Chapter 19 Further Reading: Market Surveillance: Detecting Manipulation and Abuse

This reading list is organized into five sections: primary regulatory sources, regulatory guidance and supervisory publications, enforcement case studies, academic literature, and technology resources. Practitioners at different career stages should prioritize different sections: regulatory text for those building legal foundations; supervisory guidance for those designing operational surveillance programs; academic literature for those seeking quantitative depth; and technology resources for those evaluating or implementing surveillance platforms.

Section 1: Primary Regulatory Sources

EU Market Abuse Regulation (EU MAR)

Regulation (EU) No 596/2014 of the European Parliament and of the Council of 16 April 2014 on market abuse EUR-Lex: eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32014R0596

The foundational text of the EU market abuse framework. Key articles for surveillance practitioners: - Article 7: Inside information — the four-element definition, with separate provisions for commodity derivatives and emission allowances - Article 8: Insider dealing — primary and secondary insiders, definition of "use" of inside information - Article 9: Legitimate behaviors — the pre-commitment safe harbor, market maker exemption, and execution of orders - Article 10: Unlawful disclosure — the tipping prohibition and its relationship to legitimate disclosure under Article 17 - Article 12: Market manipulation — the three main categories, with Annex I indicators that point toward manipulation - Article 13: Accepted market practices — the regulatory framework for practices that appear manipulative but are determined to be legitimate - Article 16: Obligation to detect and report — the STOR obligation, its scope, and the promptness requirement - Article 17: Public disclosure of inside information — the obligation to disclose and the MAR delay mechanism

UK Market Abuse Regulation (UK MAR)

The Market Abuse (Amendment) (EU Exit) Regulations 2019, as it forms part of UK domestic law under the European Union (Withdrawal) Act 2018 Available via legislation.gov.uk

UK MAR is substantively identical to EU MAR in its prohibition framework. Practitioners should read it alongside the FCA's MAR guidance to understand post-Brexit divergence in technical standards and supervisory approach. The FCA's primary rulebook reference for MAR obligations is MAR Sourcebook in the FCA Handbook (available at handbook.fca.org.uk).

Commission Delegated Regulations under EU MAR

ESMA has published numerous delegated regulations and implementing technical standards (ITS) under EU MAR. Key instruments for surveillance practitioners include:

Commission Delegated Regulation (EU) 2016/957: Indicators of market manipulation (Annex I indicators in operational form)
Commission Delegated Regulation (EU) 2016/908: Accepted market practices
Commission Delegated Regulation (EU) 2016/522: Exemptions for buy-back programmes and stabilisation
Commission Implementing Regulation (EU) 2016/523: Format and template for STORs

MiFID II — Recording and Surveillance Obligations

Directive 2014/65/EU (MiFID II), Article 16(7) EUR-Lex: eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX%3A32014L0065

Article 16(7) requires investment firms to record telephone conversations and electronic communications relating to the reception, transmission, and execution of client orders and own account transactions. Read alongside:

Commission Delegated Regulation (EU) 2017/565, Articles 72-76: Detailed requirements for recording, retention, and access
FCA COBS 11.8: UK recording requirements post-Brexit

Section 2: Regulatory Guidance and Supervisory Publications

ESMA Guidelines on the Market Abuse Regulation

ESMA Guidelines on the Market Abuse Regulation (ESMA70-145-111) Available at: esma.europa.eu

ESMA has published a comprehensive set of guidelines covering accepted market practices, inside information, and the delay of disclosure of inside information. For surveillance practitioners, the most directly relevant guidelines are:

Guidelines on delayed disclosure of inside information: When and how issuers may delay disclosure under Article 17(4), and its implications for surveillance
Guidelines on market manipulation indicators: Operational guidance on applying Annex I indicators to real-world trading patterns — essential reading for rule calibration
Guidelines on persons receiving STORs: Expectations for NCAs, including information-sharing with ESMA

The FCA's Market Watch newsletter is the single most important source of supervisory guidance on the practical operation of a surveillance program. Published approximately six times per year, it combines statistical analysis of STOR flows, thematic reviews of firm practices, and direct guidance on FCA expectations. Key editions for Chapter 19 topics:

