border payments facilitates monopoly rents: whether for large correspondent banks or closed-
loop systems (such as PayPal, the payment networks of savings and loan associations, or
cooperative banks). In Part III we describe the securities law principle of “best execution” and
show how best execution works for securities brokerage. In Part IV we analyse to what extent
this principle can be transferred into payment law. In Part VI we address a number of major
policy considerations.
II RENTS IN THE CROSS-BORDER PAYMENTS CHAIN

Cross-border payments are conducted today through correspondent banking or closed-loop
systems. Awareness of the need to change this system has grown strongly of late, 11 especially
with the current COVID-19 pandemic.12
A Correspondent Banking
Correspondent banking rests on a sequence of booking transactions eventually resulting in the
booking of an amount on the payee’s account equal to the amount removed from the payer’s
account (minus fees). Yet, no money is transferred as such.
Figure 1: Correspondent Banking13
11
BIS, Nexus: a blueprint for instant cross-border payments. Joint report by the Bank for International
Settlements Innovation Hub Singapore Centre and the Monetary Authority of Singapore (MAS), 28 July 2021,
<https://www.bis.org/publ/othp39.htm#:~:text=PDF%20full%20text%C2%A0(2%2C062kb)> (“The summary
report addresses the challenges that would need to be overcome and explains how a bridging platform like
Nexus could streamline the process of linking national systems. It makes recommendations for countries that are
upgrading or rebuilding their payments technology to prepare for cross-border interoperability.“ For the latest
picture of the framework, see FSB, G20 Roadmap for Enhancing Cross-border Payments. First consolidated
progress report, 13 October 2021, < https://www.fsb.org/wp-content/uploads/P131021-1.pdf>.
12
McKinsey, Accelerating winds of change in global payments, October 1, 2020,
<https://www.mckinsey.com/industries/financial-services/our-insights/accelerating-winds-of-change-in-global-
payments#download/%2F~%2Fmedia%2Fmckinsey%2Findustries%2Ffinancial%20services%2Four%20insight
s%2Faccelerating%20winds%20of%20change%20in%20global%20payments%2F2020-mckinsey-global-
payments-report-vf.pdf%3FshouldIndex%3Dfalse>.
13
Adjusted from European Banking Federation, Guidance for Implementation of the Revised Payment Services
Directive: PSD2 Guidance (Guidance, 20 December 2019) 11 <https://www.ebf.eu/wp-
content/uploads/2019/12/EBF-PSD2-guidance-Final-December-2019.pdf>.

[2021]
Rather, the process as described in Figure 1 involves:

(1) PSP1 first debits the payer’s account the amount to be transferred;
(2) PSP1 credits a mirror account (nostro account14) held in the name of PSP2 which is kept
purely for accounting purposes;
(3) PSP1 sends to PSP2 a payment message via an electronic messaging system (e.g. SWIFT
MT 202) and announces the forthcoming payment to PSP3 (via e.g. SWIFT MT 103);
(4) PSP2 debits PSP1’s account with PSP2 (vostro account15).
Then, if no (electronic) fund transfer/payment system is involved:

(5) PSP2 credits PSP3’s account with PSP2 (loro account16);
(6) PSP2 sends a payment message to PSP3 via an electronic messaging system (e.g. SWIFT
MT202);
(7) PSP3 debits PSP2’s mirror (nostro) account with PSP3 which is kept purely for accounting
purposes;
(8) PSP3 credits payee’s account with PSP3.
Then, if an (electronic) fund transfer/payment system (e.g. FedWire, EURO1/STEP1/STEP2,

Target2, INTERAC, RBI EFT) is involved:
14
The nostro account is a bank’s foreign currency account maintained by the bank in a foreign country and in
the home currency of that country. Translated from Italian: ‘our account with you’.
15
The vostro account is the local currency account maintained by a foreign bank/branch. Translated from
Italian: ‘your account with us’.
16
The loro account is an account wherein a bank remits funds in a foreign currency to another bank for credit to
an account of a third bank.

(5) PSP2 sends a payment message to the fund transfer system (often using a proprietary
messaging standard);
(6) Settlement takes place via the fund transfer system;
(7) The fund transfer system sends a payment message to PSP3 (often using a proprietary
messaging standard);
(8) PSP3 credits the payee’s account with PSP3.
On a global scale correspondent banking requires the participation of many banks, but
correspondent banking is presently characterised by a number of problematic trends.17
First, the number of financial institutions connected in a correspondent relationship is
decreasing. Costs and risk issues are driving this trend. In a correspondent banking relationship,
funds from various payment institutions flow essentially through a few, large financial
institutions which maintain a network of correspondent banks. It is estimated that a mere 200
banks function in this network as access points for all cross-border payments.18
Regulatory factors related to market integrity, such as economic sanctions, anti-money
laundering and terrorist financing, and know-your-customer requirements, have reduced
banks’ risk appetite to be connected to countries, financial institutions, businesses and
individuals with elevated risk exposures or poor risk disclosure and information. This has led
to the dominance of a few large financial institutions in the market for correspondent banking.
Second, banks see correspondent relationships more as a service than a sales channel. They are
perceived as a necessary ancillary service rather than a product on their own. This potentially
results in under-investment in innovative technologies and in exploring new service
opportunities.
Third, these two trends together have led to a higher concentration of correspondent
relationships and hence higher systemic risk and less competition. Higher systemic risk arises
because fewer institutions are available as back-ups if the main relationship experiences
operational or financial difficulties. Lower competition arises particularly where the
relationship is between entities within the same financial conglomerate.
These trends together lead to higher costs and lower liquidity in many (usually less frequently
traded) currencies that are less accessible via correspondent relationships. Banks without
relationships with those few globally active correspondent banks must pay for indirect access
to the global banks’ “best friends”, to tap into an existing correspondent banking network,
which further increases costs.
17
Committee on Payments and Market Infrastructures, Correspondent Banking (CPMI Paper No 147, July
2016) <https://www.bis.org/cpmi/publ/d147.pdf>.
18
Finteq, Clearing House Gateway and Capture Solution for Cross Border Credit Payments 2
<https://www.finteq.co.za/wp-
content/uploads/2015/12/Finteq_Payment_Gateway_SADC_brochure_FEB14.pdf>.

[2021]
B Closed Loop Systems

In a closed-loop system, both payer and payee have a relationship with the same payment
operator. Examples include the large credit card providers, payment systems which settle
transactions between saving and loans associations, and “new” entrants like PayPal and
Alipay/Ant Group, Samsung Pay, and Apple Pay.
Funds wired from the payer to the payee flow through the accounts operated by the affiliates
that together form the financial conglomerate operating the closed loop system; for instance,
funds from PayPal US in California flow to PayPal EU in Luxembourg. Clients that transfer
and receive funds from their respective PSPs’ accounts to PayPal will thus experience PayPal
as a cross-border fund transfer system.
Figure 2: Closed loop systems
To achieve this, funds are routed from the group entity in the payer’s country to the entity in
the payee’s country. In principle, since payment business generally should be a licensed activity
and licensing is limited to each jurisdiction, for each country the closed loop system must
operate as its own entity. Hence, a closed loop system de facto works similarly to correspondent
banking within a network of group affiliates, yet with one set of technology for the whole loop.
This makes the operational and legacy issues of payment technologies less prevalent and
allows, in principle, lower fees.
The developments described above – increased concentration, less competition and increasing
costs – are prevalent in closed loop systems, justified by the still high(er) costs in traditional
cross-border correspondent banking.
Economic rents accrue in both types of payment arrangements predominant in cross-border
transactions. This fact, in combination with operational complexity and regulatory
uncertainties, means that the difficulties observed above will not simply self-correct due to
market factors or technological developments.

C Regulatory Approaches to Enhancing Competition and Efficiency

The above developments have resulted in the G20/BIS/FSB/CPMI seeking to address the
increasing concentration and lesser competition in cross-border payments.19 In particular, three
measures have been discussed to remedy the situation: (a) standardisation, (b) joint
infrastructure and (c) fee regulation.
1 Standardisation
One measure to reduce concentration is enhancing competition. A precondition for increased
competition in payments, which is a technology-based, data-driven business, is the
standardisation of interfaces and/or payment processes. For instance, the CPMI (2016) has
actively promoted the standardisation of messaging standards as well as KYC/AML processes.
This has been a long-term effort, dating back to the 1970s with the establishment of SWIFT,
the FATF, the CPMI itself and the ongoing International Standards Organisation (ISO)
initiative.20
Standardisation requires significant coordination among many regulators. Payment regulators
are often not in charge of AML processes, data protection and tax rules; and there are typically
strong tensions between local and global standards. Major economies compete to promote
potentially conflicting standards because countries that are able to create global standards
closer to their own will benefit their own financial institutions, at the expense of foreign ones.
These tensions have existed for decades, with limited progress made. While highly desirable,
standardisation has proven difficult to achieve, by way of regulation, legislation, soft law,
technical standards, market forces, or technological evolution.
2 Infrastructure Provision and Operations: Reducing Infrastructure Costs

Central banks’ collaborative development of cross-border payment infrastructure could be
another strategy for reducing the cost of cross-border payments. In this case, publicly backed
infrastructure could provide the alternative to inefficient infrastructure employed by private
actors. Many regional integration projects, for instance, follow this route.21
However, the cross-border integration of national payment systems requires significant
political and financial investment. In the absence of joint central banks (such as the ECB or
BCEAO) joint infrastructure requires cooperation among central banks. If central banks find
19
Cf. Enhancing Cross-Border Payments (n 1) and Financial Stability Board, Enhancing Cross-Border
Payments: Building Blocks of a Global Roadmap: Stage 2 Report to the G20 (Report, July 2020)
<https://www.bis.org/cpmi/publ/d193.pdf>.
20
‘APIs and ISO 20022 – Whitepaper’, ISO 20022 (Web Page) <https://www.iso20022.org/about-iso-
20022/apis-and-iso-20022>.
21
Ross P Buckley, Douglas W Arner, Dirk A Zetzsche, Thomas Lammer & S[**] Gazi, ‘Regional Solutions to
Global Payments Challenges: Towards a Single Rulebook’, Banking and Finance Law Review special
Festschrift issue (forthcoming).

