November 17, 2022
Introduction
A priority of the Ontario Securities Commission (the OSC or we) is improving the investor experience
and expanding investor protection through a range of initiatives
1
. Included in this are initiatives to
support behavioural insights and policy testing capabilities. Reflecting this commitment, the Investor
Office examined gamification and other behavioural techniques that are currently being used or may
be used in the future by Order Execution Only or self-directed investment platforms as part of their
digital engagement practices (DEPs). The report arising from this work (the Report) examines how the
different gamification and other behavioural techniques may influence retail investors behaviours—
both positively and negatively. The Report is appended to this staff notice (this Notice).
Purpose
A wave of digital, mobile-friendly investing platforms has created new options for retail investors in
Canada and around the world. While these platforms have expanded market participation, there is
growing interest in some of the DEPs that they and, to a lesser extent, more traditional retail
investment platforms use and how these may raise investor protection concerns. These tactics,
sometimes referred to broadly as “gamification,” use insights from behavioural science to influence
investor behaviour.
Regulators have faced some challenges in understanding and responding to these developments,
including: a lack of common terminology and definitions currently in use; an absence of a regulatory
inventory of practices currently employed by Canadian (and US) dealers in the marketplace; and,
limited direct testing and data as to effects of DEPs on shaping investor behaviour.
To respond to the above, the Investor Office undertook a behavioural science study on gamification
and other behavioural techniques under the DEPs umbrella. The Report, which is appended to this
Notice, is a result of this work and provides:
(i) a taxonomy of gamification and other behavioural techniques that are currently being
used or may be used by online brokerages in the future and their likely impact on
retail investor behaviour—both positively and negatively; and
(ii) the results of an online randomized controlled trial (RCT) experiment that examines
the use of points and top traded lists to determine their impact on trading frequency.
The goal of this work is to assist the OSC, other regulators and stakeholders in understanding these
new developments. In responding to the developments, we encourage taking an evidence-informed
1
Ontario Securities Commission Business Plan for the fiscal years ending 2023-2025, available online at
https://www.osc.ca/sites/default/files/2022-04/pub_20220426_osc-2023-2025-business-plan.pdf at p. 23.
OSC Staff Notice 11-796
Digital Engagement Practices in Retail Investing:
Gamification and Other Behavioural Techniques
2
approach, using behavioural insights to facilitate the use of DEPs in a manner that supports good
investor outcomes.
Research findings
Gamification refers to a variety of behavioural techniques that integrate game-related elements into
non-gaming contexts and applications, with the purpose of improving user experience and
engagement. We use the term other behavioural techniques to refer to DEPs that use insights from
behavioural science in ways that can influence investor behaviour but do not meet the definition of
gamification. DEPs themselves are “the tools including behavioural techniques, differential marketing,
gamification, design elements or design features that intentionally or unintentionally engage with
retail investors on digital platforms as well as the analytical and technological tools and methods.”
2
This definition highlights a range of potential tools, such as behavioural techniques, differential
marketing, gamification, design elements, design features, and data analytics, that increase user
engagement. There are other DEP tools beyond this definition as well, such as artificial intelligence
and dark patterns.
The Report
examines five gamification techniques used on self-directed investor platforms:
1. Ga
mblification: Techniques derived from gambling, which most prominently include the use
of variable rewards. Variable rewards are economic benefits (e.g., cash payouts) where the
size, timing, or likelihood of the benefit is unpredictable to the user. Beyond variable rewards,
the gamblification category might also include language and imagery that evokes gambling
(e.g., reference to “jackpots” or scratch cards).
2. Leaderboards: Public displays of ranked information about application users’ performance.
Leaderboards enable and encourage social comparison and competition.
3. Rew
ards (negligible or non-economic rewards such as points, badges, scores): Providing
rewards for performing tasks or accomplishing goals within an online application. Our
definition includes rewards with either no economic value or with nominal economic value
that should not materially influence investor behaviour under a purely rational economic
decision-making model.
4. Goal and
Progress Framing: Design elements that i) help users set and visualize their goals,
and/or ii) strategically frame users’ performance and progress with respect to these goals to
stimulate greater levels of engagement.
5. Feedback: T
he provision of information about a user's performance on a task in (near) real-
time, including both continuous progress feedback and immediate success feedback.
The Report also examines four other behavioural techniques:
2
This definition is consistent with that of the U.S. Securities and Exchange Commission (SEC)’s definition found
in Release Nos. 34-92766 Request for Information and Comments on Broker-Dealer and Investment Adviser
Digital Engagement Practices, Related Tools and Methods, and Regulatory Considerations and Potential
Approaches; Information and Comments on Investment Adviser Use of Technology to Develop and Provide
Investment Advice at page 1, available online at https://www.sec.gov/rules/other/2021/34-92766.pdf.
3
1. Salience / Attention-inducing Prompts: Information is more likely to influence people’s
behaviour if it attracts their attention.
2. Sim
plification and Selective Deployment of Friction Costs: The design of the user experience
that reduces or introduces small barriers across the user journey, influencing the likelihood
and manner in which a user completes a specific task. We use “simplification” to refer to
reductions in small barriers and “friction costs” to refer to increases in small barriers.
3. Soci
al Interactions: Design elements that enable platform users to interact with other users
by i) generating, sharing, viewing, and reacting to content, and ii) engaging in direct
messaging.
4. Social Norms: Design features which signal social norms (i.e., information about how others
think and behave).
These techniques could be employed in a manner that has positive influences on retail investors such
as:
• encouraging deposits to investment accounts,
• encouraging greater participation and learning of investor education modules on digital
platforms,
• improving diversification of the investor’s portfolio, and
• setting and monitoring progress towards long-term retirement savings goals.
However, these techniques also could have negative influences on investor behaviour such as:
• increasing risk taking by overweighting small probabilities,
• creating habit forming behaviours,
• invoking a psychological “hot” state that influences a user’s subsequent behaviour such as
the “hot hand” fallacy, making a person more likely to gamble with a windfall or unexpected
bonus,
• increase trading frequency and risk-taking,
• inc
reasing focus on shorter-term outcomes or trading activity that reduces longer term
returns and/or undermines investment goals, and
• increasing investor’s (over)confidence, negatively impacting investor performance.
We
tested some of the techniques identified in the Report in a RCT. We conducted an online RCT with
2,430 Canadians to assess the impact of two techniques of interest on investing behaviours in a
simulated trading environment: (1) giving investors “points” with negligible economic value for buying
or selling stocks—a form of reward, and (2) showing investors a “top traded list”—a combination of
attention-inducing prompts and social norms. The experiment was conducted online in a simulated
real-world trading environment with Canadians aged 18-65 engaging through mobile, tablet or
desktop devices.
Research participants received $10,000 in simulated “money” to invest in up to six different
fictitiously-named stocks. After their initial allocation of funds, they were taken through seven
4
simulated weeks of stock price movements, with an option to buy and/or sell stocks during each week.
At the end of the experiment all participants received a fixed amount of compensation for
participating in the experiment. They also earned additional compensation based on their balance at
the end of the experiment. Participants were aware that the larger the value of their portfolio at the
end of the experiment, the more they would earn. This created an incentive for participants to trade
thoughtfully and to try to maximize their returns.
Importantly, participants who were rewarded with points made 39% more trades than participants in
the control group (i.e., those who were exposed to the same trading simulation but without any
gamification or other behavioural techniques). This statistically significant difference was found
despite the fact that the points had negligible value. This is an important finding given that there is a
strong negative impact of increased trading volume on investors’ returns (on average)
3
, and in light of
the material benefit that may be gained by digital trading platforms from increased trading volume.
Showing research participants a top traded list did not increase their trading frequency.
Furthermore, participants who saw the top traded lists were 14% more likely than participants in the
control group to buy and sell those top listed stocks. This finding suggests that showing participants a
top traded lists can affect their trading decisions, nudging them towards buying and selling the stocks
listed as top traded, which is herding. There were no differences between the points group and the
control group in terms of the buying and selling the top traded stocks.
Conclusion
These findings reinforce the importance of using behavioural science as a policy tool by regulators.
Given the statistically significant findings derived from the RCT, the Report recommends that
regulators consider the implications of the findings, including whether any of the gamification and
other behavioural techniques examined have attributes similar to recommendations and/or result in
investor behaviour that is (on average) detrimental to investor outcomes, and if so, consider possible
responses.
The Report also recommends:
1. collecting more data to see the impact of gamification and other behavioural tactics through
leveraging data collected by digital trading platforms, or through other experiments,
2. collecting evidence and data on strategies to mitigate negative impacts of DEPs to determine
if mitigation approaches are effective (e.g., adding friction points), and
3. ex
ploring positive impacts of gamification and other behavioural techniques to increase
investing knowledge and level of expertise.
W
e encourage registrants to review the findings of the Report and consider the influence that their
DEPs may have on their clients so that negative investor behaviours are not encouraged (whether
inadvertently or otherwise), and to focus their use of DEPs in a manner that supports good investor
outcomes.
3
Barber, B. M., & Odean, T. (2000). Trading is hazardous to your wealth: The common stock investment performance of individual
investors. Journal of Finance, 55(2), 773-806.
5
We look forward to engaging with investors, registrants, and other stakeholders with respect to the
Report’s findings and our broader work to improve the investor experience and expand investor
protection.
Que
stions
If you have any questions or comments about this Notice or the Report, please contact:
Tyler Fleming
Director
Investor Office
20 Queen Street West, 22nd Floor
Toronto, ON M5H 3S8
Email: [email protected]
Marian
Passmore
Senior Advisor, Policy
Investor Office
20 Queen Street West, 22nd Floor
Toronto, ON M5H 3S8
Email: mpassmor[email protected]ov.on.ca
Ma
tthew Kan
Senior Advisor, Behavioural Insights
Investor Office
20 Queen Street West, 22nd Floor
Toronto, ON M5H 3S8
Email: mkan@osc.gov.on.ca
Research Report
November 2022
Digital Engagement
Practices in Retail Investing:
Gamifcation &
Other Behavioural
Techniques
Prepared by the Behavioural Insights Team (BIT) in Collaboration with the Investor Office
Research and Behavioural Insights Team (IORBIT) of the Ontario Securities Commission
2
Table of Contents
Table of Contents...................................................................................................................2
Executive Summary ...............................................................................................................3
Key Findings.......................................................................................................................5
Introduction ............................................................................................................................7
Overview of Project Approach..............................................................................................10
Key limitations ..................................................................................................................11
Exploratory Research...........................................................................................................12
Exploratory Research Methodology..................................................................................12
Exploratory Research Findings: Taxonomy of Gamification and Other Behavioural
Techniques.......................................................................................................................15
Experimental Research........................................................................................................ 36
Experimental Research Methodology ............................................................................... 36
Experimental Research Findings ......................................................................................... 44
Conclusion: Considerations for Regulators .......................................................................... 48
Appendix A: Use of Gamification and Other Tactics on Trading Platforms ...........................50
Appendix B: Detailed Experimental Research Findings........................................................52
Appendix C: Experimental Research Analysis Technical Details..........................................62
Appendix D: Experimental Research Screens......................................................................66
Appendix E: Bibliography .....................................................................................................79
3
Executive Summary
A wave of digital, mobile-friendly investing platforms has created new options for retail
investors in Canada and around the world. While these platforms have expanded market
participation, there is growing interest in some of the digital engagement practices (DEPs)
that they and, to a lesser extent, more traditional retail investment platforms use and how
these may raise investor protection concerns. These tactics, sometimes referred to broadly
as “gamification,” use insights from behavioural science to influence investor behaviour.
Broadly, investing platforms use a wide range of DEPs to increase user engagement. They
do this for a variety of business objectives (e.g., customer acquisition and retention, revenue,
profitability) and not necessarily to improve long-term outcomes for their retail investors.
Various regulators around the world have expressed concerns that some of these tactics
may negatively impact investor outcomes. For example, the United States Securities and
Exchange Commission (SEC) has flagged concerns that these features may encourage
investors to trade more often, invest in different products, or change their investment strategy
in inappropriate ways.
1
The goal of this research report is to support the Ontario Securities Commission (OSC) and
other regulators and stakeholders in understanding and responding to these new
developments. This report aims to help chart an effective, evidence-informed path forward in
the months and years ahead as digital trading platforms continue to evolve and grow.
The Behavioural Insights Team was engaged by the OSC’s Investor Office to:
1. Generate a taxonomy of gamification and other behavioural techniques by conducting
literature and environmental scans; and,
2. Conduct an experiment that examines the effects of gamification and other related
behavioural techniques on retail investor behaviours.
We worked in close partnership with the Investor Office Research and Behavioural Insights
Team (IORBIT) to develop the research parameters for the taxonomy, design and conduct
an experiment using a randomized controlled trial (RCT), analyze the experimental data, and
prepare this report. IORBIT’s insights, advice, and feedback were critical to this project’s
success.
In this report, we examine gamification and other behavioural techniques to see how they
affect investor behaviour—both positively and negatively. We outline a taxonomy of
gamification and other behavioural techniques currently employed or with high relevance to
retail investing, and their potential implications for investor behaviour. The five gamification
techniques examined were: (1) gamblification, (2) leaderboards, (3) rewards (negligible or
non-economic rewards such as points, badges, scores), (4) goal and progress framing, and
(5) feedback. The four other behavioural techniques examined were: (1) salience / attention-
inducing prompts, (2) simplification and selective deployment of friction costs, (3) social
interactions, and (4) social norms. We also discuss the results of an experiment (an RCT)
1
U.S. Securities and Exchange Commission (2021, Aug. 27). SEC Requests Information and Comment on Broker-Dealer and
Investment Adviser Digital Engagement Practices, Related Tools and Methods, and Regulatory Considerations and Potential
Approaches; Information and Comments on Investment Adviser Use of Technology. Retrieved from:
https://www.sec.gov/news/press-release/2021-167.
4
with 2,430 investors that simulated a real-world trading environment, in which we measured
the effects of two digital engagement practices, points and top traded lists, on trading
behaviour (e.g., trading frequency).
As the use of the terms digital engagement practices, behavioural techniques, and
gamification have become increasingly popular, it is imperative for regulators to have a clear
and common definition of these terms to allow for rigorous research and potential regulatory
action. To navigate the DEP landscape, we have used the SEC’s definition of DEPs (see Key
Definitions, below) as a foundation. This definition highlights a range of potential tools, such
as behavioural techniques, differential marketing, gamification, design elements, design
features, and data analytics, that increase user engagement. There are other DEP tools
beyond this definition as well, such as artificial intelligence and dark patterns. This report
does not examine all such tools; it focuses on gamification and other behavioural techniques
used in self-directed digital trading platforms. Figure 1 illustrates the relationship among
DEPs, behavioural techniques, and gamification. Gamification techniques are a subset of
behavioural techniques, which are in turn a subset of DEPs.
Key Definitions
For the purposes of this report:
Digital Engagement Practices (DEPs) are defined, consistent with the U.S.
Securities and Exchange Commission, as “the tools including behavioural
techniques, differential marketing, gamification, design elements or design
features that intentionally or unintentionally engage with retail investors on
digital platforms as well as the analytical and technological tools and methods.”
Gamification refers to a variety of behavioural techniques that integrate game-
related elements into non-gaming contexts and applications, with the purpose of
improving user experience and engagement.
We use the term other behavioural techniques to refer to DEPs that use
insights from behavioural science in ways that can influence investor behaviour
but do not meet the definition of gamification.
Digital trading platforms are websites, portals, and applications for trading
securities that are available to retail investors through their phones, computers,
tablets.
5
Key Findings
The Experiment
We conducted an online RCT to assess the impact of two gamification techniques of interest
on investing behaviours in a simulated trading environment: (1) giving investors “points” with
negligible economic value for buying or selling stocks, a form of reward, and (2) showing
investors a “top traded list”, a combination of attention-inducing prompts and social norms.
Participants who were rewarded with points made almost 40% more trades than
participants in the control group (i.e., those who were exposed to the same trading
simulation but without any gamification or other behavioural techniques). This is despite the
fact that the points had negligible value. This is a striking finding given the strong negative
impact of increased trading volume on investors’ returns (on average) and the benefit of
increased volume that may exist for digital trading platforms. The “top traded list” did not
2
increase trading frequency in our experiment.
In addition, participants who saw the top traded lists were 14% more likely than
participants in the control group to buy and sell those top listed stocks. This finding
suggests that showing participants a top traded lists can affect their trading decisions,
nudging them towards buying and selling the stocks listed as top traded. There were no
2
Barber, B. M., & Odean, T. (2000). Trading is hazardous to your wealth: The common stock investment performance of
individual investors. Journal of Finance, 55(2), 773-806.
Figure 1: The relationship between DEPs, gamification, and other behavioural
techniques. This illustration is a simplification as some overlap between these
categories is not depicted.
6
differences between the points group and the control group in terms of the buying and selling
the top traded stocks.
Implications
Based on our findings, we recommend that regulators consider the implications for retail
investors when digital trading platforms offer points for trading activity, as well as display top
traded lists. More broadly, we encourage regulators to consider whether any of the
gamification and behavioural techniques examined have attributes similar to
recommendations and/or result in investor behaviour that is (on average) detrimental to
investor outcomes. If so, then possible responses to these techniques should be considered.
Furthermore, we encourage regulators to close the major gaps in empirical evidence by
collecting more data. Such data can be generated by conducting more experimental studies
using simulated investing platforms, and by reviewing the data from digital trading platforms
that have implemented gamification or other behavioural techniques. These actions will
enable the OSC and other regulators to set new empirically-driven regulatory strategies and
approaches.
7
Introduction
A wave of digital, mobile-friendly investing platforms has created new options for retail
investors in Canada and around the world. While these platforms can increase access and
expand market participation, there is growing regulatory interest in some of the digital
engagement practices (DEPs) that they and, to a lesser extent, more traditional retail
investment platforms use and how these may raise investor protection concerns. These
tactics, sometimes referred to broadly as “gamification,” use insights from behavioural
science to influence user (investor) behaviour. Regulators are concerned that some of these
tactics may negatively impact investor outcomes. For example, the United States Securities
and Exchange Commission, (the “SEC”) has flagged concerns that these features may
encourage investors to trade more often, invest in different products, or change their
investment strategy.