Market Watch 69 (2021): Communications surveillance — FCA expectations for firms' monitoring of voice and electronic communications; common weaknesses observed in supervisory reviews
Market Watch 71 (2022): STOR quality review — analysis of STOR submissions received, including examples of high-quality and low-quality STORs; guidance on what the FCA expects in terms of evidence and reasoning
Market Watch 73 (2022): Algorithmic trading and surveillance — expectations for firms running algorithmic strategies to surveil their own algorithms' behavior for potential manipulation
Market Watch 75 (2023): Front-office surveillance controls — expectations that surveillance is embedded in front-office culture, not just operated as a back-office compliance function

All editions available at: fca.org.uk/publications/newsletters/market-watch

FCA Guidance on STORs

FCA SUP 15.10 (Suspicious Transaction and Order Reports) FCA Handbook: handbook.fca.org.uk

Provides detailed guidance on who must report, what must be reported, and how to submit. Read alongside the technical standards in Commission Implementing Regulation (EU) 2016/523 (retained in UK law).

IOSCO Principles for Financial Market Infrastructures — Market Surveillance Provisions

IOSCO Report on Market Surveillance (FR06/2021) iosco.org

The International Organization of Securities Commissions has published extensive guidance on cross-border market surveillance, manipulation detection, and cooperation between regulators. IOSCO's Multilateral Memorandum of Understanding (MMoU) on regulatory cooperation is the framework within which the FCA and non-EU regulators share STOR information.

IOSCO Report on the Use of Artificial Intelligence and Machine Learning by Market Intermediaries and Asset Managers (2020) — addresses the use of ML in surveillance contexts, governance requirements, and explainability considerations.

FCA Algorithmic Trading Compliance in Wholesale Markets (2018)

FCA Thematic Review TR18/3 fca.org.uk/publications/thematic-reviews/algorithmic-trading-compliance-wholesale-markets

Contains detailed findings from the FCA's supervisory review of algorithm surveillance programs at major UK investment firms, including assessment of rule calibration, false positive management, and governance frameworks.

Section 3: Enforcement Case Studies

The following FCA and international enforcement actions are essential reading for calibrating surveillance judgment. Final Notices are available at fca.org.uk/news/final-notices.

FCA Enforcement Actions — Market Manipulation

Aviva Investors Global Services Limited (2019) FCA Final Notice: 4 February 2019 Penalty: £17,607,000

This is the most important MAR enforcement action in the UK fixed income space. The FCA found that a fixed income trader had placed and cancelled orders in UK government bonds (gilts) in a pattern consistent with marking the close, in order to influence the closing prices used for fund valuation. The Final Notice contains detailed analysis of the order pattern, the surveillance failures that delayed detection, and the FCA's view of what constitutes "artificial prices" under MAR Article 12. Essential reading for anyone designing fixed income surveillance.

James Alan Mellor (2017) FCA Final Notice: 19 April 2017 Penalty: £1,170,000

Insider dealing case involving the acquisition of securities in a company subject to a takeover bid. Illustrates the FCA's approach to establishing that information was precise and material, and the evidentiary use of trading patterns correlated with information access events.

Corrado Abbattista (2016) FCA Final Notice: 24 November 2016 Penalty: £100,000

Market manipulation through equity derivatives. Demonstrates how the FCA identifies manipulation in derivatives markets where the manipulative activity is designed to influence the underlying equity price.

LIBOR Enforcement Actions (International)

The LIBOR manipulation cases generated enforcement actions across multiple jurisdictions. Key reference documents:

Barclays PLC Final Notice (FCA, June 2012): £59.5 million penalty; the foundational UK LIBOR enforcement document, containing the FCA's detailed analysis of how submissions were manipulated and the communications evidence relied upon
Deutsche Bank AG — US CFTC Order (April 2015): $800 million penalty; extensive analysis of the manipulation scheme including communications evidence
UBS AG — US DOJ Deferred Prosecution Agreement (December 2012): Contains detailed factual basis including communications excerpts that are widely cited in academic and practitioner literature

Section 4: Academic Literature

Spoofing and Order Manipulation

Comerton-Forde, C., & Putniņš, T. J. (2015). "Stock price manipulation: Prevalence and determinants." Review of Finance, 19(4), 1499-1531.

This paper provides empirical evidence on the prevalence of closing price manipulation in equity markets and identifies the firm and market characteristics associated with higher manipulation risk. The methodology — detecting abnormal returns immediately following the close and correlating with end-of-day volume patterns — is directly relevant to marking-the-close surveillance design. Cited by regulators including ESMA in guidance on manipulation indicators.

