Fintech literacy among millennials: The roles of financial literacy and education

Abstract

While readily available fintech products are in rise for consumers, the lack of basic fintech literacy (FTL) may preclude them fully utilize its benefits. This study aims to investigate FTL, and then identifies if actual financial literacy, perceived financial literacy and demography predict FTL. Using millennials from Malaysia, the study reports that (a) millennials show medium level of FTL; (b) they display higher literacy on P2P lending, but lower on machine learning; (c) high fintech literacy millennials are male, younger, Chinese and highly educated; (d) actual financial literacy is positively associated with machine learning and crowdfunding literacy, whereas perceived financial literacy is negatively associated with robo-advisor’s literacy; (e) education has a significant positive impact on fintech’s definition, machine learning, blockchain, P2P lending and crowdfunding; (f) age has an inverse relationship with cryptocurrency and blockchain literacy. Ethnicity and gender also contribute to FTL. Implications are discussed for millennials and fintech service providers.

Keywords:

• Fintech literacy
• financial literacy
• millennials
• education
• Malaysia

1. Introduction

The usage and investment of fintech products become increasingly crucial decisions due to readily available, but more complex fintech products. Although these decisions are common to all people, it is particularly important for millennials in Asian countries, where fintech has higher penetration (APAC Fintech Rankings, 2022). The number of more sophisticated fintech products is continuously rising in these regions and easily accessible to everyone at lower cost (e.g., robo-advisors). However, consumers, who are the most important end users, seem to have lower adoption of these innovative products (e.g., Zhe, 2019). While there are multiple possible reasons, insufficient literacy about the availability and functionality of different fintech products is argued to be one of the reasons. Furthermore, the risk and uncertainty associated with traditional financial products increase the need for alternative financial products for better diversification and financial planning. In this regard, a vast range of fintech products could be alternative options. To fully utilize these benefits, individuals require a certain level of relevant fintech literacy before its usage and investment. Hence, fintech literacy (FTL) is important. In line with the definition of literacy stated by UNESCO (2018), FTL refers to the ability of individuals to identify the fintech products, understand and interpret their basic functions properly.

Prior studies related to FTL have either looked at literacy of various cryptocurrencies (Bannier et al., 2019; Henry et al., 2019; Khan, 2023; Steinmetz et al., 2021) or digital financial literacy, which refers to knowledge of financial services in digital platforms (Lyons & Kass-Hanna, 2021; Setiawan et al., 2020). In addition, previous studies identified the role of financial literacy in improving bitcoin literacy and adoption of fintech products (Bannier et al., 2019; Moenjak et al., 2020; Morgan & Trinh, 2019). However, there has been less focused on FTL, and its relationship with financial literacy. Financial literacy’s intervention from early childhood originates from the social cognitive theory (Bandura, 1986). In line with social cognitive theory, financial literacy can be a way to increase one’s literacy on the complex fintech products. In addition, individuals varied in different demographic characteristics may exhibit different fintech literacy due to heterogeneity. These research gaps are spotted by the following question: to what extent do financial literacy and demography influence FTL? Moreover, millennials have been ignored in existing literature as studies on FTL (e.g., bitcoin literacy) have focused on general population or household (e.g., Henry et al., 2019). They are a growing segment of the internet everywhere and tech-savvy (The Edge Malaysia Weekly, 2017). Therefore, this study aims to investigate millennials’ FTL and the effects of financial literacy and demography on FTL.

This study designs a survey to collect information from millennials in Malaysia for the following reasons. First of all, Malaysia has up-and-coming large working-class millennials (Malaysia Fintech Ecosystem Report, 2019). According to Malaysia’s demographic age distribution, over 6 million people are youth in the country (Fong, 2020); and they have greater potential to accelerate fintech growth. However, the major challenge is the adoption and usage of fintech products (Malaysia Fintech Ecosystem Report, 2019), possibly, due to inadequate literacy. Second, Malaysia’s position in the global fintech index (ranked 15 in Asia) indicates that fintech industry is growing quickly (APAC Fintech Rankings, 2022), which offers a great opportunity for millennials to take advantage of them. Third, while various fintech products relating to payment, lending, advisory and investments are at doorstep for millennials, the adoption of these products has been sluggish (Zhe, 2019). The question is if millennials have sufficient fintech literacy or not. Fourth, although a fintech playbook has been proposed to provide fintech-related education to Malaysians (Zhe, 2019), it is unknown which area of fintech needs greater attention in the fintech playbook.

To test the degree of FTL, the study uses standard 15 fintech literacy items adapted from the CFA institute (Mack & Kissell, 2018) following the true/false format of Bannier et al. (2019) and finds that millennials show medium level of FTL. Millennials have higher knowledge of P2P lending and lower on machine learning. The study reveals a specific demographic characteristic of high FTL millennials (i.e., male, younger, Chinese and highly educated). The study further investigates the effects of’ actual financial literacy, perceived financial literacy and demography on FTL. The findings show that actual financial literacy is positively associated with the knowledge of machine learning and crowdfunding, whereas perceived financial literacy is negatively associated with robo-advisor’s knowledge. Among demography, education leads to an increase in knowledge on fintech’s definition, machine learning, blockchain, P2P lending and crowdfunding. Age has an inverse relationship with the knowledge of cryptocurrency and blockchain. Ethnicity and gender also contribute to FTL. When using FTL index, actual financial literacy and education positively predict FTL.