3
Broadly, investing platforms use a wide range of Digital Engagement Practices (DEPs) to
increase user engagement. They do this for a variety of business objectives (e.g., customer
acquisition and retention, revenue, profitability) and not necessarily to improve long-term
outcomes for their retail investors. Various regulators around the world have expressed
concerns that some of these tactics may negatively impact investor outcomes. For example,
the United States Securities and Exchange Commission (SEC) has flagged concerns that
these features may encourage investors to trade more often, invest in different products, or
change their investment strategy in inappropriate ways.
4
The goal of this research report is to support the Ontario Securities Commission (OSC) and
other regulators and stakeholders in understanding and responding to these new
developments. This report aims to help chart an effective, evidence-informed path forward in
the months and years ahead as digital trading platforms continue to evolve and grow.
While there is significant interest surrounding the use of gamification and other behavioural
techniques, they are recent developments in the investing context, and there is little research
into how they are affecting investor behaviour and decision-making. In this context, the
Ontario Securities Commission (OSC) engaged the Behavioural Insights Team (BIT) to:
1. Generate a taxonomy of gamification and behavioural techniques by conducting
literature and environmental scans of:
a. Relevant research into how gamification and other behavioural techniques
can be used to influence retail investor behaviour, as well as key gaps in that
research;
b. How firms serving retail investors are currently using these techniques in
Canada and in select international markets; and
c. Other ways that firms may use gamification and other behavioural techniques
in the future, given the approaches being used in other industries.
3
U.S. Securities and Exchange Commission (2021, Aug. 27). SEC Requests Information and Comment on Broker-Dealer and
Investment Adviser Digital Engagement Practices, Related Tools and Methods, and Regulatory Considerations and Potential
Approaches; Information and Comments on Investment Adviser Use of Technology. Retrieved from:
https://www.sec.gov/news/press-release/2021-167.
4
U.S. Securities and Exchange Commission (2021, Aug. 27). SEC Requests Information and Comment on Broker-Dealer and
Investment Adviser Digital Engagement Practices, Related Tools and Methods, and Regulatory Considerations and Potential
Approaches; Information and Comments on Investment Adviser Use of Technology. Retrieved from:
https://www.sec.gov/news/press-release/2021-167.
8
2. Conduct an experiment that examines the effects of gamification and other related
behavioural techniques on retail investor behaviours.
We worked in close partnership with the Investor Office Research and Behavioural Insights
Team (IORBIT) to develop the research parameters for the taxonomy, design and conduct
an experiment using a randomized controlled trial (RCT), analyse the experimental data, as
well as prepare this report. IORBIT’s insights, advice, and feedback were critical to this
project’s success.
In this report, we examine gamification and other behavioural techniques to see how they
affect investor behaviour—both positively and negatively. We outline a taxonomy of
gamification and other behavioural techniques with high relevance to retail investing, and
their potential implications for investor behaviour. We also discuss the results of an
experiment (an RCT) with 2,430 investors that simulated a real-world trading environment, in
which we measured the effects of two digital engagement practices (i.e., points and top
traded lists) on trading behaviour (i.e., trading frequency).
As the use of the terms digital engagement practices, behavioural techniques, and
gamification have become increasingly popular, it is imperative for regulators to have a clear
and common definition of these terms to allow for rigorous research and potential regulatory
action. To navigate the DEP landscape, we use the SEC’s definition of DEPs (see Key
Definitions, below) as a foundation. This definition highlights a range of potential tools for
increasing user engagement. This report does not examine all such tools. For example, we
do not examine the use of predictive data analytics, dark patterns, or artificial intelligence.
Reflecting our expertise in behavioural science, this research focuses on gamification and
other behavioural techniques used in self-directed digital trading platforms. Figure 2
illustrates the relationship between DEPs, behavioural techniques, and gamification.
Gamification techniques are a subset of behavioural techniques, which are in turn a subset of
DEPs.
9
Key Definitions
For the purposes of this report:
Digital Engagement Practices (DEPs) are defined, following the U.S. Securities
and Exchange Commission, as “the tools including behavioural techniques,
differential marketing, gamification, design elements or design features that
intentionally or unintentionally engage with retail investors on digital platforms as
well as the analytical and technological tools and methods.”
Gamification refers to a variety of behavioural techniques that integrate game-
related elements into non-gaming contexts and applications, with the purpose of
improving user experience and engagement.
We use the term other behavioural techniques to refer to DEPs that use
insights from behavioural science in ways that can influence investor behaviour
but do not meet the definition of gamification.
Digital trading platforms are websites, portals, and applications for trading
securities that are available to retail investors through their phones, computers,
tablets, or other technology.
The image below represents the relationship between DEPs, gamification, and other
behavioural techniques, all of which can be implemented in digital platforms.
Figure 2: The relationship between DEPs, gamification, and other behavioural
techniques. This illustration is a simplification as some overlap between these categories
is not depicted.
10
Overview of Project Approach
This project was conducted in two main phases: exploratory research and experimental
research, as illustrated in the diagram below.
For the exploratory research, we conducted a scan and synthesis of relevant behavioural
science to understand how gamification and other behavioural techniques may influence
retail investor behaviour, in ways that both support and may negatively impact investor
outcomes. We also reviewed select retail investor platforms, news articles, and various
regulators’ statements and reports to understand how firms in Canada and other select
markets are currently using these approaches. Our exploratory research was summarized in
a taxonomy that listed each current or potential technique being employed, their known or
potential impact on investor behaviour, and their current use on investment platforms we
reviewed.
In the second phase, we ran an experiment to empirically test the impact of two selected
techniques on key investor behaviours. This experiment was designed to address key gaps
in the existing evidence base.
As a final step in our process, we developed a set of considerations for regulators informed
by both streams of research.
Figure 3: Overview of the exploratory and experimental research approaches
Overview of Project Approach
11
Key limitations
This section summarizes the most important limitations of this research report. Following
sections provide more detailed accounts of limitations related to each research methodology.
1. This report is not exhaustive in identifying how gamification and other behavioural
techniques are being used on investment platforms today. The proliferation of
platforms, limitations in access to platforms and ability to engage in trading activity,
and the bounded timelines for this review prevent an exhaustive report. However, we
believe it reflects a reasonable cross-section of the techniques being used by self-
directed investing platforms.
2. While we have tried to identify likely uses for gamification and other behavioural
techniques beyond what we found in our review, operators of digital trading platforms
are likely to identify further applications that this report does not consider.
3. There are a vast number of unanswered empirical questions about how DEPs
influence investor behaviour. Our experiment had to select a limited number of
techniques to test and behaviours to measure. Further, our experiment was
conducted in a controlled environment, an investing simulation. Participants in the
experiment did not use real funds and traded in fictitious equities. However, the use
of a robust experimental method (i.e., an RCT) provides us with confidence in terms
of the validity of our findings and our ability to generalize them to real-world trading. In
addition, research participants were compensated based on their returns, and other
aspects of the simulation were carefully designed to enhance its generalizability.
12
Exploratory Research
This section summarizes the methodology and findings from our exploratory research, which
included a literature scan and environmental scan.
Exploratory Research Methodology
We applied a mixed-methods approach to address two exploratory research questions:
1. How do gamification and other behavioural techniques influence investor decision-
making?
2. How are firms serving retail investors currently using or planning to use these
techniques?
Methods included a literature scan and environmental scan. The literature scan provided a
theoretical foundation by identifying and summarizing 31 items of relevant behavioural
science and economic literature related to gamification, other behavioural techniques, and
retail investing. The environmental scan conducted September 7 - October 1, 2021, provided
context on the extent to which firms serving self-directed retail investors are applying these
techniques. It included a direct observation of 12 self-directed retail investor platforms (which
have been anonymized and numbered Platforms 1 to 12 for the purposes of this report), and
a further review of 16 news articles and statements. Both methods informed the development
of a taxonomy of gamification and other behavioural techniques for retail investing platforms.
This taxonomy summarized each current identified technique, their known or potential impact
on investor behaviour, and their current use on the investing platforms we reviewed. More
information on each part of our exploratory research method is captured in the table below:
Exploratory Research Activities: Overview of Methodology
Literature Scan
● The literature scan began by conducting research database
searches for key terms (e.g., investing, gamification), then used a
“snowball method”, whereby we reviewed the sources cited by
relevant papers. The search was also expanded to identify non-peer
reviewed (grey literature), internal BIT resources, and sources
recommended by the OSC.
● Key methodological considerations for the literature scan included:
○ Defining gamification: Gamification is an umbrella term
used to describe the integration of game-related elements
into non-gaming contexts and applications, with the
purpose of improving user experience and engagement.
5
Gamification has become an increasingly popular design
component of applications that do not constitute games,
such as apps aimed at helping users keep track of their
5
Deterding, S., Dixon, D., Khaled, R., & Nacke, L. (2011, September). From game design elements to gamefulness: defining"
gamification". In Proceedings of the 15th international academic MindTrek conference: Envisioning future media environments
(pp. 9-15).
13
weight loss goals, learn new languages, or trade
securities. It is used broadly to increase engagement with
digital applications and to increase the behaviours
encouraged by those applications (e.g., studying,
exercising).
○
Exploring gamification research: Published academic
research on gamification is concentrated in two sectors,
education and health. While this may reflect greater use of
gamification in these sectors, it may also be that
researchers are more likely to conduct experiments and
receive data in these domains. While our literature scan
prioritized the more limited research related to investor
behaviour, we integrated findings from other fields where
we believed it would be relevant across user contexts.
Environmental
Scan
● To conduct the environmental scan, we collaboratively identified 12
self-directed retail investing platforms of interest (both Canadian and
international) with the OSC and went through the process of
registering for an account, keeping a record of gamification and other
behavioural techniques identified throughout the user experience
within each platform. Where it was not possible to register for
accounts with platforms based outside of Canada, we reviewed
videos on YouTube highlighting the features of each of these apps
and how to use them. We did not execute any trades on these
platforms.
● We also reviewed a range of other sources, including news articles
and statements by regulators for further information on how these
platforms may be using (or planning to use) gamification and other
behavioural techniques.
● Our scan excluded banking and financial management platforms that
do not enable trading in securities.
●
See Appendix A for a table summarizing which gamification and
other behavioural techniques were observed on each platform
reviewed.
Taxonomy of
Gamification
and Other
Behavioural
Techniques
● Synthesizing the findings from the literature and environmental
scans, we developed a taxonomy of gamification and other
behavioural techniques that outlines how each technique has been
shown to (or plausibly might) impact specific retail investor
behaviours or choices. Investor behaviours of interest included
enrolling in the platform, engagement with the platform, deposits,
and a wide range of trading-related behaviours like trading frequency
and risk-taking.
● As noted above, there is no authoritative list or common
understanding of what constitutes a gamification technique in the
context of investing platforms. The most widely accepted definition of
gamification, “the use of game-design elements in non-gaming
contexts”, is broad and does not clearly identify what counts as a
14
gamification technique. Systematic reviews of gamification
techniques employ varying taxonomies. Securities regulators tend to
understand gamification very widely, including concepts from
behavioural science (e.g., attention-inducing prompts like
notifications) that are not generally understood to be part of
gamification. To address these challenges, we developed our own
taxonomy of gamification and other behavioural techniques that are
most relevant to retail investing platforms.
● We developed an initial list (i.e., taxonomy) of gamification
techniques based on three widely cited meta-analyses / systematic
reviews of gamification.
, ,
This initial taxonomy was further refined
876
after conducting our environmental scan; we eliminated certain
techniques that did not appear to be relevant to investing platforms
(such as chatbots, avatars, or fantasy themes). Our environmental
scan also revealed that certain other behavioural techniques that do
not meet the traditional definition of gamification are often discussed
alongside gamification tactics and used widely on investing
platforms. We included those techniques, like attention-inducing
prompts, in a separate section of our taxonomy and define them as
other behavioural techniques.
● The widespread use of gamification on investing platforms is a new
phenomenon that is rapidly escalating and shifting. We believe that
new techniques are likely to be deployed by firms operating
platforms. Given the overall purpose of this report in supporting
regulatory strategy, we did not want to constrain our taxonomy and
considerations solely to techniques that have already been
implemented. Where we speculate on further potential use cases,
we clearly note that such approaches are not yet in effect. We draw
conclusions on the potential impact of both current and potential
approaches based on our theoretical and empirical findings. Given
the importance of specific implementation features and context, we
cannot draw definitive conclusions on how gamification techniques
are likely to affect investor behaviour across platforms. As supported
by the existing research and our own experiment (see section
below), we draw inferences on the likely impacts of these techniques
on trading frequency, risk appetite, and other aspects of investing
behaviour.
6
Hamari, J., Koivisto, J., & Sarsa, H. (2014, January). Does gamification work?--a literature review of empirical studies on
gamification. In 2014 47th Hawaii international conference on system sciences (pp. 3025-3034).
7
Looyestyn, J., Kernot, J., Boshoff, K., Ryan, J., Edney, S., & Maher, C. (2017). Does gamification increase engagement with
online programs? A systematic review. PloS one, 12(3), e0173403.
8
Johnson, D., Deterding, S., Kuhn, K. A., Staneva, A., Stoyanov, S., & Hides, L. (2016). Gamification for health and wellbeing:
A systematic review of the literature. Internet Interventions, 6, 89-106.
15
Exploratory Research Findings: Taxonomy of
Gamification and Other Behavioural Techniques
The following taxonomy includes nine techniques informed by behavioural science that are
relevant to investing platforms. For each technique we provide: (1) a definition, (2) a
summary of how it is being used across industries and its general impact on user behaviour,
and (3) description of how it is being used in investing platforms and its impact on investor
behaviour and/or how it might be used and affect investor behaviour.
Gamification Techniques
Gamification refers to a variety of behavioural techniques that integrate game-related
elements into non-gaming contexts and applications, with the purpose of improving user
experience and engagement. Gamification techniques represent a subset of behavioural
techniques, which are a subset of (DEPs).
Gamblification
Definition: Gamblification refers to techniques derived from
gambling, which most prominently include the use of variable
rewards. Variable rewards are economic benefits (e.g., cash
payouts) where the size, timing, or likelihood of the benefit is
unpredictable to the user. Beyond variable rewards, the
gamblification category might also include language and
imagery that evokes gambling (e.g., reference to “jackpots,”
scratch cards).
9
General use and impact on behaviour: Gamblification has
been used to encourage a broad set of behaviours ranging
from user-platform engagement to vaccination . For
11
10
instance, Google Pay gives users virtual scratch cards worth
up to $10 in cash rewards as a variable reward for using this
payment option. In Canada, Tim Hortons’ “Roll Up the Rim”
contest is a famous example of a retailer leveraging variable
rewards to motivate purchasing behaviour.
Lotteries and other variable reward interventions can be potent
drivers of behaviour for three main reasons:
9
The overall relationship between trading and gambling is outside the scope of this research. While trading is distinct from
gambling insofar as investors pursue long-term strategies (and invest in productive assets), there is a relationship between
gambling and investing. Some traders are motivated by the thrill of potential short-term returns and treat trading as gambling. In
their paper “Trading as Gambling,” Dorn, Dorn and Sengmueller (2015) provide evidence for this, showing that when there are
large lotteries, trading activity declines as people substitute lottery entries for trading.
10
Grieve, R. & Lowe-Calverley, E. (2016, Jul 16). The power of rewards and why we seek them out. The Conversation.
11
Jacobson, L. & Tamakloe, K. S. (2021, Aug 19). “Have vaccine lotteries worked? Studies so far show mixed results”.
Politifact.
Figure 4: A modified version
of Platform 7’s rewards,
which give “surprise stocks”
with variable value to first-
time users when they sign up
for an account and to users
who refer a friend
16
1. Their combination of high-impact rewards and generally low win probabilities taps into
our tendency to disproportionately focus on reward magnitude and overweight small
12
probabilities.
13
2. The inherent uncertainty of variable rewards is habit-forming. Decades of research
have found that animals trained to perform behaviours such as pressing levers or
14
seeking drugs using variable rewards learn these behaviours much more readily
15
than those trained with consistent reward schedules. Such animals are also known to
be particularly resistant to so-called behavioural extinction
,
, meaning that they
1716
continue performing their conditioned behaviours long after they are no longer being
reinforced. Indeed, variable reward schedules are often used to simulate and study
addiction in research settings, and are thought to be responsible for the
18
addictiveness of gambling.
19
3. Variable rewards and other incentives couched in language reminiscent of gambling
can invoke a psychological “hot” state that influences a user’s subsequent behaviour.
For instance, when an investment is framed as a “jackpot” entry, our decision-making
becomes dominated by considerations of reward magnitude and less sensitive to
realistic chances. The results of a gamble can further distort our judgements, with
20
one study showing that winning on a gamble makes us 80% likely to continue with the
next gamble, compared to 20% after a loss. This behaviour is thought to stem from
21
a flawed assumption that a win signals a streak where future positive outcomes are
more likely than before (also known as the hot hand fallacy).
22
Current use on digital trading platforms and impact on behaviour: Gamblification
strategies are present on several digital trading platforms. Platform 7 gives “surprise stocks”
with variable value to first-time users when they sign up for an account and to users who
refer a friend. Notably, the free stocks are presented in the form of a scratch card, where
users are presented with an option of three scratch-off tickets and must choose one to
“scratch” their fingers across the phone screen to see what they’ve won. Platform 10 users
have to click on a virtual present to reveal the prize that they have “won” for referring a friend,
the value of which also varies. The platform then presents users with a list of three potential
stocks for them to invest in with the referral bonus. This list of three stocks makes it more
likely the user will choose one of those stocks instead of other securities that could be more
12
Griffiths, M., & Wood, R. (2001). The psychology of lottery gambling. International Gambling Studies, 1(1), 27-45.
13
Burns, Z., Chiu, A., & Wu, G. (2010). Overweighting of small probabilities. In Wiley Encyclopedia of Operations Research and
Management Science (pp. 1-8). Hoboken, NJ, USA: John Wiley & Sons.
