Our bi-weekly Opinion provides you with latest updates and analysis on major capital market and financial investment industry issues.
Demographic Changes in Korea: Implications for Household Asset Holdings and Capital Market Participation
Publication date Jan. 02, 2025
Summary
Korea is undergoing a profound demographic transformation characterized by rapid changes in age and cohort structures, which could influence household asset demand. This article examines the dynamics of asset holdings and capital market participation among Korean households, distinguishing between age-specific patterns and cohort-specific trends.
The analysis of age-specific patterns, controlling for cohort-related differences, reveals that total assets, net assets, real estate, and financial asset holdings exhibit a gradual depletion trend among older populations. In contrast, capital market asset holdings experience a sharper decline as age increases among older cohorts. Notably, household participation in capital markets peaks in the early 30s but diminishes significantly thereafter with age. Cohort-specific analysis, independent of age-related differences, indicates no significant variations in capital market asset holdings across cohort groups. However, capital market participation rates show a marked decline among more recent birth cohorts.
If these age and cohort-specific patterns persist, population aging is projected to be a key factor in reducing household demand for capital market assets. Simultaneously, changes in the cohort structure are likely to further reduce the proportion of households participating in capital markets. To sustain the demand base for Korea’s capital markets and broaden the investor pool, it is crucial to promote capital market participation among younger cohorts and foster sustained engagement from middle-aged and older populations.
The analysis of age-specific patterns, controlling for cohort-related differences, reveals that total assets, net assets, real estate, and financial asset holdings exhibit a gradual depletion trend among older populations. In contrast, capital market asset holdings experience a sharper decline as age increases among older cohorts. Notably, household participation in capital markets peaks in the early 30s but diminishes significantly thereafter with age. Cohort-specific analysis, independent of age-related differences, indicates no significant variations in capital market asset holdings across cohort groups. However, capital market participation rates show a marked decline among more recent birth cohorts.
If these age and cohort-specific patterns persist, population aging is projected to be a key factor in reducing household demand for capital market assets. Simultaneously, changes in the cohort structure are likely to further reduce the proportion of households participating in capital markets. To sustain the demand base for Korea’s capital markets and broaden the investor pool, it is crucial to promote capital market participation among younger cohorts and foster sustained engagement from middle-aged and older populations.
The confluence of declining fertility and increasing life expectancy has accelerated population aging in Korea. These demographic shifts are presenting profound implications for the economy, potentially reshaping the composition of household asset holdings and the demand for risk assets. According to the life-cycle hypothesis, individuals accumulate wealth through savings and investments during their working years and deplete assets to finance consumption in retirement. If this hypothesis holds, the declining population share of younger and middle-aged individuals who are primarily responsible for asset accumulation, combined with the growing share of asset-depleting older cohorts, may result in a reduction in aggregate household asset holdings. Notably, older cohorts tend to exhibit risk-averse behavior in their financial asset allocation. As a result, the growing share of an aging population is expected to dampen the overall household demand for risk assets, such as stocks and funds.
Against this backdrop, this article analyses patterns in asset accumulation and capital market participation, with a focus on age- and cohort-based differences, using household panel data. Analyses of asset holding patterns within a given year capture not only age-related variations but also cohort-specific discrepancies. To address this limitation, this article adopts an analytical approach that distinguishes between the age effect, reflecting age-specific patterns independent of cohort-based variations, and the cohort effect, representing cohort-specific trends separated from age-related factors. Finally, based on the analysis results, it assesses the potential ramifications of demographic shifts on future asset holdings and the dynamics of capital markets in Korea.
Korea’s demographic changes
According to population projections from Statistics Korea, the proportion of individuals aged 65 and older is forecasted to more than double over the next 25 years, rising from 19.2% in 2024 to 39.8% by 2049 (Figure 1). The median age of Korea’s population is also expected to steadily rise from 44.9 years in 2024 to 55 years by 2049.