[2021]
standardisation challenging, the creation of full cross-border infrastructure (as has been done
with SWIFT and CLS) is even more difficult.
3 Fee Regulation
A third common approach is fee regulation, which has enjoyed remarkable success in domestic,
and within confined, regions. For instance, in the EU SEPA region a fixed fee is set for all
retail cross-border payments, which reduced fees, in total, by more than 70% between 2012
and 2014. In the US (regulated) debit card fees fell by 42% after fee regulations took effect in
2011.22 Further, the cap on interchange fees for consumer card payment transactions adopted
by the EU Interchange Fee Regulation (IFR)23 reduced issuers’ fees by approximately EUR 3
billion per year between 2015-16.24
However, fee regulation in multi-currency settings is problematic. Regulated fees will likely
be too high for oft-traded currencies and too low for scarcely traded currencies, given the latter
involves the risk that the payment institution will not find a counterpart at that rate. Legal,
sanctioning and enforcement risks may further reduce the willingness of participants, resulting
in even fewer banks functioning as correspondent banks for infrequently traded currencies at
the regulated rate. Finally, overly high regulated fees often disincentivise institutions from
innovating. The same can be so for overly low fees that do not cover the costs of technological
developments. Finding the right balance can be exceptionally difficult.
III INTRODUCING BEST EXECUTION: LEARNING FROM SECURITIES REGULATION

Each of the aforementioned measures has its merits, but each also brings considerable
challenges. Thus, we examine a new way forward in cross-border payments that could
complement these approaches and increase their effectiveness: the introduction of the best
execution principle established in the context of securities brokerage.
22
See Bank for International Settlements, BIS Annual Economic Report 2020 (Report, 2020) 67, 84
<https://www.bis.org/publ/arpdf/ar2020e3.pdf>.
23
Council Regulation (EU) No 751/2015 (OJ 2015 L 123 p. 1) (Regulation of the European Parliament and of
the Council of 29 April 2015 on Interchange Fees for Card-based Payment Transactions).
24
See European Commission, Study on the Application of the Interchange Fee Regulation (Final Report, 2020)
<https://ec.europa.eu/competition/publications/reports/kd0120161enn.pdf>; Santiago Carbó-Valverde, Sujit
Chakravorti and Francisco Rodriguez-Fernández, ‘The Role of Interchange Fees in Two-Sided Markets: An
Empirical Investigation on Payment Cards’ (2016) 98 The Review of Economics and Statistics 367.
11

A Best Execution: A Principle of Securities Regulation

Best execution legally requires a financial institution to provide the most advantageous order
execution for each customer given the prevailing market environment.25 This legal principle
seeks two objectives: first, the enhancement of market efficiency through reduced transaction
costs,26 and second, the protection of investors through the management of conflicts of interest
between brokers and clients.27
Key factors within a best execution obligation include price, speed of execution and risks
relating to execution. A financial institution must follow a client’s ex ante preference, if there
is one. Absent such a stipulation, the financial institution retains some discretion, but as a
fiduciary has to treat the client’s interests as paramount.
The US Securities and Exchange Commission (SEC) describes the securities law principle of
best execution as follows:
Brokers are legally required to seek the best execution reasonably available for their customers’
orders. To comply with this requirement, brokers evaluate the orders they receive from all
customers in the aggregate and periodically assess which competing markets, or electronic
communications networks (…) offer the most favorable terms of execution. Some of the factors
a broker must consider when seeking best execution of customers’ orders include: the
opportunity to get a better price than what is currently quoted, the speed of execution, and the
likelihood that the trade will be executed.28
Best execution prohibits a financial institution from putting its own interests first. For instance,
if an institution is compensated by a trading facility for additional trading volume routed
through the facility, that compensation must be disregarded when determining the routing.
B Legal Framework of Best Execution
25
Cf. Niamh Moloney, EU Securities and Financial Markets Regulation (Oxford University Press, 3rd ed, 2014)
519, Jean Pierre Casey and Karel Lannoo, The MiFID Revolution (Cambridge University Press, 2009) 58-77;
John Armour et al, ‘Financial Advice’, in John Armour et al (eds), Principles of Financial Regulation (Oxford
University Press, 2016) 226, 243-4.
26
Casey and Lannoo (n 25) 63-4 and Thomas Iseli, Alexander F Wagner and Rolf H Weber, ‘Legal and
Economic Aspects of Best Execution in the Context of the Markets in Financial Instruments Directive’ (2007)
1(4) Law and Financial Markets Review 313.
27
See, more recently, Securities and Exchange Commission, ‘Order Directing the Exchanges and the Financial
Industry Regulatory Authority to Submit a New National Market System Plan Regarding Consolidated Equity
Market Data’, Release no 34-88827, File no 4-757, 6 May 2020, 56,
<https://www.sec.gov/rules/sro/nms/2020/34-88827.pdf> at 56; CFA Institute, Regulation NMS: Review and
Recommendations (Report, September 2007) <https://www.cfainstitute.org/-/media/documents/article/position-
paper/regulation-nms-review-recommendations.ashx> (stating that ‘[i]nducements that trading markets offer to
influence routing decisions and build market share have exacerbated the potential for conflicts of interest at
broker-dealers’).
28
‘Best Execution’, US Securities and Exchange Commission (Web Page, 9 May 2011)
<https://www.sec.gov/fast-answers/answersbestexhtm.html>.

[2021]
1 US Securities Regulation
The Best Execution principle originated from “the concept of fiduciary duty in light of the
agency relation between brokers and their clients”29 requiring brokers “to act loyally for the
principal’s benefit in all matters connected with the agency relationship”.30
The principle matured, however, in the further development of US securities regulation. While
the SEC advocated for best execution in securities regulation in 1961, and this principle was
later adopted by NASDAQ,31 best execution only became mandatory law in the US in 1988,
with the adoption of s. 11A of the Securities Exchange Act of 1934.32 This change was
reinforced in 2000 when the SEC adopted the new Rule 605 (formerly Rule 11Ac1–5) and Rule
606 under the Securities Exchange Act to require monthly disclosure of execution quality. The
aim was to promote a more competitive and efficient national market system by increasing the
visibility of execution quality, supported by private law fiduciary principles. The SEC,33
supported by commentators,34 argued that public disclosure of order routing practices would
result in more vigorous competition in order routing performance.
Rule 606 was meant to open market infrastructure for tech-based competition, and did so.35
Today’s US equity market infrastructure is “highly automated, dispersed among myriad trading
centres, and more complex [in light of the] rapid and ongoing evolution of technologies for
29
Casey and Lannoo (n 25) 58-77; Onnig H Dombalagian, ‘Best Execution: An Impossible Dream?’, in A.B.
Laby (ed), The Cambridge Handbook of Investor Protection (Cambridge 2021 forthcoming), 4; citing De
Kwiatkowski v. Bear, Stearns & Co., 306 F.3d 1293, 1302 (2d Cir. 2002) (which considered, but eventually
rejected a fiduciary duty of a broker to a sophisticated client, and held whether a fiduciary duty exists is subject
to the circumstances of the case). Whether brokers are subject to fiduciary duty under English law is also a
matter for debate. See Andrew F Tuch, ‘The Weakening of Fiduciary Law’, D. Gordon Smith & Andrew S.
Gold (eds) Research Handbook on Fiduciary Law (2018) 354, Washington University in St. Louis Legal
Studies Research Paper No. 17-04-03, 354–376, also available at https://ssrn.com/abstract=2951460 (for a
comparison between US and UK law); ‘Law Commission Consultation Paper 215 Fiduciary Duties of
Investment Intermediaries: A Consultation Paper’. Key Non Parliamentary Papers Law, October 31, 2013,
http://www.lawcom.gov.uk/app/uploads/2015/03/cp215_fiduciary_duties.pdf, and Iris H-Y Chiu and Alan H
Brener, ‘Articulating the Gaps in Financial Consumer Protection and Policy Choices for the Financial Conduct
Authority- Moving Beyond the Question of Imposing a Duty of Care’ (2018), https://www.ucl.ac.uk/ethics-
law/sites/ethics-law/files/duty_of_care_article_draft_24_october_2018.pdf.
30
American Law Institute, Restatement of the Law of Agency (3d) (2006) §8.01, as reported in Casey and
Lannoo (n 25) 58.
31
Jonathan R Macey and Maureen O’Hara, ‘The Law and Economics of Best Execution’ (1997) 6 Journal of
Financial Intermediation 188.
32
15 U.S.C. § 78k-1(a) (1988). See David A Lipton, ‘Best Execution: The National Market System’s Missing
Ingredient’ (1982) 57 Notre Dame Law Review 449; D Bruce Johnsen, ‘A Transaction Cost Assessment of SEC
Regulation Best Interest’ (2018) 3 Columbia Business Law Review 695.
33
See Securities Exchange Act Release No 43590 (17 November 2000), 65 F.R. 75414, 75417 (1 December
2000).
34
See, based on pre-2001 data, Christine A Parlour and Duane J Seppi, ‘Liquidity-Based Competition for Order
Flow’ (2003) 16(2) The Review of Financial Studies 301; Hendrik Bessembinder, ‘Quote-based Competition
and Trade Execution Costs in NYSE-listed Stocks’ (2003) 70(3) Journal of Financial Economics 385.
35
Ibid 42 ff.
13

generating, routing, and executing orders, and the impact of regulatory changes”.36 Increasingly
sophisticated order execution algorithms (either aggressive or passive) 37 and smart order
routing systems affecting timing, pricing, and number of orders contribute to the high-speed
and highly efficient activity of institutional customers in equity markets. This is exactly what
we seek to achieve in the world of payments.
Rule 606 was supplemented by SEC Rule 611 of National Market System Regulation or
Regulation NMS of 2005 (the so called “Order Protection Rule” or “Trade-through Rule”).38
Regulation NMS implements the objectives set forth under s. 11A of the Securities Exchange
Act, facilitating the establishment of a national market system with a view to ensure best
execution of investors’ orders. The Rule promotes “intermarket price protection by restricting
‘trade-throughs’ – the execution of trades on one venue at prices that are inferior to publicly
displayed quotations on another venue”.39 Specifically, it requires a trading centre to implement
policies reasonably aimed at preventing “trade-throughs” with the only exceptions being those
set out in para. (b) of the Rule. Such quotations – which must be disseminated in the market
data feeds – must be displayed by an automated trading centre, namely centres that are
immediately and automatically executable, without delay. To be protected, a quotation must be
the “best bid” or “best offer” of a national securities exchange (namely, the highest-priced bid
or the lowest-priced offer).40 The Rule does not: (i) cover any additional depth-of-book prices
outside the best prices displayed by an automated trading centre (lower prices for bids and
higher prices for offers), (ii) require the routing of orders to trading centres displaying the best
prices, (iii) require orders to be routed to execute against displayed quotations before trades
could be executed at matching prices (i.e. “trade-at” restriction), or (iv) require investors to
display their own trading interest. Many mark the coming into force of Regulation NMS as the
true beginning of a best execution rule in the US.41
In 2014, the best execution framework was supplemented by FINRA Rule 5310 on best
execution and interpositioning, as part of the self-regulation of broker-dealers. It requires firms
to regularly and rigorously review the execution quality of customer orders (if the firm does
not conduct an order-by-order review), on a minimum quarterly basis and on a security-by-
36
Ibid 19 Fn 28 -9.
37
Ibid 21.
38
For an in-depth analysis, and historical outlook, see SEC, ‘Rule 611 of Regulation NMS - Memo to SEC
Market Structure Advisory Committee’ (Memorandum, 30 April 2015)
<https://www.sec.gov/spotlight/emsac/memo-rule-611-regulation-nms.pdf>.
39
SEC, ‘Disclosure of Order Handling Information’, Release No 34-78309; File No S7-14-16, 17 CFR Parts
240 and 242 [II.A] <https://www.sec.gov/rules/proposed/2016/34-78309.pdf> (‘Disclosure of Order Handling
Information’).
40
Cf. Regulation NMS Adopting Release, 70 F.R. 37496, 37534 (29 June 2005).
41
Chester S. Spatt, Improving the SEC’s Best Execution Rule. Remarks to the Securities and Exchange
Commission’s Investor Advisory Committee on June 10, 2021, June 24, 2021
<https://clsbluesky.law.columbia.edu/2021/06/24/improving-the-secs-best-execution-rule/>; and Giovanni
Petrella, ‘MiFID, Reg NMS and competition across trading venues in Europe and the USA’ (2010) 18(3)
Journal of Financial Regulation and Compliance 257.

