Investors of the twenty-first century can harness the power of publicly available data to form a view on a specific company or – more generally – on a particular investment topic. Satellite imagery, marine and air traffic trackers, keyword or search engine trends can provide critical insights on how a company is performing, sometimes unbeknownst to the company itself and most other interested investors. This contribution explores whether there are circumstances in which data extracted from public sources is or can become insider information under Swiss law.

By Ariel Ben Hattar (Reference: CapLaw-2019-13)

1) Introduction

At first, the idea that publicly available data, or the aggregation of publicly available data, could become insider information seems like a contradiction in terms. An example can however illustrate potential points of contact between the two. An investor acquires foot traffic data covering streets neighboring a sample of stores of a company. By monitoring the data at regular intervals and correlating it with financial results of the operator of the stores, the investor can form a view on how the company is doing during a particular quarter earlier than when the company releases its quarterly updates. It may even be that the company only receives information about its performance at a later date, when it receives data from its stores through its internal accounting and monitoring processes, not to mention investors who have not invested the kind of efforts and money necessary to acquire the foot traffic data.

As the above example shows, through technology and financial capacity, some investors may gain an edge over their peers by acquiring and processing vast quantities of data that is publicly available. Other investors could acquire the same data sets, which cover foot traffic in public spaces, or they could commission third parties to assess foot traffic around the stores by simply walking around them and counting the number of people walking by the stores. The costs of such an exercise would, presumably, be very significant, so much so that a majority of investors will never be able to afford them. While it can generate a sense that a level playing field between investors is lacking, does this use of publicly available data amount to insider trading? To answer this question, it is first necessary to take a step back and consider the current state of insider regulation in Switzerland.

2) From “Who” to “What”

When Switzerland first criminalized insider trading in 1987, the assumption was that insider information necessarily came from within the company. At the root of insider trading lied a breach of trust. A company had entrusted a director, employee or advisor with a confidential information and the information’s recipient breached the company’s trust by misusing the information, either to trade based on it or to communicate it to a third party.

This view did not survive the revision of the rules on insider trading that entered into force in 2013. The revision created a new category of insiders: those who do not access insider information due to their privileged ties to the company. A commonly used illustration of this new category is that of a contractor cleaning corporate offices who finds a document containing insider information in a waste bin. This example is however somewhat misleading, as the confidential information still comes from within the company. A more illustrative example would be the case of a fund manager who learns that another investor is about to launch a public takeover offer for the company. In such a situation, the company may not even be aware of the insider information, as it comes from a decision of a third party investor. 

As a result of these changes, delimitating the scope of the insider trading prohibition is now less about who the insider is vis-à-vis the company, but rather about what constitutes insider information. When looking at a particular piece of information, one must therefore determine if it fits into the definition of insider trading of article 2 (j) of the Financial Market Infrastructure Act (FMIA). 

3) The Spectrum of Publicly Available Data

Before examining the components of the definition of insider information in more details, it is helpful to establish first that publicly available data can take several forms, not all of which present the same challenges for the analysis presented herein. To simplify, I have divided publicly available data into the following three categories.

– Published data:

In its most basic form, publicly available data encompasses information that has been made available to all interested persons through actions designated to achieve such purpose, e.g. through a press release of the company or an article published in a newspaper. I refer to this type of information as “published data”. 

– Data from public sources:

A second category of publicly available data is information that can be found in or deduced from public sources, even though it is not published data. A public source, in this context, is a source of information that anyone could access (even if against a fee, upon request or by taking specific actions, e.g. being in a certain place). This category includes, for example, observing from a passing train that the main factory of a company is in flames (on this example, see Section 4.c)iii) below), examining public records available upon request (e.g. archives), or attending industry conferences.

– Aggregation of data from public sources:

Finally, publicly available data includes the aggregation of data from public sources. Examples include data derived from pictures of multiple locations taken by satellites, exit polls conducted during elections or aggregations of online reviews of a restaurant chain.

4) Publicly Available Data and Insider Information

a) Article 2 (j) FMIA

Pursuant to article 2 (j) FMIA, an insider information is (i) an information that (ii) is confidential and (iii) is susceptible, if disclosed, of having a notable influence on the price of securities admitted to a Swiss trading venue (i.e. a Swiss stock exchange or multilateral trading facility). The relevance of publicly available data for each of these criteria is examined in the following sub-sections.

b) Is Data Information?