In parallel, demographic upheaval involves not only changes in the age-specific composition but also shifts in the composition of cohort groups. As of 2024, individuals aged 34 and younger1)—those born in and after 1990—comprise 32.8% of the population. This share is projected to climb steadily to 46.1% by 2049, while that of older cohorts is set to gradually diminish (Table 1). In the future population structure, younger cohorts will account for a growing population share and their median age will steadily rise over time. How these shifts in the age structure influence household asset demand depends on age-specific asset-holding patterns, whereas the effects of changes in the cohort structure hinge on cohort-specific asset-holding trends.
Korea’s household asset holdings by age
Based on household data from the financial panel survey, this article examines the asset-holding patterns of Korean households, categorizing them by the age of the household head. The analysis adopts a breakdown of key asset categories, including total assets, net assets, real estate, financial assets, and capital market assets. Capital market assets are defined as the aggregate holdings of stocks, funds, and bonds, while financial assets encompass deposits, capital market assets, insurance and pension assets, and other financial instruments. Real estate assets represent the total appraised value of primary residences, additional residential properties, and other real estate holdings. Total assets include the aggregate holdings of financial assets, real estate, rental housing deposits (both monthly rent and jeonse), vehicle values, and other tangible assets, while net assets are derived by subtracting total liabilities (the sum of all debt categories) from total assets.
The left panel of Figure 2 illustrates the pattern of average asset holdings by the age of household heads in 2021.2) The findings reveal a gradual increase in the average holdings of total assets, net assets, real estate, and financial assets with age, reaching a peak in the 60-64 age group. Beyond this point, asset holdings begin to decline in the 65-69 age group, representing an inverted U-shaped pattern. For households headed by those aged 75 and older, asset holdings remain above 50% of the peak level observed in the 60-64 age group.
In contrast, the dynamics of capital market asset holdings follow a distinctly different pattern, with a sharper and earlier decline. As of 2021, average capital market asset holdings rise steadily with age, peaking in the 45-49 age group, and subsequently remain stagnant through the 60-64 age group before experiencing a sharp decline (right panel in Figure 2). For households headed by those aged 75 and older, capital market asset holdings account for only around 12% of the peak level observed in the 45-49 age group. The ratio of capital market assets to financial assets hits the highest level at about 14.1% in the 45-49 age group and rapidly diminishes with age, reaching a mere 1.1% among the 75 and older age group.
The age-specific asset-holding patterns shown in Figure 2 reflect averages across all households, including both asset-holding and non-holding households. Notably, capital market assets exhibit a significantly higher share of non-holding households, compared to other asset categories such as real estate or financial assets. According to the 2021 financial panel data, 63.5% of households owned real estate and 86.3% held deposits, while only 15.7% possessed capital market assets. This limited participation of Korean households in capital markets likely contributes to the smaller share and size of capital market asset holdings by households. Figure 3 depicts the share of households holding capital market assets by the age group of household heads in 2021. This share peaked at 26.4% in the 45-49 age group and then declined sharply with age, dwindling to just 2.3% for the 75 and older age group.
Age and cohort effects on asset holdings
The analysis of household panel data at a specific point in time reveals a notable trend where capital market asset holdings show a steeper decline at an earlier age compared to total assets, real estate, and financial assets. However, using the panel data has a limitation in that age-specific differences in asset holdings reflect both age effects and cohort effects. The age effect refers to the typical characteristics exhibited by individuals as they progress through specific stages of the life cycle, while the cohort effect captures the shared economic experiences of individuals born in the same period throughout their life cycle. Since older age groups at a given point of time correspond to cohorts born in earlier periods, the asset holdings of older age groups inherently reflect not only age effects but also cohort effects attributed to older cohorts. To distinguish these overlapping influences, this analysis separates the age effect from the cohort effect in examining asset holdings.
For the separation, a linear regression model3) was developed using financial panel data spanning from 2007 to 2021.4) This model sets holdings for each asset category as dependent variables and uses age-specific and cohort-specific dummy variables as explanatory variables. For this regression analysis using dummy variables, the population born in and after 1990 is designated as the reference group.
