Does Hyperbolic Discounting Mediate the Association Between Financial Literacy and Investment in Risky Assets?

Abstract

Investment in risky financial assets plays a crucial role in individual wealth accumulation and broader financial market development. However, existing research has primarily emphasized financial literacy while giving limited attention to behavioral mechanisms that may weaken its influence on investment behavior. In particular, hyperbolic discounting, reflecting time-inconsistent preferences that favor immediate rewards over long-term gains, may constrain the effective translation of financial knowledge into forward-looking financial decisions. Against this background, this study examines whether hyperbolic discounting mediates the association between financial literacy and investment in risky assets using large-scale survey data from Japan’s Money and Life survey. Employing regression-based mediation analysis within a cross-sectional framework, the results indicate that financial literacy is strongly and positively associated with risky asset investment, while hyperbolic discounting exerts a statistically significant but economically small mediating effect that slightly attenuates this relationship. The findings suggest that cognitive financial capability remains the dominant driver of participation in risky financial markets, whereas present-biased preferences play a secondary behavioral role. These results provide important implications for investors, educators, and policymakers by highlighting that policies aimed at improving financial literacy are likely to yield substantial investment benefits, while complementary interventions addressing behavioral biases may offer additional, though more modest, gains in promoting long-term, forward-looking financial decision-making.

1. Introduction

Investment in financial assets is essential for both individual financial growth and broader economic stability. Financial literacy serves as a fundamental determinant of individuals’ ability to navigate complex financial markets, particularly when engaging with equities and high-risk securities (Khan et al., 2020, 2021; Yamori & Ueyama, 2022). Previous studies underscore that financial literacy increases understanding of investment fundamentals, fosters confidence, and mitigates cognitive biases such as overconfidence and herding, which are vital for effective portfolio management (Oehler et al., 2024; Xiao et al., 2022; Cupák et al., 2022). For example, financial literacy develops individuals’ knowledge of core financial concepts such as interest compounding, inflation, and risk diversification, which underpin informed intertemporal financial decision-making (Lusardi & Mitchell, 2008). However, investment decisions extend beyond the domains of financial literacy. Behavioral tendencies, particularly hyperbolic discounting, introduce a layer of complexity by affecting the ability of individuals to prioritize long-term financial goals over immediate gratification (Frigerio et al., 2020; Laibson, 1997; Ainslie, 1991). Hyperbolic discounting, characterized by a preference for immediate rewards over delayed benefits, creates temporal inconsistencies that influence investment behavior (Bawalle et al., 2024; Ventre et al., 2024). Unlike exponential discounting, hyperbolic discounting captures dynamic inconsistency between short-term preferences and long-term intentions, making it especially relevant for investment behavior. Although financial literacy equips individuals with the ability to evaluate long-term gains, hyperbolic discounting can lead to decisions that undermine future financial stability (Katauke et al., 2023). This interaction suggests that hyperbolic discounting may mediate the relationship between financial literacy and investment behavior, which could potentially weaken the direct influence of financial literacy. Ignoring this behavioral channel may therefore lead to incomplete conclusions regarding the effectiveness of financial education and investment guidance. Despite substantial research on the role of financial literacy in fostering prudent investment behavior, the mediating effect of hyperbolic discounting remains understudied. Exploring this dynamic is essential to refine financial education and investment strategies to holistically address both cognitive and behavioral factors.

Previous research demonstrates that financial literacy influences a broad range of household financial and economic behaviors, including participation in financial market securities (Lusardi & Mitchell, 2014). Seminal evidence by van Rooij et al. (2011) shows that financially literate individuals are significantly more likely to participate in stock markets even after controlling for wealth, income, and education. Subsequent empirical studies consistently confirm that financial literacy enhances engagement with risky financial assets and improves investment decision quality across institutional contexts (Khan et al., 2020, 2021; Adil et al., 2022; Arora & Chakraborty, 2023; Oehler et al., 2024; Xiao et al., 2022). This relationship operates through several interrelated channels. Financial literacy strengthens understanding of diversification, risk tolerance, and market functioning, thereby increasing confidence in portfolio management while reducing susceptibility to behavioral biases such as overconfidence, herding, and the disposition effect (Cupák et al., 2022; Barthel & Lei, 2021; Wang & Zou, 2024; Lotto, 2020; Ullah et al., 2024). It also improves risk perception, supports informed participation in financial markets, and enhances protection against fraudulent investment schemes (Mouna & Anis, 2017; Raut, 2020; Ashfaq et al., 2024; Özen & Ersoy, 2019; Mohd Padil et al., 2022; Kasim et al., 2023). Overall, the empirical literature converges on the conclusion that financial literacy promotes participation in risky financial markets, contributes to improved portfolio outcomes, and supports broader financial stability and inclusion (Zhao & Zhang, 2021; Rodrigues & Gopalakrishna, 2023; Bucher-Koenen et al., 2023).

Over the past few decades, hyperbolic discounting has emerged as a critical behavioral phenomenon influencing household financial and economic decisions (Frigerio et al., 2020; Laibson, 1997). This intertemporal bias has been linked to lower savings, weaker portfolio diversification, and a tendency to favor high-risk financial opportunities promising rapid returns, ultimately contributing to suboptimal long-term financial outcomes (Kuramoto et al., 2024; Love & Phelan, 2015; O’Donoghue & Rabin, 1999). These patterns are particularly evident among individuals with lower financial literacy, who may lack the capacity to evaluate the long-term benefits of prudent investment strategies (Bawalle et al., 2024; Ventre et al., 2024). Empirical evidence further suggests that present-biased preferences reduce the likelihood that individuals act on financial knowledge or engage with educational programs designed to improve decision-making (Meier & Sprenger, 2013). Related research shows that hyperbolic discounting can attenuate the beneficial effects of financial literacy in other financial domains, such as debt management (Ottaviani & Vandone, 2018). Moreover, recent findings indicate a significant association between financial literacy and inconsistent intertemporal decision-making, highlighting the need to examine whether similar behavioral mechanisms influence investment in risky financial assets (Ventre et al., 2024).

