The Role of Advanced Technologies in Automated Trading Systems and Its Influence on Investor Attitudes
##plugins.themes.bootstrap3.article.main##
This paper provides an overview of algorithmic trading (robo-trading), which replaces human stockbroker recommendations to investors for buying/selling/holding stocks in their investment portfolio, and the integration of technologies like artificial intelligence and quantum computing within automated systems. The paper covers the impacts of algorithmic trading, the specific implementations of advanced technologies, and the acceptance of these technologies among college students. The cited articles show that robo-trading has existed since the 1970s. Algorithmic trading has increased stock market efficiency. Artificial intelligence allows for predictive models to adjust to the changing financial markets. On the other hand, quantum algorithms like Variational Quantum Eigensolver (VQE) aid in portfolio optimization. The paper also presents findings from an empirical study on college students’ attitudes toward advanced technologies within robo-trading. Although a few other studies exist on investor attitudes about acceptance of algorithmic trading for stock portfolios, no other study examined the attitudes of college students on the potential application of advanced technologies within robo-trading. A statistical analysis was undertaken to see how attitudes shift based on factors like class level and previous stock market experience. The paper discusses the results and policy implications for stock brokerage firms that hope to implement advanced technologies.
Downloads
Introduction
The paper begins by breaking down algorithmic trading along with its historical context. Then, it transitions to the upsides of algorithmic trading and its impact on the stock market. The paper then moves to current advancements within artificial intelligence and quantum computing, which have greatly improved the efficiency of algorithmic trading. Additionally, it addresses the societal and economic implications of these technologies, considering both the opportunities and challenges they present, including ethical concerns and regulatory frameworks. The paper also touches upon the criticisms of algorithmic trading and how those can be addressed. Because advanced technologies are broad in explanation and scope, the shifts in topics within the literature review are meant to illustrate the potential and freedom of using advanced technologies within algorithmic trading. The paper concludes with an empirical study to identify the significance of factors like gender and class level on the acceptance of advanced technologies within algorithmic trading.
Although a few other studies exist on investor attitudes about acceptance of algorithmic trading for stock portfolios, no other study examined the attitudes of college students on the potential application of advanced technologies within robo-trading. A statistical analysis was undertaken to see how attitudes shift based on factors like class level and previous stock market experience. The paper discusses the results and policy implications for stock brokerage firms that hope to implement advanced technologies.
Literature Review
Algorithmic trading is the process of using a computer program with pre-defined instructions to execute a trade. A few common algorithmic trading strategies include trend-following, arbitrage/risk-free, and mean reversion. Automated trading helps eliminate human errors from trading, and trades that use algorithmic trading programs often ensure the best possible prices for the trader. Despite its benefits, algorithmic trading does have its drawbacks. Since algorithmic trading relies on historical financial data to make educated trading decisions, unforeseen market conditions can confuse programs and result in large losses. Furthermore, increased market volatility, large shifts in the price of assets, is a direct consequence of algorithmic trading as it executes trades of large volumes at high speeds.
Algorithmic trading started with the development of electronic communication networks (ECNs) in the 1970s. These networks allowed ventures outside of traditional trading when it came to buying and selling orders. The 1990s were characterized by digitalization, the integration of digital tools within business operations, which made algorithmic trading more popular through the increased presence of online trading platforms (Garg, 2023). HFT kicked off in the early 2000s. HFT is when algorithms execute orders at large volumes and fast speeds. HFT capitalizes on arbitrage within the market, creating a profit through these price discrepancies. With its ability to learn from historical trends and analyze patterns within trading data, machine learning has been recently used in algorithmic trading. It can help make trading algorithms more adaptive to market conditions and increase profits through optimization of trading methods.
High-Frequency Trading (HFT) improves market efficiency, liquidity provision, and price discovery. HFT algorithms work to decrease price differences across different markets and make sure that asset prices are accurate. This quick data adjustment is important to provide fair prices within the stock market. In terms of liquidity provision, HFT helps firms make sure that there are always counterparties available for trades, thereby increasing overall market liquidity. Furthermore, HFT algorithms continuously monitor liquidity levels, volatility, and other market metrics in real-time, optimizing liquidity provision. With more use of HFT by firms, there is a decrease in wider bid-ask spreads and lower trading costs for everyone involved (Chughet al., 2024). HFT not only boosts liquidity but also assists in managing buy and sell orders. As mentioned before, the speed at which HFT algorithms process information allows for rapid price adjustments, reducing delay and boosting exchanges. HFT ensures that the stock market runs smoothly, providing fair and equal opportunities for investors and those who trade within the market.
