In 2023, the Bundesbank once again surveyed households in Germany about their wealth, debt and income. The data collected by the survey contribute to a detailed picture of the financial situation of households in Germany and provide valuable insights into the distribution of wealth.

The average amount of net wealth per household grew slightly in nominal terms between 2021 and 2023, but declined when adjusted for inflation. Nevertheless, net wealth remains at a higher level in both nominal and inflation-adjusted terms than in 2017, which was the last survey year before the COVID-19 pandemic.

Regarding the development of wealth inequality, no clear trend can be identified: amid high inflation rates and the significant rise in interest rates, the various measures of wealth inequality changed only marginally between 2021 and 2023.

Ownership of real estate and businesses continues to be strongly correlated with high levels of wealth. Riskier financial assets, such as shares and investment funds, are also more likely to be held by wealthier households. Less wealthy households, by contrast, primarily hold balances on savings accounts and other low-risk forms of investment. However, the proportion of households that invest in funds or shares has been rising for several years now.

The share of indebted households has fallen slightly. The proportion of income that these households spend on principal and interest payments for loans has not changed significantly.

Overall, the results show that the structures of both wealth and debt among households in Germany are fairly stable. With the data currently available, it is not yet possible to estimate the impact of the higher interest rates on wealth and its distribution over the medium term. The next survey is scheduled to take place in 2026.

1 Wealth distribution in 2023 compared with previous years

The Bundesbank’s panel on household finances in Germany continued in 2023 with another wave of the survey. This survey was conducted in an environment of high rates of inflation, rising interest rates, only marginally rising or even falling real estate prices, and only moderately rising share prices.

Holdings of wealth, measured as average net wealth per household, grew only slightly in nominal terms between 2021 and 2023, following a significant increase between 2017 and 2021.[1] On average, households in Germany had net wealth of around €324,800 in 2023, representing an increase of around 3% compared with 2021 (€316,500). Adjusted for inflation, this constitutes a decline from €268,700 to €239,200.[2] Median net wealth – which marks the middle of the wealth distribution and divides households into a wealthier half and a less wealthy half – also decreased considerably after adjusting for inflation, falling from €90,500 in 2021 to €76,000 in 2023. In nominal terms, it decreased by around €3,400 to €103,200.

Over the longer term, households’ net wealth has increased both in nominal and inflation-adjusted, or real, terms. Between 2017 – the last pre-pandemic survey year – and 2023, mean net wealth rose by 40% in nominal terms and by 13% in real terms. The median rose by 46% in nominal terms and by 18% in real terms over the same period.

Mean and median values of the net wealth distribution of households in Germany

The PHF study’s definition of wealth

The objective of the PHF study is to compile and present detailed information on the wealth of German households.[1] The PHF study’s definition of wealth is therefore designed to capture both the assets and the liabilities on households’ balance sheets. The assets side (gross wealth) consists of real assets and financial assets. The other side of the balance sheet lists liabilities, i.e. loans secured by real estate and unsecured loans. Net wealth is calculated as the difference between gross wealth and debt.

The data on asset types captured in the PHF are more granular than those in other studies examining wealth as a topic. Real assets, for example, include not only real estate and business ownership but also the value of vehicles, collections and jewellery. The data comprehensively cover financial assets as well. These consist of balances with banks, savings banks, and building and loan associations, securities, long-term equity investment and assets under management. The positive balances from private pension and life insurance policies are also included.[2]

Any future statutory pension entitlements are not included. As Germany has a pay-as-you-go pension system, those paying into the system accrue only entitlements rather than wealth. However, making a variety of assumptions about life expectancy, interest rate developments and the retirement age, it would be possible to convert future entitlements for individual types of statutory pension provision into assets (to capitalise them). Such simulations show that wealth inequality is smaller when looking at wealth including statutory pension provision than it is when it is not included.[3]

Households evaluate their assets themselves. This mainly applies to property and business ownership. In both cases, households are asked what price they think they could achieve for their property or business if they were to sell it.

Assets held abroad are also included in the calculation of a household’s total assets if the respondents report this information.

Balance sheet of a household – a schematic overview

The PHF defines households as groups of persons who live at the same address and share daily expenses. Persons who temporarily do not live at that address but regularly return there are also considered part of the household. Persons or groups of persons who live in a shared residence without having a family or partnership relationship, or domestic staff residing at that address, constitute households in their own right.
Households’ wealth includes private pension and life insurance policies in the accumulation phase or where contributions have been suspended. They are removed from households’ balance sheets once payouts from the policies commence, and the relevant flows of income are then taken into account when calculating income.
See also the supplementary information entitled “Counting statutory pension entitlements as an additional wealth component”.

The developments in interest rates and asset prices between 2021 and 2023 mentioned at the beginning of this article have not led to any substantial changes in the distribution of wealth. For the relative measures of inequality, no clear trend can be identified.[3] For example, the share of total net wealth held by the wealthiest 10% of households amounted to more than 50% in 2023, as in all other survey years since 2010. At 54%, this figure was slightly lower than in 2017 (55%) and 2021 (56%).[4]

Share of total net wealth attributable to households

The ratio between mean and median net wealth – another measure of inequality – rose marginally from 3.0 in 2021 to 3.1 in 2023. Other ratios between different components of the wealth distribution also increased. For instance, the ratio between the cut-off value above which a household belongs to the wealthiest 10% of households in Germany (known as the 90th percentile) and the median increased from 6.8 to 7.6. In 2023, this ratio was thus slightly below its level from before the COVID-19 pandemic (2017: 7.8). The higher the value, the more the net wealth of the household in the middle of the distribution would have to rise in order for it to rank among the wealthiest 10% of households.

By contrast, the Gini coefficient – a classic measure of inequality – remained virtually constant for net wealth at 72.4% (2021: 72.8%).[5] Comparing the relative measures of inequality with those in other euro area countries shows that, by international standards, Germany continues to be characterised by a high degree of wealth inequality. In Spain, for example, the Gini coefficient was 69% and the ratio between the mean and median was 2.2 in 2022.[6] In Italy, the Gini coefficient was 66% and the ratio between the mean and median was 1.9 in 2022.[7]

**Table 2.1: Indicators of net wealth distribution**
Item	2010	2014	2017	2021	2023
Mean/median	3.8	3.6	3.3	3.0	3.1
P90/median	8.6	7.7	7.8	6.8	7.6
Gini coefficient	75.8	76.2	73.9	72.8	72.4
Percentage share of total net wealth held by wealthiest 10%	59	60	55	56	54
Interquartile range (€) (nominal)	203,000	221,000	262,000	338,000	390,000
Interquartile range (€) (in 2010 prices)	203,000	207,000	238,000	287,000	287,000
Difference between P90 and P10 (€) (nominal)	442,000	468,000	555,000	725,000	779,000
Difference between P90 and P10 (€) (in 2010 prices)	442,000	437,000	504,000	615,000	574,000
Sources: PHF 2010-11, PHF 2014, PHF 2017, PHF 2021, PHF 2023.