14
Skinner, B.F. (1969). Contingencies of reinforcement: A theoretical analysis. Meredith Corporation.
15
Lagorio, C. H., & Winger, G. (2014). Random-ratio schedules produce greater demand for iv drug administration than fixed-
16
Miltenberger, R. G. (2008). “Behavioral Modification: Principles and Procedures”. Florida, Thomson.
17
Shull, R. L., & Grimes, J. A. (2006). Resistance to extinction following variable‐interval reinforcement: Reinforcer rate and
amount. Journal of the Experimental Analysis of Behavior, 85(1), 23-
39.
18
Egli, M., Schaal, D. W., Thompson, T., & Cleary, J. (1992). Opioid-induced response-rate decrements in pigeons responding
under variable-interval schedules: reinforcement mechanisms. Behavioural Pharmacology.
19
Haw, J. (2008). Random-ratio schedules of reinforcement: The role of early wins and unreinforced trials. Journal of
Gambling Issues, (21), 56-
67.
20
Chaudhry, S., & Kulkarni, C. (2021, June). Design Patterns of Investing Apps and Their Effects on Investing Behaviors. In
Designing Interactive Systems Conference 2021 (pp. 777-788).
21
Croson, R., & Sundali, J. (2005). The gambler’s fallacy and the hot hand: Empirical data from casinos. Journal of Risk and
Uncertainty, 30(3), 195-209.
22
Ayton, P., & Fischer, I. (2004). The hot hand fallacy and the gambler’s fallacy: Two faces of subjective randomness?. Memory
& cognition, 32(8), 1369-1378.
ratio schedules in rhesus monkeys. Psychopharmacology, 231(15), 2981-2988.
17
suitable, given the increased salience and reduced friction of purchasing them. Users do not
have to invest the referral bonus in these three stocks, but the approach increases the
likelihood that they will. Platform 10 also uses variable rewards for new users, offering the
cash equivalent of a stock worth up to $4500. The free stock bonus has a value between $5
and $4500 with an average of $
15.
23
Given the evidence on variable rewards more broadly, we expect that these approaches are
likely to increase platform sign-ups by offering a large potential bonus and increase the
frequency of referrals. We also believe that these experiences may increase the likelihood of
ongoing use of the application.
Beyond the immediate behaviour being rewarded, there are reasons to believe that
gamblification tactics can change subsequent financial decisions as well. We are significantly
more likely to gamble on money that feels like a windfall or unexpected bonus,
24,25
like a
large variable reward. These rewards may increase retail investor risk taking after receiving
the bonus, especially where the bonus is unexpectedly large. There are two underlying
effects. First, the “house money” effect describes how gamblers are less concerned about
losing their winnings than losing their own money, their pre-existing stake.
26
Second, those
who receive an unexpectedly large award may be influenced by the “hot hand fallacy”,
27
the
feeling that one is on a “hot streak” and that things are going to continue going well. This
might motivate higher levels of trading activity than users might otherwise engage in.
Potential use on digital trading platforms and impact on behaviour: In the future, digital
trading platforms may provide variable rewards for other behaviours, including deposits and
trading (e.g., based on the volume or type of trades made). For example, users might be
awarded an additional entry into a high-stakes lottery for every trade they carry out. If
implemented, the evidence on variable rewards suggests this could have an outsized effect
on trading frequency, surpassing the value of a fixed incentive per trade. Trading frequency
or volume is a critical behaviour of interest, given the strong negative correlation with investor
returns
28
and the incentives platforms can have to see higher trading volume. In terms of
deposits, variable rewards would likely result in increased deposit behaviours within a
specific account.
Leaderboards
Definition: A public display of ranked information about application users’ performance.
Leaderboards enable and encourage social comparison and competition.
General use and impact on behaviour: Leaderboards are one of the most common
gamification tactics across digital platforms and apps in a variety of industries.
29
By offering
23
Footnote source deleted re: Platform 10
24
Kellner, C., Reinstein, D., & Riener, G. (2019). Ex-ante commitments to “give if you win” exceed donations after a win. Journal
of Public Economics, 169, 109-127.
25
Adam, M., Roethke, K., & Benlian, A. (2021). Gamblified digital product offerings: an experimental study of loot box menu
designs. Electronic Markets, 1-16.
26
Thaler, R. & Johnson, E. (1990), Gambling with the house money and trying to break even: The effects of prior outcomes to
risky choice. Management Science, 36, (6), 643-660.
27
Croson, R., & Sundali, J. (2005). The gambler’s fallacy and the hot hand: Empirical data from casinos. Journal of Risk and
Uncertainty, 30(3), 195-209.
28
Odean, T. (1999). Do investors trade too much? American economic review, 89(5), 1279-1298.
29
Johnson, D., Deterding, S., Kuhn, K. A., Staneva, A., Stoyanov, S., & Hides, L. (2016). Gamification for health and wellbeing:
A systematic review of the literature. Internet Interventions, 6, 89-106. Also: Looyestyn, J., Kernot, J., Boshoff, K., Ryan, J.,
18
users the opportunity to see and show their peer groups where they stand relative to others,
leaderboards tap into our desire for recognition and innate tendency for social comparison
and competition.
30,31
A 2017 meta-analysis (a statistical analysis that combines the
results of multiple scientific studies that address the same
question and increases the confidence in the results)
indicates that leaderboards are among the most effective
gamification tactics, often outperforming other approaches
like points and badges and generating small-to-medium-size
effects compared to control conditions.
32
In one study
focused on educational outcomes, researchers found that
gamifying an online learning platform with leaderboards
(alongside several other tactics), resulted in a 25% increase
in student retention, as well as 23% higher average test
scores compared to those produced by control conditions.
33
Another study found that leaderboards alone led to
approximately 40% higher levels of user activity in a gamified
image annotation task compared to control conditions, which
was approximately 4% and 16% more than researchers were
able to achieve with levels and points, respectively.
34
Indeed, commercial platforms frequently use leaderboards to
enhance user participation, with particular prevalence within
the fitness app industry, where platforms such as Strava and
Nike+ track and rank users based on running mileage and
other parameters of performance. It is worth noting, however,
that inducing a competitive spirit may not have the same
effects on everyone and may in fact disadvantage the performance of individuals who are
intrinsically less competitive.
35
Current use on digital trading platforms and impact on behaviour: Leaderboards are a
relatively rare feature of digital trading platforms. US-based Platform 8 offers the option to
enable social investing, allowing users to compare how they are doing with their peers by
featuring on a leaderboard where users are ranked based on returns weighted within a
certain time frame. Users must meet certain criteria to feature on the leaderboard, including
owning at least a minimum number of holdings worth at least a certain combined valued to
Edney, S., & Maher, C. (2017). Does gamification increase engagement with online programs? A systematic review. PloS one,
12(3), e0173403.
30
Leibbrandt, A., Gneezy, U., & List, J. A. (2013). Rise and fall of competitiveness in individualistic and collectivistic societies.
Proceedings of the National Academy of Sciences, 110(23), 9305-9308.
31
Looyestyn, J., Kernot, J., Boshoff, K., Ryan, J., Edney, S., & Maher, C. (2017). Does gamification increase engagement with
online programs? A systematic review. PloS one, 12(3), e0173403.
32
Looyestyn, J., Kernot, J., Boshoff, K., Ryan, J., Edney, S., & Maher, C. (2017). Does gamification increase engagement with
online programs? A systematic review. PloS one, 12(3), e0173403.
33
Krause, M. Mogalle, M., Pohl, H. & Williams, J. J. (2015). “A playful game changer: fostering student retention in online
education with social gamification”. In Proceedings of the Second ACM Conference on Learning@Scale.
34
Mekler, E. D., Brühlmann, F., Opwis, K., & Tuch, A. N. (2013, October). Do points, levels and leaderboards harm intrinsic
motivation? An empirical analysis of common gamification elements. In Proceedings of the First International Conference on
gameful design, research, and applications (pp. 66-73).
35
Song, H., Kim, J., Tenzek, K. E., & Lee, K. M. (2010, June). The effects of competition on intrinsic motivation in exergames
and the conditional indirect effects of presence. In Annual conference of the International Communication Association,
Singapore.
Figure 5: A modified version
of Platform 8's social
investing features, which
rank users on a leaderboard
based on weighted returns
within a certain time frame.
19
dissuade members from copying deceptive results from other users (e.g., a massive return
solely from one penny stock).
Platform 2 features multiple types of leaderboards. As an example, users have access to an
“Editors’ choice” leaderboard of investors to follow and copy, as well as a leaderboard of the
most copied investors on the platform. Users are also invited to try joining the Platform 2’s
“Popular Investor Program” which allows successful joiners to generate an income from
being copied by other users. As of September 2021, we did not identify any Canadian
platforms using leaderboards.
Leaderboards have not been evaluated in a trading context in the academic literature.
However, the studies conducted in other contexts, mentioned in the section above, suggest
that leaderboards can be expected to increase user engagement with digital trading
platforms. This may increase trading frequency and risk-taking, particularly in users who are
more motivated by social comparisons
36
and competition than their longer-term financial
goals. Leaderboards may also implicitly signal a social norm (see section below) around
striving for and celebrating high financial performance. Here, frequently changing leader
names may be viewed as the culmination of an ongoing competition, and a sign that this
contest is desirable and popular. The impact of this is likely to depend on the salience of the
leaderboard, whether economic or non-economic rewards are tied to leaderboard
performance, and the type of returns or activity that the leaderboard represents.
Leaderboards that focus on shorter-term returns, like Platform 2’s 12-month returns, may
increase myopic, speculative trading. Traders with a ranking on the leaderboards may also
experience increased (over)confidence, which negatively impacts returns from trading.
37
Potential use on digital trading platforms and impact on behaviour: In the future, digital
trading platforms could implement additional leaderboards for other types of investor
behaviour, such as trading frequency or even social interactions like “posts” or “likes” (see
following section). Displaying a leaderboard that measures activity could clearly increase the
frequency of trading. As described further below, even leaderboards for social interactions
may be deceptively risky, given how strong an influence on behaviour social feedback and
recognition can be. On the other hand, leaderboards for completing investor education
modules, where offered by digital platforms, could encourage greater participation and
learning.
Rewards (e.g., points, badges, scores)
Definition: Providing rewards for performing tasks or accomplishing goals within an online
application. Our definition includes rewards with either no economic value (e.g., badges,
scores, animations) or with nominal economic value (e.g., points that can be redeemed for
an insignificant financial value) that should not materially influence investor behaviour under
a purely rational economic decision-making model. This category excludes larger financial
36
For further discussion of social comparison, see the section on Social Interactions, below.
37
Biais, B., Hilton, D., Mazurier, K., & Pouget, S. (2005). Judgemental overconfidence, self-monitoring, and trading performance
in an experimental financial market. The Review of Economic Studies, 72(2), 287-312.
20
rewards (e.g., cash bonuses or points that can be redeemed for significant financial value),
as they constitute a traditional incentive, not a “behavioural” intervention.
General use and impact on behaviour: Providing
rewards like points, badges and scores are among the
most commonly used gamification tactics.
38
While these
rewards can have little or no economic value, they can still
have a significant effect on consumer / user behaviour,
promoting engagement in online programs
39
and
influencing consumer behaviour.
40
For example, the Nike+
app awards “NikeFuel points” for completing physical
activity tasks. These techniques motivate behaviour
through social comparison (see Leaderboards, above) and
our intrinsic desire to make progress, even if the measure
is arbitrary.
Badges, in particular, are popular features of apps and
online programs. They act as publicly visible signs of status
within the network of application users. Amazon marks
individuals as “top reviewers” when enough other users
mark their reviews as helpful. A “pre and post” evaluation of
the web platform “Sharetribe” found that badges increased
user posts, page views, and transactions.
41
In a commercial
context, retailers have long offered loyalty points and
programs. While these points can generally be redeemed
for goods and can be considered a traditional economic
reward, their impact on behaviour outstrips their pure economic value as people tend to
overvalue points they collect.
42
Not only are they overvalued, the mere decision to redeem a
reward significantly increases purchase behaviour before and after the redemption event.
43
However, studies examining the effect of badges exclusively on engagement with online
programs have found only small effect sizes.
44
Current use on digital trading platforms and impact on behaviour: An American
financial platform, Platform 8, uses points to reward users. Platform 8 is differentiated from
other investing platforms by providing a wide range of financial products and services on one
platform, including investing options, credit cards, loans, insurance, bank accounts, credit
score information, budgeting tools, etc. While points are not given for investing behaviours,
users can earn “Platform 8 points” for actions like spending money with the credit card or
38
Looyestyn, J., Kernot, J., Boshoff, K., Ryan, J., Edney, S., & Maher, C. (2017). Does gamification increase engagement with
online programs? A systematic review. PloS one, 12(3), e0173403.
39
Looyestyn, J., Kernot, J., Boshoff, K., Ryan, J., Edney, S., & Maher, C. (2017). Does gamification increase engagement with
online programs? A systematic review. PloS one, 12(3), e0173403.
40
Tobon, S., Ruiz-Alba, J. L., & García-Madariaga, J. (2020). Gamification and online consumer decisions: Is the game over?.
Decision Support Systems, 128, 113167.
41
Hamari, J. (2017). Do badges increase user activity? A field experiment on the effects of gamification. Computers in Human
Behavior, 71, 469-478.
42
Van Osselaer, S. M., Alba, J. W., & Manchanda, P. (2004). Irrelevant information and mediated intertemporal choice. Journal
of Consumer Psychology, 14(3), 257-270.
43
Dorotic, M., Verhoef, P. C., Fok, D., & Bijmolt, T. H. (2014). Reward redemption effects in a loyalty program when customers
choose how much and when to redeem. International Journal of Research in Marketing, 31(4), 339-355.
44
Looyestyn, J., Kernot, J., Boshoff, K., Ryan, J., Edney, S., & Maher, C. (2017). Does gamification increase engagement with
online programs? A systematic review. PloS one, 12(3), e0173403.
Figure 6: A modified version of
Platform 8’s rewards program,
which rewards users with points
for completing various in-app
actions, such as checking your
credit score
21
signing into the app on a daily basis and on consecutive days. The economic value of the
Platform 8 points is quite limited. For example, users are rewarded with 1 point for a daily
app login, and each Platform 8 point is worth $0.01. The earned points then can be
converted into fractional shares of stocks within the investing component of Platform 8’s
digital platform, converted into cryptocurrency within the digital assets component of the
platform, turned into cash, used towards loans, or used as credit towards the Platform 8
credit card. Enabling users to apply their points on traditional securities or crypto investments
may be a particularly effective way to get non-investors to use these components of the
Platform 8 digital platform.
45
It leverages the concept of ‘mental accounting’, which describes
how people tend to treat money differently based on subjective criteria, such as its source.
They may, for example, be more willing to take on risk with “bonus money” than money from
other sources (e.g., employment income).
Another US-based personal finance app has taken a slightly different approach. They reward
users with a percentage of their debit card purchases back in stock. For example, when a
user spends money at Walmart, Amazon, or Starbucks, they earn fractional shares in these
companies. When they spend at a smaller business, such as a local restaurant, they earn an
investment of their choosing, either a stock or an ETF. The app’s company’s analysis
suggests that one-third of customers using this reward card go on to make a follow up
investment in the given stock or fund.
46
While offering users automated investments in the
market is not negative for investors, this type of reward system may reduce diversification.
In a much simpler use of rewards, Platform 7 previously showered users’ screens with digital
animations to celebrate certain actions like placing a first trade or successfully referring
friends. A 2021 experiment tested how gamification techniques, including confetti bursts,
achievement badges, and messages of encouragement influence users’ risk taking when
trading. Participants in the experiment were assigned to trade virtual assets on either a
simple experimental platform that mimics a retail investing app or a gamified version.
47
In
each round, participants were given a virtual asset that they can sell at any time. Every two
seconds, the asset price either increased by a random amount or, with a small probability
that varied each round, crashed to zero. Users who traded on the gamified version of the
platform took on significantly more risk. For example, they waited 14% longer to sell in the
gamified version. The impact of gamification was stronger for high-risk environments (i.e., for
assets that had a higher probability of crashing). Increasing the probability of a crash from
2% to 5% led to a 246% stronger impact of gamification on risk taking. In addition, the effect
was stronger for inexperienced traders with lower financial literacy; a one standard deviation
increase in a financial literacy score reduced the impact of gamification by 56%.
Potential use on digital trading platforms and impact on behaviour: Trading platforms
could introduce points / scores or badges as a way to potentially motivate a wide range of
investor behaviours. For example, badges could be awarded for purchasing different types of
securities (e.g., options) or points could be awarded on a per-trade basis. Digital platforms
might consider enabling users to “cash in” these points for small rewards (e.g., fractions of
stocks, gift certificates, etc.) or keep them purely nominal. Small rewards are likely to have a
45
Footnote source deleted re: Platform 8
46
Footnote source deleted re: a US-based platform
47
Chapkovski, P., Khapko, M., & Zoican, M. (2021). Does gamified trading stimulate risk taking?. SSRN 3971868.
22
larger impact on investor behaviour, given that they could be overvalued as being
economically significant, and would have some financial cost to the digital platforms.
There may also be opportunities for investing platforms to deploy trading-related rewards in
ways that are more likely to benefit users. For example, they could award points for
improving the diversification of the investor’s portfolio (e.g., across asset classes or sectors).
Goal and progress framing
Definition: Design elements that i) help users set and visualize their goals, and/or ii)
strategically frame users’ performance and progress with respect to these goals to stimulate
greater levels of engagement.
General use and impact on behaviour: Diverse goal and progress framing tactics are
being used across industries (e.g., air travel, food & drink, health apps) to motivate two
primary types of engagement behaviours: i) purchasing
and consumption, and ii) work and productivity. For
example, flight miles programs motivate consumption by
strategically framing their customers’ flight histories as
progress towards a particular goal. One study has found
that reminding customers of how close they are to
unlocking rewards associated with hitting an arbitrary
points target can make them 55% relatively more willing to
agree to receive marketing content in exchange for bonus
miles, compared to when they are further from that
threshold.