The paper makes the following contribution: (i) this study considers various aspects of FTL unlike one aspect (e.g., cryptocurrency); (ii) this study shows millennials’ FTL in contrast to general individuals; (iii) this study explores the role of actual financial literacy, perceived financial literacy and demography in explaining FTL; (iv) As few concepts of FTL have been studied in developed markets, the results of this study can be comparable to those from developed markets.

2. Literature review

While existing studies cover the literacy of few concepts of fintech, less studies concentrate on broader scope of FTL. According to EY Global Fintech Adoption Index, 96% of consumers are aware of at least one of the fintech services available in the market (Agarwal & Zhang, 2020). More specifically, Steinmetz et al. (2021) evaluate cryptocurrency literacy of a representative sample of German households employing an online survey and find a high cryptocurrency awareness (83%). A low level of cryptocurrency knowledge is reported when assessing self-rated cryptocurrency knowledge. Henry et al. (2019) investigate bitcoin literacy using the Bitcoin Omnibus Survey from a nationally representative sample, and report that 64% of the respondents know about bitcoin, where only 2.9% of them own this well-known virtual currency. Bannier et al. (2019) subsequently use the Bitcoin Omnibus Survey together with different surveys of US nationally representative data to measure bitcoin literacy by using the correct answers to six true/false questions and demonstrate that the mean value of the bitcoin literacy index is 3.013 indicating half of the participants have bitcoin literacy. In their study, men possess stronger bitcoin knowledge than women. Khan (2023) investigates not only bitcoin literacy but also ethereum and litecoin literacy, and documents that individuals’ bitcoin literacy is relatively higher than ethereum and litecoin. Khan asserted that higher financial literacy is associated with higher cryptocurrency literacy.

Ran et al. (2019) examine the literacy of borrowers and lenders on P2P lending platforms in China and whether financial literacy affects both borrowers’ and lenders’ participation and performance on the P2P platform. While high financially literate borrowers are less likely to default on their loans, less financially literate borrowers are more likely to fall into the debt trap and have poor financial planning. The evidence also shows that financially literate investors are more likely to get higher returns. The study of Ran et al. (2019) indicates the importance of having knowledge of P2P lending.

A handful of studies covers the individual’s knowledge on robo-advisors. Bhatia et al. (2020) conduct an in-depth interview, with experts regarding their understanding of robo-advisors in India. Education and building trust are suggested to increase awareness of robo-advisors. The firms providing robo-advisory services need to educate their clients about the automated financial robo-advisors. Gan et al. (2021) show that in Malaysia only 20.7% of consumers are aware of robo-advisors.

Prior studies attempt to examine the link between financial literacy and fintech adoption. Bannier et al. (2019) document that actual and perceived financial literacy are the key factors affecting bitcoin literacy. Both types of financial literacy explain the majority (40%) of the gender gap in bitcoin literacy. Using survey data from a low-level ICT developed country in LAOS, Morgan and Trinh (2019) find that financial literacy score is positively related to the awareness and adoption of fintech products and services. Financial literacy is crucial in using new innovative fintech products and services. Similar findings are documented by Morgan and Trinh (2020) in the context of Vietnam, where one’s level of financial literacy has a positive impact on knowledge of fintech products and its usage. In another study, Morgan (2022) finds the role of financial literacy as a key input to increase FTL and adoption of fintech products. However, their study did not measure different aspects of FTL in detail, ignored perceived financial literacy and used general population as a sample. Moenjak et al. (2020) discuss a conceptual framework connecting fintech, financial knowledge, and financial behaviors within the scope of Thailand. They argue that a lack of digital knowledge (e.g., digital transactions) causes lower usage of digital financial services. Financial education, on the other hand, leads to greater probability of fintech adoption and higher switching probability of fintech apart from trust and preferences for transparency (Jünger & Mietzner, 2020).

Across prior studies, the focus is merely on one aspect of FTL, which is not sufficient to represent one’s FTL covering various aspects of fintech. Other than that, no prior research is conducted for examining the role of actual financial literacy, perceived financial literacy and demographic variables as controls on FTL particularly among millennials- the gaps that this study intends to fill.

3. Methodology

3.1. Data collection and sample

An online survey was administered to collect the data from millennials in Malaysia. Millennials born between 1987 and 2001 (i.e., 18–32 years old at the time of data collection) were the target population. The age range of millennials was alike the surveyed millennials in prior studies (Au-Yong-Oliveira et al., 2018; Lee et al., 2020). The questionnaires were distributed to millennials using non-probability convenience sampling due to lack of sampling frame, convenient and easier accessibility of social media platforms (e.g., Facebook, WhatsApp) to reach millennials (Etikan et al., 2016). Millennials grew up with technology, and therefore, it was easy to contact them via social media platforms. A link of Google form was posted to the social media groups consisting members of Malaysian millennials. As an initial screening criterion, respondents were solicited their age range to be the potential respondents. Conducting a survey through online has several advantages (Lee et al., 2020). First, an online survey facilitates collecting survey data within a short period of time without contacting respondents face-to-face (Lazar & Preece, 1999; Wright, 2005). Second, an online survey allows to reach many respondents in broad areas at less cost (Lee & Yang, 2013; Sue & Ritter, 2012; Tierney, 2000). Third, the survey responses are automatically stored in the database together with the date and time when respondents filled up the questionnaires; and it is very efficient to proceed with analysis (Granello & Wheaton, 2004). Fourth, an online survey facilitates higher efficiency concerning cost and a high possibility of reaching the target number of respondents within the shortest time. The survey was administered for the three weeks (i.e., December 2019 to January 2020) yielding a collection of 350 responses. After exclusion of inappropriate responses, for example, choosing the same rating for all questions and those who did not meet the criteria, the final sample size was 330. This sample size is deemed to be sufficient based on the statistical output of G*Power (Hair et al., 2013). With the statistical significance of 5%, odds ratio of 1.3, and statistical power of 95%, a priori sample size with the statistical test of logistic regression shows 269 as the required sample. Before distributing the final version of the questionnaire, a pre-test with four experts was administered, and the necessary amendments were made based on their suggestions.