Counting statutory pension entitlements as an additional wealth component

The wealth concept underpinning the PHF study does not include any future entitlements to a statutory pension.[1] One reason for this is that these entitlements differ in character from non-financial assets or financial assets, which count as assets for the purposes of the PHF. Because these entitlements are claims rather than accumulated wealth components, they are illiquid and cannot normally be used as collateral. Respondents can state the value of their financial and non-financial assets straight away and can generally provide certain information on their claims to statutory pensions. This might include how many years they have contributed, the type of insurance scheme or the pension amount they are expected to draw upon retirement (based on information from the statutory pension scheme). However, they are unable to state the value of that entitlement overall. To enable statutory pension entitlements to be counted towards the household wealth recorded under the PHF, then, the first step is to determine their value, and that is done by capitalising the future income flows from pensions. This exercise requires a number of assumptions to be made concerning life expectancy, interest rate developments, the retirement age and the like. Some of the information needed to approximate the present value of these entitlements, such as the type of vested statutory pension, is collected in PHF surveys.

A rough simulation can be used to illustrate how counting statutory pension entitlements towards net wealth affects the distribution of wealth. This simulation is based, amongst other things, on the following assumptions:

Life expectancy is a key factor in estimating the pension-drawing period. Life expectancy data are taken from the mortality tables published by the Federal Statistical Office. Differences in the life expectancy of women and men are accounted for.
The rate at which future entitlements are discounted is assumed to be 2%, as presented in Bönke et al. (2020).
Individuals who are not yet drawing a pension start receiving a pension when they reach the standard retirement age.
Widow pensions are not accounted for. Not enough information on this type of pension is available in the PHF for it to be given adequate consideration.

Like other studies of a similar nature,[2] the simulation calculations show that including these entitlements reduces wealth inequality. In the rough simulation, the Gini coefficient for net wealth comes to around 58% when capitalised entitlements are included, compared with roughly 72% when those entitlements are excluded. The wealthiest 10% of households account for 39% of total wealth when capitalised entitlements are included. By comparison, that share is roughly 54% when no adjustments are made to net wealth. Accordingly, the share of the net wealth held by households in the lower half of the distribution rises from around 3% to just over 10%.[3]

The inclusion of statutory pension entitlements reduces wealth inequality mainly because these entitlements are more significant, relative to financial and non-financial assets, for low-wealth households than for high-wealth ones. For the least wealthy 10% of households, for example, average net wealth is positive when capitalised statutory pension entitlements are included, but negative when they are not.

Entitlements to statutory pensions as a share of total net wealth including capitalised statutory pension entitlements

See also the supplementary information entitled “The PHF study’s definition of wealth”.
For more information, see Bönke et al. (2020), Kuhn (2020) and Bartels et al. (2023a).
By comparison, the simulation exercise by Bartels et al. (2023a) finds that for 2017 the share of wealth attributable to the less wealthy half increases from 2% to 9%.

Alongside the relative measures of inequality, the absolute gaps between certain sections of the distribution are also used to measure inequality.[8] In nominal terms, the gap between the least wealthy 25% and the wealthiest 25% of the net wealth distribution widened to around €390,000 (2021: €338,000). However, if inflation is taken into account, this gap did not grow between 2021 and 2023, but remained at €287,000 in real terms in both years. The gap between the wealthiest 10% and the least wealthy 10% of households in terms of net wealth actually narrowed after adjustment for inflation.[9]

Using the data from the PHF study, it is possible to not only calculate the aforementioned measures of inequality but also look at the overall distribution. Here, too, it is important to distinguish between nominal and real values.

In nominal terms, the values that divide the net wealth distribution into ten equal sections (deciles) have risen over almost the entire distribution, but each of these values declines markedly after adjusting for inflation. As previously mentioned, the median in 2023 was lower, at €103,200 in nominal terms and €76,000 expressed in 2010 prices. As in previous waves of the survey, the largest absolute increase in nominal terms was observed for the cut-off value between the wealthiest 10% of households and the rest of the net wealth distribution (+7%). In 2010 prices, by contrast, it fell by around 7%. A closer look at households with net wealth below the median shows that only the cut-off value between the least wealthy 10% of the distribution and the rest of the distribution fell marginally in nominal terms.[10] At €800, however, it remains well above the figure from 2017 (€100).

Net wealth distribution of households in Germany

Methodology of the PHF study 2023

The fifth PHF survey was conducted between May 2023 and February 2024 and is comparable to previous waves. As in the surveys in 2010-11, 2014 and 2017, in 2023 household interviews could be conducted face to face with respondents on location. In the 2021 survey, telephone interviews with households had also been an option due to contact restrictions in connection with the coronavirus pandemic. Very few adjustments were made to the questionnaire compared with previous waves, and the survey is therefore comparable with previous surveys both in terms of questions and concept.

The period for conducting interviews was extended in order for a sufficient number of interviews to be completed. The survey, which was originally scheduled for May to October 2023, ultimately ran between May 2023 and mid-February 2024. Despite extending the field phase, the number of interviews conducted fell short of expectations. There was once again only limited success in getting new households that had never been interviewed before to take part in the survey. As a result, households that had already participated in at least one of the surveys in 2010-11, 2014, 2017 or 2021 make up 78% of the completed interviews. This is a similar level to that observed during the coronavirus pandemic when it was difficult to approach new households in person to take part in the study on account of contact restrictions. In the waves prior to the pandemic, the share of households that had been surveyed repeatedly had been much lower, at around two-thirds. The table below provides an overview of the composition of the completed interviews in each survey year as well as the participation rates of individual groups.

**Table 2.2: Composition of completed interviews by survey wave**
Group	Number of participants
Group	2010-11	2014	2017	2021	2023
Households (total)	3,565	4,461	4,942	4,119	3,985
Individuals aged 16 and over (total)¹	6,661	8,349	9,165	6,852	6,621
Households surveyed repeatedly (“panel” and “split”)	-	2,191	3,335	3,434	3,112
Households surveyed for the first time (“refresher”)	-	2,270	1,607	685	873
Percentage participation rates²
Total	17.3	27.9	30.8	25.9	27.7
“Panel” and “split”	-	68.4	66.6	56.8	56.8
„Refresher“	-	17.7	14.5	6.9	10.0
1 These figures correspond to the number of completed in-person interviews. The actual number of household members is higher as not all individuals in every household were available for interview. 2 Share of completed and usable interviews in the gross sample.