[2021]
security, type-of-order basis (e.g., limit order, market order and market on open order).42 This
implies consideration of the market characteristics, as well as the “size and type of transaction
and the number of markets checked”.43 Looking at the output, FINRA classified a series of
issues with some firms’ execution quality reviews. In some cases, firms did not assess the
quality of the execution order routing with the one that the firm could have obtained from
competing markets; in other cases, firms did not review orders on a type-of-order basis, or did
not consider some factors required by the Rule (speed of execution, price improvement
opportunities and the likelihood of execution of limit orders) and potential conflicts of
interests.44 Further, certain firms still did not disclose the material aspects of the non-directed
order flows routed to their own trading desk, of their relationships with each of the significant
venues identified in their reports, of payment details and profit-sharing relationships.45 At least
prior to the 2019 reforms, empirical evidence suggests some brokers did not play by the rules.46
It is unsurprising that since its inception, numerous court cases have reinforced the relevance
of best execution for investor plaintiffs,47 even though federal law often prevents securities
brokers from entering into class actions.48
The patchy compliance by brokers with the rules, prompted further action from the SEC. In
June 2019, building on both Regulation NMS and FINRA’s Rule 5310, the SEC introduced
amendments to Regulation NMS, dubbed “Regulation Best Interest”.49 Rules 606(b)(3) and
42
See also FINRA, Guidance on Best Execution Obligations in Equity, Options and Fixed Income Markets:
Regulatory Notice 15-46 (November 2015)
<https://www.finra.org/sites/default/files/notice_doc_file_ref/Notice_Regulatory_15-46.pdf>.
43
Peter Kruger Andersen, ‘Time to Reduce Complexity in a Data-Driven Regulatory Agenda – Perspectives on
the MiFID II Best Execution Regime’ (2020) 17(6) European Company and Financial Law Review 692, 721-
723. See also Xin Zhao and Kee H Chung, ‘Information Disclosure and Market Quality: The Effect of SEC Rule
605 on Trading Costs’ (2007) 42(3) Journal of Finance and Quantitative Analysis 657 (providing evidence that
transparency results in efficiency enhancing competition).
44
FINRA, 2019 Examination Findings Report: Best Execution (Report, 16 October 2019)
<https://www.finra.org/rules-guidance/guidance/reports/2019-report-exam-findings-and-observations/best-
execution>.
45
Ibid.
46
Robert Battalio, Todd Griffith and Robert Van Ness, ‘Do (Should) Brokers Route Limit Orders to Options
Exchanges That Purchase Order Flow?’ (2021) 56 Journal of Financial and Quantitative Analysis 183, 211
(‘[S]ome brokers seemingly maximize the value of their order flow by selling marketable orders and sending
nonmarketable orders to exchanges that offer large liquidity rebates. Other brokers appear to bypass liquidity
rebates by routing both marketable and nonmarketable orders to exchanges that purchase order flow’).
47
See Klein v. TD Ameritrade Holding Corp., 327 F.R.D. 283 (D. Neb. 2018); Schwab v. E*Trade Fin. Corp.,
285 F. Supp. 3d 745, 753 (S.D.N.Y. 2018); Schwab v. E*Trade Fin. Corp., 752 Fed. App’x. 56 (2d Cir. Oct. 26,
2018); Newton v. Merrill Lynch, Pierce, Fenner, Smith, 259 F.3d 154 (3d Cir. 2001); In Re Merrill Lynch
Securities Litigation, 911 F. Supp. 754 (D.N.J. 1995); Newman v. Smith, [1974-1975 Transfer Binder] FED.
SEC. L. REP. (CCH) 95,078, at 97,783 (1975); Sinclair v. SEC, 444 F.2d 399, 400 (2d Cir. 1971); In re Kidder
Peabody & Co., 43 S.E.C. 911, 915 (1968); In re Delaware Management, 43 S.E.C. 392, 397 (1967).
48
See Zola v. TD Ameritrade, Inc., 889 F.3d 920, 926 (8th Cir. 2018); Lewis v. Scottrade, Inc., 879 F.3d 850,
855 (8th Cir. 2018); Fleming v. Charles Schwab Corp., 878 F.3d 1146, 1154-55 (9th Cir. 2017); Kurz v. Fid.
Mgmt. & Research Co., 556 F.3d 639, 642 (7th Cir. 2009).
49
17 CFR § 240.15/1.
15

606(a) of Regulation NMS now require broker-dealers to publish transaction reports on a

quarterly and biannual basis to customers in either XML or pdf format.50 The new rules aim to
increase transparency around best execution by providing key benchmark metrics that are
similar between firms to enhance transparency on potential conflicts between broker-dealers’
venue selection and the best interests of the client for each order.51 The new rules are expected
to abolish the malpractice of collecting rebates from execution venues in return for executing
the customer’s order on that route.52
The developments since the introduction of Rule 606 in November 2000 confirmed the SEC’s
view. Pairing best execution with additional disclosure has reduced transaction costs and
enhanced market efficiency.53 US securities regulation sought to achieve this end by adding a
disclosure obligation as a means of enforcing an existing principle rather than writing a
regulation anew. As it turns out, in time, Europe took the same approach.54
2 Europe: MiFID
Historically, EU Member States placed only minor importance on the best execution rule in
securities transactions. The roots of best execution in Europe date back to 1986, when the
London Stock Exchange55 introduced the best execution principle for brokers licensed to
operate on the LSE. The UK regulations were more detailed, clear and certain than those of
other European countries.56 Europe sought to catch up to some extent with Article 11 of the
1993 Investment Service Directive 93/22/EEC which introduced a broad conduct of business
principle.
50
‘Disclosure of Order Handling Information’ (n 39) 38. The new rules resemble the conflict-of-interest rules,
as well as the disclosure and suitability rules under articles 23–25 of MiFID II, discussed infra, at III.B.2. As to
the output, see also Q&A on MiFID II (n 61) 26.
51
In the pre-reform US landscape, see Amber Anand et al, ‘Institutional Order Handling and Broker-Affiliated
Trading Venues’ (2021) 34(7) The Review of Financial Studies 3364.
52
See the statement by Dermot Harris that the new rules ‘make it pretty much impossible to hide the old game
of collecting rebate from venues in order to execute the customer’s order.’; ‘New Best Execution Rules Likely
Beyond the SEC’s Rule 606 Changes’, A-Team Insight Blogs (Blog Post, 23 July 2019) <https://a-
teaminsight.com/new-best-execution-rules-likely-beyond-the-secs-rule-606-changes/?brand=tti>.
53
See Robert Battalio, Brian Hatch and Robert Jennings, ‘Toward a National Market System for U.S. Exchange-
Listed Equity Options’ (2004) 59 The Journal of Finance 933; Hendrik Bessembinder, ‘Trade Execution Costs
on NASDAQ and the NYSE: A Post-Reform Comparison’ (2009) 34(3) Journal of Finance and Quantitative
Analysis 387; Zhao and Chung (n 43).
54
‘Disclosure of Order Handling Information’ (n 39) [II.A].
55
John Board and Stephen Wells, ‘Liquidity and Best Execution in the UK: A Comparison of SETS and
Tradepoint’ (2001) 1(4) Journal of Asset Management 344, 350.
56
For further discussion of the concentration rule see Gerard Hertig, ‘Mifid and the Return of Concentration
Rules’ in Stefan Grundmann, et al (eds), Festschrift für Klaus J. Hopt zum 70. Geburtstag am 24. August 2010
Festschrift (De Gruyter, 2010), 1989-2000. On the UK regulation, see David Capps, ‘Better than best
execution? The Financial Services Authority's new proposals to reform UK best execution requirements’ (2003)
11(1) Journal of Financial Regulation and Compliance 37; and the roundtable of commentators, ‘Could
Europe’s Lawmakers Drive Trading Overseas?’ (2003) 22(7) International Financial Law Review 17.

