## ANNEX A. ALTERNATIVE SCENARIO WITH MARKET PRICE ADJUSTMENTS ONLY

## Source details

**Canonical URL:** [ANNEX A. ALTERNATIVE SCENARIO WITH MARKET PRICE ADJUSTMENTS ONLY](https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026022-source-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2026/english/wpiea2026022-source-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2026/english/wpiea2026022-source-pdf.pdf.json)

---

### A.1. Alternative assumption and methodology
- Scenario scope:
  - Includes only market price adjustments on financial assets, real assets, and liabilities; excludes behavioral responses on real assets and liability contraction or liquidation.
- Financial-asset assumptions:
  - Deposits: changes use country-level average change in deposits from the European Central Bank's MIR-MFI Interest Rate Statistics, focusing on "Euro-denominated deposits with an agreed maturity of more than one year from euro area households (percentages per annum, rates on new business)."
  - Other financial assets (stocks): changes based on average returns from European financial markets (Bloomberg); assumes households can freely reallocate financial assets across integrated European markets.
- Liabilities:
  - Account for changes in interest rates at country level between the beginning and the end of the sample period.
  - Assume households do not acquire new debt or repay early to adjust to inflationary shocks.
- Real assets:
  - Initial stock valuation from Eurosystem HFCFS dataset.
  - House price valuation uses quarterly Housing Price Index from the European Central Bank, change between 2021:Q3 and 2022:Q2, discounted for inflation consistent with Equation 7 in Section II.B with 훽 assumed equal to 1.

### A.2. Wealth channel and combined well-being effect with financial assets under market-price adjustments only
- Aggregate comparisons with baseline:
  - Average wealth loss in this alternative (financial assets only) is 0.7 percentage points lower than the baseline scenario.
  - Average welfare loss (wealth channel included across countries and quintiles) is -6.4 percent of annual household income under this assumption, compared to -7 percent of annual household income in the baseline.
- Distributional heterogeneity and drivers:
  - In 8 countries (example countries: Ireland, Latvia, the Netherlands) where stock valuations exceed the European average, the average wealth loss across quintiles is lower in the baseline than in this alternative.
  - Baseline scenario generally yields a less negative wealth effect for lower income quintiles (they are typically net creditors with limited debt), while higher quintiles tend to benefit relatively more under the market-price-only alternative.
- Aggregate table means (reported values from Annex):
  - Table A1 (wealth channel, financial assets, percent of annual household income): -3.4, -1.8, -4.0, -2.3, -0.6, 0.6, 0.7, 2.3, -3.4, 0.2.
  - Table A2 (total well-being, financial assets only, percent of annual household income): -14.8, -7.7, -14.8, -8.6, -11.3, -5.6, -9.9, -3.9, -13.8, -5.9.
- Notable outcomes:
  - With predominantly negative income and consumption basket effects, total welfare results in an average loss across income quintiles for most countries, with exceptions (Cyprus and the Netherlands).
  - The average welfare effect across countries and quintiles shows an inverted U-shape when only financial assets are considered.

### A.3. Combined well-being effect including real assets under market-price adjustments only
- Aggregate comparisons with baseline:
  - Wealth channel: under alternative assumptions including real assets, average wealth channel effect across the full sample is a wealth loss of 1.2 percent of annual household income, compared to a loss of 12.4 percent in the baseline estimation.
  - Welfare: Table A4 reports an average well-being loss of 7.4 percent of annual household income under the alternative with real assets; this is less negative than the baseline estimation of -18.5 percent of annual household income.
- Distributional patterns and country-specific outcomes:
  - Adjustment in real assets (house prices) can lead to an average wealth gain for households in the poorest income quintile, improving their wealth effects relative to the financial-assets-only scenario.
  - Under the alternative assumptions, the second and third quintiles are the most affected by inflation rather than the poorest quintile.
  - Only four countries (Croatia, France, Lithuania, Slovakia) show a monotonic decrease in welfare losses for higher quintiles in this scenario.
  - In seven countries, the poorest quintile experiences either a lower welfare loss or a higher welfare gain than the richest quintile.
  - Five countries (Ireland, Luxembourg, the Netherlands, Portugal, Slovenia) report positive welfare gains across all income quintiles in response to the inflationary spike under these market-price-only assumptions.
- Aggregate table means (reported values from Annex):
  - Table A3 (wealth channel including real assets, percent of annual household income): -134.8, 0.2, -98.6, -3.2, -80.4, -2.0, -75.0, -0.1, -81.7, -1.0.
  - Table A4 (total well-being including real assets, percent of annual household income): -146.1, -5.8, -109.5, -9.5, -91.1, -8.2, -85.5, -6.3, -92.1, -7.1.
- Key implication:
  - Welfare effects of the inflation spike remain negative across scenarios but can be substantially less severe than baseline depending on assumptions about market-price adjustments and inclusion of real assets.
  - Standard narratives (for example, "inflation implies a redistribution from older households (net creditors) to younger households (net borrowers)" or "the poorest are typically the most affected by inflation") must be qualified and analyzed in each country's specific context due to sensitivity to asset and liability assumptions.