As in the 2021 survey, in 2023 a sampling concept was employed in which households resident in the eastern federal states were overrepresented in the (gross) sample (“oversampling”). As a result, wealthy households were oversampled to a lesser extent than in the waves up to 2017.[1] However, this does not mean that only a small number of wealthy households were included in the sample. In the group of households that have been surveyed repeatedly, wealthy households as a proportion of all households had been rising steadily over time, so it seemed unnecessary to significantly oversample wealthy households again for the fourth and fifth wave. The higher probabilities of selection for wealthy households and households in eastern Germany in the current waves and for wealthy households in the 2010-11, 2014 and 2017 waves were taken into account in the weighting, meaning that the results shown can be regarded as representative of households in Germany.

No adjustments were made to any other aspects of the methodology. In 2023, the target population again included households with at least one person over 18, but did not include people living in collective households (e.g. retirement homes, student halls of residence and refugee homes) or institutions (e.g. monasteries or prisons). The addresses of households approached for the first time were once again selected randomly from lists held by residence registration offices. The data were collected using electronic questionnaires. The just under 250 trained interviewers required a little over an hour on average to complete an interview. Go to bundesbank.de/phf-en for further information on the methodology and background of the PHF study.

When drawing the sample, income tax statistics were used to divide smaller municipalities with fewer than 100,000 residents into “rich municipalities” and “other municipalities”. In cities with 100,000 residents or more, wealthy street sections are identified using micro-geographic information on location and purchasing power. Finally, the proportion of households in the sample is selected such that households in wealthy municipalities and wealthy street sections are oversampled relative to their numbers in the population.

Average wealth holdings in different sections of the net wealth distribution paint a mixed picture. Whilst average wealth has fallen in both nominal and real terms for households with net wealth below the median,[11] it has seen a nominal rise in the middle of the distribution and remained virtually constant at the top end. After adjusting for inflation, however, there is a decline across the entire distribution. The observed decrease in corporate valuations coupled with only slightly higher share prices and declining real estate prices for existing properties are likely to have been an important factor in the sluggish dynamics and the slight decline in net wealth at the top end of the distribution (around -1%). These three types of asset tend to be held by wealthier households.

The underrepresentation of very wealthy households in the PHF surveys also plays a role when interpreting developments at the top end of the wealth distribution. Very wealthy households, which are represented to varying degrees in the individual waves of the survey, can have an impact on the mean measured for the wealthiest households even after weighting. According to the Bundesbank’s distributional wealth accounts, which seek to offset this underrepresentation, the top decile of the net wealth distribution shows net wealth growth of around 12% between 2021 and 2023.[12] After adjusting for inflation, however, net wealth would still fall slightly during this period for this decile, too.

Change in average net wealth between 2021 and 2023

Distributional wealth accounts for households in Germany

Microdata and macro statistics reveal gaps in the data on the wealth situation of households in Germany. There are two key sets of Bundesbank statistics which provide information on the wealth situation of households in Germany. First, there is the Panel on Household Finances (PHF) survey, which gives detailed information on the individual wealth and debt situation of the households surveyed. Second, the national accounts statistics provide aggregate information on the amount and structure of wealth in the entire household sector. Although both sets of statistics describe the wealth situation in the household sector, there is a considerable gap in wealth reporting between the aggregated and extrapolated PHF microdata and the national accounts statistics. To close these data gaps as far as possible, experts from the European System of Central Banks have been collaborating in various working groups since 2015 to explore how best to link data from the household surveys with the national accounts statistics for the household sector within a consistent analytical framework.[1]

The wealth concept needs to be adjusted so that due account can be taken of the differences (in terms of concepts and definitions) between the PHF and the national accounts statistics. Conceptual and methodological differences are one factor explaining why the two sets of statistics capture the wealth situation differently. These include differences in the definition of the population or periodicity and timeliness of the statistics. However, differing valuation concepts between the two sets of statistics can also be a decisive factor.[2] Differences in how certain types of assets and liabilities are defined play a role. The distributional wealth accounts accommodate all these differences by including only those wealth components which are sufficiently comparable between the two sets of statistics.[3] Ultimately, this approach results in a definition of wealth which – as measured by the national accounts statistics for the household sector – covers around 90% of households’ total assets.

Despite using an adjusted wealth concept, when it comes to net wealth there is still a considerable gap between the aggregate of the household survey and the corresponding data from the national accounts statistics. On average, the net wealth recorded in the household survey over the first four PHF waves is around €2,000 billion (20%) lower than the level in the national accounts statistics.[4] A major factor here is that the PHF study does not adequately cover very wealthy households. While the national accounts statistics provide aggregate information on the amount and structure of wealth in the household sector as a whole, very wealthy households are typically not sufficiently represented in the actual samples of the PHF.[5]

The distributional wealth accounts link microdata with national accounts statistics. One key component in the preparation of the distributional wealth accounts is to add the missing very wealthy households to the original PHF dataset using a “rich list”. The observations on net wealth from this list supplement the PHF dataset. Since these data only take into account the top tail of the net wealth distribution, synthetic wealthy households are also estimated, which then supplement the original PHF dataset as well. The net wealth of these synthetic households lies between that of the members of the rich list and the wealthiest households included in the wealth survey.[6] Ultimately, the distributional wealth accounts resulting from the overall adjustments contains valuable information from the combination of both sets of statistics: this dataset takes into account the distributional information from the household wealth survey at the individual household level, as well as the quarterly dynamics and levels of the national accounts statistics for the period since 2011. In this context, the distributional wealth accounts rank households based on their level of net wealth, which is broken down into the following types of assets and liabilities: deposits, debt securities, listed shares, investment funds, insurance claims, financial and non-financial business wealth, housing wealth, and liabilities in the form of mortgages and other debt. In addition, a household’s net wealth is ultimately calculated as the difference between its total assets and its liabilities. It should be noted, in particular, that entitlements to statutory pensions are not taken into consideration here under the standard approach. Thus far, neither the financial accounts nor wealth surveys have comprehensively captured these types of asset.

Net wealth distribution

Distributional wealth accounts for the household sector show that wealth inequality is high overall in Germany. Over the lifetime of the dataset – that is, since 2011 – the top 10% of the wealth distribution have, on average, held more than 60% of the total net wealth of households in Germany. During the same period, the less wealthy half of the wealth distribution accounted for an extremely small share, averaging 2.1%. That said, the distribution did shift slightly in favour of the less wealthy half of households over the observation period. The share of total net wealth held by the less wealthy 50% of households rose from around 2% in 2011 to more than 2.4% in 2023.[7] Similarly, the evolution over time of the Gini coefficient – a measure of net wealth inequality – suggests that a slight decline has taken place, especially since 2014.