To be relevant in the context of the Swiss insider rules, publicly available data must first be tangible enough to qualify as an “information”. An information is generally about a fact, and not merely about a rumor or a speculation. 

There is nothing inherently non-factual about publicly available data. If anything, publicly available data will generally tend to be fact, although this will evidently depend on the source of the relevant data. Recommendations based on, or conclusions drawn from, publicly available data may however fall short of being facts. Deducing from car traffic data around stores that a supermarket chain is having a good or bad quarter could for example amount to only a speculation if the supermarket chain also derives significant turnaround from its online activities.

c) Confidentiality and Publicly Available Data

i. The different approaches to confidentiality

Although it is mentioned in article 2 (j) FMIA, confidentiality is not further defined in the context of insider trading regulation. Precedents from the Swiss Supreme Court indicate that information is not confidential when it is almost certainly known by an extended circle of market participants (decision published in BGE 118 Ib 448, consideration 6b/aa) or when it can be discovered by third parties, even if this requires efforts (Supreme Court decision 2A.230/1999, consideration 6b). Some scholars have taken a different view and argue that information is confidential unless it is widely available through generally accessible means. As will further be explained below, to what extent publicly available data could be considered as confidential depends on which one of these views is upheld.

To account for the centrality of confidentiality to assess the status of publicly available data under insider trading regulation, I analyze it from the perspective of all three categories of publicly available data described under Section 3.

ii. Published data

Data that has been made available to the public through actions intended to achieve this purpose are, by essence, not confidential (anymore). This is the case even if the relevant information does not appear on a widely read newspaper, is not displayed prominently on the company’s website or does not appear in an “ad hoc” publicity press release published under the listing rules of SIX Swiss Exchange or in a similar document. It should indeed not be that investors have to ask themselves whether a document found e.g. on the Internet is sufficiently well advertised to be considered as not confidential anymore. The conclusion that published data is not confidential should hold true reached regardless of which of the interpretations of “confidentiality” referred to above is followed.

iii. Data from public sources

For what regards data gathered from public sources, the different approaches to confidentiality lead to different outcomes. 

If the theory according to which information is confidential if it is not available through generally accessible means were to prevail, then information obtained from public sources could sometimes be deemed confidential. For example, a scientific paper demonstrating that a particular molecule is ineffective to cure a medical condition may only be accessible to the subscribers of the journal in which it is published. Those subscribers would then not be allowed to use this information to trade the shares of a company that wishes to use the molecule, since there is probably only a fraction of all interested investors who subscribe to the journal. To regain their ability to trade the relevant shares, those who have access to the scientific article would need to either wait for a newspaper to cover the story or for the company to publish an “ad hoc” press release, or even engineer a publication of the paper’s findings on their own (which would raise a host of different challenges). Such a conclusion would evidently be problematic and cannot, in my view, be supported.

Following the reasoning of the Supreme Court, however, leads to a coherent and – in my view – correct solution. From the Supreme Court’s perspective, what matters ultimately is whether other investors could access the relevant information, even if it costs them money or efforts. In the above example, all investors could access the information contained in the scientific paper by subscribing to the journal, although they would need to pay a (potentially not insignificant) fee to do so. The fact that specialized knowledge may be required to understand the ramifications of the paper is also not relevant. Although an issuer itself may be expected to explain its business activities in layman’s terms when communicating with its shareholders, investors who have a better grasp of those activities (or who can commission third party to do the analysis for them) should not be restrained in their ability to trade by their superior knowledge. While it is true that retail investors would be unlikely to invest the kind of efforts and money needed to achieve this, the purpose of insider trading regulation is not to ensure that all investors have the same actual knowledge, which would anyway be impossible. Rather, it is to ensure that trading is based on information that all investors could access. 

A similar conclusion would be reached in the European Union under Regulation (EU) No 596/2014 of the European Parliament and of the Council of April 16, 2014 on market abuse (MAR). In its guidance on MAR, the United Kingdom’s Financial Conduct Authority (FCA) relates the following example: a passenger on a train passes a burning factory and calls his broker, asking the latter to sell all his shares in the factory’s owner. As the investor obtained the information that the factory’s owner may be in trouble by legitimate means through observation of a public event, the FCA concludes that, under MAR, the passenger is not considered to have used insider information.

