In the past few years, artificially-intelligent “trading bots” have entered the market and begun executing securities transactions for investors. Although much has been written about the function, performance, and future of such bots, very little has been said about their implications for recoveries in securities class actions in the US and non-US group actions. As FRT monitors these cases, we see how the use of such bots has the potential to complicate recovery efforts in certain jurisdictions.

Demonstrating “Reliance” in Securities Actions

Many countries have developed standards for demonstrating so-called “reliance” in securities actions as part of the process of proving claims and recovering losses. The concept of reliance is that investors must have relied directly or indirectly on the alleged misinformation by a defendant in making their investment decisions.

In the United States, courts have generally “presumed reliance” through a fraud-on-the-market theory in which it is assumed unless otherwise demonstrated, that investors relied on false or fraudulent information provided by a defendant corporation when making investment decisions. This theory is based on the belief that for stocks trading in an efficient market, prices reflect all publicly- available information including material misstatements by companies, and that investors typically trade in reliance on the accuracy of the market prices.

Outside of the U.S., courts have varied widely on if and how reliance must be shown in securities cases. Some jurisdictions have adopted an approach similar to the U.S. This includes the Netherlands, which presumes causation between a false or misleading statement and investment decisions. Other countries, like Japan, have done away with the reliance requirement altogether for securities actions involving material misstatements. Investors bringing suits there need only prove that a company issued a report containing a false statement or omission of material fact that caused investor losses, regardless of whether investors knew of them. Other countries, like Germany, require proof of reliance for some causes of action but not others. Germany does not require investors to prove reliance if seeking inflationary losses under the German Securities Trading Act. However, if investors seek rescission damages or bring claims for violations of tort laws, they must prove reliance including stating that they would not have purchased shares, or paid what they did, had the truth been known.

Other countries- apply a much more rigorous standard of “direct reliance” for certain fraud-based claims. Courts in the UK may apply a direct reliance standard in securities actions alleging false or misleading statements under Section 90A of the Financial Services and Markets Act 2000 (“FSMA”). If so, investors may need to show that they not only read but also relied upon information provided by the defendant when making investment decisions and that such reliance was reasonable. Note, however, that for brought under Section 90 of the FSMA – for false or misleading statements in a prospectus – investors only have to satisfy a more lenient presumed reliance standard.

Proving “Reliance” in AI Trading

Proving reliance can be complicated when an AI trading bot is involved in making investment decisions. Determining whether, and to what extent, an investor who uses such bots can prove reliance requires an understanding of how they function.

Traditional investment tools, such as statistical forecasting or limit transactions, have not impeded investors from proving reliance over their years of use. This is because, although these tools may execute instructions or provide recommendations without direct human input, they do not make active decisions themselves. A model provides recommendations; a limit executes an instruction set by a person. But ultimately it is the investors themselves who are making the decisions to buy and sell the relevant securities.

More advanced forms of artificial intelligence, however, can and do eliminate human decision-makers from the process. A neural network (one popular model of trading bot) empowered with the ability to buy and sell securities without direct approval from an investor is, essentially, a free agent.

Presumed reliance under a fraud-on-the-market theory can likely be met by an investor using trading bots. Trading bots typically follow the market and make decisions based on their performance. AI bots that are programmed to follow the market should, by definition, have indirectly relied on any corporate disclosures that influenced the market. Indeed, because these bots are programmatically incapable of making decisions for any reason other than market trends, it may actually be easier to prove indirect reliance on their part compared to a human who may include subjective, non-market considerations.

In principle, it is possible to design and use a trading bot that executes buys and sells truly randomly or with incomplete information on market conditions. Were such a bot in use, it could be argued that presumed reliance is absent because the bot was not purchasing based on market conditions and stock price. Realistically, however, it is unlikely anyone would use trading bots that do not incorporate market prices in some way into the bots’ decision-making process. Therefore, the use of trading bots should not impact the ability of investors to bring claims in countries where there is presumed reliance, such as the United States, or countries that do not require reliance, such as Japan.

A much more difficult question exists in jurisdictions that may require direct reliance, like the UK and Germany when rescission damages are sought.

An investor who uses trading bots could find it difficult, if not impossible, to show direct reliance. Most trading bots are unable to ingest any information outside of numbers and trends in the market. Investors using these bots cannot, therefore, show direct reliance because the bots are incapable of incorporating any specific false or material misstatements into their trading decisions. The only way investors using such bots could prove direct reliance would be to show that they had enabled, disabled, or altered a bot’s programming based on information provided by a defendant corporation, which, while theoretically possible, does not appear to be how investors use these bots.

If trading bots could be designed in such a way as to incorporate corporate disclosures to the market into their decision-making process, investors could potentially argue that the bots considered such information before trading in securities. (Such developments would certainly represent a leap in technology, but are theoretically possible even today).

However, even if trading bots with that capability were developed, it would still be difficult to prove direct reliance and show that the trading bot directly relied on any specific disclosure in making a specific trade, sell, or purchase. After all, the bots can’t provide affidavits or testimony. Investors’ use of such bots in jurisdictions with direct reliance therefore may limit their ability to bring a successful securities claim.

Conclusion

It may take some time to reach that point where courts grapple with this issue, as the number of countries requiring direct reliance is limited. Even in those countries requiring direct reliance, like the UK and Germany (for certain claims), cases will be prospected on a representative or test case basis by others. Also, claimants faced with the requirements may choose to forego recovery if there is risk their AI technology will be subject to disclosure to opposing parties. But with AI integrating itself more and more into our investment world, it seems a matter of when, not if, we reach the point where courts have to decide the question of whether and how investors using artificially-intelligent trader bots prove it.

Learn More

To learn more about how FRT can help your firm maximize recoveries in shareholder class action settlements, contact us at learnmore@frtservices.com.

About FRT

SETTLED CLAIMS  I  PASSIVE GROUP  I  ANTITRUST  I  FUTURE CLAIMS  I  OPT-IN MONITORING  I  OPT-OUT MONITORING

Founded in 2008, Financial Recovery Technologies (FRT) is the leading technology-based services firm that helps the investment community identify eligibility, file claims and collect funds made available in securities and other class action settlements. Offering the most comprehensive range of claim filing and monitoring services available, we provide best-in-class eligibility analysis, disbursement auditing and client reporting, and deliver the highest level of accuracy, accountability, and transparency available. For more information, go to www.frtservices.com.

• Follow us on LinkedIn: Financial Recovery Technologies
• Email us: learnmore@frtservices.com

This communication and the content found by following any link herein are being provided to you by Financial Recovery Technologies (FRT) for informational purposes only and do not constitute advice. All material presented herein is believed to be reliable but FRT makes no representation or warranty with respect to this communication or such content and expressly disclaims any implied warranty under law. Opinions expressed in this communication may change without prior notice. Firms should always seek legal and financial advice specific to their unique situation and objectives.