48
In a practice that has attracted some controversy, Uber
uses goal framing tactics to nudge their workers to keep
driving beyond their desired log-off times.
49
When drivers
are about to log off for the day, the app alerts them to how
close they are to their daily income target (or a target
which the company took the liberty to set for them) and
encourages them to continue working. Although an
absence of publicly available data from these interventions
makes it difficult for us to specify the magnitude of the
behavioural effect, the aggregate evidence suggests that
goals (even arbitrary ones), influence behaviour across a
range of activities.
How companies or apps choose to present an individual’s
progress toward a goal is also impactful. First, the closer we think we are to a goal, the more
effort we are willing to expend to achieve it, a concept called the goal-gradient hypothesis.
Our perceived proximity to a goal can be manipulated through “endowed progress.” People
getting a “Buy 12 coffees, get 1 free” card, with two of these coffees already pre-stamped
48
Kivetz, R., Urminsky, O., & Zheng, Y. (2006). The goal-gradient hypothesis resurrected: Purchase acceleration, illusionary
goal progress, and customer retention. Journal of Marketing Research, 43(1), 39-58.
49
Scheiber, N. (2017). How Uber uses psychological tricks to push its drivers’ buttons. New York Times.
Figure 7: Uber has used
progress framing to encourage
drivers to continue driving
23
buy more coffee than those getting a “Buy 10 coffees, get 1 free” card even though the actual
proximity to the goal, buy 10 more coffees, is the same.
50
Second, a theory known as the small area hypothesis states that “individuals in pursuit of a
goal exhibit stronger motivation when they focus on whichever is smaller in size: the area of
their completed actions or their remaining actions needed to reach a goal.”
51
This means that
if a user is 10% of the way to a goal, it’s more motivating to focus their attention on the 10%
they have accomplished than the 90% they haven’t. Conversely, if they are 90% of the way
to a goal, it is more motivating to focus on the 10% remaining than the 90% complete. One
study of the behaviours of over 90,000 members of an online Q&A community has found that
small-area progress framing accounted for a minimum of 78% increase in user activity and
engagement after the platform was restructured.
52
“Streaks” are another popular tactic used to frame progress. Derived from the concept of
“winning streaks” in sports, they are used as a measurement of how consistently a user
completes a specific action. For example, Duolingo refers to streaks in their language-
learning platform, where users grow their streak for each day in a row they complete a
lesson. An analysis of this feature has revealed that the streaks help increase users’
attention to their learning purpose when the challenges increase and improves motivation.
53
Winning streaks were also shown to increase in perceived attractiveness with greater length.
The effectiveness of this feature has been noted by other industries: mobility service
providers such as Uber and Lyft distribute so-called “streak” or “consecutive ride” bonuses to
motivate their driver employees.
54
Current use on digital trading platforms and impact on behaviour: Our environmental
scan did not identify any platforms using goal & progress framing tactics as defined in this
report.
55
Potential use on digital trading platforms and impact on behaviour: Given the clear
relevance to financial decision-making, we believe it is likely that digital trading platforms will
introduce goal-setting and goal-framing features. While platforms offering effectively self-
directed investment services are prohibited from offering recommendations, they could
permit investors and financial consumers to set out their financial goals and measure
progress.
The way in which companies choose to solicit and define their users’ goals and frame their
progress will shape the behaviours that result. Helping users set and monitor progress
against retirement savings goals, for example, could help instil a long-term investment
outlook. However, such interventions would still need to be paired with clear guidance
50
Jensen, J. D., King, A. J., & Carcioppolo, N. (2013). Driving toward a goal and the goal‐gradient hypothesis: the impact of goal
proximity on compliance rate, donation size, and fatigue. Journal of Applied Social Psychology, 43(9), 1881-1895.
51
Koo, M., & Fishbach, A. (2012). The small-area hypothesis: Effects of progress monitoring on goal adherence. Journal of
Consumer Research, 39(3), 493–509.
52
Kundisch, D., & von Rechenberg, T. (2017). Does the framing of progress towards virtual rewards matter? Business &
Information Systems Engineering, 59(4), 207-222.
53
Huynh, D., & Iida, H. (2017). An Analysis of Winning Streak's Effects in Language Course of “Duolingo". University
Kebangsaan Malaysia Press.
54
Uber (2022). “How does the Consecutive Trips promotion work?” Retrieved from: https://help.uber.com/driving-and-
delivering/article/how-does-the-consecutive-trips-promotion-work?nodeId=de983305-076a-40cf-aaf4-7b23f50a0007. See also:
Lyft (2022). "Streak Bonus". Retrieved from: https://help.lyft.com/hc/e/articles/115015748908-Streak-Bonus.
55
There are financial apps which do not enable users to trade individual securities, offer such features, but are not included in
the scope of this research.
24
regarding the impact of increased trading frequency on long-term outcomes
56
to help mitigate
the risk that people think more active trading approaches will support long-term goals.
Similarly, goal framing should be attentive to other key aspects of successful personal
investing, including diversification, savings rates, risk taking, possible need for funds for
short-term emergencies, etc. Beyond encouraging long-term thinking, platforms could help
users break down their goals into manageable sub-goals to support goal attainment.
57
On the
other hand, setting goals related to shorter-term outcomes or trading activity could be
harmful. For example, setting monthly goals for investment returns could lead to more risk-
seeking choices.
Feedback
Definition: The provision of information about a user's
performance on a task in (near) real-time, including both
continuous progress feedback and immediate success
feedback.
58
We exclude feedback from other users in this
category, as it is covered under Social interactions.
General use and impact on behaviour: Feedback has
been most commonly applied in education contexts, where
the frequency, intensity, and immediacy of feedback is
found to be helpful for learner engagement and learning
effectiveness.
59
A recent meta-analysis indicates a medium
effect of 0.48 standard mean difference (SMD) of feedback
on student learning outcomes.
60
While less common in
non-education domains, Uber provides continuous
progress feedback to its drivers (e.g., trips taken, money
earned, rating) as part of a broader package of gamification
techniques.
61
Data on the impact of feedback to Uber
drivers is not publicly available.
Feedback has been used to facilitate more responsible
gambling on online platforms. Players on the online
gambling platform Norsk Tipping were randomly selected
to receive personalized feedback, receiving details on their
losses over the last month. Compared to a control group,
the players who received the feedback reduced their
theoretical loss, the amount of money they would be
56
Odean, T. (1999). Do Investors Trade Too Much? American Economic Review, 89 (5): 1279-1298.
57
Lewis, L. K., Rowlands, A. V., Gardiner, P. A., Standage, M., English, C., & Olds, T. (2016). Small steps: preliminary
effectiveness and feasibility of an incremental goal-setting intervention to reduce sitting time in older adults. Maturitas, 85, 64-
70.
58
Johnson, D., Deterding, S., Kuhn, K. A., Staneva, A., Stoyanov, S., & Hides, L. (2016). Gamification for health and wellbeing:
A systematic review of the literature. Internet Interventions, 6, 89-106.
59
Nah, F. F. H., Zeng, Q., Telaprolu, V. R., Ayyappa, A. P., & Eschenbrenner, B. (2014, June). Gamification of education: a
review of literature. In International conference on hci in business (pp. 401-409). Springer, Cham.
60
Wisniewski, B., Zierer, K., & Hattie, J. (2020). The power of feedback revisited: A meta-analysis of educational feedback
research. Frontiers in Psychology, 10, 3087.
61
Scheiber, N. (2017, Apr 2). How Uber Uses Psychological Tricks to Push Its Drivers' Buttons. New York Times. Retrieved
from: https://www.nytimes.com/interactive/2017/04/02/technology/uber-drivers-psychological-tricks.html.
Figure 8: A modified version of a
US-based app's tool for
diversification analysis gives
users a diversification score
based on investments held in
their portfolio
25
expected to lose based on the amount wagered and the house advantage on a given bet
type, and amount of money wagered.
62
The Canadian company Sun Life previously offered Money UP, an online platform that aims
to educate consumers on retirement and investing planning by challenging users to pass
levels by demonstrating financial knowledge while receiving quick feedback about their
answers to questions on quizzes about these topics.
Feedback is effective because it serves as a self-regulation strategy, revealing progress in
relation to goals,
63
and because it motivates changes in behaviour by focusing one’s
attention on the task itself.
64
It can be effective in influencing any behaviour that is the subject
of that feedback, although in most cases feedback is used to help people improve at a task.
Current use on digital trading platforms and impact on behaviour: A US-based personal
finance app has a tool for diversification analysis that gives users a diversification score
based on their investments held in their portfolio, along with recommendations for
“handpicked” investments that can help the user diversify. The app’s evaluation of this
feature found that those who engage with this tool have portfolios that are two times more
diversified on average than those who do not.
65
A Canadian online trading platform has also recently launched a tool that examines
investors’ portfolio, which measures users’ portfolio holdings across four key indicators,
including asset allocation, diversification, security ratings, and risk, and provides a report
highlighting strengths as well as aspects the user may want to reconsider.
An US-based robo-advisor app uses timely feedback to provide users with the estimated tax
impact of a withdrawal (or allocation change) “just in time,” before the user commits to the
transaction. The company behind the app indicates that the feature is effective in reducing
allocation changes—noting that users shown an anticipated tax of $5 or more were 62% less
likely to complete an allocation change compared to those who were not.
66
This illustrates
the power of timely feedback in influencing investor behaviour.
Researchers have also shown how a feedback intervention can help investors mitigate
cognitive bias in trading decisions. In one study, investors played investing “games” in a
simulated environment hosted on an external learning platform. They received feedback on
their emotional regulation and the cognitive biases they exhibited. This training had a
significant, positive impact on reducing the observed disposition effect when these investors
later made trades in the real world.
67
The disposition effect refers to the general tendency of
investors to sell securities that have increased in value and hold on to securities that have
gone down in value. Stakeholders within the field have varying views of simulated trading
62
Auer, M. M., & Griffiths, M. D. (2016). Personalized behavioral feedback for online gamblers: A real world empirical study.
Frontiers in Psychology, 1875.
63
Carver, C. S., & Scheier, M. F. (2012). Attention and self-regulation: A control-theory approach to human behavior. Springer
Science & Business Media.
64
Kluger, A. N., & DeNisi, A. (1996). The effects of feedback interventions on performance: A historical review, a meta-analysis,
and a preliminary feedback intervention theory. Psychological Bulletin, 119(2), 254.
65
Footnote source deleted re: a US-based platform
66
Footnote source deleted re: an US-based robo-advisor app
67
Fenton-O'Creevy, M., Adam, M., Clough, G., Conole, G., Gaved, M., Lins, J. T., et al. (2015). A game based approach to
improve traders' decision-making. In: The International Gamification for Business Conference 2015: Strategic Industrial
Applications of Games and Gamification.
26
educational games; some feel they provide a good introduction to trading, while others fear
they may increase short-termism.
68
Potential use on digital trading platforms and impact on behaviour: Trading platforms
could use feedback in other ways to support investor education and decision-making. Where
platforms help users understand how trades work, how to analyze market data and risk, how
to understand their fees and performance, and other investment-related behaviours,
feedback is likely to support learning engagement and effectiveness. They could even
provide users with a simulated trading environment and provide feedback to help users learn
about common trading mistakes (e.g., frequent trading, under-diversification, myopia,
disposition effect, etc.)—an approach tested by researchers.
69
As noted above, the value of
such approaches is disputed and should be rigorously evaluated before full-scale
implementation.
Beyond investor education, platforms could implement feedback on investors’ real-world
trading behaviour. For example, they could offer continuous progress feedback through more
frequent reports on how investors are progressing against their savings goals. However,
more frequent feedback might cause investors to focus too strongly on the short-term. A
large-scale field experiment has shown that individuals who receive information about their
investments’ performance too frequently tend to underinvest in riskier assets, losing out on
potential gains in the long-term.
70
Similarly, immediate success feedback on trades, like
highlighting when investors exit a position at a profit, would likely reinforce investors’
disposition effect, reducing future returns. Where the performance / feedback has been
positive, it could also increase traders’ (over)confidence, which, as noted above, negatively
impacts investor performance. Investing is complex, and the impact of different types of
feedback will vary depending on the context and the investor.
Overall, we believe there is sufficient evidence to suggest that feedback interventions on, for
example, diversification and common investor biases can help investors make better
decisions for themselves. However, feedback on short-term performance or individual trades,
for example, could encourage investors to take actions that undermine their investment
goals.
Other Behavioural Techniques
The following sections describe and assess behavioural techniques that do not meet the
definition of gamification. However, these techniques are informed by behavioural science
and are being used by digital trading platforms in ways that influence investor behaviour.
71
68
FINRA (2021, Jun 30). "FINRA Requests Comment on Effective Methods to Educate Newer Investors". Retrieved from:
https://www.finra.org/rules-guidance/notices/special-notice-063021#comments.
69
Fenton-O'Creevy, M., Adam, M., Clough, G., Conole, G., Gaved, M., Lins, J. T., et al. (2015). A game based approach to
improve traders' decision-making. In: The International Gamification for Business Conference 2015: Strategic Industrial
Applications of Games and Gamification.
70
Larson, F., List, J. A., & Metcalfe, R. D. (2016). Can myopic loss aversion explain the equity premium puzzle? Evidence from
a natural field experiment with professional traders (No. w22605). National Bureau of Economic Research.
71
U.S. Securities and Exchange Commission (n.d.) "Request for Information and Comments on Broker-Dealer and Investment
Adviser Digital Engagement Practices, Related Tools and Methods, and Regulatory Considerations and Potential
Approaches...". Retrieved from:https://business.cch.com/srd/34-92766.pdf.
27
Salience / attention-inducing prompts
Definition: Information is more likely to influence people’s behaviour if it attracts their
attention. A very wide range of design features are included in this category, including visual
cues as well as the specific language used in prompts. Often, the most attention-inducing
language leverages other behavioural insights like social norms, loss aversion, and invoking
scarcity. In this section, we focus specifically on the way platforms seek to direct user
attention to some features and behaviours—and away from others.
General use and impact on behaviour: Strategies to increase the salience of certain
information (or decrease the salience of other information) are ubiquitous across industries
and used to influence an extraordinarily wide range of behaviours. Retailers, financial
services companies, and public sector bodies alike use design choices like colour, images,
personalization, the size, and placement of different information, and even humour to attract
people’s attention to specific information and options. These tactics can have positive effects,
such as raising awareness of privacy notices on apps and websites.
72
They can also harm
users by distracting them from other information, like additional costs,
73
and the proverbial
“fine print.” Push notifications are frequently used by app designers to boost engagement
with the application, and prompts (e.g., pop-ups) are
used to direct user attention to certain activities within the
apps.
Current use on digital trading platforms and impact
on behaviour: All trading apps make deliberate choices
about what information is most salient to the user
experience from signing up, to logging on, and to
executing a trade. The way a choice is presented often
influences what choice is made.
74
Given there is no truly
“neutral” way to present options, the way information is
displayed is necessarily influencing users’ attention and
action. Some design choices can be helpful, enabling
users to find the most relevant features and understand
key information more easily. In this section, we focus on
more potentially concerning uses of attention-inducing
prompts.
Many popular digital trading platforms prominently
feature lists of stocks on their home screens, including
“Top Movers,” and “Most Popular” stocks. The salient
placement of these lists has a dramatic impact. New
entries into Platform 7’s “Most Popular” list are five to
seven times more likely to be purchased in the days
following their listing.
75
The inclusion of a stock on
Platform 7’s “Top Mover” list is associated with it being
72
Ebert, N., Alexander Ackermann, K., & Scheppler, B. (2021, May). Bolder is Better: Raising User Awareness through Salient
and Concise Privacy Notices. In Proceedings of the 2021 CHI Conference on Human Factors in Computing Systems (pp. 1-12).
73
Cara, C. (2019). Dark patterns in the media: A systematic review. Network Intelligence Studies, 7(14), 105-113.
74
Johnson, E. J., Shu, S. B., Dellaert, B. G., Fox, C., Goldstein, D. G., Häubl, G., & Weber, E. U. (2012). Beyond nudges: Tools
of a choice architecture. Marketing Letters, 23(2), 487-504.
75
Footnote source deleted re: Platform 7
Figure 9: A modified version of
Platform 7, which prominently
features different lists on their
home screen, including “100
Most Popular”
28
traded 36 times more than the amount that it is traded on average, even when controlling for
the overall market activity.
76
These attention-inducing lists are significantly associated with
increased “herding”
77
, where retail investor choices are positively correlated with each other
(they are making the same trades). Herding appears to result in significantly poorer returns
for investors, with one study showing average 20-day abnormal returns of -4.7% for top
stocks purchased each day.
78
In contrast, research conducted using transaction-level data
from two German retail banks indicated that the prominent placement of “top movers” does
not appear to affect returns.
79
The contrasting results could be a result of different platform
interface designs, but it is also worth noting that the investors in this study were, on average,
45 years old with nine years of investing experience, which may suggest that these types of
attention-inducing lists may have a more profound effect on younger, less experienced
investors, who are the primary market for certain online trading platforms.
How information is displayed for individual stocks may also influence investor behaviour.
Research examining investor-level brokerage data from China has found that increasing the
salience of a stock’s purchase price while keeping other information unchanged on the online
trading screen increased investors’ disposition effect
80
by 17%.
81
Many popular trading apps also use push notifications to send out information to investors.
For example, they allow users to create alerts for when stocks they’re interested in hit a
certain price or percentage increase or decrease of their choice. These types of attentional
triggers can be useful for investors who need to monitor price movements as part of their
investment strategy. Similar to other salience-inducing prompts like top traded lists, they can
also influence investor behaviours in ways that are not aligned to the investor’s strategy. The
“lottery anomaly” refers to investors’ preference for stocks with large potential gains, even
when risks are proportionately higher. This anomaly is heightened when such stocks receive
more investor attention, which might be the result of more analyst coverage or extremely
positive recent returns.
82
Notifications are likely to play a similar role in heightening investor
attention.