The securities trading industry has readily adopted algorithmic trading for its efficiency, cost-reduction, and speed. The original value chain within securities trading involved the reliance of brokers to execute orders; however, the usage of algorithmic trading has allowed buy-side firms to replace brokers and their services, altering traditional roles in the market. Buy-side institutions compare the potential performance improvements, such as cost reduction and better execution quality, to the resources and efforts it takes to implement automated systems. The Task-Technology Fit (TTF) model is important as it highlights the role of technology and algorithmic trading within financial markets along with its adoptability. The performance of algorithmic trading on specific trading tasks was studied and shown to be important in the adoption of the technology. High levels of correlation represent increased usage of automated trading along with better trading performances. Institutions with a specialization in advanced technologies look for low implementation efforts but high-performance outputs for the technology to be utilized (Gsell, 2024). Nonetheless, further research on topics like organizational culture is necessary to implement algorithmic trading more efficiently and vastly within the financial market.
Roughly 53% of hedge fund trades and 58% of FOREX have been executed through algorithms, signifying that the use of algorithmic trading within financial markets has increased (Corgnetet al., 2023). It is argued that HFT may contribute to market instability through events like flash crashes. The study method employed within the study focuses on the payoffs of traders using algorithms. It compares two markets with both having human traders as well as algorithms, allowing for the analysis of types of algorithms on markets. There is a comparison of a market with only human traders to those involving either a limit-order algorithm, which adds liquidity, or a market-order algorithm, an algorithm takes liquidity. Algorithmic trading is generally identified to increase liquidity and price efficiency. However, studies focusing on welfare are rare. On the contrary, the student analyzes different types of algorithms, the cognitive ability of traders, and the impact of algorithmic trading on welfare.
High-cognitive traders earn more and contribute to price efficiency, but algorithms decrease their trading potential. Instead, algorithms level the playing field for traders with lower cognitive abilities. This is where traders’ attitudes toward algorithms come into play, measured through the Negative Attitudes Toward Robots Scale. Traders with negative attitudes towards algorithms tend to perform worse in algorithmic markets, making more mistakes and earning less. This result highlights the ability of technology to fill gaps in trading skills and decision-making.
Limit-order algorithms significantly improve welfare by participating in more mutually beneficial trades. However, there is little benefit to human traders because surpluses are not distributed to humans. Market-order algorithms, on the other hand, do not increase welfare and make use of human traders for speed advantages. Both types of algorithms improve price efficiency and reduce volatility, but limit-order algorithms are more effective in building up liquidity. This goes to show that some algorithms can help market stability. The behavior of the market is a key factor in the use of algorithms. Experimental results show that human traders can adjust to markets that use algorithms, but their cognitive and perception of algorithmic trading has an impact on their success. Traders with negative attitudes towards algorithms are more likely to make errors, leading to substandard performance.
Robo-advisors, trading platforms operated by algorithms, have grown in use since the 2008 financial crisis. These robo-advisors oversee managing portfolios and handling clients. These platforms take feedback from users for their preferred trading methods and personalized risks. The robo-advisor then uses the feedback from the client to create an investment portfolio. Generally, the benefits of robo-advisors include their low cost when compared to other financial advisors and features like portfolio rebalancing. Robo-advisors are seen as investment methods that use technology to improve trading experiences for clients (Sironi, 2016). Other research, such as Phoon and Koh (2018), highlighted the low-cost and accessible nature of investing that robo-advisors provide, especially for younger, less experienced traders.
It is important to use the Diffusion of Innovation theory and the Information Search Model to showcase how internal and external factors affect the usage of robo-advisors (Fan & Chatterjee, 2020). Internal factors include financial literacy, investment knowledge, and risk tolerance. For instance, individuals with sizable financial knowledge and risk tolerance are more likely to adopt robo-advisors. External factors, like the advice of friends and family, also can impact the trustability of robo-advisors. Participation in investment clubs increases the likelihood of using robo-advisors when trading as they are more informed about recent technologies. Attitudes toward traditional financial advisors also influence the adoption of robo-advisors. Individuals who value access to more investment options might lean towards using robo-advisors. This goes to show how the convenience and accessibility of robo-advisors can help generate a growing user base for it.
Additionally, demographic and socioeconomic characteristics affect the use of robo-advisors. Individuals under the age of sixty-five who have a high income and a good collection of assets use robo-advisors. This demographic has a higher risk threshold, making them more readily open to adopting new technologies when investing. On the other hand, older individuals with lower financial literacy may be less likely to adopt robo-advisors due to unfamiliarity with the technology and its impact on their trading platform.
The Penn-Lehman Automated Trading Project (PLAT) simulates real-world market conditions, allowing for the execution of trades with market data while being low in risk. A key component of the PLAT is the ability to demonstrate market microstructure and order execution and encourage the development of effective trading strategies. Market conditions are mimicked through concepts like liquidity, volatility, and order flow, and these conditions serve to evaluate various trading algorithms: market-making algorithms, arbitrage algorithms, and mean reversion algorithms. Automated trading strategies and algorithms developed within the PLAT project improve market efficiency, provide liquidity, and help determine prices. This is possible since the algorithms trade with high speed and precision, thus reducing the execution time and effort necessary for large orders. They ensure tight bid-ask spreads are maintained through continuous buy and sell order placements, thereby reducing market frictions (Kearns & Ortiz, 2003). Furthermore, such automated trading algorithms assist pricing by responding to current information as well as adjusting the price accordingly. As a result, all available information is factored correctly in asset prices which bring fair and transparent markets. At large, integrating automated trading strategies through projects like PLAT project has an important part of making financial markets more efficient and secure.