3.2. Measurements

Fifteen multiple-choice questions were used to measure the score of FTL. Nine questions (i.e., Q1 to Q7, Q14, and Q15) were adapted from the Chartered Financial Analyst (CFA) program curriculum developed by the CFA institute. These nine questions were taken from the multiple-choice practice problems given at the end of the chapter entitled “Fintech in Investment Management” authored by Mack and Kissell (2018). The multiple-choice practice problems were meant for CFA program level I. These questions seemed to be more appropriate and suitable to assess one’s level of FTL because this outlined the basics of fintech, fintech application to investment management, and financial application. Afterwards, multiple-choice options were transformed to true/false questions following Bannier et al. (2019) ‘s true/false format. Six additional questions (Q8 to Q13) were included to represent all the aspects of FTL, which have been indicated by Milian et al. (2019). The newly added questions fit with the broader scope of fintech. These six questions are also formulated in true/false format. A value of 1 was assigned for each correct answer, else 0. Referring to the financial literacy index measured by Van Rooij et al. (2011), the FTL index was measured by the sum of correct answers to 15 FTL questions.

Actual financial literacy was quantified by the seven financial literacy questions developed by OECD (Arceo-Gomez & Villagomez, 2017; Lusardi & Mitchell, 2006; OECD, 2013) with slight modifications to fit the Malaysian context. The actual financial literacy questions comprised of interest rate, diversification, inflation, and the risk-return concept. A value of 1 was given for a correct answer, else 0. Following the approach of financial literacy literature (Arceo-Gomez & Villagomez, 2017; Lusardi & Mitchell, 2006; Van Rooij et al., 2011), the actual financial literacy index was computed by taking the total score of financial literacy questions.

Perceived financial literacy evaluated an individual’s self-rated degree of financial knowledge regarding budgeting, savings, managing debt, investment, financial and retirement planning. Based on the study of Nguyen et al. (2017), which followed the items developed by the Australia & Financial Literacy Foundation in 2007, perceived financial literacy was measured. Six items quantified self-assessed financial literacy on a 5-point Likert scale anchoring 1= “very low” and 5= “very high”. The average value was used to compute the perceived financial literacy index (Nguyen et al., 2017; Van Rooij et al., 2011). Table 1 describes the definition of variables.