More broadly, push notifications on stock performance (e.g., “$ACME shares down over -
5.2%”) sent to investors' cell phones have been found to increase investors’ risk-taking. One
study showed that this type of attention-induced trade carried, on average, a 19-percentage-
point higher leverage than other trades.
83
This impact was stronger for male, younger, and
less experienced investors.
Potential use on digital trading platforms and impact on behaviour: Given the current
widespread use of attention-inducing prompts on trading apps, it is likely that more platforms
76
Footnote source deleted re: Platform 7
77
Footnote source deleted re: Platform 7
78
Footnote source deleted re: Platform 7
79
Kalda, A., Loos, B., Previtero, A., & Hackethal, A. (2021). Smart (Phone) Investing? A within Investor-time Analysis of New
Technologies and Trading Behavior (No. w28363). National Bureau of Economic Research. NB: this study did find that
smartphone-trading did have a negative impact overall.
80
The disposition effect refers to the well-established but “irrational” tendency for investors to sell securities that are worth more
than their purchase price and hold on to those that are worth less.
81
Frydman, C., & Wang, B. (2020). The impact of salience on investor behavior: Evidence from a natural experiment. Journal of
Finance, 75(1), 229-
276.
82
Bali, T. G., Hirshleifer, D., Peng, L., & Tang, Y. (2021). Attention, social interaction, and investor attraction to lottery stocks
(No. w29543). National Bureau of Economic Research.
83
Arnold, M., Pelster, M., & Subrahmanyam, M. G. (2022). Attention triggers and investors’ risk-taking. Journal of Financial
Economics, 143(2), 846-875.
29
will continue to employ similar tactics. Prominent lists of trading options, like “Top Mover” and
“Most Popular” lists may continue to cause “herding” and may also increase trading
frequency. The widespread use of notifications and in-app prompts is also likely to expand,
with many such messages leveraging other behavioural science techniques (e.g., invoking
scarcity, loss aversion, etc.). Depending on the focus of these messages, they could
increase trading frequency or shift traders into higher risk securities (e.g., options).
Attention-inducing prompts could also be deployed to support users in ways that are likely to
support their long-term goals. For example, prompts could encourage users to engage in
investor education, to increase deposits, or to explore options for diversifying their portfolios.
Simplification and selective deployment of friction costs
Definition: The design of the user experience that reduces or introduces small barriers
across the user journey, influencing the likelihood and manner in which a user completes a
specific task. We use “simplification” to refer to reductions in small barriers and “friction
costs” to refer to increases in small barriers.
General use and impact on behaviour: The small details that make a task appear more
challenging or effortful have a large impact on the likelihood of a person completing that task.
In fact, the large effect they have on whether someone completes the task is disproportionate
to the size of the detail. For example, simplifying the process of applying for college financial
support by pre-populating the application with known information resulted in students being
more likely to attend university by 8 percentage points.
84
Even something as minor as
reducing the number of ‘clicks’ it takes to access a tax form by linking the form directly
instead of linking to a website that includes the form has been shown to increase tax
collection rates by 4%.
85
Amazon has been a trailblazer in leveraging simplification in online
retail, implementing a “1-click” buying option to encourage impulse buying. In terms of mobile
apps, the perceived ease of use of an app significantly impacts one’s intention to use the app
among young adults.
86
Ease of use may also contribute to the perception of enjoyment of
gamified e-banking apps.
87
However, while simplifying user experiences and reducing points of friction generally have
positive effects, it can also make undesirable behaviours too easy to follow through on. In
these instances, adding friction can help people slow down and think more deeply about their
actions rather than relying on their gut instincts. For example, in a large online survey
experiment, creating a moment for reflection by asking participants to explain how they knew
that a political headline was true or false decreased their intention to share false news
headlines.
88
84
Bettinger, E. P., Long, B. T., Oreopoulos, P., & Sanbonmatsu, L. (2012). The role of application assistance and information in
college decisions: Results from the H&R Block FAFSA experiment. The Quarterly Journal of Economics, 127(3), 1205-1242.
85
The Behavioral Insights Team. (2014). EAST: Four simple ways to apply behavioural insights.https://www.bi.team/wp-
content/uploads/2015/07/BIT-Publication-EAST_FA_WEB.pdf.
86
Mehra, A., Paul, J., & Kaurav, R. P. S. (2021). Determinants of mobile apps adoption among young adults: theoretical
extension and analysis. Journal of Marketing Communications, 27(5), 481-509.
87
Rodrigues, L. F., Oliveira, A., & Costa, C. J. (2016). Does ease-of-use contribute to the perception of enjoyment? A case of
gamification in e-banking. Computers in Human Behavior, 61, 114-126.
88
Fazio, L. (2020). Pausing to consider why a headline is true or false can help reduce the sharing of false news. Harvard
Kennedy School Misinformation Review, 1(2).
30
Current use on digital trading platforms and impact on behaviour: Many trading apps,
including Platform 7, 8, 10, and others boast quite seamless user experiences, making it as
easy as possible to register for an account and begin placing trades.
1. They make it easy to sign up for a trading account by keeping the sign-up process
short and guiding the user through the process. This is likely to increase market
participation and save investors time.
2. The platforms also make it easy to deposit and access funds. For example, Platform
10 prompts users to set up automatic deposits into their account when adding funds.
They also offer ‘premium features’ for a small monthly fee ($3), allowing users to “add
up to $1,000 in seconds with no holds”, enabling more immediate trading. While there
are certainly positive elements to this, it may also increase “hot state” (non-
deliberative) trading by making it easier to make spur of the moment trades.
3. The platforms tend to present users with less data for researching investment
options. Besides basic market information, Platform 7 only provides investors with
five charting indicators,
89
while another US-based trading platform provides 489.
90
It is
unclear what impact, if any, this form of simplification has. On one hand, it may
reduce choice overload and increase investor focus on the most relevant information.
On the other hand, it may reduce the time investors spend researching and
deliberating or make pertinent information harder to find and therefore less likely to be
used in trading decisions. Our literature scan did not identify any empirical evidence
on the average impact of reducing the number of indicators. From a theoretical
perspective, it is important to balance the need for providing investors with the data
required to make informed investing decisions (e.g., deciding between similar
investment funds, for example) and avoiding information overload.
4. These platforms also appear to require fewer “clicks” to execute trades, although we
did not find unequivocal data to confirm this. While a streamlined process can save
users time, it may also provide fewer opportunities to rethink their trades and increase
speculative trading, leading to lower investment returns.
91
This is particularly relevant
when considering the full package of social interactions, attention-inducing prompts,
and other behavioural science-informed tactics present on these platforms. For
example, a social interaction may trigger an inclination to make a trade that is not
aligned with the investor’s overall strategy, and then a seamless, quick trading
interface would increase their propensity to follow through.
5. Platform 7 has received scrutiny for making it easier to trade more complex, high-risk
securities like options by simply checking a box during the account sign-up, without
providing explanation of these types of assets or warnings of the increased risk and
complexity.
92
While we do not have evidence on the impact of this, it could be
problematic in encouraging investors to make trades that do not align with their
investment strategy and risk tolerances due to a lack of understanding. Platform 7
89
Charting Indicators include: Volume, Moving Average (MA), Exponential Moving Average (EMA), Relative Strength Index
(RSI), Moving Average Convergence Divergence (MACD).
90
Footnote source deleted re: Platform 7
91
Footnote source deleted re: Platform 7
92
Footnote source deleted re: Platform 7
31
has now tightened eligibility for users to engage in options trading, requiring users to
have a certain level of experience. Our environmental scan did not determine the
approach that other platforms are taking to higher risk activities like margin or options
trading, as we did not attempt to execute any trades on the accounts we set up.
Potential use on digital trading platforms and impact on behaviour: It is likely that digital
trading platforms will continue to search for opportunities to reduce frictions across the user
experience. In some areas, it is likely that these platforms have already streamlined the
experience as much as possible within the limits set by Canadian regulation that govern
Order Execution Only activities.
93
Members of the U.S. Securities and Exchange Commission (SEC) Investor Advisory
Committee have advocated for the inclusion of frictions into the trading process to promote a
deliberative mindset.
94
Suggested frictions include building in delays, creating decision
points, and using nudges that encourage users to slow down and reflect on the longer-term
outcomes of their investment decisions by envisioning the effect of the decision in the future.
Balancing the ease of use and friction, as well as selection of where and when friction is
required, can help protect investors.
Social interactions
Definition: Design elements that enable platform users to interact with other users by i)
generating, sharing, viewing, and reacting to content, and ii) engaging in direct messaging.
Note that we exclude referrals from this section as they represent a very low level of social
engagement and are not the focus of the social interaction literature. We also exclude social-
media style “posts” or “feeds” that come from individuals who self-identify as financial /
investing experts.
General use and impact on behaviour: Techniques designed to facilitate social exchange
and feedback—such as ‘likes’ on Facebook, Instagram, and Twitter—are central elements of
the social media experience. They are also increasingly used by other applications, such as
fitness apps, to stimulate greater levels of engagement. For example, Peloton has introduced
new social features like hashtags to enable users to connect with peers with similar goals
and interests.
Brain scan studies indicate that virtual social feedback might contribute to the addictive
nature of social media platforms, as it engages brain regions that are at the core of our ability
to process rewards, acquire habits, and become addicted.
95,96
While such an internal reward
mechanism plausibly evolved to reinforce (and ensure repeats) of beneficial social
interactions, there is a growing recognition that it can generate a problematic level of
93
Investment Industry Regulatory Organization of Canada – IIROC (2019, Apr 9). "Guidance on Order Execution Only Services
and Activities". Retrieved from: https://www.iiroc.ca/news-and-publications/notices-and-guidance/guidance-order-execution-
only-services-and-activities.
94
Powell, L. (2021, September 14). SEC Investor Advisory Committee examines digital engagement practices [Blog post].
Retrieved from https://jimhamiltonblog.blogspot.com/2021/09/sec-investor-advisory-committee_14.html.
95
Montag, C., Markowetz, A., Blaszkiewicz, K., Andone, I., Lachmann, B., Sariyska, R., ... & Markett, S. (2017). Facebook
usage on smartphones and gray matter volume of the nucleus accumbens. Behavioural brain research, 329, 221-228.
96
Sherman, L. E., Hernandez, L. M., Greenfield, P. M., & Dapretto, M. (2018). What the brain ‘Likes’: neural correlates of
providing feedback on social media. Social cognitive and affective neuroscience, 13(7), 699-707.
32
attachment to the social media experience. To our knowledge, effect sizes associated with
social features have not been reported.
97
Current use on digital trading platforms and impact on behaviour: Social interactions
are present across a range of online digital trading
platforms. Users of a US-based trading platform have
the option to publicly display their portfolios and share
their investment decisions with other users. Platform 11
features a similar mix of social interaction features:
users have access to a social news feed consisting of
content posted by their peers, the opportunity to chime
into conversations highlighted under ‘Most popular
topics’ hashtags, and the chance to follow and copy
others’ investment decisions. These functions are also
replicated on Platform 2. Platform 8 also had the option
to enable ‘Social Investing’, where users can discover
and follow other members’ investment holdings,
watchlists, and investment activity, and have the ability
to comment on and react via emoji to other members’
activity. This feature has been disabled, although it is
unclear why.
98
At least one platform currently in
development will go even further, fully “enabling
dialogue and community engagement between users.”
99
Social interaction features are leveraged by Platform 5,
which has integrated a Twitter-like social media platform
for the financial community, into their mobile platform,
enabling users to follow investing ideas and discussions
from other investors.
Social interactivity has a demonstrated impact on
engagement with digital applications. This in itself may be problematic, as investors who
check the status of their investments more often trade more and have worse performance.
100
Sociality might also contribute to problematic levels of engagement in other ways. If users’
trades can be “liked”, recommended, or otherwise promoted, their desire for social
recognition and status could lead to increased trading frequency in the same way people
post more on Facebook to get social affirmation. Social interactions are also likely to
increase the disposition effect, with users holding onto losing investments for longer periods
of time or not sharing information about losing trades to “save face”.
101
The increase in
97
Hou, Y., Xiong, D., Jiang, T., Song, L., & Wang, Q. (2019). Social media addiction: Its impact, mediation, and intervention.
Cyberpsychology: Journal of psychosocial research on cyberspace, 13(1).
98
Footnote source deleted re: Platform 8
99
Cutts, J. (2021, Sep 14). “Not all brokerages should 'gamify' a la Robinhood - but others can/will go further.” Traders
Magazine
100
Barber, B.M. and Odean, T. (2000), Trading Is Hazardous to Your Wealth: The Common Stock Investment Performance of
Individual Investors. The Journal of Finance, 55: 773-806.
101
Chaudhry, S., & Kulkarni, C. (2021, June). Design Patterns of Investing Apps and Their Effects on Investing Behaviors. In
Designing Interactive Systems Conference 2021 (pp. 777-788).
Figure 10: A modified version
of Platform 2, which hosts a
social news feed consisting of
content posted by users
33
disposition effect is more likely to occur in stock trades than (actively managed) fund trades,
as traders can assign blame for poor performance to fund managers in the latter case.
102
In terms of the quality of investment, the evidence is equivocal about whether making
decisions under social influence is inherently bad for financial outcomes. One recent analysis
of over 28 million trades on an unspecified online trading platform with social features
revealed several insights that are useful for a broad risk assessment. First, investment
decisions made as a form of social mirroring represent the majority (67.6%) of decisions
compared to those that could be classified as independent. This implies that, when digital
trading platforms make social influence features available, users’ behaviours are likely
significantly affected. Secondly, the study revealed a complex picture when it comes to
assessing whether social trades make for poor decisions. Social mirroring, where a user
copies the decisions of one or more selected traders, produced more “wins” (a trade with a
net positive return) compared to independently chosen trades (84% vs 59%). However, these
wins resulting from mirroring had statistically significantly lower ROIs (0.177% vs 0.183%),
while ROIs on losses were also significantly more negative (-0.9% vs -0.38%).
103
Thus, while
social mirroring may produce more “wins,” it may also lead to lower levels of positive ROIs
and higher levels of negative ROIs. Despite the evidence that social influence can have
benign effects on investor outcomes, it is difficult to ignore prior events which point to a
dramatic impact of mass social mirroring phenomena. This is clearly demonstrated in the
2021 “meme stock” phenomenon driven by Reddit’s “wallstreetbets” forum.
Social features might interact with other gamification elements in a way that increases user
risk. For instance, while it’s unclear whether peer-to-peer sharing of investment decisions
(which is already available on apps such as Platform 11) produces negative investment
outcomes, there would likely be significant cause for concern where sharing functionalities
are coupled with leaderboards (see Leaderboards) or badges (see Non-economic rewards).
These two techniques have the potential to nudge users towards making decisions in a
competitive mode rather than based on their individual investing goals and context, which
could be heightened by embedding social interaction features.
Investors already have many options for social interactions related to investing, including
prominent message boards like Reddit. However, embedding social interactions within the
digital trading platform makes it much easier to act on the basis of that interaction without
pause for consideration.
Potential use on digital trading platforms and impact on behaviour: We believe that
most potential social interaction features are discussed in the previous sub-section, although
we certainly cannot rule out further developments. For example, many digital platforms
enable their users to engage in social interactions using avatars–digital personas that serve
as idealized visual representations of the users. While avatars are considered a classic
gamification technique and are used across a range of contexts (e.g., self-help health
102
Chang, T. Y., Solomon, D. H., & Westerfield, M. M. (2016). Looking for someone to blame: Delegation, cognitive dissonance,
and the disposition effect. The Journal of Finance, 71(1), 267-302.
103
Liu, Y. Y., Nacher, J. C., Ochiai, T., Martino, M., & Altshuler, Y. (2014). Prospect theory for online financial trading. PloS one,
9(10), e109458.
34
apps
104,105
, virtual meeting platforms
106
), we did not identify any instances of them being used
by investing platforms. Avatars could be used to promote and encourage the forms of social
interaction described above. Platforms could also create team-based competitions, as such
competitions create more engagement and referrals.
107
Social norms
Definition: Design features which signal social norms—i.e.,
information about how others think and behave. This might
involve explicitly informing individuals about statistics
pertaining to relevant group behaviours (e.g., “88% of your
fellow users have invested in green energy this year”) or
implicitly signalling “crowd approval” through features such
as “This week’s 100 most popular stocks” and “Today’s
trending industries.”
General use and impact on behaviour: As social animals,
we are influenced by norms—that is, how we perceive others
to think and behave (or how we perceive their expectations
of our own thoughts and actions). When a particular norm is
not known to us, or when we have little knowledge of the
best course of action in a specific situation, being informed of
what others are doing either implicitly or explicitly can
influence our behaviour. The impact of learning about a norm
is higher when that norm is surprising and the group whose
behaviour and/or preferences we are being informed of feels
relevant to us.
Social norm interventions are a common feature of behaviour
change programs in a variety of domains. Communications
about descriptive social norms, which inform people about
the behaviours of others, have been found to reduce the use
of plastic bags
108
and increase energy conservation
109
,
recycling
110
, and timely tax filing
111
. Retailers frequently
invoke social norms to increase product sales (e.g., “Over X million satisfied customers!”).
While less common, researchers and some organizations are starting to test the impact of
dynamic social norms, communicating norms in terms of change rather than absolute levels
104
Tuah, N. M., Yoag, A., Ahmedy, F., & Baharum, A. (2019). A gamification and avatar self-representation application for
diabetes self-management. Int. J. Adv. Trends Comput. Sci. Eng, 8, 401-407.
105
Hswen, Y., Murti, V., Vormawor, A. A., Bhattacharjee, R., & Naslund, J. A. (2013). Virtual avatars, gaming, and social media:
Designing a mobile health app to help children choose healthier food options. Journal of mobile technology in medicine, 2(2), 8.
106
Kesel, W. (2020, Apr 1). 3 Avatar-Based Virtual Event Platforms for Planners to Consider. BizBash.com.
107
Morschheuser, B., & Hamari, J. (2019). The gamification of work: Lessons from crowdsourcing. Journal of Management
Inquiry, 28(2), 145-148.