Wealth inequality in Germany is still fairly high even by international standards. According to the distributional wealth accounts for the euro area, the Gini coefficients among Member States range from 57% to 77%. With a coefficient of just over 76%, Germany is towards the top of the ranking (see Chart 2.8). However, when comparing countries, it should be noted that the underlying definition of wealth used for the distributional wealth accounts does not include entitlements to statutory pensions. If statutory pensions are factored into the calculation, Germany’s net wealth inequality is significantly lower than if they are not taken into consideration.[8]

Wealth inequality in Germany and the euro area

For further details on the concept and methodology, see Deutsche Bundesbank (2022), Engel et al. (2022), and European Central Bank (2020, 2024).
The survey data generally reflect households’ own assessments of different assets and debt, while the national accounts statistics value the individual types of wealth based on their market prices.
Wealth positions that are sufficiently recorded in only one of the two underlying sets of statistics are excluded. This means that “non-life insurance reserves”, for example, are omitted because these currently only form part of the financial accounts. Entitlements to statutory pensions are not included either.
To date, the distributional wealth accounts have been based on PHF survey waves one to four. When the fifth survey wave is published, suitable adjustments will be made to ensure that, going forward, distributional information from the fifth survey wave can also be fed into the preparation of the distributional wealth accounts.
See European Central Bank (2020, 2024).
By adding the missing highly wealthy households to the original data from the household survey, this data gap can be closed by an average of more than 50%.
This development came about because net wealth saw especially strong growth in the less wealthy half of the distribution in particular, albeit from an extremely low level. Differences in net wealth growth along the net wealth distribution are a result of unevenly distributed forms of wealth. While wealth growth among the less wealthy half of the distribution was dominated by financial portfolios, the upper middle of the distribution benefited to a comparatively large degree from an increase in the value of housing wealth. Finally, the growth in wealth seen in the top 1% of the distribution is partly attributable to increases in business wealth, alongside housing wealth. For more information, see inter alia Adam and Tzamourani (2016) and Deutsche Bundesbank (2022, 2023).
See Bartels et al. (2023b).

The decline in average net wealth at the bottom end of the distribution is due, amongst other things, to an increase in outstanding amounts for mortgage loans, which are not offset by correspondingly high real estate values.[13] This could be partly due to the decline in real estate prices for existing properties. Looking at financial asset holdings less unsecured loans, there are hardly any differences between 2021 and 2023 for the least wealthy 20% of households.[14] The additional savings accumulated by less wealthy households during the COVID-19 pandemic[15] do not appear to have been fully used up yet, and the outstanding amounts of unsecured loans among less wealthy households have also risen only slightly.[16]

**Table 2.3: Asset holdings and debt – least wealthy 20% of households, nominal**
Item	2017		2021		2023
Item	Prevalence (%)	Conditional mean (€)	Prevalence (%)	Conditional mean (€)	Prevalence (%)	Conditional mean (€)
Net wealth	100	- 6,800	100	- 3,100	100	- 10,700
Financial assets	98	2,700	99	3,400	99	3,200
Savings and current accounts	98	1,300	98	2,200	99	1,800
Real assets	45	15,500	49	17,700	55	22,200
Owner-occupied housing	4	105,500	2	109,400	3	221,500
Debt	54	29,700	54	28,000	47	55,700
Loans secured by main residence	3	147,900	–	–	3	283,900
Unsecured loans	53	11,300	52	10,400	45	13,400
Sources: PHF 2017, PHF 2021, PHF 2023.

Wealth mobility

In addition to the wealth distribution at a given point in time, households’ mobility within the distribution over time – their wealth mobility – is also an important aspect to consider when analysing household finances. Wealth mobility provides information on the degree to which households move up or down in the wealth distribution over time, including how much the composition of household types changes over time in the individual parts of the distribution. Factors such as educational decisions, labour income, entrepreneurial activity and changes in household composition can contribute to changes in position within the distribution across the life cycle (see also Chart 2.13).

The following wealth mobility analysis is based on households whose reference person was between 20 and 69 years old in the first wave of the survey in 2010-11. It divides households into ten-year cohorts based on the age of the reference person in the first wave (ages 20-29, 30-39, 40-49, 50-59, 60-69). The next step is to calculate the relative wealth position of households within these age cohorts in each wave, i.e. the age-specific wealth position. This controls for the natural life cycle, i.e. the effect whereby, on average, younger households start with a low level of wealth, which then rises.

Wealth mobility can be measured using a variety of methods. One approach is to use a household’s position in the net wealth distribution in a base year (2010-11) to determine its probability of transitioning to other positions at a later point in time (2023). To achieve this, the age-specific wealth distribution for the two years is divided into five equal sections, each containing 20% of households (“quintiles”). This analysis only includes households that participated in the survey in both 2010-11 and 2023, as these are the only households whose position in the wealth distribution can be traced over time.

Households in wealth quintiles: transitions between the surveys in 2010-11 and 2023

Most mobility occurs in the middle of the wealth distribution.[1] Chart 2.9 shows the share of households that had moved from a given wealth quintile in the 2010-11 survey to another quintile or were in the same quintile 12 years later. The x-axis shows the original quintile and the y-axis the share of households from that original quintile in each quintile in 2023. Across all starting quintiles in 2010-11, households were more likely to stay in the same quintile than move to a specific other quintile. At the same time, however, there are also differences along the wealth distribution. While for the top and bottom quintiles in 2010-11 the probability of remaining in the same quintile is above 50%, the mobility for households in the middle of the wealth distribution is higher. These households are thus more likely to rise or fall in the wealth distribution than to remain in their original quintile. In addition, along the entire distribution, households are much more likely to move to neighbouring quintiles than jump to quintiles that are further away, despite the two points in time being more than ten years apart. The probability of a household switching from the bottom 40% of the distribution to the top 20%, or vice versa (from the top 40% to the bottom 20%), is particularly low.

Wealth mobility by age cohort: rank-rank correlation

Another common method for measuring mobility is to calculate households’ wealth positions in waves 1 to 5 of the PHF and then to determine the correlation of households’ age-specific wealth positions between the individual waves (“rank-rank correlation”). A correlation of 0 means that wealth in the last wave does not play a role in today’s wealth (perfect mobility). By contrast, a value of 1 means that there were no changes in wealth positions (complete immobility – this corresponds to the 45-degree line in Chart 2.10). The chart summarises all 5 waves, broken down by age group. It turns out that wealth mobility is highest for the youngest age group (20-29) (the line is the flattest) and then decreases with age (the lines approach the 45-degree line). As well as enabling observation of mobility across the life cycle, the fact that data is now available for five waves means that wealth mobility can be analysed over time. For this purpose, the data are broken down by survey wave. The analysis shows that mobility has remained stable over time. The periods 2010-11 to 2014, 2014 to 2017, 2017 to 2021 and 2021 to 2023 all produce a rank-rank correlation of 0.82.

This is consistent with recent results for Norway (see Audoly et al., 2024).