With the growth of artificial intelligence, deep learning, a subset of machine learning that uses neural networks to make human-like decisions, has the potential to transform algorithmic trading through its ability to learn and process hidden patterns within financial datasets. Trading agents that are under deep reinforcement learning (DLR) can pick up optimal trading strategies through extensive data analysis and trial and error. Agents observe market states, take actions based on studying trends, and get feedback that helps shape and improve their next financial decision. DRL models can analyze non-linear relationships within data like price movements data, improving trading strategies to ensure the best return. The most important aspect of DRL within trading is the potential to adjust to market conditions, which works to assist in risk management and increase profits.
Artificial intelligence has many practical applications in stock market trading, specifically in areas like portfolio optimization, stock market prediction, and financial sentiment analysis. Genetic algorithms and reinforcement learning, a branch of artificial intelligence, help balance risk and return, optimizing portfolio allocations. Deep learning models like CNNs and RNNs extensively study and analyze historical stock data to make educated trading decisions. Another AI tool, natural language processing, analyzes the sentiment of financial reports and news to assess market sentiment and uncover important financial information to guide financial decisions. These applications highlight the impact of AI within financial markets and how it can improve trading strategies through predictive modeling.
The use of machine learning algorithms, such as support vector machines and decision trees, can help aid in creating predictive models for market movements. Additionally, machine learning can process historical and current data, allowing for varying and adaptive trading strategies. Deep learning techniques, especially recurrent neural networks and long short-term memory networks, can model sequential data and capture dependence in financial time series (Cohen, 2022). Optimization through reinforcement learning (RL) is growing. RL agents can learn optimal strategies by market interactions and incentive system, especially important in high-frequency trading environments.
The application of machine learning models within financial markets comes with its issues. Although machine learning models increase algorithmic trading efficiency, the complexity of the models presents risk of overfitting and less interpretability. Machine learning models used in trading can become overly complex due to their ability to learn from data and develop pattern recognition skills. This presents challenges by ensuring that models are both accurate and usable by humans. Overfitting is when models “learn” from noise in the data rather than genuine patterns, causing wrong decisions. Techniques such as regularization do not encourage complexity unless they improve the performance of the machine-learning model. Furthermore, human judgment remains crucial in the development of machine learning models. Experts are needed to keep complexity in check and that the models function as they are supposed to. Simplicity in models allows for transparency in using models within financial markets. Being able to balance simplicity and complexity in machine learning models for algorithmic trading is important to assess the use of the models in different financial settings (Hansen, 2020). An alternative technology, quantum computing, has become appealing in algorithmic trading to solve long and complex optimization problems, making the algorithms useful for risk management, portfolio optimization, and market simulation.
Quantum computing can play a huge role within the trading market. It has the potential to improve portfolio optimization, option pricing, and risk management. Portfolio optimization is the selection of certain assets or a portfolio based on criteria like expected return. Quantum algorithms, specifically Variational Quantum Eigensolver (VQE) and Quantum Approximate Optimization Algorithm (QAOA), help maximize returns and minimize risks that benefit portfolios amidst market changes (Hermanet al., 2023). For option pricing, quantum computing can improve Monte Carlo simulations through quantum amplitude estimation, reducing the number of required simulations and thus providing faster pricing models. Risk management is vital within the trading market. Risk management is managed through quantum computation and its ability to improve the calculation of risk measures like Value at Risk (VaR). This improved calculation helps to prevent losses on a given investment.
To study the application of quantum computing within portfolio optimization, the Variational Quantum Eigensolver (VQE) was employed to enhance portfolio optimization. The tests of the quantum optimization methods are performed on simulators and quantum hardware like IBM NISQ. The experiment used a small dataset of four well-known assets: Apple, IBM, Netflix, and Tesla, spanning from 2011 to 2022. The primary goal of the research tests was to improve parameters for VQE, minimize noise, and compare quantum optimization methods to classical methods.
For the set of four assets, the VQE method was able to generate optimal portfolio allocations that balanced risk and return. Quantum solutions were found to be competitive to classical optimization means, but noise and decoherence within quantum devices were still limiting factors (Buonaiutoet al., 2023). On the contrary, the quantum simulators outputs had better portfolio optimization results due to the absence of noise found commonly within quantum hardware. The small number of assets used in the study was due to the limited number of qubits and high noise levels in current quantum hardware. Along with this, noise still stands as a major problem for quantum computation mainly because decoherence could occur, stripping qubits of their quantum physical properties and making quantum computing full of errors. Ultimately, although quantum computing could be used for portfolio optimization on a smaller scale, it is much too underdeveloped for large-scale use.