While behavioral biases are widely recognized as important determinants of financial decision-making, their quantitative importance relative to financial literacy remains insufficiently understood. In particular, it is unclear whether behavioral time inconsistency meaningfully constrains the effectiveness of financial knowledge in shaping risky investment behavior. This study addresses this gap by examining whether hyperbolic discounting mediates the relationship between financial literacy and investment in risky financial assets. Accordingly, the analysis evaluates three related questions: whether financial literacy reduces the likelihood of hyperbolic discounting, how hyperbolic discounting relates to risky asset investment, and whether it mediates the association between financial literacy and portfolio risk exposure. Grounded in intertemporal choice theory and behavioral finance, the proposed framework conceptualizes forward-looking investment behavior as the joint outcome of cognitive financial capability and behavioral self-control. The central hypothesis posits that hyperbolic discounting mediates the relationship between financial literacy and investment in risky financial assets, thereby attenuating the direct influence of financial knowledge through present-biased, time-inconsistent preferences. The novelty of this study lies in explicitly modeling hyperbolic discounting as a mediating behavioral mechanism rather than treating financial literacy and behavioral bias as independent determinants and in empirically testing this integrated pathway using large-scale micro-level data on active investors. Accordingly, the study contributes to the literature in three ways: empirically, by quantifying the direct and indirect channels linking financial literacy, present bias, and risky asset allocation; theoretically, by integrating financial literacy research with dynamic intertemporal preference frameworks in a unified mediation structure; and practically, by providing evidence to inform financial education policies and behavioral interventions aimed at promoting sustainable long-term investment behavior. More broadly, the findings clarify how cognitive capability and behavioral time inconsistency jointly shape participation in risky financial markets and long-term household financial decision-making.

The remainder of this paper is structured as follows. Section 2 reviews the relevant literature, Section 3 describes the data and methodology, Section 4 presents the empirical results, Section 5 discusses the findings, and Section 6 concludes.

2. Literature Review

2.1. Theoretical Foundations Linking Financial Literacy, Behavioral Biases, and Risky Asset Investment

Classical investment theory conceptualizes portfolio choice as a trade-off between risk and return, where asset prices and allocations are determined by rational expectations and equilibrium pricing mechanisms. Within this framework, the Efficient Market Hypothesis (EMH) posits that financial markets fully incorporate available information, implying that individual behavioral deviations do not systematically affect equilibrium outcomes (Fama, 1965, 1970; Almansour et al., 2025). Consequently, traditional asset-pricing models such as the Capital Asset Pricing Model (CAPM) assume rational investors, homogeneous expectations, and optimization based solely on risk–return considerations.

However, a substantial body of behavioral finance research challenges these assumptions by demonstrating that systematic psychological biases shape information processing, preferences, and investment behavior. Behavioral finance emphasizes cognitive and emotional distortions, including overconfidence, loss aversion, mental accounting, and herding, that can generate persistent deviations from rational market predictions and influence portfolio allocation decisions (Malik et al., 2025; Lo, 2005; Kahneman & Tversky, 1979). These insights motivate alternative perspectives such as the Adaptive Markets Hypothesis, which reconciles elements of market efficiency with evolving behavioral dynamics (Lo, 2005).

Among behavioral mechanisms, hyperbolic discounting represents a central intertemporal bias characterized by present-biased preferences and declining discount rates over time (Laibson, 1997). Theoretical life-cycle portfolio models show that hyperbolic discounting can reduce desired saving, delay participation in risky asset markets, alter portfolio shares across the life cycle, and weaken incentives to accumulate financial knowledge (Love & Phelan, 2015). Experimental and cognitive research further indicates that subjective perceptions of time and emotional valuation contribute to temporal inconsistency, providing psychological micro-foundations for hyperbolic discounting behavior (Malik et al., 2025). These mechanisms imply that present bias may directly and indirectly influence risky asset investment through both saving behavior and financial knowledge acquisition.

Financial literacy theory complements behavioral finance by framing financial capability as a combination of knowledge, skills, and attitudes that improve decision quality, shape perceived risk, and facilitate participation in higher-return financial assets (Sawitri & Candraningrat, 2025; Kuramoto et al., 2025; Chu et al., 2017). Foundational contributions emphasize that financially literate individuals are more likely to engage with complex financial products, diversify portfolios, and participate in equity markets (Lusardi & Mitchell, 2014; van Rooij et al., 2011). At the same time, behavioral insights suggest that literacy alone may not ensure optimal decisions when present-biased preferences distort intertemporal evaluation. This interaction motivates behavioral reinterpretations of traditional risk-return models and asset-pricing frameworks, where psychological factors coexist with classical equilibrium structures (Malkiel et al., 2005; Palanichamy et al., 2024; Shefrin, 2005).

Integrating these strands, theoretical models indicate a joint pathway in which hyperbolic discounting reduces saving and financial knowledge accumulation while financial literacy shapes perceived risk and participation, implying a mediating mechanism through which present bias and literacy together determine risky asset investment outcomes (Kuramoto et al., 2025; Love & Phelan, 2015). This theoretical synthesis provides the conceptual foundation for examining hyperbolic discounting as a mediator in the relationship between financial literacy and risky asset allocation.