### Policy-relevant comparisons and takeaways from Annex A
- Magnitude sensitivity:
  - Including only market-price adjustments materially reduces the estimated severity of wealth-channel losses relative to the baseline that includes behavioral adjustments.
  - When real assets are allowed to adjust to market prices (without behavioral asset/liability changes), average well-being losses fall from -18.5 percent to 7.4 percent of annual household income (baseline vs alternative with real assets).
- Distributional shifts:
  - Market-price-only assumptions can shift the burden away from the poorest quintile in some countries, making the second and third quintiles most affected in the alternative scenario.
  - Country heterogeneity remains large: some countries exhibit positive welfare gains across all quintiles under market-price-only assumptions (Ireland, Luxembourg, the Netherlands, Portugal, Slovenia).
- Analytical caution:
  - Results emphasize the critical role of assumptions about price versus behavioral adjustments and the inclusion of real assets in determining distributional outcomes from inflation.
  - Policy conclusions should account for sensitivity to these assumptions and country-specific housing- and credit-market structures.

*Source: IMF Working Paper — Annex A. Alternative Scenario with Market Price Adjustments Only (2021:Q3–2022:Q2).*

### ANNEX A. ALTERNATIVE SCENARIO WITH MARKET PRICE ADJUSTMEN

### ANNEX A. ALTERNATIVE SCENARIO WITH MARKET PRICE ADJUSTMENTS ONLY

### A.1. Alternative Assumption Including Only Market Price Effects in the Wealth Channel
- Title as presented: "A.1. Alternative Assumption Including Only Market Price Eff ects in the Wealth Channel"
- Page reference: 40

### A.2. Wealth Channel and Combined Well-Being Effect with Financial Assets Under Price Adjustments Only
- Title as presented: "A.2. Wealth Channel and Combined Well-Being Effect with Financial Assets Under Price Adjustments Only"
- Page reference: 41

### A.3. Combined Well-Being Effect Including Real Assets Under Alternative Second-Round Effect Assumption
- Title as presented: "A.3. Combined Well-Being Effect Including Real Assets Under Alternative Second-Round Effect Assumption"
- Page reference: 41

### ANNEX B. DESCRIPTIVE STATISTICS
- Title as presented: "ANNEX B. DESCRIPTIVE STATISTICS"
- Page reference: 47

*Source: wpiea2026022-source-pdf - ANNEX A. ALTERNATIVE SCENARIO WITH MARKET PRICE ADJUSTMENTS ONLY*

### REFERENCES      ___________________________________________________________________________________   48

### wpiea2026022-source-pdf - REFERENCES

### Glossary and country list
- Country abbreviations used: BEL (Belgium), CYP (Cyprus), CYP (Cyprus), ESP (Spain), EST (Estonia), FIN (Finland), FRA (France), HRV (Croatia), GRC (Greece), ITA (Italy), IRL (Ireland), LTU (Lithuania), LVA (Latvia), LUX (Luxembourg), MLT (Malta), NLD (Netherlands), PRT (Portugal), SVK (Slovak Republic), SVN (Slovenia).
- Key acronyms used: COICOP (Classification of Individual Consumption by Purpose), ECB (European Central Bank), HBS (Household Budget Survey), HFCS (Household Finance and Consumption Survey), HICP (Harmonised Index of Consumer Prices), IMF (International Monetary Fund), OECD (Organisation for Economic Cooperation and Development).

