The FCA is considering whether alternative data could introduce new risks to market integrity.

By Rob Moulton, Fiona Maclean, Stuart Davis, and Charlotte Collins

The FCA’s recently published Insight article explores how alternative data might give rise to market abuse risks. The article reports a significant increase in spending on alternative data in recent years, leading to questions about whether access to such data might provide recipients of the data with an unfair informational advantage over other market participants.

While traditional sources of data, such as a company’s financial statements, may contain inside information and must be treated appropriately before they are made public, the nature of alternative data is less clear-cut. Alternative data does not come from the company itself, and may derive from (or be extrapolated from) a number of sources. Alternative data may allow those with access to know things about a company that others in the market do not know, or that the company itself does not know. This may be the case, even if, as is frequently the case, the pool of structured/unstructured data used by the analytics engine is in the public domain. Evidently, this could provide trading opportunities that put the holder of such information at an advantage, as compared with other market participants.

A key example of where alternative data has raised concerns recently is in relation to so-called “secret polling”. The government has had exchanges with the FCA concerning the potential use of private polling data to obtain a trading advantage in advance of election results. The regulator’s view is that, while the Market Abuse Regulation (MAR) might be engaged by such activities, MAR would only apply if the underlying information were to constitute inside information. This is unlikely to be the case, unless the information met the MAR recital 28 test of information “routinely expected by the market” to be published, such as weekly BBC opinion polls. Therefore, MAR does not restrict the sharing of polling information that is not inside information. However, this position clearly raises political questions of fairness, as those able to pay for and access the data may well gain an advantage in the market, and those providing the data may not understand the use to which it will be put.

Position Under the Market Abuse Regime

A key factor in whether information constitutes inside information is whether that information has been made public. Previously, the Market Abuse Directive (MAD) contained a recital that read “Research and estimates developed from publicly available data should not be regarded as inside information and, therefore, any transaction carried out on the basis of such research or estimates should not be deemed in itself to constitute insider dealing within the meaning of this Directive”. The FCA supplemented this with its own guidance, which stated that certain factors could be taken into account in determining whether or not information had been made public, including the extent to which the information could be obtained by analysing or developing other information that is generally available. The FCA added that it did not consider relevant the fact that the analysis is only achievable by a person with above average financial resources, expertise, or competence.

Essentially, this meant that information would be considered publicly available even if sophisticated analytical skills were required to reach a meaningful interpretation. Therefore, those with the nous or resources to reach conclusions that others might not reach would not be considered to have an unfair advantage in the market.

However, this recital was adjusted under MAR to emphasise that research and estimates based on publicly available data should not per se be regarded as inside information, but that market participants should consider the extent to which the information is non-public, and the possible effect on financial instruments traded in advance of its publication or distribution, to establish whether they would be trading on the basis of inside information. The FCA also removed the factor noted above (the extent to which the information could be obtained by analysing or developing other information that is generally available) from its list of factors indicating that information had been made public, suggesting that the FCA did not consider guidance in this area compatible with MAR. This may indicate a narrowing of the FCA’s approach towards data analysis and its position on when information might be considered to be made public.

An Unfair Advantage?

A key issue regarding alternative data and market abuse risk is that although the data may be obtained from publicly available sources, it is not readily accessible by the whole market. It often requires sophisticated manipulation of complex data sets, and providers will charge a fee for access. Another factor in the FCA’s guidance that may indicate whether information has been made public is that it can be obtained by observation by members of the public without infringing rights or obligations of privacy, property, or confidentiality.

Therefore, alternative data raises the question of whether it is fair for some market participants to pay for data that can only be produced by big data providers with the skills and resources to gather and process large quantities of data in a way that most market participants would not be able to. But how does this differ from paying for traditional research from a star analyst? A practice that the market has long accepted as a fair approach, even if it leads to informational asymmetries between those who can access such research and those who cannot. This is the conundrum the regulator faces.

Position under the General Data Protection Regulation

Alternative data sources are often fuelled by personal data and accordingly compliance with data protection laws is a key consideration. Variety, velocity and volume are key characteristics of the data pools used to feed alternative sources thus enabling companies to build enriched profiles on consumers, businesses, and market trends.

To the extent that any of these data sources contain personal data, data privacy laws will be applicable. Data privacy legislation, on an almost globally consistent scale, demands fair, accurate and non-discriminatory use of personal data. Nowhere more so than in the EU, where the General Data Protection Regulation (GDPR) puts the data subject at the heart of its rules, requiring — in respect of all data processing — enforceable data subject rights and effective remedies for data subjects.

Firms using alternative data must therefore undertake appropriate diligence as to the sources of all such data, to ensure it is collected and processed in accordance with data privacy laws. Data hygiene and good governance are critical to the use of alternative data, not only for the purpose of mitigating regulatory risk, but also alleviating reputational risk.

Regulatory Approach

Evidently, at this stage, the regulator does not have a settled view on how to approach the issue of alternative data. However, the increase in regulatory focus signals greater interest in the regulatory risks that might arise. The FCA identified in its Business Plan for 2019/20 that “innovation appears to be increasing firms’ ability to predict market events early. This should benefit markets, but it could also introduce new risks of collusion and new factors that will need to be considered in identifying and prosecuting market abuse”. The FCA planned to carry out a Call for Input on data in wholesale markets during the 2019/20 financial year. Although this has not yet materialised, we expect that it will appear in 2020. Information gathered from the Call for Input will help to shape the regulator’s thinking and determine what steps (if any) it wishes to take.

The Information Commissioner’s Office (ICO) has, over the years, repeatedly noted the importance of ethics in respect of the use of “big data”. In November 2019, the ICO appointed its first data ethics adviser to ensure that the ICO upholds the objectives set forth in its Information Rights Strategic Plan: namely, enhance information rights for the UK public in the digital age and increase public confidence in the use of data. The ICO has admitted that this is an innovative area, and one that is not without its challenges. This is, however, a clear marker from the ICO that strict legal compliance is not sufficient, and use of alternative data sources should also take account of the more intangible, and subjective, matters of right and wrong.

Firms using, or considering using, alternative data should bear in mind the FCA’s and ICO’s concerns, and the fact that the regulators’ views, and readiness to enforce, could change in the near term. It would be advisable to proceed with caution, and to engage with the regulator as early as possible in order to feed into any (formal or informal) regulatory consultation process. The FCA’s Insight article concludes that a debate is required between industry and regulators to help reach a consensus on whether regulatory intervention is required, and firms should stand ready to contribute to this debate.