Table 1. Definition of variables

Variables	Items
Gender	Male, Female
Age	18–22 years, 23–27 years, 28–32 years, above 32 years
Ethnicity	Malay, Chinese, Indian
Education	Secondary, Diploma, Foundation, Degree, Master’s degree and above
Fintech Literacy	Fintech literacy is measured by the 15 fintech literacy questions. Of 15 items, 9 items (Q1-Q7, Q14 and Q15) for fintech literacy are adapted from the CFA Institute (Mack & Kissell, 2018). The remaining six items are added by following the true/false questions of Bannier et al. (2019).
	Q1 (Fintech definition). A correct description of fintech is that it: a. Is driven by rapid growth in data and related technological advances. b. Increases the need for intermediaries. c. Is at its most advanced state using systems that follow specified rules and instructions. A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q2 (Machine learning). Machine learning ensures that traditional and alternative data are fully segregated. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q3 (Machine learning). In the use of machine learning (ML), training data can be learned too precisely leading to inaccurate predictions when used with different datasets. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q4 (Big data). A characteristic of Big Data is that it involves formats with diverse types of structures. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q5 (Big data). Big Data is characterized by the three Vs- volume, velocity, and variety- and includes both traditional and non-traditional datasets. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q6 (Cryptocurrency). Cryptocurrency is a distributed ledger technology (DLT) application. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q7 (Cryptocurrency). Cryptocurrency is irreversible once the transactions made. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q8 (Blockchain). A blockchain is a centralized system. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q9 (Blockchain). Financial institutions can use blockchain to optimize their payment, clearing, and settlement processes. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q10 (Peer-to-peer lending). Borrowers on peer-to-peer lending platform indicate the amount of loan they need with their desired maximum interest rates. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q11 (Peer-to-peer lending). Peer-to-peer lending is safe from default risk because it tracks post loan information of borrowers. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q12 (Crowdfunding). In reward-based crowdfunding platform, funders aim to get positive return in monetary terms. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q13 (Crowdfunding). In equity-based crowdfunding platform, a start-up company will issue securities to get funds to meet the financial needs. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q14 (Robo-advisor). Robo-advisors are free from regulation when acting as fully-automated wealth managers. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
	Q15 (Robo-advisor). Robo-advisors offer fairly conservative advice as easily accessible guidance in providing investment services. a. True; b. False A binary variable takes a value of 1 for correct answers, 0 otherwise.
Fintech literacy index	Is measured by the sum of correct answers to the 15 fintech literacy questions,, spanning from 0 to 15 correct answers
Actual Financial Literacy	1. Suppose you had RM100 in a savings account and the interest rate was 2% per year. After 5 years, how much do you think you would have in the account if you let the money to grow? a. More than RM102; b. Exactly RM102; c. Less than RM102; d. I don’t know 2. Imagine you have a savings account that offers you an annual interest rate of one percent. At the same time, you know that Department of Statistics Malaysia has reported that annual inflation will be two percent. At the end of the year you will be able to: a. Buy more than today; b. Buy less than today; c. Buy exactly the same as today; d. I don’t know 3. Investing a certain amount of money in a single stock gives you a safer return than investing the same amount of money in various different stocks. a. True; b. False; c. I don’t know 4. High inflation means that the cost of living is increasing rapidly. a. True; b. False; c. I don’t know 5. An investment with a high return is likely to be high risk. a. True; b. False; c. I don’t know 6. You lend RM25 to a friend one evening and he gives you RM25 back the next day. How much interest he has paid on this loan? a. RM25; b. No interest; c. I don’t know 7. Suppose you put RM100 into a savings account with a guaranteed interest rate of 2% per year. You don’t make any further payments into this account and you don’t withdraw any money. How much it would be in the account at the end of five years? a. More than RM110; b. Exactly RM110; c. Less than RM110; d. It’s impossible to say based on the information given; e. I don’t know Adapted from Arceo-Gomez and Villagomez (2017); Lusardi and Mitchell (2006); OECD (2013)
Actual financial literacy index	Is measured by the sum of correct answers to the actual financial literacy questions, spanning from 0 to 7 correct answers.
Perceived Financial Literacy	Please rate your financial knowledge and understanding for the following on a 5-point scale, where 1= “very low” to 5= “very high”: 1. Budget day-to-day finances. 2. Save money. 3. Manage debt. 4. Invest money. 5. Plan for the financial future. 6. Save enough money for retirement. Adapted from Nguyen et al. (2017); Financial Literacy Foundation, (2007)
Perceived financial literacy index	Is measured by the average value of the six literacy items quantified on a 5-point scale.

3.3. Econometric model

The following logistic regression model was estimated for the predictors of FTL:

$Log\left(\frac{p}{\left(1-\mathrm{p}\right)}\right)$= α₀ + α₁(Actual financial literacy index) + α₂(Perceived financial literacy index) + α₃(Gender) + α₄(Age) + α₅(Ethnicity) + α₆(Education) + μ_i

In the above equation, P denotes the probability that an individual provides the correct answer to each FTL question, which equals 1 for the correct answer and 0 otherwise. Fifteen FTL questions are defined as binary variables; a value of 1 is assigned for each correct answer. Among explanatory variables, the actual financial literacy index is a summation of number of correct answers. Perceived financial literacy is an aggregate score of six financial literacy items measuring on a 5-point Likert scale. Gender equals one if the respondent is a male, else 0; age takes an ordinal value; ethnicity is a categorical variable where Chinese equals 1 if the respondent belongs to Chinese ethnicity, else 0; Malay equals 1 if the respondent belongs to Malay ethnicity, else 0; Indian equals 1 if the respondent is from Indian ethnicity, else 0; education is an ordinal variable; and μ is the error term.

In addition to binary logistic regression, ordered logistic regression was estimated to investigate the drivers of FTL index with the same set of explanatory variables. Further, cross-tabulation of fintech literacy was conducted across demographic variables.

4. Results

4.1. Demographic information

Table 2 reports the demographic information of survey participants. The sample of millennials shows that most respondents are male (56.4%) and 28–32 years old (41.3%). Chinese ethnic group (55.5%) represents a larger proportion of respondents, following Malay (33.3%) and Indian (11.2%). The employment status indicates that 64.5% engage in full-time employment, whereas only 35.5% are in part-time employment. The education level of millennials is fairly high, with 33.9% having bachelor’s degrees or above.

Table 2. Demographic information

	Frequency	% of sample
Gender
Male	186	56.4
Female	144	43.6
Age
18–22 years	108	32.7
23–27 years	86	26.1
28–32 years	136	41.3
Ethnicity
Malay	110	33.3
Chinese	183	55.5
Indian	37	11.2
Employment status
Full-time	213	64.5
Part-time	117	35.5
Education
Secondary (e.g., SPM)	17	5.2
Diploma	105	31.8
Foundation/Pre-university	96	29.1
Degree	99	30.0
Master’s degree and above	13	3.9