Putniņš, T. J. (2012). "Market manipulation: A survey." Journal of Economic Surveys, 26(5), 952-967.

A comprehensive academic survey of market manipulation theory, empirical evidence, and regulatory response. Covers all major typologies, the theoretical conditions under which manipulation is profitable, and the empirical literature on detection. An excellent single-source introduction to the academic landscape.

Cao, C., Chen, Z., & Griffin, J. M. (2005). "Informedness and the manipulation of closing prices." Journal of Financial Markets, 8(1), 69-101.

One of the earliest rigorous empirical analyses of closing price manipulation, providing the theoretical framework for understanding why the closing auction is a preferred venue for manipulation and what price patterns distinguish manipulation from legitimate closing-period activity.

Machine Learning in Surveillance

Li, S., Zheng, W., Liu, S., & Yu, P. S. (2021). "Detecting spoofing attacks in order-driven markets using deep learning." Expert Systems with Applications, 183, 115441.

This paper applies deep learning (specifically LSTM networks) to order book data for spoofing detection, outperforming rule-based approaches on recall while maintaining comparable precision. Provides a detailed methodology for feature engineering from order flow data that is directly applicable to surveillance system design. The feature set developed — cancel ratio, size ratio, price impact, time-to-cancel distribution — maps closely to the components of the SpooingDetector implemented in this chapter.

Cao, C., Li, X., & Liu, X. (2016). "Does Stock Return Predictability Imply Improved Asset Allocation and Performance? Evidence from the U.S. Stock Market." (Note: For ML surveillance specifically, also consult:)

Micheli, A., & Ott, D. (2018). "Voice of the trading floor: NLP-based compliance monitoring." Journal of Financial Data Science, 1(3), 66-78.

Examines the application of NLP to voice transcript data for compliance monitoring, including the specific challenges of financial jargon, speaker diarization in multi-participant calls, and the high false positive rates inherent in keyword-based approaches. The paper's findings on reducing false positives through contextual modeling are directly relevant to the communications surveillance architecture in Case Study 2.

Cao, T., Chen, J., & Lee, C. F. (2020). "Machine learning applications in market manipulation detection: A review." Journal of Computational Finance, 24(1), 1-35.

A comprehensive review of ML applications in market surveillance, covering unsupervised anomaly detection, supervised classification, graph-based network analysis, and NLP. Includes a section on regulatory acceptance of ML-based alerts and the explainability requirements that affect system design.

Benchmark Manipulation

Duffie, D., & Stein, J. C. (2015). "Reforming LIBOR and other financial market benchmarks." Journal of Economic Perspectives, 29(2), 191-212.

Darrell Duffie and Jeremy Stein provide the most influential academic analysis of why LIBOR was manipulable and how benchmark design should be reformed. The paper's framework — distinguishing transaction-based from survey-based benchmarks and analyzing the incentive structures of each — is the standard reference for understanding both the LIBOR manipulation and the design principles of SOFR and other reformed benchmarks.

Abrantes-Metz, R. M., Kraten, M., Metz, A. D., & Seow, G. S. (2012). "Libor manipulation?" Journal of Banking & Finance, 36(1), 136-150.

Early empirical evidence that LIBOR submissions during the 2008 financial crisis were systematically below implied rates from other instruments, providing academic support for the subsequent regulatory findings. Methodologically useful for understanding how statistical analysis of benchmark submissions can identify manipulation.

Insider Trading

Meulbroek, L. K. (1992). "An empirical analysis of illegal insider trading." Journal of Finance, 47(5), 1661-1699.

The foundational empirical study of insider trading, examining price behavior in the period before SEC enforcement actions. Demonstrates that insider trading is detectable through price pattern analysis before the inside information becomes public — the methodological basis for surveillance-based detection of insider dealing.

Section 5: Technology Resources

Enterprise Surveillance Platforms

The following vendors provide enterprise-grade market surveillance and communications surveillance solutions. This is not a recommendation or endorsement; it is a reference list for practitioners evaluating the market.

NICE Actimize actimize.nice.com

Actimize's Surveillance solution is one of the most widely deployed in Tier 1 investment banking. Its strength is deep integration across trade surveillance, communications monitoring, and transaction monitoring (AML). The platform's Suspicious Activity Monitoring (SAM) module covers all major manipulation typologies. Actimize publishes an annual Financial Crime Report that provides useful industry benchmarking data on alert volumes, false positive rates, and STOR trends.