Figure 4 illustrates the estimated asset holdings that reflect age effects, controlling for cohort effects. With the population born in and after 1990 serving as the reference group, the age effect as observed in Figure 2 reflect the estimated asset holdings as individuals from this reference group progress through each respective age group.5) For total assets and real estate, asset holdings increase until peaking in the 65-69 age group before declining, while net assets reach the highest level in the 70-74 age group and financial assets peak earlier in the 60-64 age group. All these asset categories follow an inverted U-shaped trajectory. Even in the 75-and-older age group, asset holdings remain above 80% of the highest levels at the respective peak ages. In contrast, the age effect on capital market asset holdings displays a different pattern. Capital market asset holdings increase up to the 40-44 age group, and plateau through the 55-59 age group before plunging sharply. For individuals aged 75 and older, the age effect on capital market assets is found to be statistically insignificant.
Figure 5 presents the estimated cohort effect on asset holdings, highlighting the differences in average asset holdings across cohorts relative to the cohort group born in and after 1990. For total assets, net assets, real estate, and financial assets, more recent cohorts tend to hold greater amounts of assets, representing positive cohort effects compared to earlier cohorts. However, capital market asset holdings deviate from this trend. Across all cohorts, the estimated cohort effects for capital market asset holdings are statistically insignificant. This finding underscores the unique dynamics of capital market asset holdings, which are not influenced by positive cohort effects, compared to other asset categories such as total assets, real estate, and financial assets. Furthermore, the differences in cohort effects are not statistically significant across cohorts.6)
Finally, this article also distinguishes between the age effect and the cohort effect on household participation in domestic capital markets. To this end, a probit regression analysis was conducted, using ownership of capital market assets as the dependent variable and age-specific and cohort-specific dummies as explanatory variables. By estimating the probit regression coefficients for each dummy variable, the analysis calculated the probabilities of capital market participation attributable to age and cohort effects.
Figure 6 shows the estimated probabilities of household participation in capital markets, categorized by age and cohort effects. The age effect refers to the likelihood of capital market participation when individuals born in and after 1990 progress through respective age group, while the cohort effect reflects the probability of participation for each cohort group when they reach the 30-34 age group.7) The age effect on capital market participation increases until peaking in the 30-34 age group and declines steeply as age rises thereafter, representing an inverted U-shaped pattern. On the other hand, the cohort effect reveals a downward trend, indicating that more recent cohorts are less likely to participate in capital markets.
Implications
Korea’s future demographics are expected to involve not only a shift in the age structure driven by population aging but also a changing cohort structure influenced by the growing share of more recent birth cohorts. Thus, the impact of these demographic changes on household asset demand will depend on the distinct asset-holding patterns across various age and cohort groups. In an analysis of household panel data, which separates age effects from cohort effects, total assets, net assets, real estate, and financial assets show a gradual downward trend in older age groups, while capital market asset holdings exhibit a steeper decline as age increases in older populations. Notably, the rate of household participation in capital markets reaches its peak in the early 30s, followed by a steep decline as age increases.
Additionally, the analysis uncovers positive cohort effects for total assets, net assets, real estate, and financial assets among more recent birth cohorts. For capital market asset holdings, no statistically significant cohort effects were identified across all cohorts. Furthermore, a negative cohort effect was observed in the rates of capital market participation, indicating that more recent birth cohorts are less likely to engage in capital markets.8)
If the observed patterns by age and cohort persist, changes in Korea’s age structure are less likely to lead to a significant reduction in total household asset holdings. This can be attributed to the positive cohort effect, characterized by the gradual depletion of assets among older populations and the tendency to hold greater assets among younger cohorts. However, shifts in the age structure are projected to exert downward pressure on demand for capital market assets. This reduced demand is unlikely to be offset by changes in the cohort structure, particularly followed by the steady decline in the share of households engaging in capital markets.