2.2. Empirical Evidence on Financial Literacy, Behavioral Biases, and Hyperbolic Discounting

The substantial empirical literature documents a positive association between financial literacy and participation in risky financial assets across diverse institutional contexts. Early household-level evidence demonstrates that individuals with higher financial literacy are significantly more likely to participate in stock markets and hold diversified portfolios even after controlling for socioeconomic characteristics (van Rooij et al., 2011; Lusardi & Mitchell, 2014). Subsequent cross-country and micro-survey studies reinforce this conclusion, showing that financial literacy enhances investment participation, improves portfolio allocation, and supports long-term financial well-being (Khan et al., 2020, 2021).

More recent empirical analyses provide deeper insight into the mechanisms underlying this relationship. Household survey evidence from China shows that higher financial literacy, particularly advanced literacy, is associated with mutual fund ownership, delegation to financial experts, and improved investment returns (Chu et al., 2017). Panel evidence from Germany further indicates that financial literacy contributes positively to risky asset participation, partly through indirect channels operating via reduced risk aversion, a finding consistent with broader evidence linking literacy, confidence, and investment engagement (Oehler et al., 2024). Studies using investor-level survey data in emerging economies similarly demonstrate that financial literacy improves both investment decision quality and financial risk tolerance, with risk tolerance mediating the literacy–investment relationship (Shahid, 2025; Ahmed et al., 2021).

Beyond literacy, behavioral biases play a systematic role in shaping investment outcomes. Empirical research shows that overconfidence, herding, disposition effects, and risk aversion are significantly associated with investment choices, while financial literacy can moderate the influence of these biases on decision-making (Ahmed et al., 2021; Palanichamy et al., 2024). Importantly, objective financial knowledge and subjective confidence may diverge; overconfidence in financial literacy is linked to lower perceived risk aversion and greater engagement in risky financial assets (Shefrin, 2005). These findings collectively highlight the interaction between cognitive capability and behavioral distortions in determining portfolio behavior.

Empirical evidence specifically addressing hyperbolic discounting remains comparatively limited. Theoretical and calibrated life-cycle analyses demonstrate that present-biased preferences reduce saving, delay stock market entry, and lower wealth accumulation while also discouraging investment in financial knowledge, thereby producing inefficient portfolio outcomes (Love & Phelan, 2015). Experimental and behavioral studies further confirm that subjective time perception contributes to temporal inconsistency in intertemporal choice, offering a psychological explanation for hyperbolic discounting patterns (Malik et al., 2025). Country-specific empirical evidence, including studies of Japanese investors, suggests that present bias influences investment attitudes such as loss tolerance, although sample coverage and generalizability remain heterogeneous (Kuramoto et al., 2025).

Despite broad empirical support for the roles of financial literacy and behavioral biases in shaping investment behavior, an important gap remains in the literature. Existing empirical studies largely examine financial literacy, behavioral biases, and investment outcomes in isolation or through partial relationships, without directly integrating these elements within a unified empirical framework. In particular, there is limited empirical evidence explicitly linking hyperbolic discounting, financial literacy, and risky asset participation within a mediation structure, even though theoretical models suggest such an interconnection. Consequently, current evidence remains incomplete regarding whether present-biased preferences systematically attenuate the influence of financial literacy on investment in risky assets. These gaps motivate the present study, which empirically examines hyperbolic discounting as a mediating behavioral mechanism connecting financial literacy to risky asset investment. By integrating cognitive financial capability with intertemporal behavioral bias within a unified empirical framework, this research contributes to clarifying the joint determinants of household participation in risky financial markets.

3. Data and Methods

3.1. Data

The data for this study were collected from the Money and Life survey, a collaborative effort between Rakuten Securities, a leading online securities company in Japan, and Hiroshima University. The survey targeted individuals aged 18 and older who had logged into their accounts at least once in the past year. Conducted between November and December 2023, the survey utilized an online format, with emails containing the questionnaire sent to all Rakuten Securities account holders. The survey was designed to capture detailed information on respondents’ demographic, socioeconomic, and psychological characteristics. To ensure the reliability and accuracy of the data, respondents with missing or incomplete information were excluded. The final dataset comprised responses from 108,682 individuals. This dataset provides a valuable basis for exploring whether hyperbolic discounting mediates the association between financial literacy and investment in financial securities, shedding light on key psychological and behavioral mechanisms in investment decision-making. Given the survey’s linkage to an active securities platform in Japan, the sample is particularly suitable for examining household portfolio allocation and risky asset participation within the Japanese financial market context.

3.2. Variables

Investment in risky assets, defined as the proportion of total financial holdings invested in risk-bearing instruments, serves as the dependent variable. Prior studies have employed varying definitions of risky assets depending on the specific research objective. For instance, research focusing on stock market participation typically classifies equities, derivatives, and bonds as risky assets, whereas studies examining broader investment in risky financial instruments extend this definition to include additional asset classes such as foreign currencies and other market-exposed securities. Consistent with the objective of the present study namely, to examine investment in financial securities, risky assets are defined to include mutual funds, listed stocks, derivatives (futures and options), corporate bonds, foreign currency deposits, foreign bonds, precious metals, and crypto assets (Yamori & Ueyama, 2022; Cupák et al., 2022; Khan et al., 2021). These instruments represent market-exposed financial products whose returns are subject to price volatility and capital risk, making them appropriate proxies for risky financial investment within the Japanese institutional setting. Measuring the dependent variable as a portfolio share, rather than a binary participation indicator, allows for a more precise assessment of the intensity of risky asset exposure.

