Does the deployment of algorithms combined with direct electronic access increase conduct risk? Evidence from the LME
| DOI | https://doi.org/10.1108/JFRC-04-2022-0046 |
| Published date | 08 August 2022 |
| Date | 08 August 2022 |
| Pages | 220-236 |
| Author | Alexander Conrad Culley |
Does the deployment of
algorithms combined with direct
electronic access increase conduct
risk? Evidence from the LME
Alexander Conrad Culley
Southampton Business School, Faculty of Business and Law,
University of Southampton, Southampton, UK
Abstract
Purpose –The purpose of this paper is to examine the effectiveness of two regulatory initiatives in
developing awarenessof conduct risk associated with algorithmic and direct-electronicaccess (DEA) trading
at broker-dealers: the UK Financial Conduct Authority’s algorithmic trading compliance in the wholesale
markets and Commission Delegated Regulation 2017/589 (CDR 589) to the second Markets in Financial
InstrumentsDirective.
Design/methodology/approach –A qualitative examination of 15 semi-structured interviews with
representativesof London Metal Exchange member firms, their clientsand regulators.
Findings –This paper finds that the key conduct related messages in algorithmic trading compliance in
the wholesale markets may not yet be fully embedded at broker–dealers. This is because of a perceived
simplicity of the algorithms deployed by broker dealers or, alternatively, a lack of reflection on their
impact. Conversely , a concern exists that clients ’deploymentof algorithms on DEA channels provided by
broker–dealers increase conduct risk. However, the threat of harm posed by clients is not envisaged in
current definitions of conduct risk. Accordingly, CDR 2017/589 does not currently require firms to
evaluate clients’awareness of it.
Research limitations/implications –This study’sfindings are limitedto the insights provided by 15
participants.
Originality/value –This paper contributes to existing research by deepening understanding of
conduct risk arising from algorithmic trading and DEA. To account for the potential harm arising from
clients’activities, this paper proposes a revision to Miles’sdefinition of conduct risk. This is
complemented by a proposed amendment to CDR 2017/589 to require evaluation of clients’understanding
of conduct risk.
Keywords MiFID II, Algorithmic trading, Market abuse, Conduct risk direct electronic access,
London Metal Exchange
Paper type Research paper
1. Introduction
[...] we are seeing new complexities emerge around issues like disintermediation. [...] a split
from the principal/agent model that’s underpinned monetary transactions for most of human
history. (Wheatley, 2015)
Thank you to Professor Simon Wolfe at the University of Southampton for providing guidance and
feedback during the preparation of this paper. Thanks also to the investment professionals who
participated in this project, and who helped me recruit participants. Finally, thank you to my family
for their patience.
JFRC
31,2
220
Received3 April 2022
Revised10 June 2022
Accepted11 July 2022
Journalof Financial Regulation
andCompliance
Vol.31 No. 2, 2023
pp. 220-236
© Emerald Publishing Limited
1358-1988
DOI 10.1108/JFRC-04-2022-0046
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/1358-1988.htm
These sentiments, expressed by Martin Wheatley, then Chief Executive Officer of the UK
Financial Conduct Authority (FCA), in response to retail developments would also inform
regulatory initiatives concerning the proliferation of algorithmic trading and direct
electronic access in the wholesalemarkets [4].
In Algorithmic Trading Compliance in the Wholesale Markets [5], hereinafter
“Algorithmic Trading Compliance”the FCA differentiated between two forms of
algorithmic deployment, borrowing heavily from Commission Delegated Regulation 2017/
589 (CDR 2017/589) to thesecond Markets in Financial Instruments Directive (MiFIDII):
investment decision or “trading”algorithms: “make automated trading decisions by
determining which financial instrument should be purchased or sold”;and
order execution algorithms: “optimise order-execution processes by automatic
generation and submission of orders or quotes, to one or several trading venues
once the investment decision has been taken”.
Many market participants relyon DEA to deploy their algorithms. This is because they are
not themselves members of a trading venue such as the London Metal Exchange (LME).
DEA is defined in Article 4(1)(41)of MiFID II [7]as:
[...] an arrangement where a member or participant or client of a trading venue permits a person
to use its trading code so the person can electronically transmit orders relating to a financial
instrument directly to the trading venue.
Algorithmic Trading Compliance also encouraged firms to “identify and reduce potential
conduct risks created by their algorithmic trading strategies”. Culture Audit in Financial
Services: Reporting on Behaviour to Conduct Regulators (Miles, 2021) offers the following
definition of conduct risk:
“Subset of behavioural risk: potential cost resulting from employees or suppliers breaching
conduct rules; or more generally, business loss following staffor supplier behaviour (especially
towards customers) that undermines trust or value in the business or creates a ‘disorderly
market’. Includes managers’inaction failing to anticipate and overcome biases or asymmetry in
transactions”.
Taking a qualitative approach,this paper seeks to draw conclusions about the effectiveness
of regulatory initiatives such as Algorithmic Trading Compliance and CDR 2017/689 in
addressing potential conduct risks arisingfrom the deployment of algorithms and DEA by
broker-dealers from the analysis of 15 explorative semi-structured interviews held with
experts in trading at the LME.
The paper finds that the key conduct related messages in Algorithmic Trading
Compliance in the WholesaleMarkets may not yet be fully embedded at broker-dealers.This
is because of a perceived simplicity of the algorithms deployed by broker-dealers or,
alternatively, a lack of reflection on their impact. Conversely, a concern exists that clients’
deployment of algorithms on DEA channels provided by broker-dealers increases conduct
risk. However, the threat of harm posed by clients is not envisaged in current definitionsof
conduct risk. Accordingly, CDR 2017/589 does not currently require firms to evaluate
clients’awarenessof it.
To account for the potential harm arising from clients’activities, the paper proposes a
revision to Miles’sdefinition of conduct risk. This is complemented by a proposed amendment
to CDR 2017/589 to require evaluation of clients’in section two understanding of conduct risk.
The rest of this paper is structured as follows. First, a literature review surveys the
conduct risks that are typically associated with the deployment of algorithms and trading.
Deployment of
algorithms
221
To continue reading
Request your trial