To sustain the demand base for Korea’s capital markets amid population aging and generational shifts, it is crucial to promote capital market participation among younger cohorts and encourage middle-aged and older populations to steadily participate in capital markets. This will require continuous efforts from industry stakeholders and policymakers. They should provide tailored financial investment products designed to align with the distinct needs and preferences of various age and cohort groups. It is also necessary to develop customized financial education programs that accommodate differences in financial literacy and investment experience across demographic groups.
1) Youth (ages 19-34) and younger age groups.
2) Similar qualitative patterns are observed for other years as well.
3) For the theoretical background of this analysis method, see Poterba, J., 2001, Demographic structure and asset returns, Review of Economics and Statistics 83(4), 565-584.
4) The analysis sample includes 15,248 households, with a total number of 84,314 observation cases over the sample period. All monetary variables for all years analyzed were converted to real values using the 2021 consumer price index.
5) Even if a different cohort is set as the reference group, the graph shape is the same.
6) Differences in productivity and preferences among various cohorts may lead to variations in lifetime earnings, savings, and asset-holding patterns across cohorts. See Jappelli, T., 1999, The age‐wealth profile and the life‐cycle hypothesis: A cohort analysis with a time series of cross‐sections of Italian households, Review of Income and Wealth, 45(1), 57-75; Poterba, J., 2001, Demographic structure and asset returns, Review of Economics and Statistics 83(4), 565-584.
7) Regardless of the cohort or age group used as the reference group, the shapes of the graphs are nearly identical.
8) The negative cohort effect on capital market participation primarily stems from the negative cohort effects on fund market participation. This article estimated cohort effects separately for stock market and fund market participation. According to the findings, a positive cohort effect was observed for stock market participation, while a negative cohort effect was found for fund market participation. The latter showed a must steeper slope.
Against this backdrop, this article analyses patterns in asset accumulation and capital market participation, with a focus on age- and cohort-based differences, using household panel data. Analyses of asset holding patterns within a given year capture not only age-related variations but also cohort-specific discrepancies. To address this limitation, this article adopts an analytical approach that distinguishes between the age effect, reflecting age-specific patterns independent of cohort-based variations, and the cohort effect, representing cohort-specific trends separated from age-related factors. Finally, based on the analysis results, it assesses the potential ramifications of demographic shifts on future asset holdings and the dynamics of capital markets in Korea.
Korea’s demographic changes
According to population projections from Statistics Korea, the proportion of individuals aged 65 and older is forecasted to more than double over the next 25 years, rising from 19.2% in 2024 to 39.8% by 2049 (Figure 1). The median age of Korea’s population is also expected to steadily rise from 44.9 years in 2024 to 55 years by 2049.
In parallel, demographic upheaval involves not only changes in the age-specific composition but also shifts in the composition of cohort groups. As of 2024, individuals aged 34 and younger1)—those born in and after 1990—comprise 32.8% of the population. This share is projected to climb steadily to 46.1% by 2049, while that of older cohorts is set to gradually diminish (Table 1). In the future population structure, younger cohorts will account for a growing population share and their median age will steadily rise over time. How these shifts in the age structure influence household asset demand depends on age-specific asset-holding patterns, whereas the effects of changes in the cohort structure hinge on cohort-specific asset-holding trends.
Based on household data from the financial panel survey, this article examines the asset-holding patterns of Korean households, categorizing them by the age of the household head. The analysis adopts a breakdown of key asset categories, including total assets, net assets, real estate, financial assets, and capital market assets. Capital market assets are defined as the aggregate holdings of stocks, funds, and bonds, while financial assets encompass deposits, capital market assets, insurance and pension assets, and other financial instruments. Real estate assets represent the total appraised value of primary residences, additional residential properties, and other real estate holdings. Total assets include the aggregate holdings of financial assets, real estate, rental housing deposits (both monthly rent and jeonse), vehicle values, and other tangible assets, while net assets are derived by subtracting total liabilities (the sum of all debt categories) from total assets.
