This week in regulation


Speed Read: What's been happening this week?

  • The FCA fines an experienced bond trader £60k for market abuse
  • FCA questions whether Black Friday is contributing to consumer debt
  • Exploring the power of machine learning for regulators
  • FCA releases statement on LIBOR panels
  • FCA sets out regulatory priorities for retail banking

FCA Fines and Penalties

The FCA has fined a former experienced bond trader £60k for market abuse engagements in 12 occasions. The trader was found to have used a manipulative strategy which resulted in creating a misleading impression of the price and supply or demand of the loans, leading to market participants purchasing or selling at a worse price than they would have otherwise.

The trader carried out a series of manipulations on the Dutch State Loans (DSL) market by making bids which subsequently became the highest. As a result, other market participants also raised their bids, increasing the price of DSLs. The trader then cancelled his quotes and proceeded to sell DSLs at the higher price. Although the trader pleaded unawareness that his conduct amounted to market abuse, the FCA judged that he should have known and not knowing was negligent.


[FCA Insight] Is Black Friday putting UK consumers in the red?

With some economists blaming discount shopping events such as Black Friday on the rise in unsecured consumer debt, the FCA explores whether this is indeed true in its latest ‘Insight’ piece.

On Friday, bargain hunters galore hammered on the doors of both retail and online stores in a bid to grab the latest deal and biggest discount before Christmas. Lured in by marketing tactics such as time-limited sales and influenced by the ‘sunk cost fallacy’ and ‘transaction utility’, what hope have consumers of not succumbing to the temptation of shiny new things?

Analysis of the FCA’s data provides insight into consumer spending on Black Friday:

  • We spend £900 million more on debit cards compared to an average weekend;
  • All age groups with a high overdraft limit (i.e. high creditworthiness) spend more on Black Friday compared to an average weekend;
  • The under 30’s are the age group that spends more on Black Friday;
  • There is no significant change to arranged or unarranged overdraft use during Black Friday weekend and there’s no evidence to suggest people are using their overdraft to fund Black Friday purchases;
  • Biggest purchases from credit cards were made during November, but is most likely attributed to seasonal effect;
  • Spike in logins to online banking accounts during Black Friday suggest people are paying more attention to what they are spending.

Is Black Friday contributing to higher levels of consumer debt? Well it’s difficult to say for certain, but the FCA concludes that although spending is higher it is concentrated on a certain segment of bargain hunters and consumers seem to be keeping an eye on their finances, so it’s perhaps not as influential as may be thought.


[Speech] The power of machine learning for regulators

Stefan Hunt, Head of Behavioural Economics and Data Science at the FCA delivered a super-long speech as part of a lecture series on regulatory economics last week, arguing that much of regulation is predicated on data science and analytics. As developers of RegTech ourselves, for a long time we have been helping the industry realise the benefits of deciphering data to ensure compliance adherence and improve customer experience. There is a natural application to the FCA and other regulators that are inundated with data.

Data algorithms are increasingly shaping the world around us – what we see on social media, what adverts we see online, even the emails blocked by our junk filters. But it is not only commercial organisations that can benefit from data, algorithms and analysis – the public sector can too. From the Police Force to the Serious Fraud Office, Ofgen to the CMA, regulators and the public sector are using data science such as machine learning and artificial intelligence to derive insight from data that informs decision-making.

What is machine learning?

Data science is essentially extracting information and knowledge from data. Machine learning forms one component of data science and is the focus of Hunt’s talk. According to Hunt, “Machine learning is the ability to learn without being explicitly programmed, instead learning patterns from many examples.”

How can the regulator use machine learning?

  1. Understanding markets and consumer behaviour (unsupervised learning) – Unsupervised learning is essentially a computer discerning patterns and trends without human input. The FCA receives billions of data points from the 56,000 firms it regulates. Machine learning provides insight into this deluge of data by enabling the regulator to detect patterns and trends and segment data into groups. As a result, it can allocate its resources more efficiently and prioritise intelligently.
  2. Providing paths and options (supervised learning) – Supervised learning requires some input from humans to teach it what to discern. The FCA can make prioritisation decisions based on predictions derived from data, such as prioritisation of which markets or issues warrant higher supervision and the detection of specific issues such as fraud and collusive behaviour. Usage of supervised learning/predictive analytics will make the FCA more effective and more efficient.
  3. Using human judgement – Humans are essential for making good decisions from the information generated from unsupervised and supervised learning. Indeed, humans are also needed for the decision about when and how to apply the technology in the first place.