Nasdaq SMARTS Surveillance nasdaq.com/solutions/nasdaq-smarts

SMARTS is a market operator-grade surveillance platform used by over 50 exchanges and regulators globally, including the FCA and ESMA, to monitor markets from the regulatory side. Its availability to investment firms provides the significant advantage of applying the same detection logic as the regulator. SMARTS is particularly strong in equity market microstructure surveillance and high-frequency trading monitoring.

b-next b-next.com

A specialist European surveillance platform with strong coverage of fixed income and derivatives markets. b-next's communications surveillance module (Comply) is designed for MiFID II compliance and integrates with the major electronic communication platforms used in European investment banking. Well-suited to the tier-2 investment bank profile described in Case Study 2.

Eventus Systems eventussystems.com

An emerging vendor with a strong Python-based API architecture that has attracted interest from quantitative compliance teams. Eventus's Validus platform emphasizes customization: firms can implement their own detection algorithms alongside the platform's built-in ruleset. The vendor's approach reflects the direction of the market toward configurable, developer-accessible surveillance infrastructure.

Behavox behavox.com

A specialist in communications surveillance and behavioral analytics. Behavox's platform ingests voice, email, chat, and social media and applies ML-based behavioral pattern analysis. The platform is particularly relevant to the Case Study 2 scenario, where the primary challenge was designing a communications surveillance program from scratch against a large, diverse message population.

Open Source and Research Tools

PyThena (Python-based market microstructure analysis) github.com/QuantLib/QuantLib

For academics and quantitative practitioners building custom surveillance tools, the QuantLib library provides the pricing and risk analytics components needed to contextualize order flow data.

pandas and NumPy (Python) pandas.pydata.org | numpy.org

The foundational Python libraries for the data manipulation and statistical analysis that underlies custom surveillance implementations. The SpooingDetector and MarkingTheCloseDetector in this chapter are built on pandas DataFrames.

scikit-learn (Python ML) scikit-learn.org

Provides the unsupervised anomaly detection algorithms (IsolationForest, LocalOutlierFactor) most commonly used in surveillance applications. The library's consistent API makes it straightforward to prototype anomaly detection modules that integrate with a pandas-based data pipeline.

NLTK and spaCy (Python NLP) nltk.org | spacy.io

The primary open-source NLP toolkits for the kinds of text processing — tokenization, entity recognition, sentiment analysis — that underlie communications surveillance. Practitioners building first-generation NLP surveillance modules typically begin with spaCy's pre-trained models and fine-tune on domain-specific training data.

Practitioner Associations and Continuing Education

Futures Industry Association (FIA) — Market Regulation and Compliance fia.org

FIA's compliance-focused publications and working group outputs are particularly strong on derivatives market manipulation, including cross-product manipulation involving futures and underlying cash instruments.

Association of Certified Anti-Money Laundering Specialists (ACAMS) — Market Integrity Track acams.org

ACAMS offers certification and training that covers market abuse within the broader financial crime compliance framework. Useful for practitioners whose roles span AML and market abuse surveillance.

Chartered Institute for Securities and Investment (CISI) — Market Integrity Qualification cisi.org

The CISI's MAR-specific qualification provides a practitioner-level grounding in the UK MAR framework. Recommended for analysts joining a market surveillance function without prior regulatory background.

UK Finance — Market Abuse Forum ukfinance.org.uk

The UK Finance Market Abuse Forum publishes guidance notes on STOR quality, surveillance calibration, and emerging manipulation typologies. Membership-based, but selected publications are available publicly.

Note on Currency

The regulatory landscape for market surveillance is subject to active development. As of the writing of this chapter (2025), the following areas were subject to ongoing regulatory review or expected legislative change:

EMIR Refit / MiFIR Review: Amendments to transaction reporting requirements under EMIR and MiFIR affect the data available to surveillance programs. Firms should monitor FCA and ESMA consultation papers on reporting format changes.
AI Act (EU) implications for surveillance technology: The EU AI Act's classification of high-risk AI systems may affect the governance requirements for ML-based surveillance tools deployed by firms in the EU. ESMA is expected to publish guidance on AI Act compliance for financial sector AI applications.
FCA Discussion Paper on AI in financial services: The FCA's ongoing engagement with AI in compliance functions is expected to produce guidance specifically relevant to ML-based surveillance by 2026.

Practitioners should supplement this reading list with current regulatory publications and monitor relevant regulator websites for updates.