### Introduction — scope and core innovations
- Sample and episode:
  - 18 European economies analyzed: Belgium, Croatia, Cyprus, Estonia, Finland, France, Greece, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Portugal, Slovakia, Slovenia, Spain.
  - Inflation episode studied: 2021:Q3 to 2022:Q2; year-on-year HICP inflation averaged 12 percent across the sample.
- Three welfare transmission channels analyzed:
  - Consumption basket channel.
  - Income channel.
  - Wealth channel (expanded here to include real assets).
- Three innovations relative to prior literature:
  - Expanded household wealth definition to include real assets (e.g., dwellings).
  - Incorporation of behavioral effects on income and wealth channels.
  - Larger country coverage (18 economies, including emerging European markets).

### Key quantitative findings (aggregate and distributional)
- Average welfare loss associated with the inflation surge across the 18 economies: 18.5 percent of annual household income.
- Decomposition of the 18.5 percent loss into sufficient statistics (share of annual household income):
  - Consumption basket channel: -0.04 percent.
  - Income channel: -6.1 percent.
  - Wealth channel: -12.4 percent.
- Distributional heterogeneity:
  - Average welfare losses for households in the lowest income quintile: 26.7 percent of their annual income.
  - Losses for richest quintile: less than half the losses of the poorest (implied from text: "more than double the average losses experienced by households in the richest quintile").
- Cross-country heterogeneity when real assets are included:
  - In 13 of 18 economies, welfare losses remain negative across all income quintiles.
  - Italy example: average welfare loss about 72 percent of annual household income even after accounting for real assets.
  - Five countries exhibit positive welfare effects for at least some quintiles; Ireland, Slovakia, and Slovenia: all quintiles recorded net welfare gains during the inflationary spike.
  - Luxembourg and the Netherlands: gains for their richest households; Netherlands: from a small loss for the poorest quintile to a gain exceeding 50 percent of annual household income for the richest quintile.
- Interpretation:
  - Wealth channel is the predominant factor driving overall welfare changes.
  - Real-asset revaluation (notably house prices rising more than inflation in some countries) can produce net gains for some households.
  - Findings complicate a simple lender-to-borrower redistribution narrative and call for nuanced study of intergenerational and income-group impacts.

### Methodology — analytical framework and estimations
- Analytical decomposition:
  - Household real wealth change decomposed into three components corresponding to the three channels; formal expressions given in Equations (1)–(7).
  - Primary normalized reporting: each channel’s effect presented as a share of annual household disposable income.
- Two estimation approaches:
  - Mechanical (day-after) effect: assume no nominal adjustment in income or wealth (훼 = 0 and 훽 = 0); captures immediate erosion of purchasing power.
  - Behavioral-adjustment effect: allow 훼 ≠ 0 and 훽 ≠ 0 to capture nominal adjustments in income and wealth during the 4-quarter horizon.
- Two wealth definitions:
  - Standard: financial/liquid assets only.
  - Expanded: total wealth including real assets (dwellings and other tangible assets).

### Data sources, assumptions, and construction
- Time horizon: 2021:Q3 to 2022:Q2.
- Inflation data: OECD price change data by COICOP.
- Household income:
  - Base microdata: 2019 EU-SILC.
  - Uprated to 2021 and simulated with EUROMOD to obtain household disposable income and components for 2021 and 2022 (EUROMOD simulations based on rules/policies as of June of each year).
- Household consumption:
  - 2015 EU Household Budget Survey (HBS) used to derive COICOP expenditure shares by disposable income quintile.
  - Nominal consumption values adjusted to 2021 using observed inflation rates between 2015 and 2021; assumption: relative composition of consumption remained stable between 2015 and 2021.
- Household wealth:
  - 2021 HFCS provides financial and real asset stocks by household.
  - Wealth variables derived as share of household gross income and averaged by gross income quintile; merged with income datasets by country and quintile.
- Behavioral wealth adjustments:
  - Changes in asset and liability stock values between 2021:Q1 and 2022:Q2 estimated using ECB Distributional Wealth Accounts (Blatnik and others, 2024) reported by wealth quintile.
  - EUROSTAT experimental statistics used to align income and wealth distributions.
- Resulting dataset:
  - Household-level dataset per country: EU-SILC 2019 + EUROMOD 2021/2022 simulations + average COICOP consumption by quintile + HFCS-based wealth 2021/2022 by quintile.
  - Final analysis uses country- and disposable-income-quintile averages.