[2021]
But it was not until 2004 that the breakthrough on best execution in Europe, at least in theory,
was to occur. Article 21 of the Markets in Financial Instruments Directive (“MiFID I”) required
investment firms to “take reasonable steps to obtain, when executing orders, the best possible
result for their clients taking into account price, cost, speed, likelihood of execution and
settlement, size, nature or any other consideration relevant to the execution of the order”.57 Yet
even such a clear rule was to be somewhat neutered in practice by a splintering of the market
into multiple venues so intermediaries could avoid seeking the truly best venue – a move which
left clients still facing the risks associated with the lower liquidity of segregated markets. In
addition, the best execution standards’ overly prescriptive and opaque requirements58 led to a
narrowing of execution choices, hindering development of genuine competition among trading
venues.59 An analysis of transaction costs revealed that “the role of back-office providers in
the MiFID Best Execution chain is key, as they remain in a unique position to capture and
process transaction data independently from venues and intermediaries on behalf of the
investors of their representatives”.60
The EU thus subsequently sought to improve the functioning of best execution with articles
27(1) and (7) MiFID II, adopted in 2014, and with further rules adopted in 2017. While MiFID
I required firms to take all reasonable steps to obtain, when executing orders, the best possible
result for their clients, MiFID II requires firms to take all sufficient steps.61 Further, MiFID II
clarified that the best possible result for retail clients must be determined in terms of total price
of execution (spread plus fees).62
Even so, there is still no uniform procedure applicable to all categories of financial instruments.
Financial institutions are required to craft execution strategies that identify venues by type of
transactions,63 to update that strategy annually, and notify clients accordingly.64 Further, the
selection of the best execution conditions is limited to those that the intermediary selects on
57
Council Directive 2004/39/EC (OJ 2004 L 145 p. 1) (‘MiFID I’) art 21. The MiFID is still in force in the UK:
see Financial Conduct Authority (FCA), FCA Handbook – Conduct of Business Sourcebook 11.2A.1-11.2A.39
<https://www.handbook.fca.org.uk/handbook/COBS/11/2A.html> (‘COBS’).
58
European Securities and Markets Authority, ‘Best Execution Under MiFID’ (2015) Peer Review Report 7
<https://www.esma.europa.eu/sites/default/files/library/2015/11/2015-
494_peer_review_report_on_best_execution_under_mifid_0.pdf>.
Guido Ferrarini, ‘Contract Standards and the Markets in Financial Instruments Directive (MiFID): An
59
Assessment of the Lamfalussy Regulatory Architecture’ (2005) 1 European Review of Contract Law 19, 38.
60
Catherine D’Hondt and Jean-René Giraud, Transaction Cost Analysis A-Z: A Step towards Best Execution in
The Post-MiFID Landscape (EDHEC-Risk Institute Report, 2008) 86
<https://risk.edhec.edu/publications/transaction-cost-analysis-z-step-towards>.
61
European Parliament and Council Directive 2014/65/EU (OJ 2014 L 173 p 349), arts 27(1) and (7) (‘MiFID
II’); European Securities and Markets Authority, Questions and Answers on MiFID II and MiFIR Investor
Protection and Intermediaries Topics (Report, 28 May 2021) 19
<https://www.esma.europa.eu/sites/default/files/library/esma35-43-
349_mifid_ii_qas_on_investor_protection_topics.pdf> (‘Q&A on MiFID II’).
62
Stefan Grundmann and Philipp Hacker, ‘Conflict of Interest’, in Danny Busch and Guido Ferrarini (eds),
Regulation of the EU Financial Markets, MiFID II and MiFID (Oxford University Press, 2017) 195.
63
MiFID II art 27 and Council Regulation (EU) No 751/2015 (OJ 2015 L 123 p 1) art 66.
64
Council Regulation (EU) No 751/2015 (OJ 2015 L 123 p. 1), art 66(3).
17

behalf of the client for a given strategy and type of transaction.65 Within the limits provided by
the strategy, the financial institution determines the importance of each execution factor.66 At
the same time, firms must monitor and ensure the effectiveness of their execution arrangements
and remedy any deficiencies.67 Particular duties attach to retail clients. Sophisticated wholesale
clients can negotiate the best execution terms with the firms. Regulatory scrutiny focuses on
the needs and interests of retail clients. The firm’s discretion is limited by regulation: the factor
determining best execution for retail clients is the total consideration, i.e., the price of the
financial instrument and the costs of execution.
MiFID II’s rules on pre- and post-trade transparency, conflicts of interests and inducements,
promote the effectiveness of the best execution rule.68 In particular, MiFID II subjects
execution venues to extensive data transmission requirements regarding order flow and
execution rates that allow regulators to review, ex post, to what extent the brokers chose the
best execution venue.69 Further, article 34 requires investment firms to identify, prevent, and
manage conflicts of interest, and implement effective organisational arrangements to prevent
conflicts that adversely affect their clients. MiFID II contains two rules on inducements, a
general and a specific one. The general rule prohibits investment firms from paying benefits
to, or receiving benefits from, third parties, unless the benefits are designed to enhance the
quality of the relevant service to the client, and do not impair compliance with the investment
firm’s duty to act honestly, fairly, and professionally in accordance with the best interests of
its clients. The specific rule only concerns investment advisers and portfolio managers and
prohibits accepting benefits relating to the investment firm’s services, other than certain
specified non-monetary benefits of negligible size. Finally, Article 53(3)(c) requires
investment firms providing execution services to identify separate charges for execution, and
to apply separately identifiable charges to other benefits or services, such as investment advice,
research and data provision.70
Overall, MiFID II best execution provisions prohibit paying the investment firm for order flow
(while clients may benefit from volume-based rebates). The prohibition applies to the
investment firm receiving the payment, and will apply to an investment firm routing underlying
client orders to a venue or executing broker in return for fees or other benefits.71
In the UK, the FCA’s best execution requirements are set out in the Conduct of Business
sourcebook (COBS) in the FCA Handbook, primarily in COBS 11.2A. These have been
derived from MiFID II and related legislation. While the provision remains unchanged to a
65
Council Regulation (EU) No 751/2015 (OJ 2015 L 123 p. 1), recital 104.
66
Council Regulation (EU) No 751/2015 (OJ 2015 L 123 p. 1), recital 99.
67
MiFID II art 27(7); Moloney (n 25) 522.
68
Guido Ferrarini, ‘Market Transparency and Best Execution: Bond Trading Under MiFID’, in Michel Tison et
al (eds), Perspectives in Company Law and Financial Regulation (Cambridge University Press, 2009) 477. Cf
Peter Yeoh ‘MiFID II Key Concerns (2019) Journal of Financial Regulation and Compliance 110, 113-17.
69
See Q&A on MiFID II (n 61).
70
See generally on MiFID II, Danny Busch, ‘MiFID II: Stricter Conduct of Business Rules for Investment
Firms’ (2017) 12(3) Capital Markets Law Journal 340.
71
See MiFID II art 27(2).

[2021]
large extent and still reflects the version applicable since January 3, 2018, some sections have
been amended and are applicable since January 1, 2021 (COBS 11.2A.1, regarding the scope
of application, COBS 11. 2A.8, COBS 11. 2A.25, COBS 11. 2A.34).
More specifically COBS 11.2A.2 and 11.2A.3 largely repeat art. 27 MiFID II, as they require
that
(1) A firm must take all sufficient steps to obtain, when executing orders, the best possible
results for its clients taking into account the execution factors. (2) The execution factors to be
taken into account are price, costs, speed, likelihood of execution and settlement, size, nature
or any other consideration relevant to the execution of an order. (3) The obligation to take all
sufficient steps to obtain the best possible result for its clients should apply where a firm owes
contractual or agency obligations to the client.
The FCA redefined the best execution criteria in 2021:

[w]hen executing client orders, investment firms shall take into account the following criteria
for determining the relative importance of the factors referred to in [COBS 11.2A.2R]: (a) the
characteristics of the client including the categorisation of the client as retail or professional;
(b) the characteristics of the client order, including where the order involves a securities
financing transaction (SFT); (c) the characteristics of financial instruments that are the subject
of that order; (d) the characteristics of the execution venues to which that order can be
directed.72
In particular, investment firms should take all sufficient steps to obtain the best possible result
for a client to the extent that it executes an order, and “not structure or charge their commissions
in such a way as to discriminate unfairly between execution venues”.73
Additional amendments to the provision may be expected in the near future, as the FCA in
April 2021 opened a consultation on the removal of two sets of reporting obligations on firms,
namely the “obligation on execution venues to publish a report on a variety of execution quality
metrics to enable market participants to compare execution quality at different venues (known
as RTS 27 reports), [and the] obligation on investment firms who execute orders to produce an
annual report setting out the top 5 venues used for executing client orders and a summary of
the execution outcomes achieved (known as RTS 28 reports)”.74 The removal of these reporting
obligations requires amendments in both legislation and the FCA Handbook. On June 28, 2021,
the Markets in Financial Instruments (Capital Markets) (Amendment) Regulations 2021, SI
2021/774 were released, providing the necessary legislative changes to revoke the RTS 27 and
RTS 28 reporting requirements.75
On September 24, 2021, ESMA opened a 3 months consultation, aimed at improving the
MiFID II framework on best execution reporting with the goal of implementing effectiveness,
72
COBS (n 57) 11.2A.8.
73
Ibid 11.2A.2.
74
Financial Conduct Authority (FCA), Changes to UK MIFID’s Conduct and Organisational Requirements
(Consultation Paper CP21/9, 2021) 18-20 <https://www.fca.org.uk/publication/consultation/cp21-9.pdf>.
75
Explanatory Memorandum to Markets in Financial Instruments (Capital Markets) (Amendment) Regulations
2021, SI 2021/774, 2–3.
19

consistency and investor protection, by virtue of more granular reporting.76 Whilst the
outcomes of the consultation will not amount to an overnight change to the existing RTS 27
and 28,77 as the Technical Proposals of ESMA can only be executed after the relevant
provisions of MiFID II have been amended, ESMA will assist the European Commission in its
assessment of the adequacy of best execution reporting standards.
C Lessons from the Securities Context

The first economic analysis of the best execution principle appears to date back to 1982.78
Initially, the principle was received with some criticism. For instance, some authors argued that
the principle of best execution (“a legal fiction”) could lead to suboptimal market efficiency79
due to increased market fragmentation and reduced liquidity.80
Other commentators emphasised the difficulty of easily determining the best execution venue
ex ante, while ex post enforcement is costly and will rarely lead to results beneficial for
clients.81 It was argued that if markets were perfectly competitive, retail investors would be
unable to monitor best execution; in turn, brokers could execute poorly and use order flow
inducements to lower their commission fees.
76
ESMA, Consultation Paper. Review of the MiFID II framework on best execution reports. ESMA35-43-2836,
24 September 2021, <https://www.esma.europa.eu/sites/default/files/library/esma35-43-2836_cp_-
_best_execution_reports.pdf>.
77
“[I]n the context of ESMA’s mandate in accordance with Article 1(5) of the ESMA founding regulation
(Regulation (EU) No. 1095/2010), this paper aims at identifying the reasons for these shortcomings. Moreover,
the paper proposes possible improvements to the regime which could be adopted in the future to ensure effective
and a consistent level of regulation and supervision and enhance investor protection in this area. Those possible
improvements include both potential technical changes to RTS 27 and 28 and potential amendments to the
legislative framework (in particular, to Articles 27(3) and (6) of MiFID II) to enable the possible technical
changes to RTS 27 and 28 to come into effect at a future time. The possible improvements proposed by ESMA
also account for potential future legislation to establish a Consolidated Tape under MiFID II” (ESMA,
Consultation Paper. Review of the MiFID II framework on best execution reports. ESMA35-43-2836, 24
September 2021, https://www.esma.europa.eu/sites/default/files/library/esma35-43-2836_cp_-
_best_execution_reports.pdf, at 4).
78
Kenneth D. Garbade and William L Silber, ‘Best Execution in Securities Markets: An Application of
Signaling and Agency Theory’ (1982) 37(2) The Journal of Finance 493. For a comprehensive overview, cf.
Ferrarini, ‘Best Execution and Competition Between Trading Venues’ (n Error! Bookmark not defined.) 410-
11.
79
Casey and Lannoo (n 25) 60.
80
On the first point, see both Macey and O’Hara (n 31) and Jean-René Giraud and Catherine D’Hondt, MiFID:
Convergence Towards a Unified European Capital Markets Industry (Risk Books, 2006); on the second one,
see only Macey and O’Hara (n 31). See also Daniel Aghanya, Vineet Agarwal and Sunil Poshakwale, ‘Market
in Financial Instruments Directive (MiFID), Stock price Informativeness and Liquidity’ (2020) 113 Journal of
Banking and Finance 105730.
81
See I. H-Y. Chiu and A.H. Brener, “Articulating the Gaps in Financial Consumer Protection and Policy
Choices for the Financial Conduct Authority – Moving Beyond the Question of Imposing a Duty of Care”
(2019) 14 Capital Markets L. 217, 222, 225, 234, 242 –3. See also J. Benjamin, “The Narratives of Financial
Law” (2010) 30 O.J.L.S. 787, 800.