The left panel of Figure 2 illustrates the pattern of average asset holdings by the age of household heads in 2021.2) The findings reveal a gradual increase in the average holdings of total assets, net assets, real estate, and financial assets with age, reaching a peak in the 60-64 age group. Beyond this point, asset holdings begin to decline in the 65-69 age group, representing an inverted U-shaped pattern. For households headed by those aged 75 and older, asset holdings remain above 50% of the peak level observed in the 60-64 age group.
In contrast, the dynamics of capital market asset holdings follow a distinctly different pattern, with a sharper and earlier decline. As of 2021, average capital market asset holdings rise steadily with age, peaking in the 45-49 age group, and subsequently remain stagnant through the 60-64 age group before experiencing a sharp decline (right panel in Figure 2). For households headed by those aged 75 and older, capital market asset holdings account for only around 12% of the peak level observed in the 45-49 age group. The ratio of capital market assets to financial assets hits the highest level at about 14.1% in the 45-49 age group and rapidly diminishes with age, reaching a mere 1.1% among the 75 and older age group.
The age-specific asset-holding patterns shown in Figure 2 reflect averages across all households, including both asset-holding and non-holding households. Notably, capital market assets exhibit a significantly higher share of non-holding households, compared to other asset categories such as real estate or financial assets. According to the 2021 financial panel data, 63.5% of households owned real estate and 86.3% held deposits, while only 15.7% possessed capital market assets. This limited participation of Korean households in capital markets likely contributes to the smaller share and size of capital market asset holdings by households. Figure 3 depicts the share of households holding capital market assets by the age group of household heads in 2021. This share peaked at 26.4% in the 45-49 age group and then declined sharply with age, dwindling to just 2.3% for the 75 and older age group.
The analysis of household panel data at a specific point in time reveals a notable trend where capital market asset holdings show a steeper decline at an earlier age compared to total assets, real estate, and financial assets. However, using the panel data has a limitation in that age-specific differences in asset holdings reflect both age effects and cohort effects. The age effect refers to the typical characteristics exhibited by individuals as they progress through specific stages of the life cycle, while the cohort effect captures the shared economic experiences of individuals born in the same period throughout their life cycle. Since older age groups at a given point of time correspond to cohorts born in earlier periods, the asset holdings of older age groups inherently reflect not only age effects but also cohort effects attributed to older cohorts. To distinguish these overlapping influences, this analysis separates the age effect from the cohort effect in examining asset holdings.
For the separation, a linear regression model3) was developed using financial panel data spanning from 2007 to 2021.4) This model sets holdings for each asset category as dependent variables and uses age-specific and cohort-specific dummy variables as explanatory variables. For this regression analysis using dummy variables, the population born in and after 1990 is designated as the reference group.
Figure 4 illustrates the estimated asset holdings that reflect age effects, controlling for cohort effects. With the population born in and after 1990 serving as the reference group, the age effect as observed in Figure 2 reflect the estimated asset holdings as individuals from this reference group progress through each respective age group.5) For total assets and real estate, asset holdings increase until peaking in the 65-69 age group before declining, while net assets reach the highest level in the 70-74 age group and financial assets peak earlier in the 60-64 age group. All these asset categories follow an inverted U-shaped trajectory. Even in the 75-and-older age group, asset holdings remain above 80% of the highest levels at the respective peak ages. In contrast, the age effect on capital market asset holdings displays a different pattern. Capital market asset holdings increase up to the 40-44 age group, and plateau through the 55-59 age group before plunging sharply. For individuals aged 75 and older, the age effect on capital market assets is found to be statistically insignificant.
