Stock market participation remains low, especially among the older, less affluent, and less educated households, despite its clear benefits for wealth preservation and growth. Although digitisation has reduced some barriers, many households remain disengaged. Using data from a Chinese investment platform, this column shows that digital nudges and financial education via the platform increase stock market participation, promote portfolio diversification, and improve risk-adjusted returns, even among the older, less educated, and less affluent individuals. Digital platforms can help democratise finance by delivering scalable, low-cost financial education.
The stock market participation puzzle (Campbell 2006) – the observation that few households invest in the stock market – has attracted attention for decades. Standard models of household finance suggest that, given historically high equity premia, more households should allocate more of their wealth to stock-related products than they currently do. Explanations range from transaction costs and liquidity constraints to behavioural biases and limited financial literacy (Lusardi and Mitchell 2014, Vissing‐Jørgensen 2002).
Modern digital technology has the potential to democratise capital market participation (Bian et al. 2023, Foucault et al. 2025). Retail investors today can access equities, bonds, exchange-traded funds, mutual funds, and protected instruments on digital platforms with very low costs and entry barriers (Hvide et al. 2024, Vissing‐Jørgensen 2002). In India, for example, the launch of the Unified Payments Interface – an open, interoperable payment system – reduced frictions and increased retail stock market participation (Ayyagari et al. 2025).
Digital technology has also transformed information access: social media can provide free financial information and low-barrier investment opportunities (Farrell et al. 2022), while robo-advisers can offer affordable, automated guidance that enhances diversification and risk-adjusted returns, especially for less affluent investors (Reher and Sokolinski 2024, Rossi and Utkus 2024). One would therefore expect that investing in capital markets via digital platforms would become a popular approach, particularly among older individuals and those with limited financial resources, for preserving and growing their savings.
Yet, households’ direct investment in capital markets remains limited. Participation rates are exceptionally low in developing countries and even advanced Asian economies, with rates of just 15% in Japan, 7% in China, and 6% in India. Paradoxically, mobile devices have high penetration rates in all of these locations, indicating that access to technology alone is insufficient to bridge the participation gap. Moreover, evidence from both China and India suggests that retail investors exhibit speculative behaviours, characterised by low diversification and negative long-term excess returns (Agarwal et al. 2025, Ayyagari et al. 2025, Jones et al. 2023).
This paradox highlights that while mobile technology reduces traditional financial barriers such as transaction costs and minimum investment thresholds, they do not directly address users’ psychological frictions and knowledge gaps. Behavioural inertia and limited financial literacy remain key barriers to capital market participation, particularly among older, less educated, and less wealthy individuals (Lusardi and Mitchell 2014, Merkoulova and Veld 2022, Thaler 2015).
Digital platforms can help overcome behavioural inertia and close knowledge gaps by using mobile technology to nudge users toward accessible, diversified investment portfolios while simultaneously delivering financial education through integrated information channels. By linking financial institutions, asset managers, and clients, these platforms provide timely, scalable, and low-cost educational content. Compared to traditional financial education programmes, they offer broad, unrestricted access across user demographics, time, and location at minimal cost (Kaiser et al. 2022).
In recent research (Guo et al. 2025), we provide evidence for this, using granular data from a large Chinese investment platform that draw potential investors’ attention towards its financial products and information services through randomised digital nudges. This setting enables us to both identify the causal effect of nudging and circumvent endogeneity concerns, thereby clearly distinguishing the causal impact of digitally transmitted financial knowledge on stock market participation and portfolio decisions.
We draw our sample from a population of 550,000 users who, despite being registered on the platform for at least ten months, exhibit behavioural inertia by not engaging with either its capital market products or financial information services. Our treatment sample comprises 62,293 individuals, who received a randomly distributed message about free access to financial information and knowledge on the platform during our study period (18 January-17 February 2023), alongside a similar-sized randomly drawn control group constructed using coarsened exact matching.
Three key findings emerge. First, digital nudges sharply increase engagement with financial information services: 2.1% of nudge recipients engaged, compared with just 0.05% of non-recipients. Crucially, meaningful behavioural changes occur only among users who not only receive the nudges but also respond to acquire digital financial knowledge, whom we refer to as treated respondents. Figure 1 illustrates this distinction: while non-recipients and non-responding recipients show little change, treated respondents display higher capital market participation, increased investment in mutual funds, greater portfolio diversification, and improved risk-adjusted returns during the treatment month.
Figure 1 Impact of digital transmission of financial knowledge: Treated respondents
Notes: These graphs display the impact of digital nudging through financial knowledge dissemination on four key investment metrics: (1) capital market participation, measured using two binary indicators – (a) ‘stock’ is defined as individuals who made any investment that includes equity exposure (i.e. investments in equity-based mutual funds or other products containing equity components), and (b) ‘mutual fund’ capturing capital market participation via mutual funds), (2) mutual fund balance, (3) portfolio diversification (measured by the platform’s allocation score), and (4) risk-adjusted performance (measured by the abnormal Sharpe ratio) during the treatment month. Each bar represents one of the three groups: non-recipients, non-responding recipients, and responding recipients. ‘Diff’ reports the difference in investment outcomes between non-recipients and responding recipients, along with the corresponding p-value.
These differences between non-recipients and treated respondents capture the effect of nudging-induced learning. The economic gain is not only for the short-term: improved investment performance persists over time (3- and 6-month horizons).
Second, while the initial engagement with digital prompts was modest, reflecting behavioural inertia, repeated nudging proved effective. Figure 2 shows that the response rate rose steadily with each monthly prompt, reaching a cumulative 28.9% response rate for treated respondents after six iterations. Importantly, newly responding recipients in each month of continued nudging showed meaningful improvements in the economic efficiency of their investments, characterised by greater diversification and sustained favourable expected return-risk performance. These results underscore the need for persistent targeted nudging to overcome behavioural inertia.
Figure 2 Response rates to digital nudging across repeated nudging horizons
Notes: The graph shows the cumulative proportion of treated users (62,293 recipients) who ever responded to consecutive digital nudging messages over six consecutive monthly prompting cycles. The vertical axis denotes the cumulative rate of recorded responses, while the horizontal axis indicates the number of monthly nudges received. Recorded response rates cumulated steadily with consecutive digital nudging.
Third, the benefits are pronounced among older, less educated, and less affluent individuals – groups typically underserved in capital markets and prone to behavioural inertia (see Guo et al. 2025). These results highlight the potential for digital platforms to deliver financial education at scale and provide a cost-effective complement to traditional policy tools aimed at broadening financial inclusion.
Together, these findings have important implications for policy and practice. While lower entry barriers and user-friendly digital platforms make stock investing more accessible, this expanded access also risks exposing households to harmful speculative behaviour if not accompanied by adequate financial knowledge. Traditional financial education programmes, however, remain costly and limited in scale.
Our evidence suggests that facilitating financial education about stock investing knowledge via digital platforms with persistent nudging is a low-cost, scalable, and effective alternative to traditional approaches. This is particularly important given the potential investors’ behaviour inertia and a widespread lack of financial literacy. Regulators and policymakers could harness digital platforms to foster more inclusive financial participation, especially among vulnerable groups, while also considering how educational content might be tailored to the needs of different investor groups.
A measure of caution is warranted. While the results are encouraging and intuitively appealing, potential costs and unintended consequences remain underexplored. For instance, over time, self-interested service providers could exploit digital platforms to take advantage of unsuspecting users by turning them towards unsound investments. Thoughtful policy guidance and safeguards are therefore essential.
Source : VOXeu