[2021]
Yet, the weight of this criticism from the 1990s has been countered with experience. Applying
Coasian reasoning, the SEC (in Reg NMS) acknowledged that the principle serves a very valid
and important objective: “encouraging trading and market participation by lowering the costs
of transacting”.82 Further, it addresses the principal-agent dilemma between broker and client
by establishing a clear rule in case of doubt.83
As to the enforcement difficulties which were frequently raised against the best execution
principle84, securities regulators have: (a) requested expansive disclosures, (b) built systems to
collect large scale datasets on data gathering ex ante, transaction reporting, order routing and
post trade transparency,85 and (c) paired strong penalties with private enforcement. Over time,
this approach of technology based and enabled regulation and compliance (RegTech and
SupTech) is allowing data-driven analysis of whether brokers have in fact chosen the best
execution venue for their clients.
All in all, the argument that best execution increases transaction costs was always likely to be
inaccurate. Today, we often see even (apparently) zero fee order execution. Although the (still
existing) execution costs86 are likely extracted from some other financial intermediary
somewhere in the financial services value chain,87 order execution costs charged to clients have
ceased to be a major impediment to market efficiency. While being accompanied and even
promoted by insightful IOSCO work88, the progress on best execution was achieved without
major multi-lateral efforts nor binding agreements. Rather, financial regulation with regional
scope limited to financial institutions registered for brokerage services in two very large
financial markets (the US and the EU) prompted an upward spiral in innovation and a
downward spiral for costs.89
82
Casey and Lannoo (n 25) 61; and also Ronald H Coase ‘The Nature of the Firm’ (1937) 16 Economica 386.
Cf: Niamh Moloney, How to Protect Investors (Cambridge University Press, 2010) 358.
83
Stephen A Ross, ‘The Economic Theory of Agency: the Principal’s Problem’ (1973) 62 The American
Economic Review 134.
84
See Macey and O’Hara (n 31); Sandro Casal, Matteo Ploner and Alec N Sproten, ‘Fostering The Best
Execution Regime: An Experiment About Pecuniary Sanctions and Accountability in Fiduciary Money
Management’ (2019) 57 Economic Inquiry 600.
85
See Regulatory Technical Standard no. 27.
86
See Chris Jaccard, ‘The Cost of Zero Commission Trading’, Financial Alternatives (Web Page, 23 October
2019) <https://www.financialalternatives.com/financial-alternatives-inc/2019/10/22/the-cost-of-zero-
commission-trading>; for the same issue in the context of ETFs, see William A Birdthistle and Daniel J Hemel,
‘Next Stop for Mutual-Fund Fees: Zero’, The Wall Street Journal (online, 10 June 2018)
<https://www.wsj.com/articles/next-stop-for-mutual-fund-fees-zero-1528652532>.
See Dirk A Zetzsche et al, ‘Digital Finance Platforms – Toward a New Regulatory Paradigm’ (2020) 23(1)
87
University of Pennsylvania Business Law Journal 273.

88
See OICV-IOSCO, Methodology For Assessing Implementation of the IOSCO Objectives and Principles of
Securities Regulation (May 2017), at 173-174; OICV-IOSCO, Order routing incentives. Final report. FR
08/2017 (June 2017), 8-12 [may we use passim? It’s really widespread]
<https://www.iosco.org/library/pubdocs/pdf/IOSCOPD564.pdf>.
89
In light of the mandatory nature of the best execution rules, binding all brokerage firms registered and
licensed under the U.S. and EU securities regulation, the private international law matter as to which law or
21

This is a state we aspire to achieve for payments. Whether and how this might best be done,
we explore in the next section.
courts applies, or have been chosen by the parties, respectively, did not play a major role in prompting the
developments we described in this section.

[2021]
IV BEST EXECUTION FOR CROSS-BORDER PAYMENTS?

Two major questions present themselves. First, how might the “best execution” principle be
adapted effectively from securities regulation and applied to payments? Second, if there are
significant material differences between securities brokerage and payments, how do we adjust
to, and overcome, these differences?
A Agency Law Concepts in Payments

PSPs and customers, under existing private law, are subject to some duties resulting from
agency law.
The international transfer of funds relies on a chain of agency relationships.90 In common law
systems, the PSP is the customer’s agent, and thus not free to act according to its own priorities.
PSPs in the chain have contractual responsibility to their principal, i.e., the first PSP to the
customer, and later PSPs to their previous PSP. As an agent, the PSP must act within their
authority, for instance, pay the right person the right amount at the right time, while adhering
to the principal’s instructions. In essence, the same principles apply to payment contracts under
German and French civil law, which treat payment relationships as a “mandate”.91 While
different in detail, the outcome is similar to the common law agency relationship.
Yet, the original private law relationship is often superseded by statutory law. For instance,
Article 4a of the US Uniform Commercial Code (UCC4a), on Transfer of Funds, replaces
obligations under the original agency relationship. In the same vein, the level of detail
stipulated by EU Member States implementing Payment Services Directive I and II,92 leaves
little room for principles of agency law.93 In particular, while there are rules on cost
transparency and cost reimbursement, there is no statutory best execution requirement. This
leaves all PSPs with the option to define, in their contractual terms, the “best friends”
relationship within their network as the sole one through which they will route payments. In
summary, while PSPs subject to common law and civil law are prohibited from explicitly
looking for the most expensive way, or the most beneficial way for themselves, to achieve
cross-border payments, nothing in statutory nor private law, so far, requires PSPs to look for
and create the best payment path in light of costs, speed and risks.
90
See Royal Products Ltd v Midland Bank Ltd [1981] 2 Lloyd’s Rep 147 (Webster J.); Midland Bank Ltd v
Seymour [1955] 2 Lloyd’s Rep 147 (Devlin J.); Dovey v Bank of New Zealand [2000] 3 NZLR 641.
91
See German Civil Code [Bürgerliches Gesetzbuch] §662; French Commercial Code [Code du Commerce],
art. L134-1, L134-4.
92
See European Parliament and Council Directive 2015/2366 (OJ 2015 L 337 p 35) (‘Payment Services
Directive II’), art 107.
93
See German Civil Code [Bürgerliches Gesetzbuch] §675c to §676c. See also, Benjamin Geva, ‘The EU
Payment Services Directive: An Outsider’s View’ (2019) 54 Texas International Law Journal 211.
23

B Contrasting Cross-border Payments and Securities Transactions

Regardless of the fact that both brokerage and payments contracts rest on agency law,
significant differences exist between securities brokerage and payments, from a legal, risk and
transactional perspective.
1 Legal Differences
In most jurisdictions, securities settlement is analogised to the law of property and possession,
or in legal terms it is structured as an ius ad rem (a right in relation to a thing). This applies
even where the book-entry has replaced property and possession as part of a system of
dematerialised securities. Payments, by contrast, are in principle claims-based. Funds in a bank
account and cross-border payments each represent an unsecured claim against the financial
institution.
Legal certainty on the law applicable to issues relating to the ownership, holding, transfer and
collateralisation of book-entry securities owned by end-investors and held with intermediaries
across borders is crucial for the efficient operation of the system and the prevention of systemic
issues.94 The choice therefore involves three options for resolving conflicts.
The lex contractus, based on the free choice of law principle, governs the relationship between
the parties to the securities custody contract. Failing a choice, the applicable law is established
on the basis of connecting factors laid down in the law or regulation.
The lex societatis regulates the existing relationship between issuers of shares and bonds and
their shareholders and bondholders and determines the rights and obligations contained in a
“security”.
The lex rei sitae specifies the law applicable to the property effects of securities transactions
and is determined by the location of the register or account.
In 2018 the European Commission published a proposal for a Regulation on the law applicable
to the third-party effects of assignments of claims, which was adopted by the European
Parliament in 2019 with 24 amendments.95 More generally, the original communication
94
Communication from the Commission to the European Parliament, the Council, the European Economic and
Social Committee and the Committee of the Regions on the applicable law to the proprietary effects of
transactions in securities. COM/2018/089 final, <https://eur-lex.europa.eu/legal-
content/en/TXT/?uri=CELEX%3A52018DC0089> (“If there is legal uncertainty over who owns the asset,
depending on which Member State's courts or authorities assess a dispute concerning the ownership of a claim
or a security, the cross-border transaction may be enforceable or not, or might confer the expected legal title on
the parties or not. In case of insolvency, when the questions of ownership and enforceability of transactions are
put under judicial scrutiny, legal risks stemming from legal uncertainty may result in unexpected losses. The
objective of this Communication is to help increase cross-border transactions in securities by providing legal
certainty on the conflict of laws rules at Union level. Enhanced legal certainty will promote cross-border
investment, access to cheaper credit and market integration”).
95
European Parliament, Law applicable to the third-party effects of assignments of claims. European Parliament
legislative resolution of 13 February 2019 on the proposal for a regulation of the European Parliament and of the
Council on the law applicable to the third party effects of assignments of claims (COM(2018)0096 – C8-

[2021]
intended to clarify the law applicable to the property effects of securities transactions by
advising intermediaries to choose the law applicable to property rights based on a law outside
the EU jurisdiction when the securities settlement system is located outside the EU.
Prior to the 2018 communication, the application of lex rei sitae automatically resulted from
the choice of a Member State's law governing the securities settlement agreement. The
reintroduction of the Hague Convention's free choice of law principle may lead parties to the
custody agreement to contemplate the choice of law of a third country through a general choice
of law clause covering all aspects of their relationship, including proprietary aspects, or a
standard clause governing their respective rights and duties exclusively, but it potentially raises
additional issues.96
However, the most interesting element in this context is highlighted in the recitals to this report.
If, on the one hand, “[t]his Regulation concerns the third-party effects of the assignment of
claims […] it covers the transfer of the contracts (such as derivative contracts), in which both
rights (or claims) and obligations are included, and the novation of contracts including such
rights and obligations”97, and, on the other hand, it is clarified that it does not impact on
Directive 2002/47/EC, Directive 98/26/EC, Directive 2001/24/EC and Commission Regulation
(EU) No 389/2013, since the scope of application of the conflict of laws rules contained in the
Regulation and that of the conflict of laws rules contained in those three Directives do not
overlap”.98
These differences have legal consequences.
The first pertains to insolvency. In the case of the financial institution’s insolvency, the owner
of the security (i.e. the client with a security deposit) has a property right in the security that
survives and supersedes all claims by creditors of the insolvent institution. By contrast, in the
case of payments (absent deposit protection schemes or other statutory preferences), the claim
will only be satisfied after those of secured creditors and then pari passu with all other
unsecured creditors (subject to distinctions owed to depositors’ protection rules).
0109/2018 – 2018/0044(COD)). P8_TA(2019)0086, <https://www.europarl.europa.eu/doceo/document/TA-8-