The independent variable is financial literacy, assessed using the widely recognized three-question measure of Lusardi and Mitchell’s (2008). These questions assess fundamental financial literacy, including concepts of interest rates, inflation, and risk diversification. Specifically, respondents are asked: (i) whether money in a savings account grows faster than prices when the interest rate exceeds inflation (interest compounding); (ii) whether purchasing power declines when inflation exceeds nominal returns (inflation); and (iii) whether holding a single stock is riskier than holding a diversified mutual fund (risk diversification). This measure has been extensively validated in international household finance research and is commonly used to capture core cognitive financial capability relevant to investment decision-making (Lusardi & Mitchell, 2008).

The mediator variable is hyperbolic discounting, which reflects the tendency to disproportionately favor immediate rewards over future ones, even when the future rewards are significantly larger. Following the methodologies of Katauke et al. (2023) and Bawalle et al. (2024), hyperbolic discounting was measured using two questions.

Question 1 offered monetary choices between two and nine days.

Question 2 provided monetary choices between 90 and 97 days.

For each question, the respondents selected between 8 options A and 8 corresponding options B, where the hypothetical monetary rewards increased incrementally with each option. Hyperbolic discounting was determined by calculating discount rates at the switching point where respondents transitioned from option A to option B. The discount rates for the two questions (DR1 and DR2) were calculated based on the interest rates at these switching points. This switching-point approach is widely used in experimental and survey-based elicitation of time preferences and allows identification of present-biased (time-inconsistent) behavior rather than general impatience (Ikeda et al., 2016; Katauke et al., 2023).

Using a binary classification approach, individuals were identified as hyperbolic discounters (assigned a value of 1) if DR1 > DR2, indicating a stronger preference for immediate rewards. Non-hyperbolic discounters were assigned a value of 0. This binary operationalization is consistent with the theoretical definition of hyperbolic discounting as dynamic inconsistency and is therefore appropriate for mediation analysis focused on present bias rather than continuous impatience. This approach aligns with the methodologies employed by Fukuda et al. (2023) and Katauke et al. (2023) in previous studies. Although hyperbolic discounting is operationalized as a binary indicator, this measure is designed to capture time inconsistency rather than general impatience, following established approaches in the literature.

To ensure robust analysis, we controlled for a range of socioeconomic and behavioral factors, allowing us to isolate the association between financial literacy, hyperbolic discounting, and investment behavior. The inclusion of control variables follows the established household finance and behavioral investment literature, where demographic characteristics (gender, age, marital status, children), socioeconomic status (education, employment, income, and financial assets), and psychological traits (risk aversion) are consistently shown to influence portfolio choice and risky asset participation. Controlling for these factors reduces omitted-variable bias and improves identification of the relationships among financial literacy, hyperbolic discounting, and investment behavior (Yamori & Ueyama, 2022; Cupák et al., 2022; Khan et al., 2021). Comprehensive descriptions of all variables included in the analysis are provided in Table 1.

Table 1. Variable definitions.

Variables	Definition
Investment in risky assets	The continuous variable represents the percentage of total investment allocated to risky financial assets, including mutual funds, stocks, futures/options, corporate bonds, foreign currency deposits, foreign bonds, gold/silver/platinum, and cryptocurrencies.
Hyperbolic discounting	The binary variable equals 1 if DR1 > DR2, and 0 otherwise.
Financial literacy	Average score in the three financial knowledge questions
Gender	Binary variable, 1 = Male and 0 = Female
Age	Age of the respondent
Married	The binary variable equals 1 if the respondent is married and 0 otherwise
Divorce	Binary variable equals 1 if respondent is divorced and 0 otherwise
Children	The binary variable equals 1 if the respondent has at least one child and 0 otherwise
University education	Binary variable equals 1 if respondent has at least a university degree and 0 otherwise
Employment	The binary variable equals 1 if the respondent is employed and 0 otherwise.
Household income	Natural log of household annual gross income in 2023
Household asset	Natural log of household net financial asset
Risk aversion	A continuous measure of the risk aversion levels of the respondents extracted from the question: Usually when you go out, how high must the probability of rain be before you take an umbrella?

3.3. Descriptive Statistics

Table 2 reports descriptive statistics for the main study variables. The average share of investment allocated to risky financial assets is 41.94% (SD = 27.29), spanning the full range from 0% to 100%, which reflects considerable variation in portfolio risk exposure among participants. The mediator, hyperbolic discounting, has a mean of 0.111, suggesting that 11.1% of respondents exhibit hyperbolic discounting tendencies, with values ranging from 0 (non-hyperbolic discounters) to 1 (hyperbolic discounters). The financial literacy score averages 0.743, with a standard deviation of 0.220, and spans between 0 and 1, reflecting varying levels of financial knowledge. Among the demographic variables, gender indicates that 60.8% of respondents are male, while age averages 43.86 years, ranging from 18 to 90. The squared age variable (AgeS) reflects an average value of 2060.80. Marital status reveals that 67.98% of participants are married, while 6.19% are divorced. About 45.23% have children. Regarding education, 69.0% hold at least a university degree. Employment shows that 94.97% of respondents are employed. Household variables include the natural log of income (mean = 1.583) and assets (mean = 1.281). Finally, risk aversion has a mean of 0.536, with values ranging between 0 and 1, indicating differing attitudes toward risk among participants.

Table 2. Descriptive Statistics.