Note: N = 330

4.2. Response to FTL questions

The level of FTL is assessed by the millennials’ response to 15 fintech literacy questions. As reported in Table 3, 55.5% of individuals provide correct answers to the first question, which defines the basic description of fintech. The second and third questions are related to machine learning. For the Q2, 40.0% of individuals provide correct answer for a false statement, while only 33.0% give correct answer for a true statement on machine learning for the Q3. The next two questions are related to big data literacy. More than 50% of millennials give correct answers on characteristics of big data. Almost the same percentage (53.3%) of millennials answer correctly that the three Vs of big data are volume, velocity and variety, which include both traditional and non-traditional datasets. Afterward, a cryptocurrency literacy question is asked, where 58.5% of millennials correctly answer that cryptocurrency is a distributed ledger technology (DLT) application. A slightly higher percentage of millennials (62.1%) provide correct answers on cryptocurrency transactions. Then, Q8 asks if “a blockchain is a centralized system,” and the correct answer is provided by 50.6% of millennials. More than 50% of respondents answer correctly that financial institutions can use blockchain to optimize their payment, clearing, and settlement processes. Relating to P2P lending, Q10 asks whether borrowers indicate the amount of loan and their desired maximum interest rates on P2P platform. The response to this true statement is relatively high (66.7%). However, only 38.5% of individuals know the default risk of P2P lending. For the crowdfunding literacy question, merely 36.7% of the respondents know the function of reward-based crowdfunding. Nearly 50% of millennials know that equity crowdfunding is for start-up companies to raise funds by issuing securities. The last two questions on robo-advisor’s literacy show that 45.2% of millennials know that robo-advisors are not free from regulation; and 50.6% of respondents provide correct answer for robo-advisor’s advisory process. The mean values show that millennials answer on average 7.40 questions correctly of the total 15 FTL questions.

Table 3. Responses to fintech literacy questions

	Correct	Incorrect
[1] A correct description of fintech is that it is driven by rapid growth in data and related technological advances.	55.5%	44.5%
[2] Machine learning ensures that traditional and alternative data are fully segregated (false).	40.0%	60.0%
[3] In the use of machine learning (ML), training data can be learned too precisely leading to inaccurate predictions when used with different datasets (true).	33.0%	67.0%
[4] A characteristics of big data is that it involves formats with diverse types of structures (true).	51.8%	48.2%
[5] Big Data is characterized by the three Vs- volume, velocity, and variety- and includes both traditional and non-traditional datasets (true).	53.3%	46.7%
[6] Cryptocurrency is a distributed ledger technology (DLT) application (true).	58.5%	41.5%
[7] Cryptocurrency is irreversible once the transactions made (true).	62.1%	37.9%
[8] A blockchain is a centralized system (false).	50.6%	49.4%
[9] Financial institutions can use blockchain to optimize their payment, clearing, and settlement processes (true).	51.5%	48.5%
[10] Borrowers on peer-to-peer lending platform indicate the amount of loan they need with their desired maximum interest rates (true).	66.7%	33.3%
[11] Peer-to-peer lending is safe from default risk because it tracks post loan information of borrowers (false).	38.5%	61.5%
[12] In reward-based crowdfunding platform, funders aim to get positive return in monetary terms (false).	36.7%	63.3%
[13] In equity-based crowdfunding platform, a start-up company will issue securities to get funds to meet the financial needs (true).	46.4%	53.6%
[14] Robo-advisors free from regulation when acting as fully-automated wealth managers (false).	45.2%	54.8%
[15] Robo-advisors offer fairly conservative advice as easily accessible guidance in providing investment services (true).	50.6%	49.4%
	Mean	S.D.
Fintech literacy index	7.40	1.87

4.3. Distribution of FTL across demographics

Table 4 shows the distribution of FTL across millennials’ demographics. The results indicate that male millennials have higher FTL than female counterparts for machine learning, big data, cryptocurrency, crowdfunding, and robo-advisor. Notwithstanding, females show higher FTL regarding fintech definition, big data, cryptocurrency, blockchain, and P2P lending. Millennials aged 18–22 years possess more knowledge considering six items of FTL, which are fintech definition, machine learning, cryptocurrency, blockchain, P2P lending and robo-advisor. Millennials in the age category of 23–27 have higher knowledge only for three items, which are machine learning, cryptocurrency and P2P lending. Those who are in the age category of 28–32 have more knowledge than other age groups considering big data, blockchain, crowdfunding and robo-advisor. FTL across ethnic groups indicates that Chinese millennials possess more knowledge on fintech in most of the literacy items, including machine learning, cryptocurrency, blockchain, crowdfunding, and robo-advisor. Comparing education levels, millennials with degree or above could correctly answer most FTL items. Those who completed bachelor’s degree or above possess higher knowledge on fintech definition, machine learning, big data, cryptocurrency, blockchain, P2P lending, crowdfunding and robo-advisor. Respondents having completed their diploma and pre-university degree have relatively less knowledge on most of the FTL items.