2019-0086_EN.pdf>.
96
The issues raised are mentioned, inter alia, in this paper by European Issuers (‘Position on conflict of law
applicable to securities & claims, 13 July 2018, <http://europeanissuers.eu/positions/files/view/5b487225c9221-
en>). Another clearly dissenting position comes from the UK, namely from the City of London Law Society
(‘Proposed EU Regulation on law applicable to the third party effects of assignment of claims – Why the UK
should opt-out and work to get this proposal changed or scrapped’ (May 24, 2018),
<https://www.citysolicitors.org.uk/storage/2018/05/Proposed-EU-Regulation-on-law-applicable-to-the-third-
party-effects-of-assignment-of-claims-24-05-18.pdf>.
97
<https://www.europarl.europa.eu/doceo/document/TA-8-2019-0086_EN.html>, (17)(Am. 9).
98
<https://www.europarl.europa.eu/doceo/document/TA-8-2019-0086_EN.html>, (18)(Am. 10).
25

The second consequence pertains to the applicable law: while there is some legal uncertainty
as to which law applies to securities settlement, in principle, property law follows the lex rei
sitae (i.e. the law of the place where the property lies).99
With respect to the law applicable to securities held through a securities settlement system in
the EU, Article 9(2) of the Settlement Finality Directive contains rules that create legal
certainty in this respect. More precisely, “the traditional lex rei sitae rule is clarified in relation
to book-entry securities in such a way that the law applicable shall be the law of the so-called
place of the relevant intermediary (PRIMA). According to the PRIMA principle, the rights of
a holder of securities provided as collateral will be governed by the law of the country where
the right to the securities collateral is legally recorded on a register, account or centralised
deposit system”.100 The principal is applicable to all EU countries, but is generally embraced
elsewhere, such as in the US and Japan, and is based on very clear transaction conduct,
including for the purpose of obtaining protection from the effects of insolvency, through the
creation of a valid security interest or an absolute transfer of ownership agreement (e.g., a
repurchase agreement).101
However, the Credit Institutions Settlement Directive brings an exception to this principle, by
applying the lex contractus principle of the agreement containing the netting clause as far as
financial contracts concerning credit institutions are involved. This is an attitude that gives de
facto predictability, confirming that which parties have agreed upon in entering into the
contract.102
As to the applicable private law on claims, there is no harmonized EU law provision, but
contractual obligations stemming from an assignment of claims are ruled by the 1980 Rome
Convention and the Rome I Regulation, namely a uniform set of rules “with regard to (i) the
relationship between the parties to the assignment contract - the assignor and the assignee, and
(ii) the relationship between the assignee and the debtor, [but which] does not include conflict
of laws rules with regard to the proprietary effects of the assignment of the claim, that is, the
effects on third parties of such assignment”.103 As a consequence, such provisions are still
99
The principle applies to both bearer and registered securities, and direct or indirect securities holdings
(through intermediaries). Yet, in indirect securities holding systems, it is far from easy to determine the place
where ‘property lies’, given that in a pyramidal structure the same security may be booked in several accounts
simultaneously. See James Steven Rogers, ‘Conflict of Laws for Transactions in Securities Held through
Intermediaries’ (2006) 39(2) Cornell International Law Journal 285, 303-316; Matthias Lehmann,
Finanzinstrumente (Tuebingen, 2009).
100
BIS, Cross-border aspects of insolvency [could not find a specific date], A22,
<https://www.bis.org/publ/gten06c.pdf>.
101
BIS, Cross-border aspects of insolvency [could not find a specific date], A23, fn 35 and 36,
102
BIS, Cross-border aspects of insolvency [could not find a specific date], A23,
103
Commission Staff Working Document, Impact Assessment, Accompanying the document: Proposal for a
Regulation of the European Parliament and of the Council on the law applicable to the third-party effects of the
assignment of claims. Communication From The Commission To The European Parliament, The Council, The

[2021]
regulated by Member State laws104, notwithstanding the attempts to include a “provision on the
law applicable to the third-party effects of assignments [] in the final text of the Regulation due
to the complexity of the matter and the lack of time to deal with it in the required level of
detail”105, notwithstanding the fact that this leads to significant costs.
Two significant attempts have sought to solve the issue: the Hague Securities Convention and
the UN Convention on the Assignment of Receivables in International Trade. 106 The former is
aimed to removing legal uncertainties for cross-border securities transactions and “[i]n the
absence of a valid choice of law, it is the law of the state where the account is maintained that
shall govern questions of ownership in book-entry securities”.107 On the contrary, the latter
European Economic And Social Committee And The Committee of the Regions on the applicable law to the
proprietary effects of transactions in securities. {COM(2018) 96 final} - {COM(2018) 89 final} - {SWD(2018)
53 final}, Brussels, 28.3.2018. SWD(2018) 52 final/2,
< https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52018SC0052(01)&from=EN>, at 13.
104
“Private law, i.e. the law of contract and property, as well as securities law have developed along national
lines. Therefore, Member States have different legal solutions to address transactions in claims and securities.
When the transaction has a cross-border element, conflict of laws rules apply to determine which national law of
all those potentially applicable should apply.
In cross-border transactions in claims and securities two elements are governed by conflict of laws rules: (1) the
contractual element, which refers to the parties’ obligations towards each other under the transaction; and (2) the
proprietary element, which refers to the transfer of rights in property and which therefore affects third parties.
EU conflict rules relating to the contractual element exist in relation to claims. EU conflict rules relating to the
contractual and proprietary elements exist in certain areas relating to securities. However, no EU conflict of
laws rules exist on the proprietary aspect of assignment of claims. The main difference between the areas of
claims and securities is that, while there are no EU conflict of laws rules on the proprietary element of
assignments of claims, three Directives include conflict of laws rules on the proprietary element of transactions
in securities which, however, are not identically worded.” (Commission Staff Working Document, Impact
Assessment, Accompanying the document: Proposal for a Regulation of the European Parliament and of the
Council on the law applicable to the third-party effects of the assignment of claims. Communication From The
Commission To The European Parliament, The Council, The European Economic And Social Committee And
The Committee of the Regions on the applicable law to the proprietary effects of transactions in securities.
{COM(2018) 96 final} - {COM(2018) 89 final} - {SWD(2018) 53 final}, Brussels, 28.3.2018. SWD(2018) 52
final/2,
< https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52018SC0052(01)&from=EN>, at 11).
105
< https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52018SC0052(01)&from=EN>, at 14, fn
28 and 32.
106
< https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52018SC0052(01)&from=EN>, at 15-17.
107
27

“provides that the law of the State in which the assignor is located governs the priority of the
right of an assignee in the assigned receivable over the right of a competing claimant”108
(Article 22).
The third consequence pertains to the type of assurance implicitly given. Securities of the same
type are fungible. Following the property analogy, the seller only promises to deliver securities
of the specified type in the specified amount, not a specific security. This fungibility of
securities is essential for liquid securities markets. Likewise, in payments, the payer owes the
value agreed on, which has only to be delivered in any form that is legal tender in the country
of the payee, or her PSP.
The fourth consequence pertains to the role of the financial institution involved. Securities
brokers are “fiduciaries”, i.e. their duty is to put their client’s interests first. Securities
brokerage as such is an “off-balance sheet business”. By contrast, funds in a bank account
accrue to the bank’s balance sheet, so bank deposits, in principle, are “on-balance business”.
In case of payments, however, when payment-specific regulation requires certain (non-bank)
payment institutions to segregate client funds and treat them separately from the institution’s
own funds based on rules aiming at maintaining the nominal value of the funds used for
payment purposes, the treatment at law of securities and payments will be largely the same.
2 Types of Risk
We see two major differences in the types of risk.
First, as a result of their claims-based nature, payments are characterised by financial
counterparty or payment risk (referred to as Herstatt risk), whereas securities transfers do not
come with financial risk, but operational settlement risk (i.e., the risk the promised securities
will never be delivered).
Second, while there is no FX risk in securities title transfers, there is FX risk in cross-border
payments effected between currencies. Depending on the currency in which the contract is
written, the payee or payer will bear the FX risk.
3 Transactional Differences
In securities settlements, something needs to be transferred. The same thing, at least in legal
terms, changes hands. In the case of cross-border payments there is no transfer of “the same

< https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52018SC0052(01)&from=EN>, at 16.
108
< https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:52018SC0052(01)&from=EN>, at 16-7.