On one aspect, the FCA’s view and Swiss law may differ. The FCA notes that the information must have been obtained through legitimate means by the investor. From a Swiss law perspective, the key requirement is that third parties should be able to obtain the same information. In my view, an investor who broke the law to access information will only possess insider information if other investors cannot access the same information without also breaking the law. To put it differently, if there is a legitimate way to obtain the information, it should not be considered as confidential.

iv. Aggregation of data from public sources

From a theoretical perspective, aggregating data from the public sources is no different than making individual observations based on those sources. The same conclusions (described under section iii)4.c)iii) above) should therefore apply. 

Aggregating multiple data sources may however require efforts so significant that most investors, even professional ones, will not have the resources to replicate the data aggregation effort. In addition, contrary to single use of public sources, an investor cannot hope to simply be lucky, e.g. being in the vicinity of a factory when it is engulfed in flames, or by having expert knowledge in the company’s business. Furthermore, aggregating multiple sources of data may allow one to form a view on a matter affecting a company long before the company itself becomes aware of it. The same is generally not true with respect to one-time events like the publication of a scientific paper or a burning factory, where the time gap between the moment the first investor becomes aware of their occurrence and the moment the company announces them will tend to be short.

The particularities of the aggregation of publicly available data are nonetheless not such that they justify a departure from the Supreme Court’s theory. Those specificities however call for a closer examination of the relevant facts, and particular attention should be paid to the possibility for third party investors to replicate the data aggregation. In this respect, the following should be considered.

– The requirement set forth by the Supreme Court is that third parties should be able to gain access to the information. There is no implication that the access should be on the same terms, or that it should involve the use of the same technical means. As a practical matter, investors conducting data aggregation analyses should be wary of agreements giving them exclusive access to data sources. By the same token, investors creating their own tools to find and aggregate publicly available data may prefer to make these tools available to third parties (against a fee). It will indeed be easier to demonstrate that information is not confidential if third parties could have used the same source, as opposed to proving that other sources would have been available.

– There is no explicit or implicit threshold for the costs beyond which investors would be presumed not to have the ability to replicate a data aggregation. The fact that most investors would not be able to bear the costs of an analysis is not, in and of itself, conclusive to hold that the result of the data aggregation is confidential information. 

The use of aggregated data to produce financial research can also raise questions, which are addressed in more details under Section 5 below.

d) Price-Sensitiveness

Finally, article 2 (j) FMIA provides that an information is considered as insider information if it is price-sensitive, i.e. if its disclosure would be susceptible of having a notable influence on the price of securities in scope of the Swiss insider trading rules. To me, there are two possible ways to address this element.

– The first is to consider that if an information is based on data that is publicly available, then the whole question is moot. It would indeed not be necessary to ask how the market will react if the information is disclosed because by nature it has already been disclosed. In my view, this approach is artificial. The price-sensitiveness of an information is linked to the process through which the market at large reacts to and prices in the information as it becomes aware of it. The fact that an information is publicly available does not mean that it has already been processed and priced in by the market, and is thus not susceptible to have a notable influence on the price of securities anymore. On the contrary, as we have seen, there may be a significant time gap between the moment one investor has processed publicly available data and the moment the company and the wider market become aware of it. In that sense, the “disclosure” to which article 2 (j) FMIA alludes is different from the concept of “confidentiality”. An information can thus remain undisclosed although it is not confidential anymore.

– The second, which I hold to be the correct approach, is to consider that publicly available data is also susceptible of being price-sensitive. In most cases, however, there will be no need to answer the question of the price-sensitiveness of publicly available data. If the information is not confidential (and publicly available data should generally not be confidential), then it is not insider information and its possible influence on the price of securities is irrelevant.

5) Research Products Based on Publicly Available Information

a) The Mosaic Theory

One area where the use of publicly available information has been discussed in a Swiss insider trading regulation context is that of financial research, and in particular the “mosaic theory”. Originally a U.S. legal construct, the mosaic theory posits that the use in financial research of a mix of publicly available and not publicly available (but not individually price-sensitive) information is permissible and that the result of the analysis of such information should not be considered as price-sensitive. 

In its report on the consultation on FINMA Circular 2013/8 – “Market conduct”, FINMA noted that “in the normal case” Swiss law does not contradict the mosaic theory. FINMA’s view is to be approved. On the one hand, publicly available information should generally not be confidential and, hence, not qualify as insider information. On the other hand, confidential information that is not price-sensitive is not insider information either. Combining both should therefore not create insider information. To us, this is the case even if an analyst does more than just aggregating data but adds a layer of personal opinion or an intellectual reasoning to the analysis. 

b) When Does a Research Product Become Insider Information?