Variable	Obs	Mean	Std. Dev.	Min	Max
Investment in financial assets	108,682	41.9433	27.2870	0	100
Hyperbolic discounting	108,682	0.1114	0.3147	0	1
Financial literacy	108,682	0.7427	0.2203	0	1
Gender	108,682	0.6076	0.4883	0	1
Age	108,682	43.8578	11.7174	18	90
AgeS	108,682	2060.8020	1083.7770	324	8100
Married	108,682	0.6798	0.4666	0	1
Divorce	108,682	0.0619	0.2410	0	1
Children	108,682	0.4523	0.4977	0	1
University education	108,682	0.6900	0.4625	0	1
Employment	108,682	0.9497	0.2186	0	1
lHH_Income	108,682	1.5833	0.4227	0	2.484907
lHH_Asset	108,682	1.2810	0.7237	0	2.302585
Risk_Aversion	108,682	0.5355	0.2280	0	1

3.4. Methods

Mediation analysis requires examining the interconnections among the variables under study. Previous studies have established that financial literacy positively influences investment in financial market securities (Khan et al., 2020, 2021). Conversely, hyperbolic discounting has been shown to negatively impact investment in financial market securities (Love & Phelan, 2015) and financial literacy (Bawalle et al., 2024; Katauke et al., 2023). This raises a critical question: does accounting for impulsivity, as represented by hyperbolic discounting, diminish or mediate the observed relationship between financial literacy (the independent variable) and investment in financial securities (the dependent variable)?

Accordingly, this study employs a mediation analysis framework to decompose the association between financial literacy and risky asset investment into direct and indirect pathways operating through hyperbolic discounting. Mediation models are widely used in behavioral finance and household finance research to evaluate whether psychological mechanisms transmit the influence of cognitive factors to financial outcomes, making this approach appropriate for the present research objective (Sajid et al., 2024; Curry, 2025). The mediation model was implemented in STATA using linear regression-based estimation of mediator and outcome equations. This regression-based causal-steps framework provides transparent interpretation of pathway coefficients and enables decomposition of total effects into direct and indirect components, which is particularly useful for testing theoretically motivated behavioral transmission mechanisms (Sajid et al., 2024; Jung, 2021; Rijnhart et al., 2017). The analysis involved the following steps:

Step 1: Estimation of the relationship between financial literacy and hyperbolic discounting (mediator model). This mediator model explains how financial literacy influences hyperbolic discounting while controlling for covariates. The following model was used to estimate the mediator model:

$$\begin{array}{l}\mathrm{H}\mathrm{y}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{l}\mathrm{i}\mathrm{c} \mathrm{d}\mathrm{i}\mathrm{s}\mathrm{c}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g}=\mathsf{\alpha }+{\mathsf{\beta }}_1{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{i}\mathrm{a}\mathrm{l}\mathrm{l} \mathrm{i}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{y}}_{\mathrm{i}}+{\mathsf{\gamma }}_1{\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{r}}_{\mathrm{i}}+{\mathsf{\gamma }}_2{\mathrm{M}\mathrm{a}\mathrm{r}\mathrm{r}\mathrm{i}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{i}}+{\mathsf{\gamma }}_3{\mathrm{A}\mathrm{g}\mathrm{e}}_{\mathrm{i}}+{\mathsf{\gamma }}_4{\mathrm{E}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{y}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}}_{\mathrm{i}}+\\ {\mathsf{\gamma }}_5{\mathrm{D}\mathrm{i}\mathrm{v}\mathrm{o}\mathrm{r}\mathrm{c}\mathrm{e}}_{\mathrm{i}}+{\mathsf{\gamma }}_6{\mathrm{C}\mathrm{h}\mathrm{i}\mathrm{l}\mathrm{d}\mathrm{r}\mathrm{e}\mathrm{n}}_{\mathrm{i}}+{\mathsf{\gamma }}_7{\mathrm{E}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}}_{\mathrm{i}}+{\mathsf{\gamma }}_8{\mathrm{H}\mathrm{o}\mathrm{u}\mathrm{s}\mathrm{e}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d} \mathrm{I}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{e}}_{\mathrm{i}}+{\mathsf{\gamma }}_9{\mathrm{H}\mathrm{o}\mathrm{u}\mathrm{s}\mathrm{e}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d} \mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}}_{\mathrm{i}}+\\ {\mathsf{\gamma }}_{10}{\mathrm{R}\mathrm{i}\mathrm{s}\mathrm{k} \mathrm{A}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}_{\mathrm{i}}+{\mathsf{\epsilon }}_1\end{array}$$

(1)

Step 2: Estimation of the relationship between financial literacy, hyperbolic discounting, and investment in risky assets. This outcome model explains how financial literacy and hyperbolic discounting influence investment in risky assets while controlling for covariates. The following model was used to estimate the outcome model:

$$\begin{array}{l}\mathrm{I}\mathrm{n}\mathrm{v}\mathrm{e}\mathrm{s}\mathrm{t}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t} \mathrm{i}\mathrm{n} \mathrm{r}\mathrm{i}\mathrm{s}\mathrm{k}\mathrm{y} \mathrm{a}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}\mathrm{s}=\mathsf{\alpha }+{\mathsf{\beta }}_2{\mathrm{F}\mathrm{i}\mathrm{n}\mathrm{a}\mathrm{n}\mathrm{c}\mathrm{i}\mathrm{a}\mathrm{l} \mathrm{l}\mathrm{i}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{c}\mathrm{y}}_{\mathrm{i}}+{\mathsf{\beta }}_3{\mathrm{H}\mathrm{y}\mathrm{p}\mathrm{e}\mathrm{r}\mathrm{b}\mathrm{o}\mathrm{l}\mathrm{i}\mathrm{c} \mathrm{d}\mathrm{i}\mathrm{s}\mathrm{c}\mathrm{o}\mathrm{u}\mathrm{n}\mathrm{t}\mathrm{i}\mathrm{n}\mathrm{g}}_{\mathrm{i}}+{\mathsf{\delta }}_1{\mathrm{G}\mathrm{e}\mathrm{n}\mathrm{d}\mathrm{e}\mathrm{r}}_{\mathrm{i}}+\\ {\mathsf{\delta }}_2{\mathrm{M}\mathrm{a}\mathrm{r}\mathrm{r}\mathrm{i}\mathrm{a}\mathrm{g}\mathrm{e}}_{\mathrm{i}}+{\mathsf{\delta }}_3{\mathrm{A}\mathrm{g}\mathrm{e}}_{\mathrm{i}}+{\mathsf{\delta }}_4{\mathrm{E}\mathrm{m}\mathrm{p}\mathrm{l}\mathrm{o}\mathrm{y}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}}_{\mathrm{i}}+{\mathsf{\delta }}_5{\mathrm{D}\mathrm{i}\mathrm{v}\mathrm{o}\mathrm{r}\mathrm{c}\mathrm{e}}_{\mathrm{i}}+{\mathsf{\delta }}_6{\mathrm{C}\mathrm{h}\mathrm{i}\mathrm{l}\mathrm{d}\mathrm{r}\mathrm{e}\mathrm{n}}_{\mathrm{i}}+{\mathsf{\delta }}_7{\mathrm{E}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{a}\mathrm{t}\mathrm{i}\mathrm{o}\mathrm{n}}_{\mathrm{i}}+\\ {\mathsf{\delta }}_8{\mathrm{H}\mathrm{o}\mathrm{u}\mathrm{s}\mathrm{e}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d} \mathrm{I}\mathrm{n}\mathrm{c}\mathrm{o}\mathrm{m}\mathrm{e}}_{\mathrm{i}}+{\mathsf{\delta }}_9{\mathrm{H}\mathrm{o}\mathrm{u}\mathrm{s}\mathrm{e}\mathrm{h}\mathrm{o}\mathrm{l}\mathrm{d} \mathrm{A}\mathrm{s}\mathrm{s}\mathrm{e}\mathrm{t}}_{\mathrm{i}}+{\mathsf{\delta }}_{10}{\mathrm{R}\mathrm{i}\mathrm{s}\mathrm{k} \mathrm{A}\mathrm{v}\mathrm{e}\mathrm{r}\mathrm{s}\mathrm{i}\mathrm{o}\mathrm{n}}_{\mathrm{i}}+{\mathsf{\epsilon }}_2\end{array}$$

(2)

Step 3: Breakdown of the total effect of financial literacy on investment into direct and indirect effects, the latter representing the mediation effect of hyperbolic discounting.

The total effect (TE) of financial literacy on investment in risky assets is:

TE = β₂ + (β₁ × β₃)

(3)

The indirect effect (IE) mediated by hyperbolic discounting is:

IE = β₁ × β₃

(4)

The direct effect (DE) not mediated by hyperbolic discounting is:

DE = β₂

(5)

This structured mediation design aligns with prior empirical studies examining behavioral channels in financial decision-making and allows direct evaluation of whether present-biased preferences attenuate the investment effects of financial literacy.

4. Results

Table 3 shows the results of the mediator model. The results reveal a significant negative relationship between financial literacy and hyperbolic discounting, indicating that individuals with higher financial literacy are 6.87% less likely to exhibit impulsive decision-making tendencies. Among demographic variables, gender is significant, with males more prone to hyperbolic discounting compared to females. Age has a significantly negative association, suggesting that older individuals are less impulsive. However, the squared age term points to non-linear effects, indicating that the decrease in impulsivity with age may plateau or reverse at higher ages.

Table 3. Results of the mediator analysis.

	Dependent Variable: Hyperbolic Discounting
Variables	Coef.	Significance (z Values)
Financial literacy	−0.0687	−14.99 ***
Gender	0.0198	9.77 ***
Age	−0.0045	−7.65 ***
AgeS	0.0001	9.21 ***
Married	0.0077	2.79 ***
Divorce	0.0080	1.76 *
Children	0.0000	0.02
University education	0.0014	0.63
Employment	−0.0012	−0.24
lHH_Income	0.0026	0.91
lHH_Asset	0.0019	1.18
Risk_Aversion	0.0028	0.65
Constant	0.2137	16.31 ***
Obs.	108,682
F	46.03 ***
Adj. R2	0.0050

* and *** indicate significance level at 10% and 1%, respectively.

Marital status shows a small but significant positive effect, as married individuals are slightly more likely to exhibit hyperbolic discounting, while divorce has a weak positive association. No significant relationship is observed between hyperbolic discounting and factors such as having children, educational attainment, employment status, household income, household assets, or risk aversion.

The constant term reflects a baseline tendency toward hyperbolic discounting when all predictors are held constant. These results highlight the critical role of financial literacy in reducing impulsive decision-making, while demographic characteristics such as gender, age, and marital status also contribute to variations in hyperbolic discounting. Other variables, such as education and income, appear to be less influential in this context. The findings emphasize the need for focused policy and educational initiatives that improve financial literacy and thereby help limit impulsive financial behavior.

Table 4 shows the results of the outcome model. The results highlight several significant relationships between the variables examined and the percentage of investment in risky assets. Hyperbolic discounting exhibits a small yet statistically significant positive association, indicating that individuals with more impulsive decision-making tendencies are marginally more inclined to allocate funds to risky financial assets. In contrast, financial literacy demonstrates a strong and economically meaningful positive effect, whereby a one-unit increase in financial literacy corresponds to a 7.98-percentage-point rise in risky asset investment, highlighting the central role of financial knowledge in shaping informed investment behavior. Gender also plays a key role, as males are significantly more inclined to invest in risky assets, with an average increase of 9.06 percentage points compared to females.

Table 4. Results of the outcome model.