Figure 5 presents the estimated cohort effect on asset holdings, highlighting the differences in average asset holdings across cohorts relative to the cohort group born in and after 1990. For total assets, net assets, real estate, and financial assets, more recent cohorts tend to hold greater amounts of assets, representing positive cohort effects compared to earlier cohorts. However, capital market asset holdings deviate from this trend. Across all cohorts, the estimated cohort effects for capital market asset holdings are statistically insignificant. This finding underscores the unique dynamics of capital market asset holdings, which are not influenced by positive cohort effects, compared to other asset categories such as total assets, real estate, and financial assets. Furthermore, the differences in cohort effects are not statistically significant across cohorts.6)
Figure 6 shows the estimated probabilities of household participation in capital markets, categorized by age and cohort effects. The age effect refers to the likelihood of capital market participation when individuals born in and after 1990 progress through respective age group, while the cohort effect reflects the probability of participation for each cohort group when they reach the 30-34 age group.7) The age effect on capital market participation increases until peaking in the 30-34 age group and declines steeply as age rises thereafter, representing an inverted U-shaped pattern. On the other hand, the cohort effect reveals a downward trend, indicating that more recent cohorts are less likely to participate in capital markets.
Korea’s future demographics are expected to involve not only a shift in the age structure driven by population aging but also a changing cohort structure influenced by the growing share of more recent birth cohorts. Thus, the impact of these demographic changes on household asset demand will depend on the distinct asset-holding patterns across various age and cohort groups. In an analysis of household panel data, which separates age effects from cohort effects, total assets, net assets, real estate, and financial assets show a gradual downward trend in older age groups, while capital market asset holdings exhibit a steeper decline as age increases in older populations. Notably, the rate of household participation in capital markets reaches its peak in the early 30s, followed by a steep decline as age increases.
Additionally, the analysis uncovers positive cohort effects for total assets, net assets, real estate, and financial assets among more recent birth cohorts. For capital market asset holdings, no statistically significant cohort effects were identified across all cohorts. Furthermore, a negative cohort effect was observed in the rates of capital market participation, indicating that more recent birth cohorts are less likely to engage in capital markets.8)
If the observed patterns by age and cohort persist, changes in Korea’s age structure are less likely to lead to a significant reduction in total household asset holdings. This can be attributed to the positive cohort effect, characterized by the gradual depletion of assets among older populations and the tendency to hold greater assets among younger cohorts. However, shifts in the age structure are projected to exert downward pressure on demand for capital market assets. This reduced demand is unlikely to be offset by changes in the cohort structure, particularly followed by the steady decline in the share of households engaging in capital markets.
To sustain the demand base for Korea’s capital markets amid population aging and generational shifts, it is crucial to promote capital market participation among younger cohorts and encourage middle-aged and older populations to steadily participate in capital markets. This will require continuous efforts from industry stakeholders and policymakers. They should provide tailored financial investment products designed to align with the distinct needs and preferences of various age and cohort groups. It is also necessary to develop customized financial education programs that accommodate differences in financial literacy and investment experience across demographic groups.
1) Youth (ages 19-34) and younger age groups.
2) Similar qualitative patterns are observed for other years as well.
3) For the theoretical background of this analysis method, see Poterba, J., 2001, Demographic structure and asset returns, Review of Economics and Statistics 83(4), 565-584.
4) The analysis sample includes 15,248 households, with a total number of 84,314 observation cases over the sample period. All monetary variables for all years analyzed were converted to real values using the 2021 consumer price index.
5) Even if a different cohort is set as the reference group, the graph shape is the same.
6) Differences in productivity and preferences among various cohorts may lead to variations in lifetime earnings, savings, and asset-holding patterns across cohorts. See Jappelli, T., 1999, The age‐wealth profile and the life‐cycle hypothesis: A cohort analysis with a time series of cross‐sections of Italian households, Review of Income and Wealth, 45(1), 57-75; Poterba, J., 2001, Demographic structure and asset returns, Review of Economics and Statistics 83(4), 565-584.
7) Regardless of the cohort or age group used as the reference group, the shapes of the graphs are nearly identical.
8) The negative cohort effect on capital market participation primarily stems from the negative cohort effects on fund market participation. This article estimated cohort effects separately for stock market and fund market participation. According to the findings, a positive cohort effect was observed for stock market participation, while a negative cohort effect was found for fund market participation. The latter showed a must steeper slope.