108
De Groot, J. I., Abrahamse, W., & Jones, K. (2013). Persuasive normative messages: The influence of injunctive and
personal norms on using free plastic bags. Sustainability, 5(5), 1829-1844.
109
Allcott, H., & Rogers, T. (2014). The short-run and long-run effects of behavioral interventions: Experimental evidence from
energy conservation. American Economic Review, 104(10), 3003-37.
110
Ceschi, A., Sartori, R., Dickert, S., Scalco, A., Tur, E. M., Tommasi, F., & Delfini, K. (2021). Testing a norm-based policy for
waste management: An agent-based modeling simulation on nudging recycling behavior. Journal of Environmental
Management, 294, 112938.
111
Behavioral Insights Team (2014). EAST: Four simple ways to apply behavioural insights. Retrieved from:
https://www.bi.team/wp-content/uploads/2015/07/BIT-Publication-EAST_FA_WEB.pdf.
Figure 11: A modified version
of Platform 10, which has
several lists showing which
stocks other users are
trading and holding in high
volumes
35
of behaviour (e.g., “78% more Canadians are doing X each year” versus “78% of Canadians
are doing X”).
112
Current use on digital trading platforms and impact on behaviour: Social norms are
commonly leveraged on digital trading platforms. Platform 7 displays a daily ‘Top 100 stocks’
list on its homepage, which sends an implicit signal about the stocks that are receiving most
attention, and thus represent a “normal” position. Platform 11 presents users with a list of
‘Most popular topics’, capturing the most-discussed industries and investments of the day.
Similar tactics are used by platforms with a presence in the Canadian market. Platform 10
features dynamic social norms through elements such as ‘Top 100 on Platform’ and a ‘Top
100 Canadian/US Stocks’ list based on the most actively traded stocks on the Canadian/US
exchanges on any given day. Such lists tend to have a small sub-heading briefly explaining
the list to amplify the lists’ appeal (e.g., “The most popular stocks on Platform 10”). These
features have concrete implications for how retail investors make decisions. As noted above,
Platform 7’s daily ‘Top 100 stocks’ list promotes “herding,” whereby disproportionately large
numbers of retail investors buy or sell particular stocks at the same time.
113
In turn, herding
appears to decrease investment returns, on average. It is unclear how much of this effect is
driven by the salience of the information compared to the fact that it represents a social norm
that might be loosely summarized as: ‘if everyone else is doing / saying it, it must be right’.
114
We imagine that both factors contribute to the effects observed.
Potential use on digital trading platforms and impact on behaviour: We anticipate that
the types of lists of commonly traded securities noted above will remain the most common
form of social norm intervention deployed on digital trading platforms. More specific and
personalized social norm prompts could certainly be pushed to users (e.g., “80% of investors
like you are buying ACME”), although such prompts may not be permitted on self-directed
digital trading platforms under IIROC rules. Were they implemented, we believe that these
explicit prompts would be just as, if not more, impactful than the lists of top traded stocks.
Summary
The preceding taxonomy outlined five gamification techniques and four other behavioural
techniques relevant to retail investing on digital trading platforms. For each technique, we
provided a definition, outlined how the technique is used in general, how it is used on digital
trading platforms today, and how it might be deployed on those platforms in the future. We
also explored the positive and negative impacts these techniques may have on various retail
investor behaviours. It is challenging to assess the relative impact that each of these
techniques may have on investor outcomes given the novelty of these approaches and
corresponding gaps in empirical research. However, based on current evidence and theory,
we believe that social interactions, rewards, and gamblification techniques are likely to have
the largest impact on behaviour.
112
Sparkman, G., & Walton, G. M. (2017). Dynamic norms promote sustainable behavior, even if it is counternormative.
Psychological science, 28(11), 1663-1674.
113
Footnote source deleted re: Platform 7
114
Sherif, M. (1935). A study of some social factors in perception. Archives of Psychology, 27: 187.
36
Experimental Research
Experimental Research Methodology
We conducted a randomized controlled trial (RCT) to assess the impact of two behavioural
techniques on investing behaviours: (1) giving investors points with a very low value for
buying or selling stocks (rewards) and (2) showing investors a “top traded list” (social norms
and attention-inducing prompts). There were no fees for trading in any of the conditions. Our
primary interest was in whether these techniques increased trading frequency relative to a
control group not exposed to points nor to a top traded list. We also explored outcomes
related to diversification of holdings, disposition effects, and, for the top traded list, stock
selection.
RCTs examine the causal relationship between a condition and outcomes by eliminating the
effect of potentially correlated external factors. What sets RCTs apart from other methods is
that participants are randomly assigned to experiment groups, as defined below. With a
sufficiently large sample, random assignment ensures that—on average—individuals across
groups only differ in terms of their group assignment. By eliminating differences across
groups, we ensure that any difference in outcomes can be attributed to group conditions, as
opposed to other confounding factors.
Condition Description
Control
(n=803)
Participants in the Control condition traded six fictitiously-named stocks (based on
real stocks) over seven simulated weeks of investing on a platform that did not
include any behavioural techniques.
Points
(n=811)
Participants in the points condition traded simulated stocks on a platform that
included point rewards, which were prominently displayed.
Participants received 100 points each time they bought or sold a stock, and every
1,200 points were worth $0.01 in additional compensation. As a result, the
maximum value of these points was only about $0.08 if they traded every stock
every week, a negligible but non-zero economic value.
Top Traded
List
(n=816)
Participants in the top traded list condition traded stocks on a platform that
included a top traded list showing three stocks labelled as most actively traded on
the investing simulation platform each week. The top traded list rotated each week
based on the historical data on trading volumes for selected stocks, as described
further below.
The experiment was conducted online in a simulated trading environment using Predictiv,
BIT’s proprietary platform for online experimentation. Our sample included 2,430 Canadian
residents from all provinces aged 18-65 years engaging with the experiment on mobile
(55%), tablet (4%), or desktop (41%) devices. Over 50% of our sample were investors.
115
Our sample was well balanced on gender (50% female) and age (median age: 35-44).
115
Investors were defined as such by holding at least one of: individually held stocks, ETFs, securities or derivatives, bonds or
notes other than Canada Savings Bonds, mutual funds, or private equity investments.
37
Additional demographic details are available in Appendix B (Detailed Experimental Research
Findings).
Research participants received $10,000 in simulated “money” to invest in up to six different
fictitiously-named stocks. After their initial allocation of funds, they were taken through seven
simulated weeks of stock price movements, with an option to buy and/or sell stocks between
each week. At the end of the experiment all participants received a fixed amount of
approximately $1.00 for participating in the experiment. They also earned up to $1.70
additional compensation based on their balance at the end of the experiment.
116
Participants
were aware that the larger the portfolio at the end of the experiment, the more they would
earn. This created an incentive for participants to trade thoughtfully and to try to maximize
their returns.
The following graphic provides a more detailed overview of the steps participants went
through to buy and sell stocks on the simulated stock market. All participants went through
the same steps regardless of which trial condition they were randomly assigned to.
Figure 12: Overview of research experiment
116
Participants in the points condition also earned compensation based on the number of trades they made, with each trade
earning 100 points and 1,200 points being worth $0.01 in additional compensation. As a result, the maximum value of these
points was only $0.075, a negligible but non-zero economic value. This reflects how points are likely to be used by trading
platforms based on our environmental scan.
38
Price movements for the fictitiously-named stocks used in our simulation were based on real
stocks. To do so, we randomly selected stocks that met the following criteria: (1) listed on the
Nasdaq exchange website, (2) within the technology sector, (3) with a micro ($50M-$300M)
or nano (<$50M) market capitalization, and (4) an opening price below $4 per share. For the
selected stocks, we used historical opening price data for eight consecutive Mondays in 2021
to determine the opening prices for each of the weeks of the experiment. The purpose was to
create stocks with a similar risk profile and a medium-high level of volatility. Participants were
informed at the onset of the simulation that all stock price movements were based on real,
historical data.
Our experiment was designed to maximize its generalizability to real-world trading. Beyond
the use of variable compensation and modelling stocks on randomly selected real equities,
our RCT methodology enabled us to isolate the effects of the behavioural techniques we
examined, controlling for potentially confounding factors. Overall, the trial implementation
was successful:
● Recruitment was effective and the study was well-powered, with more than 800
participants in each condition.
● There were no differences in attrition across groups.
● Participants engaged meaningfully with the trading simulation, with a median
completion time above 12 minutes.
Notwithstanding the quality of the research design and execution, the nature of the study as
a trading simulation is an important limitation. It is likely that some aspect of participant
behaviour would be different in a real-world setting. For example, participants may be less
risk averse in a simulated environment than they would be if their own savings were on the
line. Further, the compression of trading activity into a single session cannot replicate the
effects of time on investor decision-making. Real-world events like the explosion of interest in
GameStop in early 2021 could prompt shifts in how behavioural techniques influence trading
behaviours.
39
Figure 13: Selected experiment screenshots
Control condition
40
Points condition
41
42
Top Traded List condition
43
44
Experimental Research Findings
Primary Results: Trading Frequency
Our primary outcome of interest was trading frequency, given the potential incentive for
digital trading platforms to encourage more trading and the negative impact of higher trading
frequency on investor returns. We measured trading frequency as the number of times
participants either buy or sell a stock over the course of the trading simulation.
As shown on Figure 14 below, participants in the points group made 39% more trades
than the control group, a statistically significant difference.
117
Showing research participants a
top traded list did not increase their trading frequency.
Figure 14: Trading frequency by experiment group
Primary analysis, controlling for age, gender, education, income, objective investing knowledge, and risk
preference.
118
117
Our threshold for significance (p < 0.008) includes a Benjamini-Hochberg correction for multiple comparisons and represents
a more conservative approach to statistical significance compared to the usual threshold of 0.05. Any results with p-values
above 0.008 should be considered suggestive as they do not meet the pre-specified threshold for statistical significance. Error
bars represent a 95% confidence interval.
118
Note on interpreting bar graphs in this report: The height of the control group’s bar represents the average outcome observed
for participants in the control condition. The height of the experimental groups’ bars represent what we think would have
happened to control group participants if they had been treated in the way the experimental groups were. In other words, the
heights of the Points and Top Traded List groups represent the height of the control group average plus our estimate of the
effect of the treatment (based on our analysis, which controls for covariates about participants in all groups).
45
These findings suggest that reward-based gamification tactics can meaningfully influence
investor behaviours by encouraging greater frequency of trading. This can occur even when
rewards have a negligible economic value for investing, as in our simulation. This finding is
important because more frequent trading has been shown to yield poorer investment
outcomes for retail investors, especially among those with less investment knowledge or
experience.
119
Secondary Results: Diversification
Our secondary outcome of interest was participants’ diversification, defined as the extent to
which investors diversify their money across the six different stocks investors could buy and
sell throughout the experiment.
120
Diversification is a critical risk mitigation strategy for
investors.
There were no differences across experiment groups on participants’ diversification.
On average, participants’ diversification scores across groups ranged from 0.69 to 0.71 on a
0-1 index, where 1 represents an equal allocation of funds across all investment options.
These findings suggest that exposure to these two techniques is not associated with
significantly less diverse investment portfolios. We had hypothesized that the top traded list
may reduce diversification by concentrating trading in the listed stocks. While we did not
observe that in our trial, diversification outcomes may be influenced by other factors, such as
the limited number of available stocks and their characteristics as well as the limited number
of 'weeks' of trading, that we could not fully model in our experiment.
Exploratory Results
The experiment provided rich data on a wide range of other investor behaviours and choices.
This section summarizes the additional analysis that we did using this additional data that
went beyond what was required to answer our core research questions.
Impact of top traded list on stock selection
Participants in the top traded list group bought and sold more of the stocks on that list than
participants in the other groups. As shown in Figure 15 below,
121
44% of the total value of the
stocks bought or sold in a given week in the control group (and points group) were from the
top traded list, but that value was 50% in the top traded list group. This is a 14% relative
increase. This effect was somewhat stronger for buying stocks than it was for selling stocks,
and as a result we see participants holding more of the top traded list stocks in the top traded
group than in the control group.
122
119
Odean, T. (1999). Do investors trade too much? American economic review, 89(5), 1279-1298.
120
This was measured using the inverse of an indicator of deviation (in terms of sum of squared errors) from a uniform
distribution of resources across all available investments, which we labelled a ‘Diversification Index’. A value of 1 would mean a
completely equal distribution of funds across the stocks. A value of 0 would mean that only one stock was held. Additional
details on this measure are available in Appendix C.
121
The number of participants analyzed (n) varies for these analyses as they only includes participants who helds stocks across
each week in the experiment.
122
Supplemental analyses focused on stock selling and holding further elucidate these findings. On average, participants who
saw the top traded lists had a 49% likelihood of selling a top traded stock in a given week, an 11% increase relative to
participants in the control group (who had a 44% chance of selling a top traded stock). Overall, 51% of the stocks held by the
group that saw the top traded list were on that list, a 7% relative increase compared to the control group (48%).
46
These findings reinforce analysis conducted by Barber et al. (2020) showing that a salient list
of stocks frequently traded by other investors shifts trading activity toward those stocks.
123
That study also shows that this negatively influences investor returns.
Figure 15: Amount of top traded stocks bought and sold as a percentage of total amount traded.
Exploratory analysis, controlling for age, gender, education, income, objective investing knowledge, and risk
preference.
Final fund values
There is a robust evidence base linking increased trading volume to lower retail investor
returns over time.
124
This can be attributed to the timing of retail investor trades relative to
other (e.g., institutional) traders, as well as transaction fees and other costs related to
trading. In our experiment, there were no transaction fees or other related charges, which
would have decreased final fund values for participants with greater trading frequency. Other
aspects of our experimental design – including the single-session limit on decision time
horizons, number of stocks available, and stock types – are different from real-world trading.
Given these factors, we did not hypothesise that either gamification conditions would impact
participants’ final fund balance at the end of the trading session. Indeed, we find statistically
similar balances at the end of the session across all three groups.
125
123
Barber, B. M., Huang, X., Odean, T., & Schwarz, C. (2020). Attention induced trading and returns: Evidence from Robinhood
users. Journal of Finance (Forthcoming).
124
Barber, B.M. and Odean, T. (2000), Trading Is Hazardous to Your Wealth: The Common Stock Investment Performance of
Individual Investors. The Journal of Finance, 55: 773-806.
125
All participants started the simulation with $10,000 and, on average, ended with fund values ranging from $12,395 to $12,516
across groups.
47
Disposition effect analysis
The disposition effect refers to the tendency of investors to sell assets that have increased in
value while keeping assets that have dropped in value. This tendency can negatively impact
investor returns, and we wanted to measure whether either of the behavioural techniques
had an impact on the disposition effect.
To measure the disposition effect, we created a variable for each stock, at each experiment
screen (i.e., week), indicating if price increased or decreased. We then assessed the effect of
price increase or decrease on participants’ likelihood of purchasing or selling the stock. We
found that, on average, there was not a disposition effect observed in any of the groups, and
there were no differences in disposition effect across groups.
126
The lack of disposition effect,
which is present in most investing contexts, was driven by investors in our experiment
holding onto winning investments at a high rate. We hypothesize that this is driven by lower
risk aversion by our research participants than real-world investors.
We also assessed the extent to which disposition effects impact trading frequency.
127
We
found a statistically significant correlation. This suggests that people who exhibit a greater
disposition effect are also more likely to make more trades. A 10% increase in
participants’ disposition effect index was associated with approximately one additional trade
being made. The positive correlation between participants’ disposition effects and trading
frequency suggests that a higher disposition effect may increase trading frequency as
individuals make a greater number of buy and sell decisions in response to market price
fluctuations. Alternatively or additionally, there may be one or more factors that affect both
outcomes. For example, individuals who are more prone to the disposition effect are more
likely to be making trades based on intuitive, rather than deliberative, decision-making
processes, which in turn may yield greater trading frequency.
126
For each participant, we calculated a disposition effect index that varied between 0 and 1, where 1 represented a total
disposition effect – that is, a hypothetical situation in which a participant always retains losing stocks and always sells winning
stocks. A value of 0.5 was indicative of neutral behaviour – i.e., participants being no more likely to lean towards buying or
selling as a result of a winning or losing outcome. We do not observe a disposition effect across groups as the average
disposition effect index was below 0.5 for all groups. Additional details on our disposition effect calculations are available in
Appendix C (Experimental Research Analysis Technical Details).
127
To do so, we regressed each participants’ disposition effect coefficient on their frequency of trading.
48
Conclusion: Considerations for Regulators
A wave of digital, mobile-friendly self-directed investing platforms has created new options
for retail investors in Canada and around the world. While these platforms have expanded
market participation, there is growing concern over some of the digital engagement practices
used by them and (to a lesser extent) by more traditional retail investment platforms.
The goal of this research is to support the Ontario Securities Commission (OSC) and other
regulators and stakeholders in understanding and responding to these new developments: to
help chart an effective, evidence-informed path forward as digital trading platforms continue
to evolve and grow.
Across the research activities, we found significant gaps in the empirical literature; very few
studies have examined the impact of these techniques on various investor behaviours. There
are notable exceptions. For example, there is good evidence that top traded lists can induce
“herding” behaviour by concentrating investor attention on salient stocks,
128
and that an
assortment of gamification tactics used in a simulated trading environment can increase
investor risk-taking.
129
Our research on behalf of the OSC adds another critical finding, that offering rewards (points)
with negligible economic value may dramatically increase trading frequency. Based on this
finding, we recommend that regulators consider whether to limit digital trading
platforms from offering points or other rewards for trading activity. We also believe that
regulators should consider collecting more data on how top traded lists (whether it be most
actively traded, top movers or otherwise) influence retail investor behaviour including investor
allocations to those stocks on the “top lists”.
While our research program produced insightful findings of the effects of points and top
traded lists on trading frequency, we should be cognizant of the effects of other gamification
behavioural techniques on investor behaviours. Thus, we further recommend that the OSC
and other regulators gather more data, especially for other techniques (e.g., gamblification,
feedback, social interactions, etc.). To do so, we recommend more studies be conducted on
simulated investing platforms, akin to our work as well as that of Chapkovski et al.