Along with its application in portfolio optimization and risk management, quantum computing can improve statistical arbitrage strategies in HFT (Zhuanget al., 2022). Statistical arbitrage, the process of using mean reversion analysis and large portfolios to exploit pricing inefficiencies, is a strategy to maximize profit on an investment. Within the paper, quantum linear regression is employed to ensure proper data fitting, capitalizing on Moore-Penrose pseudoinverse and quantum principal component analysis to manage datasets. Two algorithms are showcased for statistical arbitrage. The first algorithm picks portfolios prematurely based on a condition number threshold, while the second algorithm works to handle unknown condition numbers, filtering portfolios to ensure that they are co-integrated as the algorithm runs. Quantum number comparison techniques are applied to verify cointegration among portfolios. These quantum methods can aid in improving time complexity within HFT environments.
Static volatility, a common issue with trading, doesn’t capture the dynamic nature of the market. Algorithmic trading systems need to utilize dynamic profiling, adjusting trading based on market data to prevent static volatility (Shen, 2021). However, current trading systems do not have high enough computational power for dynamic profiling. To add, current trading systems are not capable of handling “special days” in trading where trade patterns differ to usual days. Trading systems need more customized and varied trading strategies, allowing for adaptability to these irregular days. As the market shifts to real-time trading of derivatives, the prices associated with computation within automated trading systems increase significantly. Traditional end-of-day pricing models do not mirror real-time trading environments.
Algorithmic trading has become important in HFT environments where lots of data needs to be processed quickly. Algorithms now analyze structured and unstructured data sources, including news articles, social media posts, etc. These algorithms are known for their accuracy in trading predictions by incorporating data from various sources. For example, studies have shown how the media and its sentiment have affected stock prices. For example, negative media could predict negative stock returns (Tetlock, 2007). Meanwhile, it has been seen that sentiment in financial news could predict stock price movements (Schumakeret al., 2012). The sentiment expressed in online media has been linked to increased market volatility and stock returns. Investor attention is the idea that a few stock options stand out to investors due to the buzz of media. Investors are more likely to purchase stocks that capture their attention (Barberet al., 2008).
It is also important to consider investor attention along with media sentiment. Based on the attention a company receives; stock prices can change dramatically with outside influence. High investor attention can increase positive or negative sentiment on stock returns. While the integration of investor attention into algorithmic trading strategies shows promise, there are challenges. One major issue is that correlations cannot be applied to small datasets or less popular companies, leading to economic loss. Furthermore, the rapid incorporation of information into stock prices by high-frequency traders can limit opportunities to capitalize on these stocks (Claphamet al., 2019).
Despite its criticism for increasing market volatility, algorithmic trading has been shown to improve liquidity and overall market quality. HFT directly improves market liquidity by reducing the bid-ask spread. As a result, transaction costs for smaller trades are reduced, making the market more accessible for all. Furthermore, HFT contributes to a deeper order book, which means liquidity is present at large, typically at various price levels. Although HFT can increase short-term volatility due to the speed of the trades, over time, the increased liquidity created by HFT can prevent extreme price changes, making the market more stable (Dubeyet al., 2021). The increased liquidity and price discovery associated with HFT make it an attractive means for algorithmic trading.
Various market capitalizations, including small, medium, and large caps, show algorithmic trading to reduce trade sizes significantly. On average, trade size declines by 2375.0115 basis points when trading with a higher algorithmic trading presence (Dubey, 2022). The use of smaller trade sizes by algorithmic trading can lead to market fragmentation which makes large order sizes impact asset prices at a higher magnitude. While algorithmic trading plays a role in improving liquidity, it also is partly responsible for market volatility, especially during periods of high market uncertainty. This increase in market instability can lead to flash crashes and a rapid price decline in a market with a fast recovery period.
The growing application of algorithmic trading within financial markets has made regulation of algorithmic trading important. Various regulatory frameworks have been used within the US and UK, focusing on market intervention and human supervision to quickly mitigate issues with automated trading. For example, MiFID II in the EU and the SEC’s guidelines in the US clearly define restrictions for algorithmic trading, emphasizing pre-trade controls and risk management (Lee & Schu, 2021). Direct market interventions, like circuit breakers, can help reduce the short-term market volatility caused by algorithmic trading. Additionally, regulators and enforcement action ensure liability and promote the integrity of algorithmic trading systems.
In an earlier study, EMI partnered with Boston Research Technologies to conduct a national survey of investors aimed at understanding their attitudes toward robo-advisors. The survey captured input from over 700 respondents, distributed across age and wealth segments (EMI, Boston, 2016). The EMI study did not focus on college students and had older respondents than the study conducted for this paper.
Study Methodology
In addition to the literature review, a survey questionnaire was developed (see Appendix 1) to assess attitudes of college students’ acceptance of using algorithmic trading instead of a human stockbroker offering recommendations for buy/sell/hold decisions with their stock market investment portfolio. Questions were asked as to gender, class level, and level of prior experience with stock market investment to see if statistically significant differences existed in attitudes about acceptance of algorithmic trading and advanced technologies like quantum computing within automated trading systems.
The survey questionnaire was administered to 168 business students at the Master’s degree graduate and Bachelor’s degree undergraduate levels at a private university in Southern California. Business students were selected as it would be expected that they would have greater knowledge of newer quantitative advances in finance and investing.