	Dependent Variable: Investment in Risky Financial Assets
	Coef.	Significance (z Values)
Hyperbolic discounting	0.5657	2.21 **
Financial literacy	7.9801	20.66 ***
Gender	9.0592	53.12 ***
Age	0.0405	0.82
AgeS	−0.0013	−2.38 **
Married	−4.2597	−18.28 ***
Divorce	0.5757	1.51
Children	−0.6178	−3.25 ***
University education	0.3733	2.03 **
Employment	−3.7446	−8.85 ***
lHH_Income	4.5247	18.90 ***
lHH_Asset	4.1687	30.98 ***
Risk_Aversion	−2.8992	−8.11 ***
Constant	26.8008	24.25 ***
Obs.	108,682
F	546.34 ***
Adj. R2	0.0612

** and *** indicate significance level at 5% and 1%, respectively.

Age does not show a significant linear effect, but the negative squared age term suggests that risky investment decreases at higher ages, reflecting diminishing risk preferences over time. Marital status has a notable effect, with married individuals investing significantly less in risky assets, while divorce shows a positive but statistically insignificant effect. Having children reduces risky investment by 0.62 percentage points, suggesting that parental responsibilities may lead to more conservative financial decisions. University education has a small but significant positive impact, while being employed is associated with a decrease in risky investment by 3.74 percentage points. Household income and household assets are strong predictors of increased investment in risky assets, and each unit increase in their respective logged values leads to substantial increases in risky investment. Risk aversion has a significant negative effect, indicating that more risk-averse individuals allocate less to risky investments.

Overall, the findings suggest that financial literacy, household income, and assets are the strongest predictors of higher-risk investments, while hyperbolic discounting and demographic factors such as gender, marital status, and parental responsibilities play important roles in shaping investment behavior. Risk aversion, as expected, drives individuals towards safer financial choices.

Table 5 presents the results of the mediation analysis, offering insight into the relationship among financial literacy, hyperbolic discounting, and investment in risky assets. The direct effect of financial literacy on risky asset investment is positive and sizable (7.9802), indicating that financial literacy strongly encourages engagement with risky financial assets, independent of the mediating influence of hyperbolic discounting. In contrast, the indirect effect (−0.0388) is negative, suggesting that hyperbolic discounting slightly diminishes the positive impact of financial literacy on risky investments. This implies that individuals who exhibit impulsive tendencies, as measured by hyperbolic discounting, may counteract some of the beneficial effects of financial literacy.

Table 5. Decomposition of Total, Direct, and Indirect Effects.

Effect	Magnitude
Indirect Effect	−0.0388
Direct Effect	7.9801
Total Effect	7.9412
Proportion Mediated	−0.0049

The total effect, combining both direct and indirect effects, remains positive at 7.9413, underscoring the overall strong association between financial literacy and investment in risky assets. However, the mediated proportion (−0.0049) is minimal and negative, indicating that hyperbolic discounting mediates only a very small fraction of the total effect (−0.49%) and slightly detracts from the overall relationship. These findings highlight that, while hyperbolic discounting has some influence, the direct pathway from financial literacy to investment behavior is dominant and remains highly significant. This suggests that addressing behavioral biases such as hyperbolic discounting could further enhance the positive effects of financial literacy on investment decisions.

5. Discussion

This study investigates the mediating role of hyperbolic discounting in the linkage between financial literacy and investment in risky assets, providing insights into how cognitive and behavioral factors interact to shape financial decisions. The findings confirm the dominant role of financial literacy in promoting informed investment behavior, with a strong direct effect on risky asset allocation. Financial literacy enables individuals to navigate complex financial markets, assess risks, and adopt strategies that prioritize long-term financial growth. From an economic perspective, this magnitude implies that improvements in financial knowledge can meaningfully shift household portfolio composition toward higher-return assets and support long-term wealth accumulation. These results are consistent with previous studies demonstrating the positive association between financial literacy and prudent investment behavior, including portfolio diversification and enhanced market participation (Khan et al., 2020, 2021; Oehler et al., 2024; Xiao et al., 2022).

The modest magnitude of the mediation effect is theoretically intuitive for several reasons. First, the sample consists of active investors, for whom basic participation barriers have already been overcome. Second, hyperbolic discounting primarily affects short-term decision-making and self-control, rather than risk preferences per se. Accordingly, present-biased preferences may influence the timing and consistency of investment actions without substantially altering long-run portfolio allocation once participation is established. Consequently, while time inconsistency may influence whether individuals act on their financial knowledge, it plays a limited role in determining the proportion of risky assets held once market participation is established.

The mediation analysis reveals a nuanced interaction between financial literacy and hyperbolic discounting. Financial literacy negatively correlates with hyperbolic discounting, suggesting that individuals with higher financial literacy are less prone to impulsive decision-making (Bawalle et al., 2024; Katauke et al., 2023; Ottaviani & Vandone, 2018). This finding underscores the potential of financial education to mitigate behavioral tendencies that favor immediate rewards over delayed benefits. Economically, this relationship indicates that financial knowledge may enhance self-control mechanisms or forward-looking expectations, thereby reducing the behavioral frictions associated with present bias. However, hyperbolic discounting introduces a subtle but significant mediating effect that slightly diminishes the direct influence of financial literacy on investment behavior. This suggests that impulsive tendencies, though less pronounced among financially literate individuals, can still interfere with the alignment of investment decisions with long-term financial goals.