130
Regulators should also seek to leverage data collected by digital trading platforms. This
could include data from A/B tests of new digital engagement practices (DEPs). If DEPs have
not been A/B tested, historical data on key investor outcomes, including trading frequency,
types of transactions (e.g., margin trades, options), and other behaviours before and after
new DEPs have been introduced would be almost as valuable. This additional data will
enable the OSC and other regulators to set new, empirically-driven strategies based on high-
quality evidence.
We particularly recommend further evidence generation on social interactions, rewards, and
gamblification. Based on our literature scan, these are the behavioural techniques in our
taxonomy that are likely to have the largest impact on behaviour.
128
Barber, B. M., Huang, X., Odean, T., & Schwarz, C. (2020). Attention induced trading and returns: Evidence from Robinhood
users. Journal of Finance (Forthcoming).
129
Chapkovski, P., Khapko, M., & Zoican, M. (2021). Does gamified trading stimulate risk taking?
130
Chapkovski, P., Khapko, M., & Zoican, M. (2021). Does gamified trading stimulate risk taking?
49
In addition to collecting data focused on DEPs, we encourage regulators to generate and
collect evidence and data on potential strategies to mitigate potentially negative impacts of
DEPs on investor choices to determine if mitigation approaches are effective. For example,
there are theoretical reasons to believe that imposing moments of frictions in executing a
trade could, to a certain extent, counterbalance the tendency of some gamification
techniques to encourage less deliberative trading decisions.
Last, we encourage more exploration of the positive impacts that gamification and other
behavioural techniques can have on investor behaviours. For example, simplification of the
user experience is likely to increase market participation, reduce confusion, and save
investors time. Feedback techniques have proven very effective in educational contexts;
digital trading platforms could use these same approaches to enhance their users’ investing
knowledge and expertise.
1
Appendix A: Use of Gamification and Other Tactics on Trading Platforms
Table 1.
Gamification and behavioural tactics identified on retail investing platforms surveyed in the period of 1–30 September 2021.
GAMIFICATION & OTHER BEHAVIOURAL TACTICS
Location*
How platform
was
reviewed
Gamblifica-
tion
Leader-
boards
Rewards
Social
interactions
Social norms
Salient
prompts
Progress
framing
Feedback
Simplifica-
tion
Platform 1
Canada
Android
(demo)
No
No
No
Yes
Yes
No
No
No
Yes
Platform 2
USA
Android
(demo)
Yes
Yes
No
Yes
Yes
Yes
No
No
No
Platform 3
Canada/
USA
Web
(demo)
No
No
No
No
Yes
No
No
No
Yes
Platform 4
Canada iOS (demo)
No
No
No
No
No
No
No
No
Yes
Platform 5
Canada iOS (demo)
No
No
No
Yes
No
No
No
No
No
Platform 6
Canada
Android
(account)
No
No
No
No
No
No
No
No
No
Platform 7
USA Web search
Yes
No
No
No
Yes
Yes
No
No
Yes
Platform 8
USA Web search
Yes
Yes
Yes
Yes
Yes
No
No
No
No
2
Platform 9
USA Web search
No
No
No
No
No
No
No
No
No
Platform 10
Canada
iOS
(account)
Yes
No
Yes
No
Yes
Yes
No
No
Yes
Platform 11
USA
Android
(demo)
No
No
Yes
Yes
Yes
Yes
No
No
No
Platform 12
USA iOS (demo)
No
No
No
No
Yes
No
No
No
Yes
Legend and N
otes
● Platforms were reviewed on Android, iOS or Web.
● Some platforms enabled us to review features without creating an account. These platforms are labelled “(demo)” in the table above.
● Where we needed to create an account to review the features, the platforms are labelled “(account)”.
● If we could not create an account (e.g., for certain US platforms), we conducted a web search (e.g., review of videos, news articles) to
try to identify gamification and other behaviourally-informed features. These platforms are labelled “web search.”
● We attempted to review a Canadian bank’s platform, but investigating any features required provision of banking details. We attempt
ed
a w
eb search but the information available was quite limited. To the best of our understanding, it does not include any gamification or
other behavioural tactics, but because of the limitations we did not want to include it in this table due to the limitations of our review.
**These are the jurisdictions where these platforms are available to retail investors.
52
Appendix B: Detailed Experimental Research
Findings
Results by Experiment Group
Primary analysis: Trading frequency
Total Trades
Group Mean
Standard
Deviation
Minimum
value
Maximum
value
Observations
Control 18.44 16.84 1 90 803
Points 25.51 20.06 1 90 811
Top Traded List 18.90 16.38 1 90 816
N = 2,430
Regression analysis:
The regression coefficient on the Points condition, 7.19 (CI: 5.4, 9.0), shows that random
assignment to the Points group caused participants to, on average, conduct approximately 7 more
trades compared to participants in the control group (p=0.00). Therefore, exposure to the Point-
based rewards increased trading frequency by 39%. The regression coefficient on the Top Traded
condition, 0.57 (CI: -1.0, 2.2), was not statistically significant (p=0.49).
Average Diversification Index
Group Mean
Standard
Deviation
Minimum
value
Maximum
value
Observations
Control 0.70 0.27 0 1 803
Points 0.69 0.27 0 1 811
Top Traded List 0.70 0.27 0 1 816
N = 2,430
53
Regression analysis:
Regression coefficients were similar and not statistically significant for the Points condition (-0.01;
CI: -0.03, 0.02; p=0.62) and the Top traded condition (0.01; CI: -0.02, 0.03; p=0.54).
Exploratory analyses
Average Ratio of Top Traded Stocks Held
Group Mean
Standard
Deviation
Minimum
value
Maximum
value
Observations
Control 0.48 0.13 0.08 1 780
Points 0.48 0.12 0.04 1 802
Top Traded List 0.50 0.14 0 1 791
N = 2,373
Regression analysis:
The OLS regression coefficient, 0.035 (CI: 0.02, 0.05; p=0.00) suggests that exposure to a list
labelling a number of stocks as “top traded” was associated with a 7% increase in the ratio of these
stock holdings. In contrast, the regression coefficient on the Points condition, 0.01 ( CI: -0.01, 0.02),
was not statistically significant (p=0.41).
Average Ratio of Top Traded Stocks Sold
Group Mean
Standard
Deviation
Minimum
value
Maximum
value
Observations
Control 0.44 0.26 0 1 670
Points 0.45 0.22 0 1 725
Top Traded List 0.49 0.26 0 1 696
N = 2,091
Regression analysis:
The OLS regression coefficient, 0.049 (CI: 0.02, 0.08; p=0.00) suggests that exposure to a list
54
labelling a number of stocks as “top traded” was associated with an 11% increase in the ratio of
these stock sold. In contrast, the regression coefficient on the Points condition, 0.005 ( CI: -0.02,
0.03), was not statistically significant (p=0.72).
Average Ratio of Top traded Stocks Bought and Sold
Group Mean
Standard
Deviation
Minimum
value
Maximum
value
Observations
Control 0.44 0.19 0 1 803
Points 0.45 0.17 0 1 811
Top Traded List 0.50 0.21 0 1 816
N = 2,430
Regression analysis:
The OLS regression coefficient, 0.065 (CI: 0.05, 0.08; p=0.00) suggests that exposure to a list
labelling a number of stocks as “top traded” was associated with a 14% increase in the ratio of these
stock sold. In contrast, the regression coefficient on the Points condition, 0.08 ( CI: -0.01, 0.03), was
not statistically significant (p=0.35).
Final Fund Value
Group Mean
Standard
Deviation
Minimum
value
Maximum
value
Observations
Control 12,394.63 1952.09 5379 22,448 803
Points 12,420.80 1991.44 5695 22,197 811
Top Traded List 12,515.81 2071.25 6784 22,047 816
N = 2,430
Regression analysis:
Regression coefficients were not statistically significant for the Points condition (44.5; CI: -149.4,
238.3; p=0.65) and the Top Traded condition (133.4; CI: -64.5, 331.3; p=0.19).
55
Average Disposition Effect
Group Mean
Standard
Deviation
Minimum
value
Maximum
value
Observations
Control 0.33 0.15 0 0.86 803
Points 0.34 0.17 0 1 811
Top Traded List 0.34 0.15 0 1 816
N = 2,430
Regression analysis:
Regression coefficients for the Points (0.01; CI: -0.01, 0.02; p=0.36) and Top Traded (0.01; CI: -
0.01, 0.02; p=0.45) conditions were not statistically significant.
Background Questions and Demographics
Investor Status
Reports holding an investment product
(i.e., individually held stocks, ETFs,
securities or derivatives, bonds or notes
other than Canada Savings Bonds, mutual
funds, or private equity investments.
55.27%
Does not report holding an investment
product
44.73%
N = 2,430
Gender
Female 49.26%
Male 50.00%
Other 0.74%
N = 2,430
56
Province
Ontario 46.30%
Alberta 13.17%
British Columbia 12.72%
Quebec 10.53%
Manitoba 4.94%
Nova Scotia 3.79%
Saskatchewan 3.25%
New Brunswick 3.21%
Newfoundland and Labrador 1.40%
Prince Edward Island 0.45%
Northwest Territories 0.12%
Nunavut 0.08%
Yukon 0.04%
N = 2,430
Age
18-24 10.62%
25-24 24.90%
35-44 25.80%
45-54 18.89%
55-65 19.79%
N = 2,430
Ethnicity
Asian (including Chinese, Japanese, Korean,
etc.)
10.25%
Black / African Canadian / Caribbean 5.57%
57
British / Scottish / Irish / Welsh 16.09%
Canadian only 29.55%
Eastern European, including Russia 2.43%
Hispanic / Latin American 2.26%
Middle Eastern 5.27%
South Asian (including India, Pakistan, Sri
Lanka, Bangladesh, Nepal)
2.96%
Southeast Asian (including Burma, Thailand,
Vietnam, Laos, Cambodia, Philippines,
Singapore, etc.)
12.30%
Western European 3.42%
First Nations / Métis / Inuit 6.38%
Other 3.62%
N = 2,430
Household Annual Income Before Taxes
Less than $14,999 8.07%
$15,000 to $19,999 3.74%
$20,000 to $22,499 3.91%
$22,500 to $24,999 1.93%
$25,000 to $27,499 3.09%
$27,500 to $29,999 2.63%
$30,000 to $32,499 3.13%
$32,500 to $34,999 2.63%
$35,000 to $37,499 2.67%
$37,500 to $39,999 2.51%
$40,000 to $42,499 3.00%
$42,500 to $44,999 2.51%
$45,000 to $47,499 2.76%
$47,500 to $49,999 2.96%
58
$50,000 to $52,499 3.87%
$52,500 to $54,999 1.73%
$55,000 to $59,999 4.40%
$60,000 to $69,999 6.83%
$70,000 to $79,999 7.08%
$80,000 to $89,999 5.06%
$90,000 to $99,999 4.65%
$100,000 to $124,999 8.19%
$125,000 to $149,999 3.83%
$150,000 and above 4.77%
Prefer not to say 4.03%
N =
2,430
Employment Status
Employed full-time 48.77%
Employed part-time 11.03%
Self-employed full-time 4.98%
Self-employed part-time 3.46%
Active military 0.25%
Inactive military/Veteran 0.04%
Temporarily unemployed 8.68%
Full-time homemaker 5.60%
Retired 7.13%
Student 5.68%
Disabled 3.58%
Prefer not to answer 0.82%
N = 2,430
59
Highest Level of Education Completed
3rd Grade or less 4.90%
Middle School - Grades 4 - 8 19.05%
Completed some high school 14.65%
High school graduate 26.46%
Other post high school
vocational training
4.16%
Completed some college, but
no degree
17.49%
College/University Degree 12.31%
Prefer not to say 0.99%
N = 2,430
Device used to complete activity
Desktop computer 41.40%
Mobile phone 54.84%
Tablet 3.76%
N = 2,430
Self-reported overall knowledge of financial matters
Very low 12.35%
Low 23.09%
Average 48.81%
High 12.30%
Very high 3.46%
N = 2,430
Objective knowledge score
0/3 questions correctly 13.46%
60
answered
1/3 questions correctly
answered
26.05%
2/3 questions correctly
answered
43.50%
3/3 questions correctly
answered
17.00%
N = 2,430
Years of experience holding an investment account
0 46.30%
Between 0 and 5 28.31%
Between 5 and 10 9.59%
Between 10 and 15 3.99%
Between 15 and 20 4.98%
Between 20 and 25 2.30%
Between 25 and 30 2.31%
More than 30 2.22%
N = 2,430
Type of management of primary investment account
I work with, or have, an advisor or portfolio
manager or exempt market dealer
22.14%
I use an online investment adviser/robo-
adviser that selects investments on my
behalf (e.g., Wealth Simple, Questrade, Nest
Wealth)
10.16%
I am a self-directed investor, I do not work
with an advisor and I manage my own
investments through a discount brokerage
(order execution only account), mostly
through an online platform
on my computer
12.47%
I am a self-directed investor, I do not work
with an advisor and I manage my own
investments through a discount brokerage
(order execution only account), mostly
8.31%
61
through an app
on my phone or tablet
I only have investments through my
employer’s
pension plan
2.18%
N/A 44.73%
N = 2,430
Attitude towards risk when making investing decisions
I am very conservative and try to minimize
risk and avoid the possibility of any loss
26.21%
I am I am conservative but willing to accept
a small amount of risk and possibility of loss
39.67%
I am willing to accept a moderate level of risk
and tolerate moderate losses to achieve
potentially higher returns
28.68%
I am aggressive and typically take on
significant risk. I can tolerate large losses for
the potential of achieving higher returns
5.43%
N = 2,430
62
Appendix C: Experimental Research Analysis
Technical Details
Trading Frequency
To assess differences in the number of trade decisions made between the control condition
(Group 1) and each of the two experimental conditions (Groups 2 and 3), we examined the
treatment effect of condition using an OLS regression model:
Freq = α + βT
i
+ e
Where:
i represents the individual, with Var
i
representing the value of hypothetical variable Var for the
i
th
participant
Freq is our outcome variable, which represents trade number counts and takes on values
ranging from 1 to 90 (the number of trading instances in the experiment and thus the
maximum possible trade count a participant can have).
T is a categorical indicator for the treatment groups, which is to be used as a predictor of the
outcome variable and can take on the value of 0, 1, or 2
β is the coefficient of interest – a scalar which represents the impact that being assigned to
each of the three groups has on the number of trades a participant makes
α is the regression constant
e is the error term
Diversification Index
Our diversification index (DI) involves summing the squares of deviation from uniform
allocation for each given stock, and then dividing the sum by the score participants would get
if they concentrated all their funds into a single stock. Doing this yields a value that varies
continuously between 0 and 1, where 1 = anti-diversification. Subtracting this amount from 1
then allows us to flip this value into a 0 to 1 scale where 0 represents anti-diversification and
1 represents maximal diversification (as defined by equal allocation of funds across all
investment options).
131
As defined for a single simulated week, the DI was estimated as follows:
DI = 1 – (sum [(M/6 – (X_n)]^2) / H
Where:
131
Note: While diversification would usually also take into account some measure of riskiness (e.g. having a good balance of
riskiness of stocks), as well as the overall quantity of investments, ours purely takes into account the “spread” of money across
different stock options. This is because: (1) stocks included in this experiment have been designed to have very similar risk
profiles, and (2) the experiment specifies a low ceiling for quantity of investments and thus requires a metric of diversification
that is ‘blind’ to the total number of possible investments.
63
M equals the total amount of money held by each investor at each experiment screen.
M/6 represents the amount that would be invested in that stock if the investor engaged in
uniform distribution of funds across all the available stocks (e.g., at experiment onset,
$10,000/6=$1666.67).
n represents each stock, 1-6
X_n is the amount invested in stock option n
H represents the sum of squared errors that a participant would receive if they allocated all
their funds into a single stock to the detriment of all other options. This differs across
simulated weeks due to the variation in available funds.
To assess differences in the Diversification Index (DI) between the control condition (Group
1) and each of the two experimental conditions (Groups 2 and 3), we examined the treatment
effect of condition using an OLS regression model:
DI
i
= α + βT
i
+ e
Where:
i represents the individual, with Var
i
representing the value of hypothetical variable Var for the
i
th
participant
DI is our outcome variable, which varies continuously between 0 and 1
T is a categorical indicator for the treatment groups, which is to be used as a predictor of the
outcome variable and can take on the value of 0, 1, or 2
β is the coefficient of interest – a scalar which represents the impact that being assigned to
each of the three groups has on a participant’s diversification index
α is the regression constant
e is the error term
Disposition Effects
For each participant, we calculated a disposition effect coefficient as follows:
Here, fractions marked in blue served the purpose of differentially weighting decisions made
in the loss and gain domains based on the relative frequencies of gains and losses in the
total pool of investment outcomes.
The resulting index varies between 0 and 1, where 1 represents complete disposition – that
is, a hypothetical situation in which a participant always retains losing stocks and always
sells winning stocks. A value of 0.5 should be indicative of neutral behaviour – i.e.,
participants being no more likely to lean towards buying or selling as a result of a winning or
losing outcome. Each participants’ disposition effect coefficient was calculated for each week
of trading and averaged to yield a single measure.
To assess for potential differences in the disposition effect (DE) between the control
condition (Group 1) and each of the two experimental conditions (Groups 2 and 3), we
64
examined the treatment effect of condition using an OLS regression model:
DE = α + βT
i
+ e
Where:
i represents the individual, with Var
i
representing the value of hypothetical variable Var for the
i
th
participant
T is a categorical indicator for our treatment groups, which is to be used as a predictor of the
outcome variable and can take on the value of 0, 1, or 2
β is the coefficient of interest – a scalar which represents the impact that being assigned to
each of the three groups has on a participant’s disposition effect
α is the regression constant
e is the error term
Further, to assess for a potential impact of disposition effect on overall trading frequency, we
used an OLS regression mode as follows:
Freq= α + βDE
i
+ e
Where:
i represents the individual, with Var
i
representing the value of hypothetical variable Var for the
i
th
participant
Freq is our outcome variable, which represents trade number counts and takes on values
ranging from 0 to 90 (the number of trading opportunities in the experiment and thus the
maximum possible trade count a participant can have).