Findings/Results
The following are preliminary results from the study, as additional statistical analysis is still being undertaken to get a deeper understanding of the results. Appendix 2 shows a tally of responses to each question in the survey and the number of total responses to each question.
The results of the study was given below.
Students with Stock Market Investments
60% of respondents had investments in the stock market, with most (93%) not using robo-trading due to not being familiar with it (72%) but willing to consider using it (77%).
Students without Stock Market Investments
For the 40% of respondents with no investments in the stock market, most (90%) were not familiar with robo-trading, but willing to consider using it (73%).
All Respondents
When selecting an investment manager, from the list of five selection factors, the most prioritized factor (30% of respondents) was the innovative use of AI and Quantum Computing for portfolio optimization.
When selecting a robo-trading platform, the most prioritized factor (39% of respondents) was a strong financial reputation, particularly in AI and advanced trading systems.
Only 3% of respondents rated their knowledge of investments as high, and 56% rated having a low knowledge of investments.
A high percentage (84% of respondents) would consider investing in assets driven by or secured with advanced technologies.
A high percentage (72% of respondents) would consider investing in cryptocurrencies. Only 14% of respondents selected a high level of familiarity with AI. The respondents were split in their class level from graduate to undergraduate freshmen. The respondents were nearly evenly split between males (46%) and females (54%).
Discussion of Findings
Students with Stock Investments
As far as having investments in the stock market or not, 60% of respondents reported that they had investments in the stock market. This high percentage of investing students can be explained by business students having greater awareness of the long-term yield and return on investment in stocks compared to alternate investments. Also, many students do not have full faith that Social Security monthly payments will cover their full retirement financial needs; thus, they need to start investing at a younger age to secure their financial futures. As many students at the private university, with high tuition, come from families with higher incomes and higher net worth compared to families with children attending public universities, one would expect they may have observed parents with investments in the stock market leading to a higher rate of investing themselves. There were factors that were statistically significant in association with students having investments in the stock market or not: ratings of higher knowledge of investments were associated with having investments in the stock market (p = 0.001 level), being more familiar with AI was associated with investments in the stock market (p = 0.003 level), and gender was significant at a p = 0.032 level with males having more investments in the stock market than females.
For those students who have investments in the stock market, only 7% used robo-trading. Due to the overall low level of users of robo-trading among students with stock investments, no statistically significant difference was found with their use of robo-trading based upon gender or class level.
Among students who had investments in the stock market, only 28% were familiar with what robo-trading is. There were statistically significant differences in familiarization with what robo-trading is among students invested in the stock market based upon gender and class level. Males were more familiar with what robo-trading is compared to females (significant at the p = 0.001 level). Students with higher class Master’s level were more familiar with what robo-trading is compared to lower class Freshmen level business students (significant at the p = 0.041 level).
Among students who had investments in the stock market and were not using robo-trading, 82% were willing to consider it. Given the high overall willingness to consider using robo-trading, there was no statistically significant difference based on gender or class level.
Students Without Stock Investments
For students who did not have investments in the stock market, 90% were not familiar with what robo-trading is. Given the overall high lack of familiarization with what robo-trading is, there was no statistically significant difference based on gender or class level.
For students who did not have investments in the stock market, 78% stated that they would consider using robo-trading. Given the overall high willingness to consider using robo-trading, there was no statistically significant difference based on gender or class level.
All Respondents
When selecting an Investment Manager, the following shows results for the listed five factors:
1. When selecting an Investment Manager, which factor do you prioritize the most when considering their ability to leverage advanced technologies?
• [31%] Innovative use of AI and Quantum Computing for portfolio optimization
• [17%] Low-cost, algorithm-driven strategies
• [23%] Historical performance using traditional and AI-driven models
• [15%] Personalized investment strategies utilizing machine learning
• [15%] Reputation for ethical use of technology and transparency
There was no statistically significant difference in these results based on gender or class level. A prior study of 700 investors asked a similar question, and these results differed as the respondents of that study were older and placed the highest priority on reputation for ethical use (EMI, Boston, 2016).
When selecting a Robo-trading platform, the following results were found:
2. When selecting a Robo-trading platform, which factor do you prioritize the most?
• [12%] Innovation in AI-driven trading algorithms
• [21%] Tech reputation in AI and quantum financial modeling
• [7%] Competitive pricing in algorithmic trading services
• [22%] User-friendly interface enhanced by AI-based decision-making support
• [39%] Strong financial reputation, particularly in AI and advanced trading systems
There was no statistically significant difference in these results based upon gender or class level. A prior study of 700 investors asked a similar question, and these results differed as the respondents of that study were older and placed less importance on a strong financial reputation (EMI, Boston, 2016).
Many respondents, 56%, rated their knowledge of investments, particularly in the application of quantum computing and AI in the financial sector, as low, with 41% moderate rating, and only 3% of respondents rated their knowledge as high. There was no statistically significant difference in these results based upon gender or class level. One would expect that business students would have higher-rated knowledge either from their business classes and/or their personal readings on investing.
