Wealth, financial access, and the paradox of inequality

Purpose This paper investigates the relationship between household wealth, financial market participation, and asset returns in China. While classic portfolio models predict wealth neutrality in risky asset allocation, empirical evidence from Western economies challenges this view. We aim to document whether similar wealth gradients exist in China's distinct institutional context and to develop a theoretical framework capable of explaining the joint patterns of participation and return heterogeneity. The analysis also seeks to evaluate the distributional consequences of policies affecting financial access, providing insights for financial reform in emerging economies where expanding market participation remains a central policy objective. Design/methodology/approach Using four waves of the China Household Finance Survey (CHFS, 2013–2019), we employ nonparametric portfolio analysis and panel regressions with household fixed effects to establish stylized facts about wealth, participation, and returns. We then develop a continuous-time heterogeneous-agent general equilibrium model featuring endogenous market participation. Households choose between direct finance (incurring a fixed cost) and indirect finance through intermediaries. The participation threshold is determined endogenously by value-matching and smooth-pasting conditions in a two-dimensional state space of wealth and productivity. The model is calibrated to match key moments from the data, including participation rates, wealth shares, and return spreads. Findings We document three robust patterns: (1) wealthier households are more likely to participate in direct finance and allocate a larger share of assets to it; (2) the wealth gradient steepens at the top, indicating concentration of direct finance among the richest; and (3) wealthier households earn systematically higher returns, a relationship that survives household fixed effects. The model shows that lowering participation costs boosts stock market access but exacerbates wealth inequality, whereas reducing intermediation costs benefits poorer households through higher deposit returns. Policies promoting stock market participation may disproportionately harm bank-dependent populations, highlighting a crucial trade-off in financial reform. Originality/value This paper advances the literature in three ways. Methodologically, it extends contingent claims methods to a two-dimensional heterogeneous-agent setting, endogenizing both participation and return heterogeneity through value-matching and smooth-pasting conditions. Theoretically, it provides a unified framework explaining why wealth-participation and wealth-return gradients emerge jointly from financial frictions. Practically, it reveals a policy paradox: reducing participation costs widens inequality, while lowering intermediation costs benefits the poor. These insights caution against one-sided financial liberalization and emphasize the redistributive role of banking systems, offering guidance for emerging economies navigating financial deepening and inclusive growth.