2 The wealth of particular groups of households

The previous section addressed developments for households in individual sections of the wealth distribution. The PHF data also allow wealth to be analysed for different types of household.[17]

Past surveys have already established that ownership of real estate and businesses is a good indicator for a household’s level of wealth. This continues to be the case. Households that own real estate have significantly higher net wealth than tenant households and are therefore found more frequently above the middle of the wealth distribution. This applies both to households that are still servicing mortgages as well as to those that have already paid them off.[18]

Share of owner and tenant households by net wealth

In 2023, the median net wealth of households that own real estate was significantly higher than the median for tenant households. Households that own real estate and had already paid off their mortgages had a median net wealth of €450,200, while households that own real estate but still had outstanding mortgage loans had a median net wealth of €379,900. By contrast, this figure stood at €18,300 for tenant households. All of these figures have risen in nominal terms compared with 2021. Households that own a business[19] also had significantly higher net wealth than households that do not own a business. On average, households that own a business had more than €1 million in net wealth, with the median amounting to around €585,000. The majority of these households’ wealth is also directly attributable to their business ownership.

When looking at the breakdowns of households by only one characteristic, it should be noted that a given characteristic may often also be associated with other characteristics that are relevant to wealth. For instance, real estate owners and tenant households generally differ not only in terms of real estate ownership, but also with regard to aspects including household size, income, age and marital status. This should also be taken into account when looking at wealth in individual regions.

Significant differences in net wealth between eastern and western Germany remain apparent. However, there are also marked differences within western Germany. Although mean wealth in eastern Germany rose particularly sharply in nominal terms between 2021 and 2023 compared with the other regions of the country, standing at €170,100 in 2023, it was still well below the mean in western Germany (€364,900). At €35,900, the median also remains considerably lower than in western Germany (€143,200). The different rates of home ownership (western Germany 45%; eastern Germany 29%) certainly play an important role in this. Within western Germany, the southern federal states stand out. For the group of federal states consisting of Bavaria, Baden-Württemberg and Hesse, mean and median net wealth came to €442,800 and €188,800 respectively.

Net wealth of households by region

The PHF survey analyses the wealth of a household and not the wealth of the individuals within that household. The composition of a household and its size are therefore generally closely related to the household’s wealth. For example, mean and median net wealth of single-person households (€221,800 and €37,700, respectively) are significantly lower than those of households comprising multiple persons. Looking at multi-person households somewhat more closely, it can be seen that households comprising couples with and without children have similar amounts of wealth (€409,700 without children, €453,500 with children). Compared with households comprising couples and single persons without children, the wealth of single parents with at least one child is significantly lower, standing at €96,900.

Although the PHF study focuses on household wealth, it also examines structures that are related to the characteristics of individuals. In order to classify a household, a reference person is used. This person is defined as the person with the highest income in the household.[20]

A breakdown of households’ net wealth by the age of the reference person defined in this way reveals the typical life cycle pattern: wealth grows over the life cycle and then decreases again in old age. Chart 2.13 also highlights the relationship with income. During the years in which the majority of persons are in employment and incomes tend to rise, they build up wealth, which is then consumed in old age, when incomes fall. However, the composition of households and other household characteristics correlated with wealth may change with age as well.

Net wealth and income of households by age of reference person

Another way of breaking down net wealth is to look at the characteristics of households in certain sections of the wealth distribution. As an example, the table below illustrates three sections of the wealth distribution and the rates of real estate and business ownership as well as the age and gender of the reference persons. Here, the structures show a high degree of stability over time. As mentioned above, high wealth is strongly correlated with ownership of real estate and businesses in all years. In addition, older reference persons tend to live in high-wealth households. As wealth increases, there is a decline in the share of female reference persons, i.e. the person with the highest income in the household.

**Table 2.4: Selected characteristics of households in various sections of the net wealth distribution**
Group	Share of home owners (%)					Share of households with business assets (%)
Group	2010	2014	2017	2021	2023	2010	2014	2017	2021	2023
Bottom 20%	4	6	4	2	3	1	2	2	1	3
Middle 60%	42	42	42	42	38	7	7	7	6	5
Top 20%	92	88	90	93	91	23	24	22	22	22
Total	44	44	44	45	42	9	9	9	8	8
Group	Age of reference person (mean in years)					Share of female reference persons
Group	2010	2014	2017	2021	2023	2010	2014	2017	2021	2023
Bottom 20%	45	48	47	49	47	46	46	47	47	43
Middle 60%	53	53	53	53	54	33	35	34	37	35
Top 20%	59	59	59	60	59	23	28	24	27	25
Total	53	53	53	54	54	34	36	35	37	35
Sources: PHF 2010-11, PHF 2014, PHF 2017, PHF 2021, PHF 2023.

3 Structure of wealth

The composition of net wealth in individual sections of the wealth distribution can influence wealth developments and measures of inequality. The structures of wealth and debt are important for the analysis of monetary policy transmission and questions relating to financial stability. In addition, different types of investment generate different returns, which can have an impact on the growth of wealth holdings.

The share of households that make use of particular forms of investment and hold non-financial assets is changing only slowly overall. The slight trend towards increased ownership of shares observed during the coronavirus pandemic continued between 2021 and 2023. The percentage of households that own shares and the percentage of households that own fund units increased by 3 percentage points each, from 15% to 18% for ownership of shares[21] and from 21% to 24% for ownership of fund units (excluding private retirement provision). At the same time, the share of households that deposit money into savings accounts decreased from 71% to 67%. The share of households that have concluded at least one private pension contract (including cash value life insurance policies) from which they are not yet receiving payments also went down (-3 percentage points). Following this further decline, this share now stands at 39%, which is around 8 percentage points below the level seen in the first PHF survey in 2010-11 (47%).[22]

Between 2021 and 2023, the percentage of households that own their main residence declined from 45% to 42%.[23] At the same time, the mean and median values of real estate have risen only marginally. When assessing the value of their own real estate, it is likely that property-owning households had not yet gained a full picture of the aggregate decline in real estate prices for existing properties.[24]