The majority of respondents, 84%, would consider investing in assets driven by or secured with advanced technologies like AI, blockchain, or quantum-resistant algorithms (e.g., cryptocurrencies, quantum-secured assets). Only 16% were not willing to consider in assets driven by or secured with advanced technologies. There was no statistically significant difference in these results based upon gender or class level.
The majority of respondents, 72%, would consider investing in cryptocurrencies. There was a statistically significant difference (p = 0.004), with males more willing to consider investing in cryptocurrencies compared to females. There was no statistically significant difference based on class level.
Only 14% of respondents indicated a high level of being familiar with AI, 72% indicated moderate, and 14% indicated a low level of being familia with AI. There was no statistically significant difference based upon class level and gender. These results were contrary to the belief that many faculty members have that all their college students have a high level of being familiar with AI and the belief that they are all familiar with AI to cheat on assignments requiring writing.
Matrix of Four Types of Students
To gain a deeper understanding of student knowledge of items in the survey and how it affected their willingness to try newer investment approaches, a transformation of the data was undertaken. For each respondent, a score was calculated based upon their summed score of knowledge: being familiar or not with what robo-trading is (either as a student with Question 2 or without investments Question 4), their knowledge indicated about investments Question 8, and their stated knowledge about what AI is Question 11. For each respondent, a score was calculated based on their willingness to try newer investment approaches such as robo-trading, willingness to consider investing in assets driven or secured with Advanced Technologies Question 9, and willingness to consider investing in cryptocurrencies Question 10. Each respondent, based upon these calculated scores, was placed into a cell in a two by two matrix: lower knowledge, higher knowledge, lower willingness to try newer investment approaches, and higher willingness to try newer investment approaches. Table I shows the number of respondents in each cell and the percentages in each cell. Overall, there is a high percentage of respondents (78%) who are willing to consider investing in the use of newer technologies. A cross-tabulation and chi-square indicated that there was no statistically significant association between the level of knowledge and willingness to try newer investment approaches. This result is primarily the result of only 22% of respondents who were not willing to try newer investment approaches. Although there was not a statistically significant association between level of knowledge and willingness to try investments with newer advanced technology approaches, only 5% of more knowledgeable respondents were not willing to try investments with newer advanced technology approaches.
Low level of knowledge | High level of knowledge | |
---|---|---|
More willing | 96 | 32 |
Less willing | 28 | 8 |
Policy Implications and Conclusion
As far as having investments in the stock market or not, 60% of respondents reported that they had investments in the stock market. Ratings of higher knowledge of investments were associated with having investments in the stock market, and being more familiar with AI was associated with investments in the stock market. For those students who have investments in the stock market, only 7% used robo-trading. Among students who had investments in the stock market, only 28% were familiar with what robo-trading is. Males were more familiar with what robo-trading is compared to females, and students with higher class levels were more familiar with what robo-trading is compared to lower-class students. Among students who had investments in the stock market and were not using robo-trading, 82% were willing to consider it. For students who did not have investments in the stock market, 90% were not familiar with what robo-trading is. For students who did not have investments in the stock market, 78% stated that they would consider using robo-trading. There was no statistically significant difference in these results based upon gender or class level as to which factor is most important in selecting an Investment Manager or in selecting a robo-trading platform. Most respondents, 56%, rated their knowledge of investments, particularly in the application of quantum computing and AI in the financial sector, as low, with 41% moderate rating, and only 3% of respondents rated their knowledge as high. One would expect that business students would have higher rated knowledge either from their business classes and/or their personal readings on investing. The majority of respondents, 84%, would consider investing in assets driven by or secured with advanced technologies like AI, blockchain, or quantum-resistant algorithms (e.g., cryptocurrencies, quantum-secured assets). The majority of respondents, 72%, would consider investing in cryptocurrencies. Only 14% of respondents indicated a high level of being familiar with AI, 72% indicated moderate, and 14% indicated a low level of being familiar with AI.
Overall, there is a high percentage of respondents (78%) who are willing to consider investing with the use of newer technologies. Although, there was not a statistically significant association between level of knowledge and willingness to try investments with newer advanced technology approaches, only 5% of more knowledgeable respondents were not willing to try investments with newer advanced technology approaches. The results of this study indicate that there is an overall need to educate all students more about both newer technologies (AI, Quantum Computing, Robo-trading) and the long-term higher yield benefits of investing in the stock market. The students showed a willingness to consider newer technologies as applied to investments (cryptocurrencies, robo-trading) but had a lack of knowledge regarding these technologies and yields from use. The results indicate that a gender gap exists where females, who have a lower level of investment and knowledge than male classmates, would benefit from more education about these newer technologies and yields.
With fears among college students about the predicted insolvency of Social Security to cover their financial needs to support their retirement, they are aware of the critical need to invest in stocks for their retirement financial well-being. Despite recognizing the need to invest in stocks, a higher number of students should be investing in the stock market, and all college students need greater knowledge about investing and advanced technological approaches that can assist with investments. Stock brokerage firms, personal finance advisors, and business schools should take steps to educate all students to a higher level, as results indicate a desire by students to gain more knowledge in these areas.