[2021]
thing”. In cross-border payments there is no physical transfer of currencies. The funds are not
sent across borders; instead, an account is credited in one jurisdiction and a corresponding
amount debited in the other.109 This is true for both the system of correspondent banks and
closed-loop systems (such as PayPal), except that in the latter case different legal entities of
the same financial conglomerate interact.
This system enables payment institutions to exchange book positions in FX-adjusted terms.
The respective amount is then credited and debited to the correspondent institution’s clients so
that these clients experience the transaction as “payments” in foreign currency. The more
correspondent institutions involved in a transaction, the more intermediate booking of
transactions is necessary, the longer the transaction will take, the greater the Herstatt risk along
the payment intermediary chain, and the higher the total cost. Naturally, closed-loop systems
have speed advantages since they can reduce time by virtue of technical standardisation.
However, clients bear the counterparty risk of being connected to “that group”. Under certain
conditions, liabilities to some group entities may put the finalisation of the transaction at risk.
C Implications
In our view, the differences laid out above do not prevent the introduction of a “best execution”
principle for payments for three reasons. First, payment-specific legislation has reduced the
differences in the legal treatment of payments and securities. Second, the structure of modern
digital transactions aims to mimic securities settlement. Third, the risks for clients and
intermediaries relating to cross-border payments have become more similar to that of securities
brokerage, partly due to the activity of correspondent banks, and partly due to the first and
second developments.
1 Impact of Payment-specific Legislation

In jurisdictions with specific regulations for specialised payment services providers (usually
separate from those applying to banks and credit institutions), there are typically legal
requirements to segregate clients’ funds from other funds, and safeguard the former, thereby
inhibiting the use of funds for purposes other than payments.110 Further, in e-money schemes,
an often used substitute for payments services, e-money service providers are increasingly
requested to
“safeguard customer funds by: (i) placing the relevant funds in segregated accounts111
or (ii) setting up appropriate measures to protect customer funds. With regard to the
109
For an explanation of different methods of cross-border payments, ‘Cross-border Payments’, Bank of
England (Web Page, 15 June 2021) <https://www.bankofengland.co.uk/payment-and-settlement/cross-border-
payments>.
110
See Payment Services Directive II, art. 10: ‘Safeguarding requirements’; The FCA has implemented the
PSD2: FCA, Implementation of the revised Payment Services Directive (PSD2): Approach Document and Final
Handbook Changes (Policy Statement PS17/19, 2017) <https://www.fca.org.uk/publication/policy/ps17-
19.pdf>.
111
The concept of segregated accounts here is considered as the consequence of maintaining full reserves.
29

latter, [there are] two typical types of requirement: (i) funds must be placed in a trust
account, administered by a trustee, solely for the benefit of the customers; or (ii)
customer funds must be covered by insurance or a comparable guarantee from an
insurer or a bank.”112
In the US, the FDIC offers deposit insurance pass-through protection on prepaid transaction
accounts, including prepaid cards and e-money products issued in the US, if:
• the prepaid transaction service is open-loop,
• the underlying funds are held in a segregated deposit account at a FDIC-covered
depository institution,
• the e-money account holders (not the payment institution) are the principal owners of
the funds in the account, and
• up-to-date records on the identity of the e-money account holders and the amount of
funds in the accounts are kept.113
Because US federal government payments, such as tax refunds and social security payments,
can only be routed to prepaid accounts that meet these requirements, there is de facto pressure
to comply with them. At least one major US retailer modified the design of its prepaid card
account service in order to qualify for FDIC pass-through insurance.114
While this does not assign a certain fraction of value on a per-client basis to each client, as in
the case of securities brokerage, the safeguarding rules render the collective funds of all clients
a separate item on the payment institution’s balance sheet, both legally and economically.
While held in the name of the payment institution, the funds become subject to entirely different
duties and obligations. They must be maintained and managed in the interest of the clients
collectively, similar to the management of collective investment schemes. For managers of
collective investment schemes, the legal position of a fiduciary is clear. Essentially, legislation
relating to payments and payments’ substitutes has diminished the legal difference between
claims and possession.
2 Structure of Contemporary Cross-Border Payment Systems

Cross-border payment systems further resemble securities brokerage, through the use of
securitisation. For instance, several payment systems translate the transaction first into a
derivative (often a credit default swap or cross-currency swap) in which one currency (being a
112
Committee on Payments and Market Infrastructures, Payment Aspects of Financial Inclusion (n 10) 26-7.
See, for instance, Council Regulation (EC) No 110/2009 (OJ 2009 L 267 p 7) art 7 (‘EU E-Money Directive’) as
well as Article 60 of the Proposal for a Regulation of the European Parliament and of the Council on Markets in
Crypto-assets, and amending Directive (EU) 2019/1937. COM/2020/593 final. Further, the Alliance for
Financial Inclusion recommends the adoption of similar rules to regulators of developing countries, see AFI,
Policy Model for E-Money, 6-7 <https://www.afi-global.org/wp-
content/uploads/2019/10/AFI_DFS_Emoney_AW_digital_0.pdf>.
113
FDIC Law, Regulations, Related Acts, §1020.220 <https://www.fdic.gov/regulations/laws/rules/8000-
1600.html>.
114
Committee on Payments and Market Infrastructures, Payment Aspects of Financial Inclusion” (n 10) 27.

[2021]
monetary value only within the jurisdiction of that central bank) is then reflected in the recipient
jurisdiction where its value is assessed and converted into the currency of the recipient
country.115 The next step in the evolution of the structure of payments has been tokenisation, in
which the right to be paid is tokenised, and property in the token is transferred. This is the basis
for many sovereign or private digital currency projects. Tokenisation is substantially equivalent
to the securitisation of a cash-flow. The issuance process legally turns the funds (as mere claims
against the institution) into an ius ad rem in the form of the token. This replaces the unwanted
features of payments, in particular Herstatt risk (counterparty and payment risk), with the
operational (settlement) risk characteristics of securities. Again, we see the obligations that
arise from the means of payments legally restructured into an approach more akin to that of
securities.
3 Risks and Incentives

The points discussed above– legislation and legal structuring of payment systems – result in an
incentive structure on the side of payment service providers and e-money schemes that is
similar to securities brokerage, in three respects:
(1) these payment institutions do not benefit from an increase or decrease of funds used for
payments;
(2) these payment institutions are not concerned about costs related to the transfer and
management of the funds because they can pass the costs to the payee or payer; and
(3) these payment institutions retain some discretion as to the way in which they route the
client’s funds to achieve the objective of cross-border payments; in particular, they can pick
their preferred counterparty, similar to a broker being allowed to pick their preferred trading
venue – in the absence of a best execution obligation.
Likewise, the risks for clients have become similar to that of securities brokerages:
(1) clients do not bear the risk of the payment institutions’ solvency, but the insolvency risk
related to that of the segregated account only, which is (often) in the name of a different
institution;116
(2) given that the investment of the safeguarded funds is strictly regulated with a view to
ensuring the maintenance of the nominal value of the funds, the interest in the financial state
of the payment institution is limited. Instead, clients become interested in the effectiveness of
115
Jessie Cheng and Benjamin Geva, ‘Understanding Block Chain and Distributed Financial Technology: New
Rails for Payments and an Analysis of Article 4A of the UCC’, Business Law Today
<https://www.americanbar.org/groups/business_law/publications/blt/2016/03/05_cheng/> (highlighting that
only robust payment rules, as the ones adopting the Uniform Commercial Code (UCC), allow payments using
distributed financial technology to discharge the underlying obligation to pay).
116
Committee on Payments and Market Infrastructures, “Payment Aspects of Financial Inclusion” (n 10) 26-27.
See also Tom Kokkola (ed), The Payment System: Payments, Securities and Derivatives, and the Role of the
Eurosystem (European Central Bank Report, 2010) 120-22
<https://www.ecb.europa.eu/pub/pdf/other/paymentsystem201009en.pdf>.
31

the institutions’ operations in managing the clients’ collective funds. In economic terms,
payment (counterparty) risk has been translated into settlement (operational) risk; and
(3) besides these operational risks, clients bear agency risk, that is the risk that the payment
institutions route the payer’s funds to the payee’s accounts in a poor manner (e.g. at high costs
or low speed), because the institutions made a poor choice in good or bad faith. This is the
same risk brokerage clients face when the broker exercises its discretion about the appropriate
execution venue.
V TRANSFORMING CROSS-BORDER PAYMENTS THROUGH BEST EXECUTION?

As the transactional structure and risks of cross-border payments more closely replicate those
of securities brokerages, it makes more sense to regulate these matters analogously and
implement best execution as part of the law of payments. Why this is so, we explore next.
A From One to Many Links?

Best execution could potentially enhance the number of links each institution establishes for
settling transactions. This addresses the issues that arise because some bank customers are
effectively locked into their payment relationship, negating incentives provided by
competition. Moreover, some banks have no access to certain currencies required for their
clients’ payment transactions, either for lack of correspondent links or lack of currency trading
liquidity, creating dependencies on globally active PSPs for rarely traded currencies. Yet, best
execution, as a non-contractible, mandatory rule of payments regulation, could require PSPs to
look for cheaper, faster, and less risky means to execute transactions, compared to the
conditions offered by their correspondent banking network. Options stimulated by best
execution could include using non-bank service offerings (e.g. TransferWise, PayPal), or
enhancing the efficiency of the existing correspondent banking network through better
technical integration to maintain liquidity in that network (for instance, this could be the effect
of lower fees and higher speed as a result of increasing digitalisation).
However, this requires a version of best execution that requires banks to look for, or even
create, choice and variety on behalf of their clients. Introducing best execution as a legal
principle would require banks to look for the best way to effect cross-border payments from
their client’s perspective.
Naturally, if banks only look at the individual client relationship they will not establish a new
link given that the establishment and maintenance of a link is costly especially for one
transaction. Hence, the principle must be designed in a way that best execution means offering
choice to clients between different order routing channels. For instance, we suggest a rule that
requires banks that engage in cross-border transactions to maintain at least three correspondent
banking channels for any given transaction.
A counterargument may be that, in practice, fewer and fewer links will be established. Hence,
asking banks to set up more links (adding to costs of cross-border transactions) could lead to
more banks focusing on national payment business only, and thus further enhancing

[2021]
concentration. This counterargument could be addressed by way of regulation. Regulators

could make a minimum number of correspondent relationships – for example five – mandatory
as a precondition for entering into cross-border banking, and ask for a justification if payment
institutions maintain and offer to their clients fewer banking alternatives, thereby enhancing
the costs of monopoly structures. Large, globally active banks would have to comply with such
a requirement, since abandoning their ability to provide global payments would not be a viable
alternative.
B New Technological Platforms

An alternative to mandating more links of the current type would be for regulators to develop
further technical infrastructure that facilitates many links among many banks for the purposes
of cross-border banking. This would be challenging with the current payments’ platforms and
an entirely new multilateral network may provide the better option.
While various alternatives exist, one technical option that could readily enhance banking links
is the use of DLT. This would allow the simultaneous access of many banks to the same data
stream. If best execution was introduced as a legal principle, then a (regulatorily sponsored or
privately established) distributed ledger system could be used to create competition among
payment institutions, relying on the information distribution function inherent to DLT, and
again similar to the role regulated stock exchanges play in securities brokerage.
Assume PSP1, located in country A, wants to transfer funds to country B. PSP1 announces the
payment via the DLT using an announcement algorithm. Now two types of PSPs may respond
(again by way of algorithms): the first group consists of PSPs with a direct representation in B,
interested in receiving the A currency. The second group consists of PSPs engaging in multi-
aggregate operations, for instance PSPs in country C with links to PSPs in both A and B, that
are interested in swapping their position in C into positions in A and B currency. Both groups
disclose their currency transfer rates, any additional costs, and the offered settlement time by
way of DLT. PSP1 then accepts the offer that represents best execution.
Figure 3: DLT as Best Execution network
33

To establish a distributed ledger system as a best execution network, attention must be given
to, inter alia, system governance, participation, contributions, messaging standards, ensuring
equal access to promote competition at equal terms. We have examined these details
elsewhere.117 DLT paired with a well-designed best execution principle could certainly lead to
a far greater number of links and greatly enhanced competition.
VI POLICY CONSIDERATIONS
There are a number of considerations relevant to drafting a best execution rule for cross-border
payments.
A Institutional Coverage?
The institutions to be covered by the best execution rule must be identified. As the purpose of
introducing best execution is to spur competition between the system of correspondent banks,
closed-loop systems and any potential new tech-driven payment solution (including digital
currencies), any rule should optimally cover:
(a) all banks engaged in cross-border payments,
(b) all payment and e-money service providers,
(c) all closed-loop systems that do not meet the aforementioned definitions, and
(d) all functional payment substitutes, including systems relating to payment-oriented crypto
assets.
Interestingly, if best execution becomes part of payments regulation, it can be mandated from
all holders of licenses active within a currency area; and the principle would then govern all
choices on payment routes and methods made by license holders for their clients’ transactions
within that jurisdiction, and would influence the offers by PSPs further up the chain. In turn,
jurisdictions can implement a best execution principle unilaterally and do not need to wait for
a multilateral policy recommendation. Each jurisdiction can require its own institutions to
execute at best terms for their clients, and thereby prompt a legal requirement to search for the
best, rather than the best friends’, option. In turn, implementation of best execution neither
requires large-scale international standard setting, nor technical implementation projects. PSPs
can be required to engage in searches for the best execution by simple amendments of payments
regulations.
117
Dirk A Zetzsche, L Anker-Sørensen and Maria Lucia Passador, DLT-based Enhancement of Cross-border
Payment Efficiency – a Legal and Regulatory Perspective (2021, BIS Working Paper, forthcoming).