**Table 2.5: Portfolio structure of households in Germany**
Position	Share of households (%)					Mean (conditional) in €					Median (conditional) in €
Position	2010	2014	2017	2021	2023	2010	2014	2017	2021	2023	2010	2014	2017	2021	2023
Real assets	80	81	83	83	84	218,600	229,500	232,800	323, 400	323,200	89,200	90,900	106,900	135,300	147,700
Ownership of main residence	44	44	44	45	42	205,800	231,400	258,800	343,200	376,300	168,000	162,000	199,200	278,800	299,500
Vehicles and valuables	73	75	78	78	79	13,000	13,300	13,600	15,300	19,100	7,780	7,000	8,000	8,900	10,800
Business assets	10	10	10	8	8	333,600	338,800	309,900	502,800	359,400	20,000	21,600	26,600	48,700	45,700
Financial assets	99	99	99	100	100	47,400	54,200	56,800	77,900	87,400	17,100	16,450	16,900	25,900	27,400
Current accounts	99	99	99	99	100	3,400	4,300	7,100	12,700	12,200	1,200	1,100	1,800	3,000	3,000
Saving accounts (excl. private retirement provision)	78	72	70	71	67	22,500	29,400	27,600	30,900	35,500	9,700	8,900	9,900	11,800	13,300
Private retirement provision (incl. life insurance policies)	47	46	43	42	39	27,200	28,300	33,200	42,100	44,500	11,400	13,500	15,400	20,000	20,300
Mutual fund shares (excl. private retirement provision)	17	13	16	21	24	29,000	39,800	37,500	44,600	58,000	10,000	14,800	12,900	15,900	19,300
Shares	11	10	11	15	18	29,100	38,700	43,700	65,100	62,400	8,600	9,800	9,900	14,400	14,500
Debt	47	45	45	41	39	56 900	57,000	65,200	72,400	83,600	12,620	15,200	19,800	17,800	19,700
Mortage debt	21	20	21	18	18	110,200	111,100	125,100	148,400	166,100	80,000	76,400	81,000	84,500	93,600
Unsecured loans	35	33	33	29	26	9,600	9,500	10,800	11,300	12,700	3,170	3,500	4,900	5,500	6,000
Sources: PHF 2010-11, PHF 2014, PHF 2017, PHF 2021, PHF 2023.

Lastly, the significant decline in mean business wealth is noteworthy. In nominal terms, mean business wealth has now roughly returned to its pre-pandemic level. It had risen significantly between 2017 and 2021. The decline in mean business wealth between 2021 and 2023 dampened developments in net wealth at the upper end of the distribution. It was also largely responsible for the fact that there was no growth in non-financial assets between 2021 and 2023 in nominal terms, despite the rise in real estate wealth. By contrast, at €45,700, median business wealth remained at a similar level as during the pandemic and was thus still significantly higher than before its outbreak.[25]

Both mean and median household financial assets rose in nominal terms. The large current account balances accrued between 2017 and 2021 have therefore not yet been fully used up. Balances on savings accounts were also up slightly for households that used this type of investment, although the share of households with savings accounts had fallen. The significant rise in interest rates since 2021 may have contributed to this. The value of households’ share portfolios has barely declined in nominal terms, while, on average, households hold higher nominal amounts in investment funds than in 2021.

The importance of real estate wealth and business ownership for the development of wealth in the top half of the wealth distribution is illustrated by the following two charts. The importance of real estate ownership increases moving along the wealth distribution from households with low wealth to those with high wealth. Only for the wealthiest 10% of households does business wealth account for a significant portion of net wealth. For households in the bottom half of the net wealth distribution, non-financial assets (excluding real estate) are also important, although these consist mainly of vehicles and valuables and not business wealth.

Breakdown of household wealth by amount

Breakdown of gross household wealth along the net wealth distribution

There are significant differences in investment behaviour along the distribution of net wealth. Investments with low risk and low yields, such as balances on savings and current accounts or fixed-term deposits, dominate among less wealthy households. As wealth increases, there is an additional focus on riskier forms of investment, such as shares and silent participations in enterprises. These types of investment are generally able to generate higher returns. However, even for the top 10% of households, the value of shares and silent participating interests amounts to only around 43% of total financial assets. In other words, households in Germany tend to invest the majority of their financial assets in liquid and low-risk forms of investment.

Inflation and wealth

The academic literature often uses what is known as the net nominal position to estimate how strongly households may be affected by inflation due to their wealth and debt structure.[1] The first step in calculating the net nominal position is to add up the balances on savings and current accounts and the value of bonds, and then subtract the outstanding debt from this total (“direct nominal position”). The next step is to add amounts held indirectly via securities ownership (“indirect nominal position”). The ratio of the net nominal position to net wealth (NNP/NW) then serves as a measure of the inflation risk that households face.[2] This method is also used by Adam and Zhu (2016), whose approach for the first PHF wave in 2010-11 is applied here to the data from the 2023 survey.

Overall, in 2023, only around 9% of households have no or low inflation risk due to their portfolio structures. Inflation risk is assumed to be low if the ratio (NNP/NW) is between -0.05 and 0.05. Around 15% of households have a negative net nominal position (NNP/NW below -0.05) or have negative net wealth, meaning that they would potentially benefit from inflation. By contrast, the majority of households (76%) have positive net nominal positions (NNP/NW higher than 0.05). For around 16% of households, almost all assets held in 2023 are exposed to inflation risk (NNP/NW between 0.95 and 1).

Along the wealth distribution, there are differences in the inflation risk resulting from households’ wealth structures. For households in the lower half of the net wealth distribution, the net nominal position makes up around 25% of net wealth in 2023. This share is thus higher than for households in the fifth to ninth decile (18%). The latter group of households probably faces lower inflation risk because a large part of their wealth is tied up in real estate, which is set against mortgage debt. For the wealthiest 10% of households, inflation risk (NNP/NW: 22%) lies between that of the lower half and the fifth to ninth decile of the wealth distribution.

**Table 2.6: Net nominal position relative to the average net wealth in 2023, according to position in the net wealth distribution, as a percentage**
Percentiles of the net wealth distribution	Percentage share
Upper 10%	22.3
50 - 90%	17.7
Lower half	24.7
Total	21.7
Sources: PHF 2010-11, PHF 2023, Bundesbank calculations.

See Doepke and Schneider (2006); Meh and Terajima (2011) and Adam and Zhu (2016).
If the share is zero, households are not exposed to any inflation risk on average. The closer the share is to one, the higher the risk. The net nominal position, and thus also the share, can also turn negative. This then indicates that households would potentially benefit from inflation.

4 Saving

Developments in wealth holdings are influenced by both gains and losses in value for certain types of asset as well as by households’ saving and debt behaviour.

In 2023, the majority of households in Germany reported that they saved at least occasionally. At 83%, this share is only slightly below the figure from 2021. During the coronavirus pandemic, the share of households that saved at least occasionally rose to 85%. Before the pandemic, in 2017, this figure was 80%. In 2021, there was a particular decrease in households that claimed they could not save due to limited financial leeway. This share has now risen again somewhat, from 11% in 2021 to 13% in 2023.

Saving behaviour of households in Germany

The most important motives for saving among households in Germany remained fairly stable overall, both compared with 2021 and over the long term. The three most important motives for saving in all years were “safety net for emergency situations”, “major purchases” and “retirement”. In 2023, 68% of households cited one of these three motives for saving.

Looking at the motives for saving on an individual basis, there have been slight changes compared with 2021. Consistent with the decline in the share of households with private pensions, the share of households that cited retirement as their main motive for saving also decreased somewhat (2021: 22%, 2023: 20%). The other motives for saving also saw small changes. As expected, the share of households that reported saving for travel and holidays as their main motive rose again after the pandemic. In 2021, around 7% cited this motive. In 2023, this share had increased to almost 10%, thus returning to its level from 2017.