Not falling for the hype

Hunt suggests there is much hype around machine learning and its capabilities and the regulator must stick to what can be reasonably expected of the technology and its application. There is not yet enough quantitative data on the impact and benefits of machine learning on improving the effectiveness and efficiency of regulation, although Hunt suggests a 25% increase in efficiency seems plausible. However, it is clear that making sense of the reams of data regulators and, indeed firms, now have at their fingertips can untangle many problems faced today in financial services. In essence, many regulatory problems are prediction problems, and machine learning simply enables better prediction.


FCA releases statement on LIBOR panels

All 20 of the panel banks have confirmed support of the LIBOR benchmark, reports the FCA. This agreement was made to ensure future sustainability and is set to run until 2021, by which time a smooth transition can be made to alternative rates (something the FCA expects to focus on now the agreement has been made).

There have been two changes to the panel compositions, but no further are expected:

  • Societe Generale will no longer submit to the US Dollar Panel;
  • Credit Agricole Corporate and Investment Bank will cease submissions to the Japanese Yen panel.


[Speech] Regulatory priorities for retail banking

Karina McTeague, Director of Retail Banking Supervision delivered a speech last week setting out the regulatory priorities for the retail banking sector.

McTeague started by setting out the areas of focus for the sector, namely:

  • Culture and governance;
  • Financial crime and anti-money laundering;
  • Promoting competition and innovation;
  • Technological change and resilience;
  • Treatment of existing customers;
  • Consumer vulnerability and access;
  • Ring-fencing;
  • Senior Managers & Certification Regime (SM&CR);
  • Remedies proposed by the Competition and Markets Authority (CMA);
  • PPI complaints handling.

In her speech, McTeague discussed in detail two areas: the FCA’s strategic review of retail banking business models and the second payment directive (PSD2), summaries of both can be found below:

Business models

Although competition issues continue to exist, McTeague foresees a transformation in the very near future due to increased digitisation, new market entrants, social trends and regulatory interventions. However, there is the risk that competition and innovation may be stifled and firms may focus more on profitability than consumer outcomes as the new paradigm places pressure on business models. The FCA’s strategic review of retail banking business models therefore focusses on business models, customers and the regulatory approach. The information the FCA will glean from this assessment will allow it to be proactive in taking pre-emptive action based on a robust analysis of the impact of changing forces on firms’ business models and profitability.

The review is made up of two phases:

  • Phase 1: discovery;
  • Phase 2: evaluation and scenario analysis.

PSD2 and Open Banking

Coming into force on 13th January 2018, PSD2 is expected to support the FCA’s objectives by improving competition and innovation, providing consumer protection and enhancing market integrity. However, for the benefits of PSD2 to be realised it can be paired with Open Banking which will enable consumers to take control of their own data, information from their online payment accounts and the financial products they use.  Firms need to educate consumers so that they can truly take advantage of the benefits PSD2 and Open Banking have to offer.

McTeague urges firms to read the FCA’s PSD2 Approach Document and a joint FCA/HMT document setting out their expectations of firms during the transitional period awaiting Regulatory Technical Standards.

McTeague goes on to say that the regulator is expanding the proactive supervision of payment services institutions to check that:

  • Culture is customer-centric and focused on treating customers fairly;
  • Systems and controls are managing financial risks and financial crime risks.

Customer are the key to PSD2 and Open Banking working as it relies upon consent, so the FCA will be concerning itself with helping them to understand what they are consenting to and the benefits of doing so.

Latest Insights


The latest insights and regulatory analysis straight into your inbox

Find out more

How can we help?

Speak to a culture and conduct risk expert today

Contact us

Find out more about our unique RegTech solution at