[2021]
B Design of a Best Execution Rule

Four building blocks underpin effective best execution rules: (1) the fiduciary status of the
service provider, (2) disclosure rules vis-à-vis clients and regulators on the execution terms and
priorities of individual transactions, (3) market intelligence on the side of regulators on the pre-
and post-execution environment, based on large scale data transfer from financial institutions
offering, accepting, and executing transactions, and (4) robust public and private enforcement.
Most notably, items (2) to (4) together aim at efficient enforcement in an environment where
the fiduciary principle (1) necessarily comes with discretions.
Reflecting this approach, in October 2021, the G20 committed, inter alia, in respect of
transparency in cross-border payments:
“All payment service providers to provide at a minimum the following list of
information concerning cross-border payments to payers and payees by end-2027: total
transaction cost (showing all relevant charges, including sending and receiving fees
including those of any intermediaries, FX rate and currency conversion charges); the
expected time to deliver funds; tracking of payment status; and terms of service.)“118
Drafting the principle itself is inexpensive. Reporting on execution terms of payments, to date,
only includes costs; whereas PSPs need to disclose why they executed a transaction in a specific
way – costs are not the entire story. Market intelligence on payments is on the rise with more
central banks seeking to obtain transaction data by introducing transaction reporting
requirements.119 The robust enforcement environment rests, on one hand, upon sophisticated
clients negotiating execution terms (which is already happening), and on the other hand, upon
supervision, sector-wide inquiries and sanctions to ensure that retail and SME clients also
benefit from advanced order routing systems.
C Allocating Volume-based Commissions?

The best execution principle must consider the practice of commissions for routed volume.
These commissions incentivise banks to bundle liquidity and enhance efficiency. Yet the
liquidity used is their clients’, and thus there may well be a disincentive to look for the best
route for individual payments.
We propose a two-pronged approach to these commissions. First, commissions received and
paid by the payment institutions should be disclosed to both regulators and the public. This
disclosure could then empower new entrants to provide better offers if their technologies allow.
Second, we propose a default rule that allocates volume-based commissions pro rata to clients,
based on transaction volume, yet a fair share of the commission (which could perhaps be 25-
50% of the commission received) would be allocated to the payment institution for its efforts
to bundle liquidity, as bundling of liquidity itself requires technical efforts and investments.
118
FSB, “Targets for Addressing the Four Challenges of Cross-border Paymetns“ (October 2021).
119
See Richard Comotto, Recommendations for Reporting Under SFTR (ICMA Report, 24 February 2020)
<https://www.icmagroup.org/assets/documents/Regulatory/Repo/SFTR/ICMA-recommendations-for-reporting-
under-SFTR-240220.pdf>.
35

Sophisticated clients may negotiate a different rate, while a greater share could be required to
be paid to retail clients.
D Openness to Innovation
Payment institutions must consider costs, risks and speed when deciding upon the order route.
With respect to risks, payment institutions would need to be free to choose the best route in
light of the technology available, after considering at least three different routes. Their former
correspondent network may or may not be “best execution” in that sense. Further, the standard
must be implemented in a way that ensures openness to innovation, so that no new routes in
terms of technology or routing systems should be disqualified as “too risky” as long as all
participants have received regulatory approval to operate.
E Minimum Number of Order Routes to be Considered

We propose to limit freedom of contract in one important respect: by requiring that payment
institutions must look for, and compare, at least three different offers in terms of costs, speed,
and risks. For that purpose, several offers received through one distributed ledger would count
as multiple offers. An exemption from this requirement may be appropriate for rarely traded
currencies, subject to regulatory approval.
F Facilitating Enforcement: A RegTech Approach

As with any fiduciary principle, enforcement will be key. Among clients, only institutional
ones will have the means to enforce best execution. We suggest that payment regulators and
central banks will be best placed to enforce best execution if they are empowered to use the
same means as securities regulators: disclosure, automated standardised data transfer on pre-
and post-execution offers, and algorithms and AI. Only through this RegTech approach will
regulators be positioned to meaningfully argue against pay-for-volume arrangements that
benefit “best friends”.
VII CONCLUSION
The recently introduced safeguarding and segregation requirements for specialised payment
services and e-money providers have reduced the differences in the legal treatment of payments
vis-à-vis securities, and some of the incentives that originally characterised payments business
by financial institutions. While these rules do not yet apply to banks, which serve in developed
countries as the most important PSPs, they apply to new types of payment providers, including
innovative closed-loop systems and the mobile payment providers that underpin payments for
billions of people in developing countries. The result of these requirements is an incentive
structure for cross-border payments that begins to resemble that of securities brokers. PSPs do
not benefit from an increase, nor suffer from a decrease, of assets provided to them by clients
when engaged in payment operations, and clients do not bear the financial insolvency risk of

[2021]
the PSPs as they did previously. However, PSPs may be indifferent to fees charged by other
payment institutions in the chain, given their ability to pass these fees onto their clients (plus a
profit margin), and have a strong incentive to consider their own interests, such as benefits
from their “best friends” network (such as commissions and kick-backs), when determining
order routing.
In such a setting, introducing a “best execution” obligation will be potentially game changing.
This legal standard requires the payment institution to exclusively consider their clients’
interests when choosing the route the order is to take. In turn, we would expect that PSPs will
develop digital routing systems (including possibly by way of distributed ledger technology)
to seek the best cross-border liquidity among multiple offers. Furthermore, if PSPs are required
to consider multiple offers for order routing (we propose a minimum of three), potentially more
links between correspondent banks, new services providers from the FinTech space, and public
payment networks (including regional integration systems) will be established, assisting the
identification of excess liquidity in less frequently traded currencies.
In securities brokerage, best execution has led to lower fees, higher speed, and more choice of
execution venues. A best execution principle in payments could be similarly transformative in
reducing fees, accelerating transactions, and driving greater competition, particularly by way
of technological innovation, in cross-border payments.
37

SSRN Id3834335

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

SSRN Id3834335

Uploaded by

Copyright:

Available Formats

UNSW

THE UNIVCRSITY Of NCW SOUTH WAL[S

University of New South Wales Law Research Series

From ‘Best Friends’ to ‘Best

Dirk A. Zetzsche, Ross P. Buckley, Douglas W.

Electronic copy available at: https://ssrn.com/abstract=3834335

DIRK A ZETZSCHE*, ROSS P BUCKLEY**, DOUGLAS W ARNER***

Cross-border payments suffer from slowness, poor transparency and limited

Electronic copy available at: https://ssrn.com/abstract=3834335

imposed unilaterally by any jurisdiction’s regulator interested in reducing costs of

Keywords: Best execution, international payments, correspondent bank.

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

F Facilitating Enforcement: A RegTech Approach .......................................................................... 36

Electronic copy available at: https://ssrn.com/abstract=3834335

The dysfunctional state of cross-border payments results in higher-than-necessary transactions

Electronic copy available at: https://ssrn.com/abstract=3834335

II RENTS IN THE CROSS-BORDER PAYMENTS CHAIN

Electronic copy available at: https://ssrn.com/abstract=3834335

Rather, the process as described in Figure 1 involves:

Then, if no (electronic) fund transfer/payment system is involved:

Then, if an (electronic) fund transfer/payment system (e.g. FedWire, EURO1/STEP1/STEP2,

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

B Closed Loop Systems

Electronic copy available at: https://ssrn.com/abstract=3834335

C Regulatory Approaches to Enhancing Competition and Efficiency

2 Infrastructure Provision and Operations: Reducing Infrastructure Costs

Electronic copy available at: https://ssrn.com/abstract=3834335

III INTRODUCING BEST EXECUTION: LEARNING FROM SECURITIES REGULATION

Electronic copy available at: https://ssrn.com/abstract=3834335

A Best Execution: A Principle of Securities Regulation

B Legal Framework of Best Execution

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

606(a) of Regulation NMS now require broker-dealers to publish transaction reports on a

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

The FCA redefined the best execution criteria in 2021:

Electronic copy available at: https://ssrn.com/abstract=3834335

C Lessons from the Securities Context

Electronic copy available at: https://ssrn.com/abstract=3834335

University of Pennsylvania Business Law Journal 273.

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

IV BEST EXECUTION FOR CROSS-BORDER PAYMENTS?

A Agency Law Concepts in Payments

Electronic copy available at: https://ssrn.com/abstract=3834335

B Contrasting Cross-border Payments and Securities Transactions

Electronic copy available at: https://ssrn.com/abstract=3834335

0109/2018 – 2018/0044(COD)). P8_TA(2019)0086, <https://www.europarl.europa.eu/doceo/document/TA-8-

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

Electronic copy available at: https://ssrn.com/abstract=3834335

proprietary effects of transactions in securities. {COM(2018) 96 final} - {COM(2018) 89 final} - {SWD(2018)

Electronic copy available at: https://ssrn.com/abstract=3834335

1 Impact of Payment-specific Legislation

Electronic copy available at: https://ssrn.com/abstract=3834335

2 Structure of Contemporary Cross-Border Payment Systems

Electronic copy available at: https://ssrn.com/abstract=3834335

DIRK A ZETZSCHE, ROSS P BUCKLEY, DOUGLAS W ARNER