Most important motive for saving

5 Households’ debt situation

The key metrics and structures regarding household debt in Germany remained unchanged between 2021 and 2023. Much like the developments in nominal wealth, the developments in debt are also characterised by a lack of dynamism.

The share of households with outstanding debt fell by just over 2 percentage points between 2021 and 2023. In 2023, 39% of households had outstanding debt, putting this share just below the 40% mark for the first time since the start of the PHF surveys. It was 41% in 2021 and 45% in 2017. The decline between 2021 and 2023 is mainly attributable to households with unsecured loans. Around 26% of households held unsecured loans in 2023, compared with 29% in 2021. This development could be a result of the coronavirus pandemic, during which less wealthy households built up balances on savings and current accounts. Households therefore seem to make use of these balances first instead of taking out consumer credit and other unsecured loans. The higher level of interest rates is also likely to have reduced the attractiveness of additional loans.

The outstanding amounts for unsecured loans have remained virtually unchanged. In 2023, the mean outstanding amount for these loans stood at €12,700 (2021: €11,300) and the median at €6,000 (2021: €5,500) in nominal terms. The mean outstanding amounts in 2023 are equivalent to around €11,000 in 2021 prices. The same is true for mortgages. In nominal terms, both the mean and the median rose between 2021 and 2023, but fell slightly when adjusted for inflation. The number of households with secured loans remains lower than the number of households with unsecured loans. The average outstanding amounts for mortgage loans, however, significantly exceed those for unsecured loans.

For debt, it is not only the analysis along the distribution of wealth that is relevant, but also the analysis along the distribution of income. Thus, when assessing financial stability, it is important to know whether households are able to service their outstanding debt reliably. Otherwise, lenders will need additional provisions or even write-downs. If such cases increase and affect a large number of banks, this can impair financial stability.

The share of households with debt increases with gross household income. In 2023, around one-quarter of households in the bottom quintile of the income distribution had outstanding credit debt, compared with more than half (55%) of households in the top quintile. The differences are not as pronounced along the net wealth distribution. The proportion of indebted households is marginally higher among less wealthy households (46%) than among wealthier households (41%). While these shares are very similar, the structure of their loans differs significantly. Unsecured loans, such as consumer credit or overdraft loans, are the dominant type of loan among lower-income and less wealthy households. By contrast, mortgages predominate among higher-income and wealthier households. Households with higher incomes and higher net wealth generally also have correspondingly larger amounts of outstanding loans. Taken together, these structures suggest that larger amounts of loans are also correlated with larger amounts of assets and that households with high outstanding debt generally have sufficient financial resources to service the associated principal and interest payments.

Since the start of the PHF surveys in 2010, the share of principal and interest payments in the net household income of indebted households has averaged between 17% and 23%. This figure was 18% in 2023 and 17% in 2021. As shown by the chart below, not only the mean but also the overall distribution of the share of debt service in net income is very stable. Only the very top of the distribution has seen a slight decline since 2021. The share of households that spend more than 30% of their income on principal and interest payments has also declined, falling from 15% of all indebted households to 13%. The rise in nominal income certainly contributed to these developments. A significant rise in interest rates would otherwise have led to an increase in debt service as a share of net income, at least for loans without longer-term interest rate fixation periods. In addition, while mortgage interest rates for new business saw a particular rise in interest rates, households with existing mortgage loans and ongoing interest rate fixation are likely to still be benefiting from the low interest rates of the 2010s.

Distribution of debt service as a share of net income for indebted households

6 Conclusion and outlook

The results for 2023 confirm many findings from previous years regarding the distribution of net wealth in Germany. Overall, the distribution is very stable and inequality remains high, even by European standards. The amount and composition of wealth has changed in individual sections of the distribution. However, when looking at the distribution as a whole, these different developments have more or less offset each other.

On balance and over the longer term, investment behaviour and the structure of household debt in Germany are fairly stable. Even unexpected events such as the coronavirus pandemic or high inflation rates have not fundamentally changed the wealth distribution and portfolio structures so far. Over the long term, the share of households with shares and fund holdings has increased continuously. It remains to be seen what impact these events and developments will have in the medium and long term. The effects that the high inflation rates and the associated real decline in wealth holdings will have on the consumption behaviour of different households also remain to be seen. This is another reason why the PHF surveys will continue, with a further survey planned for 2026.

The results of the PHF study help us to obtain a better understanding of the financial situation of households in Germany. However, other factors, such as income, the design and performance of the social security system, and the education system, also play a role in a more comprehensive analysis of the financial situation or (financial) well-being of households in Germany.

Table appendix

List of references

Adam, K. and P. Tzamourani (2016), Distributional consequences of asset price inflation in the euro area, European Economic Review, Vol. 89, pp. 172-192.

Adam, K. and J. Zhu (2016), Price level changes and the redistribution of nominal wealth across the euro area, Journal of the European Economic Association, Vol. 14 (4), pp. 871-906.

Audoly, R., R. McGee, S. Ocampo and G. Paz-Pardo (2024), The life-cycle dynamics of wealth mobility, IFS Working Paper, No 24/12.

Banca d'Italia (2025), Survey on Household Income and Wealth – 2022, Historical Tables, 10 January 2025.

Banco de España (2025), Spanish Survey of Household Finances (EFF)udfs_dwh_r2025jun_revised_UNGÜLTIG.

Bartels, C., T. Bönke, R. Glaubitz, M. M. Grabka and C. Schröder (2023a), DIW Berlin: Rentenvermögen macht Großteil des Vermögens der ärmeren Bevölkerungshälfte in Deutschland aus, DIW Wochenbericht No 45/2023, pp. 625-631.

Bartels, C., T. Bönke, R. Glaubitz, M. Grabka, and C. Schröder (2023b), Accounting for pension wealth, the missing rich and under-coverage: a comprehensive wealth distribution for Germany, Economics Letters, Vol. 231.

Bönke, T., M. M. Grabka, C. Schröder and E. N. Wolff (2020), A Head-to-Head Comparison of Augmented Wealth in Germany and the United States, The Scandinavian Journal of Economics, Vol. 122(3), pp. 1140-1180.

Deutsche Bundesbank (2025), Balance Sheets – Tables: Distributional Wealth Accounts (DWA) for households in Germany, 19 February 2025.

Deutsche Bundesbank (2024), Distributional wealth accounts: timely data on the distribution of household wealth, Monthly Report, April 2024.

Deutsche Bundesbank (2023), Household wealth and finances in Germany: results of the 2021 survey, Monthly Report, April 2023, pp. 25-58.