### Consumption basket channel — composition and incidence
- General inflation range across countries (2021:Q3–2022:Q2, year-on-year):
  - From 5.9 percent in Malta to 23.3 percent in Estonia.
- Category-specific peak changes noted:
  - Housing and utilities costs surged by as much as 66 percent in Estonia.
  - Transport-related prices (including fuels) rose by more than 30 percent in Estonia, Latvia, Lithuania, and Luxembourg.
  - Food inflation surpassed 10 percent in most countries and reached 28 percent in Lithuania.
- Consumption budget shares by category (cross-country patterns):
  - Food budget shares: range from 33 percent in Lithuania to 12 percent in Luxembourg; decline with income in nearly all countries.
  - Housing and utilities expenditure shares: highest in Slovakia at 22 percent, lowest in Malta at 9 percent; generally decline with income.
  - Transport expenditure shares: positively associated with income; highest observed at 20 percent in Sweden (note: Sweden referenced in discussion of patterns).
  - “Other Consumption” share rises with income; variation across countries (e.g., high for wealthier households in Cyprus, Luxembourg, Malta; constrained even for high-income households in Croatia, Slovakia).
- Household-specific inflation by income quintile:
  - Relatively uniform across income groups in Luxembourg, Slovakia, Finland, and France.
  - In most other countries, lower-income quintiles faced higher household-specific inflation, driven by food and housing/utilities price increases.
  - Estonia and Lithuania show the largest gap in inflation between top and bottom of income distribution: differential of approximately 3 percentage points.
- Welfare effects from consumption basket channel (summary):
  - Table 1 (referenced): In all countries except Finland and Luxembourg, inflation reduced welfare for the lowest income quintiles, with effects becoming progressively positive with higher income.
  - Largest consumption-basket-channel welfare loss: poorest households in Belgium, approximately 2 percent of their annual income.
  - Top 20 percent in Italy experienced the greatest welfare gains from this channel (Table 1 referenced).

### Caveats and scope limits noted by authors
- Attribution challenges:
  - Difficulty disentangling inflation effects from other contemporaneous market forces (e.g., house prices influenced by factors other than inflation).
  - Authors note statistical evidence that other drivers were likely subdued during the inflationary spike, citing correlations with inflation trends (Annex references).
- General-equilibrium effects are abstracted from:
  - Cannot identify which agents gain from household welfare losses (central bank, government, firms, rest of the world) or whether aggregate welfare declined.
  - Examples cited: central banks capture a small share through erosion of money’s real value; governments benefit temporarily through surprise reduction in public debt.
- Data limitations and assumptions highlighted:
  - Use of 2015 HBS to infer 2021 consumption composition; potential bias due to pandemic-era HBS 2020 timing discussed.
  - Behavioral adjustments captured via EUROMOD and ECB DWA; alignment across datasets relies on experimental matrices.

*Source: IMF Working Paper "Distributional Impacts of Inflation Accounting for Behavioral Effects and Real Assets" (content unit: wpiea2026022-source-pdf - REFERENCES).*

### 0.9 percent of their annual income. On average, households at the bottom of the income distribution

### Distributional Impacts of Inflation Accounting for Behavioral Effects and Real Assets

### Consumption Basket Channel — key findings
- Households at the bottom of the income distribution experienced a welfare loss equivalent to 0.6 percent of their annual disposable income due to inflation transmitted through the consumption basket channel.
- The wealthiest (top quintile) households registered an average welfare gain of 0.3 percent of their annual income.
- These results are consistent with the regressive nature of inflation: lower-income households allocate a larger share of their budget to goods and services with above-average price increases, while higher-income households spent relatively more on goods and services with below-average price inflation and thus benefited net.