Appendix 1
Survey on Student Attitudes About Investing
The following questions are part of a research study on student attitudes about investing; the responses are anonymous.
For those who have investments in the stock market: (if you don’t, skip to Question 4):
1. Do you use robo-trading or algorithmic trading? [ ] Yes [ ] No
2. Are you familiar with what robo-trading is? [ ] Yes [ ] No
3. If you are not using robo-trading or algorithmic trading, would you consider using it?
[ ] Yes [ ] No
For those who do not have investments in the stock market:
4. Are you familiar with what robo-trading is? [ ] Yes [ ] No
5. If you are not using robo-trading or algorithmic trading, would you consider using it?
[ ] Yes [ ] No
For All Respondents:
6. When selecting an Investment Manager, which factor do you prioritize the most when considering their ability to leverage advanced technologies?
[ ] Innovative use of AI and Quantum Computing for portfolio optimization
[ ] Low-cost, algorithm-driven strategies
[ ] Historical performance using traditional and AI-driven models
[ ] Personalized investment strategies utilizing machine learning
[ ] Reputation for ethical use of technology and transparency
7. When selecting a Robo-trading platform, which factor do you prioritize the most?
[ ] Innovation in AI-driven trading algorithms
[ ] Tech reputation in AI and quantum financial modeling
[ ] Competitive pricing in algorithmic trading services
[ ] User-friendly interface enhanced by AI-based decision-making support
[ ] Strong financial reputation, particularly in AI and advanced trading systems
8. How would you rate your knowledge of investments, particularly in the application of quantum computing and AI in the financial sector?
[ ] High—I have a strong understanding of AI/quantum computing’s role in modern investments
[ ] Moderate—I understand the basic concepts and their potential impact
[ ] Low—I am unfamiliar with AI and quantum technologies in finance
9. Would you consider investing in assets driven by or secured with advanced technologies like AI, blockchain, or quantum-resistant algorithms (e.g., cryptocurrencies, quantum-secured assets)?
[ ] Yes
[ ] No
10. Would you consider investing in cryptocurrencies? [ ] Yes [ ] No
11. How familiar are you with AI?: [ ] High [ ] Moderate [ ] Low
12. Class Level: [ ] Graduate [ ] Senior [ ] Junior [ ] Sophomore [ ] Freshman
13. Gender: [ ] Male [ ] Female [ ] Other
Appendix 2
Tally of Responses and Number of Cases Per Question
The following questions are part of a research study on student attitudes about investing; the responses are anonymous.
For those who have investments in the stock market: (if you don’t, skip to Question 4)
1. Do you use robo-trading or algorithmic trading? [7] Yes [93] No n = 100
2. Are you familiar with what robo-trading is? [28] Yes [72] No n = 100
3. If you are not using robo-trading or algorithmic trading, would you consider using it?
[73] Yes [20] No n = 93
For those who do not have investments in the stock market:
4. Are you familiar with what robo-trading is? [7] Yes [61] No n = 68
5. If you are not using robo-trading or algorithmic trading, would you consider using it?
[48] Yes [18] No n = 66
For All Respondents:
6. When selecting an Investment Manager, which factor do you prioritize the most when considering their ability to leverage advanced technologies?
[50] Innovative use of AI and Quantum Computing for portfolio optimization
[27] Low-cost, algorithm-driven strategies
[38] Historical performance using traditional and AI-driven models
[24] Personalized investment strategies utilizing machine learning
[25] Reputation for ethical use of technology and transparency
n = 164
7. When selecting a Robo-trading platform, which factor do you prioritize the most?
[19] Innovation in AI-driven trading algorithms
[33] Tech reputation in AI and quantum financial modeling
[11] Competitive pricing in algorithmic trading services
[35] User-friendly interface enhanced by AI-based decision-making support
[63] Strong financial reputation, particularly in AI and advanced trading systems
n = 161
8. How would you rate your knowledge of investments, particularly in the application of quantum computing and AI in the financial sector?
[5] High—I have a strong understanding of AI/quantum computing’s role in modern investments
[70] Moderate—I understand the basic concepts and their potential impact
[94] Low—I am unfamiliar with AI and quantum technologies in finance
n = 169
9. Would you consider investing in assets driven by or secured with advanced technologies like AI, blockchain, or quantum-resistant algorithms (e.g., cryptocurrencies, quantum-secured assets)?
[139] Yes
[26] No
n = 165
10. Would you consider investing in cryptocurrencies? [118] Yes [47] No n=165
11. How familiar are you with AI?: [23] High [120] Moderate [23] Low n = 166
12. Class Level: [16] Graduate [45] Senior [62] Junior [14] Sophomore [30] Freshman n = 167
13. Gender: [77] Male [90] Female [0] Other n = 167
References
-
Barber, B. M., Odean, T., & Zhu, N. (2008). Do Retail Trades Move Markets? Proceedings of the National Academy of Sciences, 105(8), 2991–6. https://doi.org/10.1073/pnas.0711579105.