Deutsche Bundesbank (2022), Distributional Wealth Accounts for households in Germany – results and use cases, Monthly Report, July 2022, pp. 15-38.

Deutsches Aktieninstitut (2024), Aktionärszahlen 2024: Zeit für politische Impulse.

Doepke, M. and M. Schneider (2006), Inflation and the Redistribution of Nominal Wealth, Journal of Political Economy, Vol. 114 (6), pp. 1069-1097.

Engel, J., P. G. Riera, J. Grilli, and P. Sola (2022), Developing Reconciled Quarterly Distributional National Wealth – Insight into Inequality and Wealth Structures, ECB Working Paper Series, No 2687.

European Central Bank (2024), Introducing the Distributional Wealth Accounts for euro area households, pp. 94-113.

European Central Bank (2020), Understanding household wealth: linking macro and micro data to produce distributional financial accounts, ECB Statistics Paper Series, No 37.

Federal Ministry of Labour and Social Affairs (2024), Alterssicherungsbericht 2024 – Ergänzender Bericht der Bundesregierung zum Rentenversicherungsbericht 2024 gemäß § 154 Abs. 2 SGB VI.

Kuhn, U. (2020), Augmented wealth in Switzerland: the influence of pension wealth on wealth inequality, Swiss Journal of Economics and Statistics, Vol. 156(19).

Meh, C. A. and Y. Terajima (2011), Inflation, nominal portfolios, and wealth redistribution in Canada, Canadian Journal of Economics Vol. 44(4), pp. 1369-1402.

Since the 2017 and 2021 surveys were four years apart and the 2021 and 2023 surveys were only two years apart, the smaller increase was to be expected. However, the differences are also apparent when calculating annual rates of growth. These stood at around 8% between 2017 and 2021 and only around 1.3% between 2021 and 2023.
Here, adjustment for inflation was carried out using the Harmonised Index of Consumer Prices (HICP). The HICP took the following values (base year 2010=100): 2010: 100.0; 2014: 107.1; 2017: 110.1; 2021: 117.8; 2023: 135.8. In order for the nominal values from the surveys conducted between 2014 and 2023 to be expressed in 2010 prices, these were divided by the price index for the respective year. For the sake of simplicity, all assets and debt in a given wave of the survey were adjusted for inflation by dividing by the same price index. See the supplementary information entitled “Inflation and wealth”.
Relative measures of inequality do not change when adjusted for inflation. Absolute measures of inequality, however, can change as a result of this adjustment and are therefore presented in both nominal and real terms.
The share of wealth attributable to the wealthiest households is underestimated here. It is rare for very wealthy households to participate in the surveys and thus be included in the sample. This underreporting is also not offset by weighting the data. The distributional wealth accounts (see the supplementary information entitled “Distributional wealth accounts for households in Germany”) attempt to compensate for this underreporting. According to these statistics, the wealthiest 10% of households held 60.4% of total net wealth in 2011, 62.5% in 2014, 58.8% in 2017, 60.4% in 2021 and 60.5% in 2023 (in the third quarter in each case).
The Gini coefficient generally takes a value between 0% and 100%, where 0% represents perfect equality – all households have equal amounts of net wealth – and 100% represents maximum inequality – all wealth belongs to one household. The closer the figure is to 100%, the more unequal the wealth distribution. If negative values are also included in the calculation (e.g. negative net wealth), the Gini coefficient can theoretically also assume a value of more than 100%.
See Banco de España (2025).
See Banca d’Italia (2025).
These measures have the drawback that, even if the wealth of all households were to increase by the same factor, the gaps between different sections of the distribution would grow while the relative measures would remain unchanged.
Adjusting all assets and debt by only one factor/percentage per survey year results in wealthy households “losing” more wealth in absolute terms than less wealthy households. For example, assuming an inflation rate of 10%, a household with nominal net wealth of €1,000 would have real net wealth of €909, whilst a household with nominal net wealth of €100,000 would have real net wealth of €90,909. The difference between nominal and real net wealth thus amounts in absolute terms to around €90 for the first household and around €9,000 for the second.
From the cut-off value between the least wealthy 20% of households and the rest of the distribution onwards, the respective cut-off values (with the exception of the median) rose in nominal terms.
For illustrative purposes, the bottom five deciles have not been reported separately. Some of the corresponding (nominal) mean values can be found in the table appendix at the end of this article.
Own calculations based on Deutsche Bundesbank (2025).
This does not necessarily mean that less wealthy households have taken out more mortgage loans. In 2023, there were actually households in this section of the distribution whose real estate had decreased in value, which meant that the outstanding loans then exceeded the value of the real estate (see the supplementary information entitled “Wealth mobility”).
In nominal terms, the unconditional mean for “net financial assets” in the bottom quintile of the net wealth distribution stood at -€2,100 in 2021 and -€2,900 in 2023. The fact that the nominal increase in unsecured loans (+€3,000) has not had a larger impact on “net financial assets” is because the share of households with unsecured loans in this section of the net wealth distribution declined from 52% to 45%.
See Deutsche Bundesbank (2023).
It should be noted here that the composition of households in individual sections of the distribution can change. In particular, individual households can change their position in the wealth distribution. See also the supplementary information entitled “Wealth mobility”.
Unless stated otherwise, this section and the following sections refer to nominal values.
Just some of the possible breakdowns are shown here. Further metrics for different groups of households can be found in the table appendix.
Here, business owners constitute only households with at least one member who is self-employed or occupies an active position in management, but not households that only have (silent) participating interests or similar interests in a business.
If several persons in the household have equally high incomes, one of these persons is selected at random as the reference person for the purposes of the breakdowns shown here.
According to Deutsches Aktieninstitut, the number of shareholders rose by 7% from around 12.1 million to 12.9 million between 2021 and 2022, but then fell to 12.3 million in 2023. From 2021 to 2023, the number of shareholders thus rose by around 2%. See Deutsches Aktieninstitut (2024).
This development could be attributable in part to the decline in the number of Riester contracts. See Federal Ministry of Labour and Social Affairs (2024).
The ownership rate calculated in the Federal Statistical Office’s microcensus is taken into account when weighting the PHF data. The percentages shown here therefore largely correspond to those from the microcensus.
According to the Federal Statistical Office, prices for newly constructed residential real estate rose slightly between 2021 and 2023 (the index increased from 140 to 143.5), but prices for existing residential real estate fell (the index decreased from 157.5 to 149.7). The Bundesbank’s residential real estate index shows an increase from 198.1 to 204.7 for the period under review.
Developments in the mean and median suggest that the high mean in 2021 could also be attributable to a handful of very high-value firms. Owing to the small share of households that own a business and the pronounced heterogeneity of corporate valuations (the Gini coefficient for business wealth stands at 83%), this cannot be ruled out entirely.

Monatsbericht – November 2024 76. Jahrgang, Nr. 11