Table 4. Distribution of fintech literacy across demographic variables

	[1]	[2]	[3]	[4]	[5]	[6]	[7]	[8]	[9]	[10]	[11]	[12]	[13]	[14]	[15]
	Correct	Correct	Correct	Correct	Correct	Correct	Correct	Correct	Correct	Correct	Correct	Correct	Correct	Correct	Correct
Gender
Male	51.6%	40.9%	33.3%	54.3%	50.0%	61.8%	56.5%	47.3%	50.0%	64.5%	37.6%	37.6%	49.5%	48.4%	54.3%
Female	60.4%	38.9%	32.6%	48.6%	57.6%	54.2%	69.4%	54.9%	53.5%	69.4%	39.6%	35.4%	42.4%	41.0%	45.8%
Age
18–22 years	58.3%	36.1%	35.2%	50.9%	50.0%	61.1%	72.2%	61.1%	51.9%	65.7%	38.9%	34.3%	45.4%	47.2%	45.4%
23–27 years	55.8%	44.2%	31.4%	47.7%	54.7%	65.1%	65.1%	41.9%	50.0%	68.6%	38.4%	31.4%	40.7%	46.5%	46.5%
28–32 years	52.9%	40.4%	32.4%	55.1%	55.1%	52.2%	52.2%	47.8%	52.2%	66.2%	38.2%	41.9%	50.7%	42.6%	57.4%
Ethnicity
Malay	46.4%	35.5%	33.6%	47.3%	55.5%	54.5%	60.0%	47.3%	50.9%	72.7%	40.9%	40.0%	42.7%	49.1%	46.4%
Indian	64.9%	37.8%	29.7%	62.2%	43.2%	56.8%	51.4%	37.8%	56.8%	48.6%	35.1%	45.9%	45.9%	37.8%	48.6%
Chinese	59.0%	43.2%	33.3%	52.5%	54.1%	61.2%	65.6%	55.2%	50.8%	66.7%	37.7%	32.8%	48.6%	44.3%	53.6%
Education
Secondary (e.g., SPM)	23.5%	11.8%	23.5%	17.6%	23.5%	23.5%	29.4%	29.4%	35.3%	29.4%	5.9%	5.9%	5.9%	17.6%	41.2%
Diploma	57.1%	35.2%	34.3%	57.1%	55.2%	58.1%	66.7%	38.1%	47.6%	66.7%	37.1%	35.2%	45.7%	43.8%	49.5%
Foundation/Pre-university	58.3%	42.7%	38.5%	49.0%	54.2%	65.6%	68.8%	61.5%	47.9%	68.8%	41.7%	37.5%	43.8%	47.9%	47.9%
Degree	51.5%	45.5%	28.3%	56.6%	56.6%	56.6%	57.6%	56.6%	61.6%	72.7%	40.4%	41.4%	56.6%	51.5%	56.6%
Master’s degree/above	92.3%	53.8%	30.8%	38.5%	46.2%	69.2%	53.8%	53.8%	53.8%	53.8%	53.8%	46.2%	46.2%	23.1%	46.2%

Notes: [1]= fintech definition, [2] [3]= machine learning, [4] [5]= big data, [6] [7]= cryptocurrency, [8] [9]= blockchain, [10] [11]= peer-to-peer lending, [12] [13]= crowdfunding, [14] [15]= robo-advisor

4.4. Regression results of FTL

Table 5 reports the regression results of FTL for fifteen (15) literacy questions. The results show that neither actual financial literacy nor perceived financial literacy is significant to predict the basic definition of fintech (Column 1). Instead, education is positively related to the fintech definition. Besides, Malay ethnicity is negatively associated with the definition of fintech. Actual financial literacy positively predicts machine learning’s literacy, which is machine learning’s ability to segregate traditional and alternative data (Column 2). The education level of millennials is positive and significantly related to this machine learning literacy. None of the variables including demography is significant for the second question of machine learning knowledge (i.e., machine learning’s use with the training data) (Column 3). Similarly, all the variables are found to be insignificant to impact the big data’s two literacy questions (Columns 4 and 5), which are characteristic of big data and three Vs of big data. Moving forward to cryptocurrency literacy, Column (6) reports that age is only significant, but negatively related to the knowledge of cryptocurrency (i.e., a distributed ledger technology application). Age is also negatively related to the second question of cryptocurrency’s literacy (Column 7), which is transaction made with cryptocurrency. This indicates that the higher millennials’ age is, the lower the tendency to possess cryptocurrency’s literacy.