Google Scholar
1
-
Buonaiuto, G., Gargiulo, F., De Pietro, G., Esposito, M., & Pota, M. (2023). Best practices for portfolio optimization by quantum computing, experimented on real quantum devices. Scientific Reports, 13(1), 19434. https://doi.org/10.1038/s41598-023-45392-w.
Google Scholar
2
-
Chugh, Y., Agrawal, S., Shetty, Y., & Guruprasad, M. (2024). Algotrading and its impact on stock markets. International Journal of Research in Engineering, Science and Management, 7(3), 48–55. https://doi.org/10.5281/zenodo.10849607.
Google Scholar
3
-
Clapham, B., Siering, M., & Gomber, P. (2019). Popular news are relevant news! How investor attention affects algorithmic decision-making and decision support in financial markets. Information Systems Frontiers, 22(4), 899–920. https://doi.org/10.1007/s10796-019-09950-w.
Google Scholar
4
-
Cohen, G. (2022). Algorithmic trading and financial forecasting using advanced artificial intelligence methodologies. Mathematics, 10(18), 3302. https://doi.org/10.3390/math10183302.
Google Scholar
5
-
Corgnet, B., DeSantis, M., & Siemroth, C. (2023). Algorithmic trading, price efficiency and welfare: An experimental approach. SSRN Electronic Journal. https://doi.org/10.2139/ssrn.4419304.
Google Scholar
6
-
Dubey, R. K. (2022). Algorithmic trading: The intelligent trading systems and its impact on trade size. Expert Systems with Applications, 202, 117279. https://doi.org/10.1016/j.eswa.2022.117279.
Google Scholar
7
-
Dubey, R. K., Babu, A. S., Jha, R. R., & Varma, U. (2021). Algorithmic trading efficiency and its impact on market-quality. Asia-Pacific Financial Markets, 29(3), 297–315. https://doi.org/10.1007/s10690-021-09353-5.
Google Scholar
8
-
EMI, Boston. (November 22, 2016). EMI survey of investors reveals attitudes towards robo-advisors. https://emiboston.com/emi-survey-of-investors-reveals-attitudes-towards-robo-advisors/.
Google Scholar
9
-
Fan, L., & Chatterjee, S. (2020). The utilization of robo-advisors by individual investors: An analysis using diffusion of innovation and information search frameworks. Journal of Financial Counseling and Planning, 31(1), 130–145. https://doi.org/10.1891/jfcp-18-00078.
Google Scholar
10
-
Garg, I. (2023). Algorithmic trading: A comprehensive review of technological advancements and market implications. International Journal of Social Science and Economic Research, 8(8), 2252–2247. https://doi.org/10.46609/ijsser.2023.v08i08.012.
Google Scholar
11
-
Gsell, M. (2024). Technological Innovations in Securities Trading: The Adoption of Algorithmic Trading. AIS Electronic Library (AISeL). https://aisel.aisnet.org/pacis2009/54/.
Google Scholar
12
-
Hansen, K. B. (2020). The virtue of simplicity: On machine learning models in algorithmic trading. Big Data & Society, 7(1), 205395172092655. https://doi.org/10.1177/2053951720926558.
Google Scholar
13
-
Herman, D., Googin, C., Liu, X., Sun, Y., Galda, A., Safro, I., Pistoia, M., & Alexeev, Y. (2023). Quantum computing for finance. Nature Reviews Physics, 5(8), 450–465. https://doi.org/10.1038/s42254-023-00603-1.
Google Scholar
14
-
Kearns, M., & Ortiz, L. (2003). The Penn-Lehman automated trading project. IEEE Intelligent Systems, 18(6), 22–31. https://doi.org/10.1109/mis.2003.1249166.
Google Scholar
15
-
Lee, J., & Schu, L. (2021). Regulation of algorithmic trading: Frameworks for human supervision and direct market interventions. SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3765882.
Google Scholar
16
-
Phoon, K. F., & Koh, C. C. F. (2018). Robo-advisors and wealth management. Journal of Alternative Investments, 20(3), 79–94. https://doi.org/10.3905/jai.2018.20.3.079.
Google Scholar
17
-
Schumaker, R. P., Zhang, L., & Huang, Z. (2012). Evaluating sentiment in financial news articles. IEEE Intelligent Systems, 27(6), 6–15. https://doi.org/10.1109/MIS.2012.129.
Google Scholar
18
-
Shen, J. (2021). Nine challenges in modern algorithmic trading and controls. Algorithmic Trading and Controls, 1(1), 1–9. https://arxiv.org/pdf/2101.08813v1.
Google Scholar
19
-
Sironi, P. (2016). FinTech Innovation: From Robo-Advisors to Goal-Based Investing and Gamification. Wiley.
Google Scholar
20
-
Tetlock, P. C. (2007). Giving content to investor sentiment: The role of media in the stock market. Journal of Finance, 62(3), 1139–68. https://doi.org/10.1111/j.1540-6261.2007.01232.x.
Google Scholar
21