The findings also highlight the small but notable influence of hyperbolic discounting on risky investment. Although impulsivity can lead to short-term financial decisions (Frigerio et al., 2020; O’Donoghue & Rabin, 1999; Laibson, 1997), it may occasionally drive engagement with high-risk financial products that offer quick returns (Chhabra, 2018). This pattern can be interpreted through intertemporal choice theory, where present-biased agents may overweight immediate payoff potential even when long-run expected returns are uncertain. This apparent paradox suggests that hyperbolic discounting does not uniformly undermine the benefits of financial literacy. This complexity underscores the importance of addressing behavioral biases to fully harness the potential of financial literacy to foster optimal investment strategies (Meier & Sprenger, 2013).

Demographic factors further contextualize the interaction between financial literacy, hyperbolic discounting, and investment decisions. The analysis shows that males are more likely to engage in risky investments, while married individuals and parents adopt more conservative investment approaches. These patterns are consistent with life-cycle portfolio theory and household risk-sharing models, which predict greater precautionary behavior when financial responsibilities increase and future consumption commitments become more binding. The observed differences also reflect broader social and economic dynamics that influence risk tolerance and investment preferences (Barthel & Lei, 2021; Adil et al., 2022; Khan et al., 2020, 2021). Age also plays a role, with evidence of declining risk preferences at older ages, consistent with theories that associate life stages with changing financial priorities (Khan et al., 2020, 2021; Ullah et al., 2024). From an economic standpoint, these demographic effects suggest that structural household conditions shape portfolio allocation alongside cognitive and behavioral mechanisms. However, these demographic characteristics do not mediate the relationship between financial literacy and hyperbolic discounting to the same extent as impulsivity, reinforcing the unique role of hyperbolic discounting in this framework.

The results of the mediation analysis emphasize the dominance of the direct pathway from financial literacy to investment behavior. Hyperbolic discounting mediates only a small fraction of the total effect, suggesting that while behavioral biases introduce temporal inconsistencies, their influence is secondary to the strong direct effects of financial literacy. This finding implies that policies aimed at improving financial literacy may generate larger economic gains in risky asset participation than interventions targeting behavioral bias alone, although the latter may still provide complementary benefits. This finding aligns with previous research suggesting that addressing behavioral tendencies can complement financial education to improve investment outcomes (Ventre et al., 2024; Ottaviani & Vandone, 2018).

While this study provides valuable insights into the mediating role of hyperbolic discounting, several limitations should be considered. First, the cross-sectional nature of the data restricts the ability to draw definitive causal inferences regarding the relationships among financial literacy, hyperbolic discounting, and investment behavior. Although the mediation framework reports direct and indirect effects following standard terminology in the mediation analysis literature, these estimates should be interpreted as associational rather than causal. Future research may further explore dynamic behavioral mechanisms using alternative empirical settings or experimental approaches to better understand temporal decision processes in financial contexts. Second, the measurement of hyperbolic discounting is based on self-reported preferences, which may not fully capture the complexity of impulsive decision-making in real-world contexts. Experimental approaches could provide a more nuanced understanding of how hyperbolic discounting operates in financial decisions (Bawalle et al., 2024). Finally, the study focuses primarily on risky financial assets, which may not represent the full spectrum of investment options available to individuals. Expanding the scope to include other asset classes, such as real estate or retirement funds, could offer a more comprehensive view of how financial literacy and hyperbolic discounting influence overall financial decision-making (Mouna & Anis, 2017). By addressing these limitations, future research can build on the findings of this study to refine theoretical frameworks and develop more targeted interventions that integrate both cognitive and behavioral dimensions of financial decision-making. Future research could explicitly compare hyperbolic discounting with exponential time discounting to further disentangle the roles of dynamic inconsistency and general patience in shaping investment behavior.

6. Conclusions

This study investigates the mediating role of hyperbolic discounting in the relationship between financial literacy and investment in risky assets. Using data from 108,682 active investors collected through the Money and Life survey, the analysis combines insights from financial literacy assessments, behavioral tendencies, and investment behaviors. Mediation analysis was employed to decompose the effects of financial literacy on risky investment into direct and indirect components, with hyperbolic discounting acting as the mediator.

The results underscore the significant direct effect of financial literacy on investment in risky financial assets, highlighting its critical role in enabling individuals to navigate complex financial markets and make informed decisions. Overall, the findings indicate that financial literacy remains the dominant determinant of risky asset investment, whereas hyperbolic discounting only marginally distorts the translation of financial knowledge into actual behavior. Meanwhile, hyperbolic discounting introduces a subtle mediating effect that slightly diminishes this direct relationship. Specifically, financially literate individuals are less prone to impulsive decision-making, but those who exhibit hyperbolic discounting tendencies may still deviate from optimal long-term investment strategies. Demographic variables such as gender, marital status, and age further contextualize these relationships but do not substantially alter the observed mediation effects.

The findings carry important implications for policymakers, educators, and financial institutions. Importantly, behavioral interventions should be viewed as complements rather than substitutes for financial education. Financial literacy remains the primary driver of investment behavior, while addressing behavioral biases such as hyperbolic discounting can help ensure that financial knowledge is more consistently translated into action. First, they reaffirm the need to promote financial literacy as a fundamental tool to improve investment behavior. Financial education programs should emphasize practical applications, such as risk assessment and portfolio diversification, to empower individuals in making long-term investment decisions. Second, the mediating role of hyperbolic discounting suggests that financial education alone is insufficient. Behavioral interventions, such as commitment devices, nudges, or tailored counseling, can complement cognitive skills by addressing impulsive tendencies that undermine prudent financial decision-making. Moreover, the demographic differences observed highlight the need for targeted interventions. For example, programs designed to reduce gender disparities in investment behavior should aim to enhance women’s financial confidence and mitigate behavioral biases. Similarly, strategies tailored to specific life stages, such as younger individuals prone to impulsivity or older investors with declining risk tolerance, can further enhance the effectiveness of financial education.