Table 5. Regression results of fintech literacy

	Q1 (Fintech definition)	Q2 (Machine learning)	Q3 (Machine learning)	Q4 (Big Data)	Q5 (Big Data)	Q6 (Cryptocurrency)
	(1)	(2)	(3)	(4)	(5)	(6)
Actual financial literacy index	0.094 (0.097)	0.354*** (0.104)	0.113 (0.102)	0.052 (0.095)	−0.044 (0.096)	0.141 (0.098)
Perceived financial literacy index	0.017 (0.270)	−0.047 (0.279)	−0.072 (0.282)	0.192 (0.267)	0.132 (0.268)	0.329 (0.274)
Gender (Male)	−0.345 (0.238)	−0.002 (0.244)	0.046 (0.246)	0.186 (0.233)	−0.387 (0.235)	0.389 (0.240)
Age	−0.092 (0.146)	0.150 (0.151)	−0.005 (0.151)	0.057 (0.143)	0.120 (0.144)	−0.249* (0.148)
Malay ethnicity	−0.801** (0.403)	−0.028 (0.411)	0.215 (0.416)	−0.610 (0.393)	0.491 (0.388)	−0.139 (0.393)
Chinese ethnicity	−0.447 (0.393)	−0.022 (0.392)	0.151 (0.401)	−0.463 (0.380)	0.393 (0.373)	−0.015 (0.380)
Education	0.205* (0.120)	0.259** (0.125)	−0.060 (0.125)	0.073 (0.118)	0.104 (0.119)	0.180 (0.120)
Pseudo R square	0.051	0.082	0.008	0.021	0.025	0.056
	Q7 (Cryptocurrency)	Q8 (Blockchain)	Q9 (Blockchain)	Q10 (P2P lending)	Q11 (P2P lending)	Q12 (Crowdfunding)
	(7)	(8)	(9)	(10)	(11)	(12)
Actual financial literacy index	0.033 (0.100)	0.069 (0.097)	0.040 (0.096)	0.094 (0.102)	0.056 (0.098)	0.271*** (0.103)
Perceived financial literacy index	0.135 (0.276)	−0.083 (0.272)	0.377 (0.269)	−0.342 (0.297)	−0.047 (0.273)	−0.069 (0.280)
Gender (Male)	−0.421* (0.245)	−0.230 (0.238)	−0.217 (0.235)	−0.262 (0.251)	−0.117 (0.239)	−0.066 (0.247)
Age	−0.381** (0.151)	−0.300** (0.147)	−0.028 (0.144)	0.046 (0.153)	−0.051 (0.148)	0.214 (0.154)
Malay ethnicity	0.313 (0.395)	0.345 (0.404)	−0.291 (0.391)	1.070*** (0.400)	0.225 (0.402)	−0.193 (0.397)
	Q7 (Cryptocurrency)	Q8 (Blockchain)	Q9 (Blockchain)	Q10 (P2P lending)	Q11 (P2P lending)	Q12 (Crowdfunding)
	(7)	(8)	(9)	(10)	(11)	(12)
Chinese ethnicity	0.473 (0.382)	0.482 (0.387)	−0.465 (0.380)	0.645* (0.377)	−0.049 (0.389)	−0.804** (0.388)
Education	0.049 (0.121)	0.390*** (0.124)	0.285** (0.120)	0.213* (0.126)	0.274** (0.125)	0.265** (0.127)
Pseudo R square	0.066	0.079	0.037	0.054	0.025	0.071
	Q13 (Crowdfunding)	Q14 (Robo-advisor)	Q15 (Robo-advisor)	Fintech literacy index
	(13)	(14)	(15)	(16)
Actual financial literacy index	−0.059 (0.096)	0.000 (0.096)	−0.070 (0.096)	0.175*** (0.082)
Perceived financial literacy index	0.408 (0.273)	−0.626** (0.274)	0.075 (0.267)	0.025 (0.229)
Gender (Male)	0.240 (0.236)	0.402* (0.237)	0.285 (0.233)	−0.027 (0.200)
Age	−0.018 (0.145)	−0.192 (0.146)	0.174 (0.143)	−0.059 (0.123)
Malay ethnicity	−0.212 (0.390)	0.435 (0.397)	−0.098 (0.387)	0.036 (0.331)
Chinese ethnicity	−0.036 (0.375)	0.257 (0.384)	0.248 (0.373)	0.023 (0.319)
Education	0.298** (0.121)	0.183 (0.121)	0.049 (0.118)	0.344*** (0.101)
Pseudo R square	0.046	0.048	0.029	0.161

Notes: Standard errors are reported in brackets below the coefficients. *10%, **5%, ***1

Concerning blockchain literacy, education is found to be significant and positively related to the knowledge of blockchain’s decentralized system, while age is appeared to be negatively related to this knowledge of blockchain (Column 8). For the second literacy question of blockchain, only education predicts blockchain’s literacy related to the use of blockchain to optimize payment, clearing and settlement processes (Column 9). Regarding P2P lending literacy, significant effects are observed for Malay ethnicity, Chinese ethnicity and education for the first question of P2P lending literacy (Column 10). Both ethnicity and education are positively associated with the P2P lending knowledge on borrower’s amount of loan and desired interest rates. Education continues to be positive and significant for the second question of P2P lending literacy, which is default risk of P2P platform (Column 11). Column (12) reports that actual financial literacy positively predicts reward-based crowdfunding literacy. While education has a significant positive impact on reward-based crowdfunding, Chinese ethnicity is negatively associated with this crowdfunding’s literacy. Education level of millennials predicts equity-based crowdfunding literacy (Column 13). Turning to robo-advisor literacy, while perceived financial literacy is negatively related to robo-advisor’s literacy, gender is positively related to robo-advisor’s literacy regarding regulatory issue (Column 14). Robo-advisor’s fairly conservative service is unexplained by any of the predictors of this study (Column 15).

4.5. Regression results of FTL index

After computing FTL index, the study further investigates whether the FTL index can be predicted by the actual financial literacy index and perceived financial literacy index. As reported in Column (16) of Table 5, a positive relationship is observed between actual financial literacy index and FTL index. A one-unit increase in actual financial literacy increases the likelihood of FTL by 0.175 points, assuming other effects remain constant. A significant impact is also found for education, where education is positive and significantly related to the FTL index. Education increases the probability of ascending FTL.

5. Discussion

Millennials’ responses to the 15 FTL questions show that they have medium level of FTL. In particular, millennials score higher on P2P lending literacy, whereas they score lower on machine learning. This might be because P2P lending’s definition indicates lending and borrowing between two parties in digital platforms. The function of P2P is similar to the bank lending system. On the other hand, machine learning concept is slightly technical, and its application may not be widely known in financial services. This may cause less knowledge on machine learning. These findings add to the existing studies, which reported individual’s cryptocurrency literacy (Bannier et al., 2019; Henry et al., 2019; Khan, 2023; Steinmetz et al., 2021).