In connection with the mosaic theory, also FINMA noted that the publication of a research analysis may itself be insider information, even if the analysis does not contain insider information. FINMA gave very limited guidance on when exactly it considers that this is the case, and a few additional explanations are in order.

i. The analysis versus its publication

An analysis not based on insider information cannot per se become insider information. As FINMA notes, it is the analysis’ publication that can be relevant in an insider trading regulation context. Analyses based on publicly available (and potentially also on confidential but not price-sensitive information) that are not published are thus never insider information. Those can be freely communicated to the investor who commissioned the research, for example, or be used to trade by the entity that produced them.

Treating the publication of an analysis as insider information inter alia means that the analysis should not be communicated selectively to third parties in advance of its release. Still, it does not mean that the only way to avoid committing a breach of insider trading rules is to make the analysis freely available to all investors (i.e. to make it published data, to use the terminology described under section 3). Rendering the analysis available to interested investors, even if they are required to pay a fee to access it (i.e. making it publicly available), is sufficient.

ii. Determining when the publication of an analysis is insider information

Far from all publications of analyses based on publicly available data should in practice be insider information. To identify those that may qualify as such, one may point to the author of the analysis as a key criterion: the word of a famous hedge fund manager or large bank is known to weight more than that of a retail investor. The identity of the person authoring or publishing the analysis is however only one part of the equation. An analyst famous for his coverage of company A should not be restricted in the way she talks about company B, an issuer she does not usually cover. Similarly, large banks produce copious amounts of research, some of which contributes only to a limited extent to the price formation process. 

In line with the solution found in the European Union, the decisive factors to assess whether the publication of a research product is insider information lies in the market’s expectations. Specifically, according to recital 28 to MAR the publication of an analysis can become insider information in two situations: 

– if the analysis contains the views of a recognized market commentator or institution which may inform the prices of related financial instruments; or

– if the analysis is routinely expected by the market (e.g. because it is published every quarter) and known to contribute to the price formation process of financial instruments.

These two cases are closely linked. To me, it would be incorrect to treat the publication of any view of a known commentator or prestigious financial institution as insider information. What recital 28 to MAR means is that, if a commentator or financial institution produce a kind of analyses that regularly trigger notable price movements, the publication of analyses of the same kind should be considered as insider information. 

For example, if an analyst authors publishes monthly research notes on pharmaceutical companies that are known to trigger sharp increases in the price of those companies’ shares, then future iterations of those notes may need to be treated as insider information. If the same analyst produces a separate research note on a car company, then the publication of this latter analysis is not insider information because it is an analysis of a different kind, for which the market has no or different expectations. Similarly, the research coverage of an issuer by a large investment bank may be closely followed by the market, to the point where updates by the investment bank trigger significant price movements in the shares of the issuer. Although those updates will be treated as insider information, research material published by the same bank regarding other issuers will not be contaminated and will not automatically be considered as insider information.

iii. Impact beyond financial research

Interestingly, the ramifications of recital 28 to MAR extend beyond financial research stricto sensu and, in particular, also affect benchmarks. For example, some “rig counts” (i.e. recensions of the number of active oil or gas rigs) are based on publicly available data and are known to move the share price of companies active in the oil and gas drilling sectors. Although their content is based on publicly available data that is not insider information, the publication of these benchmarks should be treated as insider information because the market has come to expect them and they are known to have a notable impact on the price of certain securities.

6) Conclusion

Since the revision of the Swiss rules on insider trading which entered into force in 2013, insider information does not need to come from “inside” the company to which the information relates anymore. Depending on how the concept of confidentiality is interpreted, large swaths of publicly available data could be considered confidential, and potentially insider information. According to the view described in this contribution, publicly available data should however generally not be considered as confidential, and should consequently not qualify as insider information. 

Investors willing to use technology to analyze publicly available data on a large scale should however be mindful that a possible characterization as insider information largely depends on the replicability of their analysis by other investors. For this reason, it may be advisable to avoid contracting third party service providers (e.g. satellite imagery services) on an exclusive basis. Data suppliers, producers of benchmarks and analysts should also consider that the publication of their products can become insider information, even if these products are based on publicly available data and are not themselves insider information. 

Ariel Ben Hattar (ariel.benhattar@lenzstaehelin.com)