DE is a continuous measure of the disposition effect, which takes on values between 0 and 1
and serves as our independent variable
β is the coefficient of interest – a scalar which represents the impact that each additional unit
on the disposition effect scale has on a participant’s frequency of trading
α is the regression constant
e is the error term
65
Exposure to Top Traded List
We compared the proportion of individuals’ stock ownership that consists of stocks featured
on a top traded list in Group 3 with the proportions derived using those same stocks in
Groups 1 and 2. This fraction was computed separately for each simulated trading week as
shown below, prior to being averaged across these weeks
Individual participants’ fractions (which we termed stock ratios; SR) then served as inputs for an OLS
regression model as follows:
SR= α + βT
i
+ e
Where:
i represents the individual, with Var
i
representing the value of hypothetical variable Var for the
i
th
participant
SR is our outcome variable, which represents a fraction varying continuously between 0 and 1
(where 1 represents 100% of all stock ownership being confined to the top traded list)
T is a binary indicator for the treatment groups, which is to be used as a predictor of the
outcome variable and can take on the value of 0 (given to all participants in Groups 1 and 2)
or 1 (given to participants in Group 3, which are ‘positive’ for the top traded stock list
exposure)
β is the coefficient of interest – a scalar which represents the impact of the presence or
absence of a top traded list on purchasing decisions
α is the regression constant
e is the error term
66
Appendix D: Experimental Research Screens
Introduction screens (all conditions)
67
Trading screens
Control condition
Control condition initial allocation of funds
68
Control condition transition screen (included between all ‘weeks’ of trading)
69
Sample control condition selling screen
70
Sample control condition buying screen
71
Points condition
Points condition initial allocation of funds
72
Points condition transition screen (included between all ‘weeks’ of trading)
73
Sample control condition selling screen
74
Sample control condition buying screen
75
Top Traded List condition
Top traded condition initial allocation of funds
76
Top traded condition transition screen (included between all ‘weeks’ of trading)
77
Sample top traded condition selling screen
78
Sample top traded condition buying screen
79
Appendix E: Bibliography
Adam, M., Roethke, K., & Benlian, A. (2021). Gamblified digital product offerings: an experimental
study of loot box menu designs. Electronic Markets, 1-16.
Allcott, H., & Rogers, T. (2014). The short-run and long-run effects of behavioral interventions:
Experimental evidence from energy conservation. American Economic Review, 104(10), 3003-37.
Arnold, M., Pelster, M., & Subrahmanyam, M. G. (2022). Attention triggers and investors’ risk-taking.
Journal of Financial Economics, 143(2), 846-875.
Auer, M. M., & Griffiths, M. D. (2016). Personalized behavioral feedback for online gamblers: A real
world empirical study. Frontiers in Psychology, 1875.
Ayton, P., & Fischer, I. (2004). The hot hand fallacy and the gambler’s fallacy: Two faces of subjective
randomness?. Memory & cognition, 32(8), 1369-1378.
Bali, T. G., Hirshleifer, D., Peng, L., & Tang, Y. (2021). Attention, social interaction, and investor
attraction to lottery stocks (No. w29543). National Bureau of Economic Research.
Barber, B. M., & Odean, T. (2000). Trading is hazardous to your wealth: The common stock
investment performance of individual investors. Journal of Finance, 55(2), 773-806.
Barber, B. M., Huang, X., Odean, T., & Schwarz, C. (2020). Attention induced trading and returns:
Evidence from Robinhood users. Journal of Finance (Forthcoming). SSRN 3715077.
Behavioral Insights Team. (2014). EAST: Four simple ways to apply behavioural insights. Retrieved
from: https://www.bi.team/wp-content/uploads/2015/07/BIT-Publication-EAST_FA_WEB.pdf
Bettinger, E. P., Long, B. T., Oreopoulos, P., & Sanbonmatsu, L. (2012). The role of application
assistance and information in college decisions: Results from the H&R Block FAFSA experiment. The
Quarterly Journal of Economics, 127(3), 1205-1242.
Biais, B., Hilton, D., Mazurier, K., & Pouget, S. (2005). Judgemental overconfidence, self-monitoring,
and trading performance in an experimental financial market. The Review of Economic Studies, 72(2),
287-312.
Burns, Z., Chiu, A., & Wu, G. (2010). Overweighting of small probabilities. In Wiley Encyclopedia of
Operations Research and Management Science (pp. 1-8). Hoboken, NJ, USA: John Wiley & Sons.
Cara, C. (2019). Dark patterns in the media: A systematic review. Network Intelligence Studies, 7(14),
105-113.
Carver, C. S., & Scheier, M. F. (2012). Attention and self-regulation: A control-theory approach to
human behavior. Springer Science & Business Media.
Ceschi, A., Sartori, R., Dickert, S., Scalco, A., Tur, E. M., Tommasi, F., & Delfini, K. (2021). Testing a
norm-based policy for waste management: An agent-based modeling simulation on nudging recycling
behavior. Journal of Environmental Management, 294, 112938.
Chapkovski, P., Khapko, M., & Zoican, M. (2021). Does gamified trading stimulate risk taking?. SSRN
3971868.
Chang, T. Y., Solomon, D. H., & Westerfield, M. M. (2016). Looking for someone to blame: Delegation,
cognitive dissonance, and the disposition effect. The Journal of Finance, 71(1), 267-302.
Chaudhry, S., & Kulkarni, C. (2021, June). Design Patterns of Investing Apps and Their Effects on
Investing Behaviors. In Designing Interactive Systems Conference 2021 (pp. 777-788).
80
Croson, R., & Sundali, J. (2005). The gambler’s fallacy and the hot hand: Empirical data from casinos.
Journal of Risk and Uncertainty, 30(3), 195-209.
Cutts, J. (2021, Sep 14). “Not all brokerages should 'gamify' a la Robinhood - but others can/will go
further.” Traders Magazine.
De Groot, J. I., Abrahamse, W., & Jones, K. (2013). Persuasive normative messages: The influence of
injunctive and personal norms on using free plastic bags. Sustainability, 5(5), 1829-1844.
Dorn, A. J., Dorn, D., & Sengmueller, P. (2015). Trading as gambling. Management Science, 61(10),
2376-2393.
Ebert, N., Alexander Ackermann, K., & Scheppler, B. (2021, May). Bolder is better: Raising user
awareness through salient and concise privacy notices. In Proceedings of the 2021 CHI Conference
on Human Factors in Computing Systems (pp. 1-12).
Egli, M., Schaal, D. W., Thompson, T., & Cleary, J. (1992). Opioid-induced response-rate decrements
in pigeons responding under variable-interval schedules: reinforcement mechanisms. Behavioural
Pharmacology.
Fazio, L. (2020). Pausing to consider why a headline is true or false can help reduce the sharing of
false news. Harvard Kennedy School Misinformation Review, 1(2).
Fenton-O'Creevy, M., Adam, M., Clough, G., Conole, G., Gaved, M., Lins, J. T., et al. (2015). A game
based approach to improve traders' decision-making. In: The International Gamification for Business
Conference 2015: Strategic Industrial Applications of Games and Gamification.
FINRA (2021, Jun 30). "FINRA Requests Comment on Effective Methods to Educate Newer
Investors". Retrieved from: https://www.finra.org/rules-guidance/notices/special-notice-
063021#comments
Frydman, C., & Wang, B. (2020). The impact of salience on investor behavior: Evidence from a natural
experiment. Journal of Finance, 75(1), 229-276.
Grieve, R. & Lowe-Calverley, E. (2016, Jul 16). The power of rewards and why we seek them out. The
Conversation.
Griffiths, M., & Wood, R. (2001). The psychology of lottery gambling. International Gambling Studies,
1(1), 27-45.
Hamari, J., Koivisto, J., & Sarsa, H. (2014, January). Does gamification work?--a literature review of
empirical studies on gamification. In 2014 47th Hawaii international conference on system sciences
(pp. 3025-3034).
Hamari, J. (2017). Do badges increase user activity? A field experiment on the effects of gamification.
Computers in Human Behavior, 71, 469-478.
Haw, J. (2008). Random-ratio schedules of reinforcement: The role of early wins and unreinforced
trials. Journal of Gambling Issues, (21), 56-67.
Hou, Y., Xiong, D., Jiang, T., Song, L., & Wang, Q. (2019). Social media addiction: Its impact,
mediation, and intervention. Cyberpsychology: Journal of psychosocial research on cyberspace, 13(1).
Hswen, Y., Murti, V., Vormawor, A. A., Bhattacharjee, R., & Naslund, J. A. (2013). Virtual avatars,
gaming, and social media: Designing a mobile health app to help children choose healthier food
options. Journal of mobile technology in medicine, 2(2), 8.
Huynh, D., & Iida, H. (2017). An Analysis of Winning Streak's Effects in Language Course of
“Duolingo". University Kebangsaan Malaysia Press.
81
Investment Industry Regulatory Organization of Canada – IIROC (2019, Apr 9). "Guidance on Order
Execution Only Services and Activities". Retrieved from: https://www.iiroc.ca/news-and-
publications/notices-and-guidance/guidance-order-execution-only-services-and-activities
Jacobson, L. & Tamakloe, K. S. (2021, Aug 19). “Have vaccine lotteries worked? Studies so far show
mixed results”. Politifact.
Johnson, E. J., Shu, S. B., Dellaert, B. G., Fox, C., Goldstein, D. G., Häubl, G., & Weber, E. U. (2012).
Beyond nudges: Tools of a choice architecture. Marketing Letters, 23(2), 487-504.
Johnson, D., Deterding, S., Kuhn, K. A., Staneva, A., Stoyanov, S., & Hides, L. (2016). Gamification
for health and wellbeing: A systematic review of the literature. Internet interventions, 6, 89-106.
Kalda, A., Loos, B., Previtero, A., & Hackethal, A. (2021). Smart (Phone) Investing? A within Investor-
time Analysis of New Technologies and Trading Behavior (No. w28363). National Bureau of Economic
Research.
Kellner, C., Reinstein, D., & Riener, G. (2019). Ex-ante commitments to “give if you win” exceed
donations after a win. Journal of Public Economics, 169, 109-127.
Kesel, W. (2020, Apr 1). 3 Avatar-Based Virtual Event Platforms for Planners to Consider.
BizBash.com
Kluger, A. N., & DeNisi, A. (1996). The effects of feedback interventions on performance: A historical
review, a meta-analysis, and a preliminary feedback intervention theory. Psychological Bulletin,
119(2), 254.
Kivetz, R., Urminsky, O., & Zheng, Y. (2006). The goal-gradient hypothesis resurrected: Purchase
acceleration, illusionary goal progress, and customer retention. Journal of Marketing Research, 43(1),
39-58.
Koo, M., & Fishbach, A. (2012). The small-area hypothesis: Effects of progress monitoring on goal
adherence. Journal of Consumer Research, 39(3), 493-509.
Krause, M., Mogalle, M., Pohl, H., & Williams, J. J. (2015). “A playful game changer: Fostering student
retention in online education with social gamification”. In Proceedings of the Second ACM conference
on Learning@Scale (pp. 95-102).
Lagorio, C. H., & Winger, G. (2014). Random-ratio schedules produce greater demand for iv drug
administration than fixed-ratio schedules in rhesus monkeys. Psychopharmacology, 231(15), 2981-
2988.
Liu, Y. Y., Nacher, J. C., Ochiai, T., Martino, M., & Altshuler, Y. (2014). Prospect theory for online
financial trading. PloS one, 9(10), e109458.
Lyft (2022). "Streak Bonus". Retrieved from: https://help.lyft.com/hc/e/articles/115015748908-Streak-
Bonus
Leibbrandt, A., Gneezy, U., & List, J. A. (2013). Rise and fall of competitiveness in individualistic and
collectivistic societies. Proceedings of the National Academy of Sciences, 110(23), 9305-9308.
Lewis, L. K., Rowlands, A. V., Gardiner, P. A., Standage, M., English, C., & Olds, T. (2016). Small
steps: preliminary effectiveness and feasibility of an incremental goal-setting intervention to reduce
sitting time in older adults. Maturitas, 85, 64-70.
Looyestyn, J., Kernot, J., Boshoff, K., Ryan, J., Edney, S., & Maher, C. (2017). Does gamification
increase engagement with online programs? A systematic review. PloS one, 12(3), e0173403.
82
Mehra, A., Paul, J., & Kaurav, R. P. S. (2021). Determinants of mobile apps adoption among young
adults: theoretical extension and analysis. Journal of Marketing Communications, 27(5), 481-509.
Mekler, E. D., Brühlmann, F., Opwis, K., & Tuch, A. N. (2013, October). Do points, levels and
leaderboards harm intrinsic motivation? An empirical analysis of common gamification elements. In
Proceedings of the First International Conference on gameful design, research, and applications (pp.
66-73).
Miltenberger, R. G. (2008). “Behavioral Modification: Principles and Procedures”. Florida, Thomson.
Montag, C., Markowetz, A., Blaszkiewicz, K., Andone, I., Lachmann, B., Sariyska, R., ... & Markett, S.
(2017). Facebook usage on smartphones and gray matter volume of the nucleus accumbens.
Behavioural brain research, 329, 221-228.
Morschheuser, B., & Hamari, J. (2019). The gamification of work: Lessons from crowdsourcing.
Journal of Management Inquiry, 28(2), 145-148.
Kundisch, D., & von Rechenberg, T. (2017). Does the framing of progress towards virtual rewards
matter?. Business & Information Systems Engineering, 59(4), 207-222.
Nah, F. F. H., Zeng, Q., Telaprolu, V. R., Ayyappa, A. P., & Eschenbrenner, B. (2014, June).
Gamification of education: a review of literature. In International conference on hci in business (pp.
401-409). Springer, Cham.
Odean, T. (1999). Do investors trade too much? American economic review, 89(5), 1279-1298.
Powell, L. (2021, September 14). SEC Investor Advisory Committee examines digital engagement
practices [Blog post]. Retrieved from https://jimhamiltonblog.blogspot.com/2021/09/sec-investor-
advisory-committee_14.html
Rodrigues, L. F., Oliveira, A., & Costa, C. J. (2016). Does ease-of-use contributes to the perception of
enjoyment? A case of gamification in e-banking. Computers in Human Behavior, 61, 114-126.
Scheiber, N. (2017, Apr 2). How Uber Uses Psychological Tricks to Push Its Drivers' Buttons. New
York Times. Retrieved from: https://www.nytimes.com/interactive/2017/04/02/technology/uber-drivers-
psychological-tricks.html
Sherif, M. (1935). A study of some social factors in perception. Archives of Psychology, 27: 187.
Sherman, L. E., Hernandez, L. M., Greenfield, P. M., & Dapretto, M. (2018). What the brain ‘Likes’:
neural correlates of providing feedback on social media. Social cognitive and affective neuroscience,
13(7), 699-707.
Shull, R. L., & Grimes, J. A. (2006). Resistance to extinction following variable-interval reinforcement:
Reinforcer rate and amount. Journal of the Experimental Analysis of Behavior, 85(1), 23-39.
Skinner, B.F. (1969). Contingencies of reinforcement: A theoretical analysis. Meredith Corporation.
Sparkman, G., & Walton, G. M. (2017). Dynamic norms promote sustainable behavior, even if it is
counternormative. Psychological science, 28(11), 1663-1674.
Song, H., Kim, J., Tenzek, K. E., & Lee, K. M. (2010, June). The effects of competition on intrinsic
motivation in exergames and the conditional indirect effects of presence. In Annual Conference of the
International Communication Association, Singapore.
Thaler, R. H., & Johnson, E. J. (1990). Gambling with the house money and trying to break even: The
effects of prior outcomes on risky choice. Management science, 36(6), 643-660.
83
Tobon, S., Ruiz-Alba, J. L., & García-Madariaga, J. (2020). Gamification and online consumer
decisions: Is the game over? Decision Support Systems, 128, 113167.
Tuah, N. M., Yoag, A., Ahmedy, F., & Baharum, A. (2019). A gamification and avatar self-
representation application for diabetes self-management. Int. J. Adv. Trends Comput. Sci. Eng, 8,
401-407.
Uber (2022). “How does the Consecutive Trips promotion work?” Retrieved from:
https://help.uber.com/driving-and-delivering/article/how-does-the-consecutive-trips-promotion-
work?nodeId=de983305-076a-40cf-aaf4-7b23f50a0007.
U.S. Securities and Exchange Commission (2021, Aug. 27). “SEC Requests Information and
Comment on Broker-Dealer and Investment Adviser Digital Engagement Practices, Related Tools and
Methods, and Regulatory Considerations and Potential Approaches; Information and Comments on
Investment Adviser Use of Technology.” Retrieved from: https://www.sec.gov/news/press-
release/2021-167
U.S. Securities and Exchange Commission (n.d.). "Request for Information and Comments on Broker-
Dealer and Investment Adviser Digital Engagement Practices, Related Tools and Methods, and
Regulatory Considerations and Potential Approaches...". Retrieved from:
https://business.cch.com/srd/34-92766.pdf
Van Osselaer, S. M., Alba, J. W., & Manchanda, P. (2004). Irrelevant information and mediated
intertemporal choice. Journal of Consumer Psychology, 14(3), 257-270.
Wisniewski, B., Zierer, K., & Hattie, J. (2020). The power of feedback revisited: A meta-analysis of
educational feedback research. Frontiers in Psychology, 10, 3087.
Ontario Securities Commission –
Investor Ofce Research and
Behavioural Insights Team (IORBIT) Behavioural Insights Team (BIT)
Tyler Fleming
Director, Investor Office
tfl[email protected]v.on.ca
Marian Passmore
Senior Advisor, Policy
mpas
[email protected]v.on.ca
Matthew Kan
Senior Advisor, Behavioural Insights
[email protected]v.on.ca
Sasha Tregebov
Director, BIT Canada
Sebastian Salomon-Ballada
Senior Advisor
Laura Callender
Advisor