Millennials high in FTL represent a distinct demographic profile. Male millennials aged 18–22 years, belong to the Chinese ethnicity, and hold a bachelor’s degree or above exhibit a higher literacy in most FTL questions. The male possesses higher literacy regarding machine learning, big data, cryptocurrency, crowdfunding, and robo-advisors. Young millennials (i.e., 18–22 years old) are more familiar with fintech definition, machine learning, cryptocurrency, blockchain, P2P lending, and robo-advisors. Because they may view these fintech products and services as good value for investments. Chinese ethnic groups show higher FTL literacy than other ethnic groups. Millennials’ education level, particularly degree or above, might assist them to provide correct answers for the definition of fintech, machine learning, big data, cryptocurrency, blockchain, peer-to-peer lending, crowdfunding and robo-advisors.

The results show that actual financial literacy is likely to increase knowledge on machine learning and crowdfunding. Those who have basic financial knowledge can easily learn and understand the concepts related to machine learning and crowdfunding. This finding extends prior literature on financial literacy and bitcoin literacy (Bannier et al., 2019; Jünger & Mietzner, 2020). On the other hand, perceived financial literacy explains only the relationship between perceived financial literacy and robo-advisors. This might be due to the perception that those who perceive to hold higher financial knowledge are less likely to acquire fintech knowledge. Among demographics, millennials’ education level is observed to positively predict most of the FTL questions. Education tends to increase knowledge of basic definition of fintech, machine learning, blockchain, P2P lending, and crowdfunding. The higher the education level, the more the possibility of gaining knowledge on fintech. Individuals higher in age have less knowledge of cryptocurrency and blockchain. While male exhibits lower knowledge on cryptocurrency, they possess higher knowledge on robo-advisors. In addition, both Malay and Chinese ethnic groups know more about P2P lending, but they have less knowledge about the basic definition of fintech and crowdfunding. Overall, the effects of actual financial literacy and education remain positive and significant when estimating the predictors of FTL index. These findings add to the literature on financial literacy and FTL.

6. Implication

As millennials in Malaysia show a medium level of FTL, they are suggested to improve some of the concepts of fintech, for example, machine learning, crowdfunding, blockchain, and robo-advisor in order to be benefited using fintech products and services. The cross-tabulation implies that a specific group of individuals needs to increase their understanding of fintech terminologies. One way might be through improving their financial literacy because financial literacy increases FTL. Millennials are encouraged to attend seminars or workshops related to fintech and financial literacy. Furthermore, focusing on education level improves FTL. Education may play a key role in understanding and acquiring knowledge on various areas of fintech. Having higher FTL, the upside potential is huge when it comes for investment, diversification, and financial planning by relevant fintech products, which are less costly and faster compared to other financial products.

The findings provide implications for financial institutions, companies and start-ups which offer fintech products related to lending, borrowing, payment, financial planning and investment. FTL of millennials provides better insight for regulators to design policies for fintech inclusion as fintech evolves to serve underserved and unbanked ones. In order for greater acceptance, fintech service providers and companies may initiate a campaign on FTL to a specific demographic profile with low FTL. Such type of campaign may increase millennials’ confidence and create self-protection against fraud. Additionally, improving FTL will likely increase demand for more fintech products for payment, investments, borrowing and saving. Furthermore, fintech companies may develop a financial app containing information on FTL to educate millennials and specific demographic groups.

As financial technology enabler group (FTEG) and Bank Negara Malaysia (BNM) are responsible for promoting fintech products and services in Malaysia, the outcome of this study may help them in achieving their goal.

7. Conclusion and future research

FTL is considered as a pre-requisite for the adoption and usage of fintech products and services. This study, therefore, investigates FTL of millennials as they play a key role in fintech’s core market and transforming fintech industry. This study assesses millennials’ response to 15 FTL questions and then investigates if actual financial literacy, perceived financial literacy, and demographic variables predict FTL in Malaysia. The results show that millennials have medium level of FTL, and they possess higher literacy on P2P lending, but lower on machine learning. Male, younger, Chinese ethnic group and highly educated millennials show relatively higher FTL. Millennials’ actual financial literacy, education and age contribute significantly to determining FTL. Actual financial literacy improves literacy on machine learning and crowdfunding. While education increases literacy of basic fintech definition, machine learning, blockchain, P2P lending and crowdfunding, age is negatively associated with cryptocurrency and blockchain literacy. Malay and Chinese ethnicity are associated with P2P lending and crowdfunding literacy. While male has lower literacy on cryptocurrency, they have higher literacy on robo-advisors.

Future research may be conducted using individual financial literacy items instead of index in predicting fintech literacy. Also, extending the context of this study to other developing countries can be a good avenue for future research. Gen Z or older respondents’ FTL may be assessed in future studies.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/23311886.2023.2281046

Additional information

Notes on contributors

Mohammad Tariqul Islam Khan

Mohammad Tariqul Islam Khan is a senior lecturer at the Faculty of Business, Multimedia University, Malaysia. His research interests are financial technology, digital financial literacy, digital finance, household finance and behavioral finance.

Tze Wei Liew

Tze Wei Liew is a senior lecturer at the Faculty of Business, Multimedia University, Malaysia. His research domain includes cognitive psychology, learning science, media psychology, cyberpsychology, human-computer interaction and human-agent interaction.

Xue Ying Lee

Xue Ying Lee graduated from the Faculty of Business, Multimedia University, Malaysia. Her research interest lies in financial literacy and fintech.