### IV. Income Channel — overview and methodology
- Two estimations are used:
  - First estimation: nominal incomes remain constant (mechanical effect).
  - Second estimation: incorporates changes in nominal income (wages, pensions, social transfers) using the tax benefit microsimulation model EUROMOD to capture behavioral effects and income dynamics.
- EUROMOD simulates tax-benefit rules in place as of June 30th of each year.

### IV.A Mechanical Effect — results
- Under the mechanical effect (nominal incomes constant), price increases lead to well-being losses uniform across income quintiles within each country.
- Losses are computed via the decline in real income expression of Equation (5) (mechanical erosion of real purchasing power).
- Cross-country magnitudes:
  - Estonia: welfare loss equivalent to 19.4 percent of household annual income.
  - Malta: welfare loss of 5.7 percent of household annual income.
- Average: welfare losses amounted to 10.7 percent of annual household income.
- Income-channel losses are substantially larger than consumption-basket-channel losses.

### IV.B Behavioral Effects — results and heterogeneity
- Incorporating observed changes in household income between 2021 and 2022 using EUROMOD reduced welfare losses substantially.
- On average:
  - Welfare losses declined from 10.7 percent to 5.4 percent of the annual household income at the bottom of the income distribution.
  - Welfare losses declined to 6.4 percent at the top of the distribution.
- Cross-country examples:
  - Estonia: losses fell by 12.9 to 11.6 percentage points of household annual income when accounting for income adjustments.
  - Malta: the increase in household income nearly offset the adverse effects of inflation across all quintiles.
- Drivers of heterogeneity:
  - Degree of changes in market remuneration.
  - Level and design of social protection and public transfers (Amores and others, 2025).
  - Examples: Ireland, Portugal, and Slovakia show relatively uniform income-channel adjustment across quintiles; Cyprus shows larger reduction in welfare losses for the lowest quintile due primarily to increased public transfers.
- Policy implication: targeted transfers and progressive income adjustments can provide meaningful relief, but effectiveness depends on magnitude and timing of interventions.
- Note on causality/caveat: analysis is agnostic on causality of observed income changes; in some countries (notably the Baltic countries) wage growth complicates interpretation. A robustness check would assume wages grow with the real GDP natural growth rate instead of zero (current assumption).

### V. Wealth Channel — definitions and approach
- Wealth (Fisher) channel estimated under two definitions:
  - (i) liquid financial assets (cash and readily accessible financial instruments).
  - (ii) total assets (financial + real assets).
- Behavioral effects considered in baseline include changes in values and volumes (sales/purchases of assets, contraction/liquidation of liabilities). Appendix A shows results when only prices move.
- Sensitivity of results to behavioral assumptions noted (Section VII for alternative assumptions).

### V.A Wealth Channel — Financial Wealth Only
#### A.1 Mechanical Effect (financial assets only)
- Mechanical assumption: no adjustments to assets/liabilities beyond direct inflationary shock; real value erosion of nominal assets/liabilities.
- Results highlight negative wealth effects in most countries, particularly among lower-income households.
- Examples:
  - Estonia, first income quintile: welfare loss of -22.3 percent of annual income.
  - Malta, second income quintile: decline of -14.2 percent of annual income.
- In Cyprus and the Netherlands (households net debtors), inflation-induced debt devaluation produced net positive wealth effects.
- On average:
  - Fourth income quintile benefited from the mechanism (inflation eroded liabilities).
  - Wealthiest households (Q5) experienced losses averaging -3.4 percent of annual household income.
- Mechanical effects show a nonlinear pattern: fourth quintile is the only quintile on average exhibiting a welfare gain from inflation.

#### A.2 Behavioral Effects (financial assets only)
- Behavioral effects incorporate changes in values of financial assets/liabilities and household responses using ECB Distributional Wealth Accounts (Blatnik and others, 2024).
- Incorporating behavioral effects tends to reduce wealth loss; more countries show positive net outcomes.
- Aggregate outcome: average annual income loss of 0.9 percent for households (financial-wealth behavioral scenario).
- Quintile-level shifts:
  - Poorest quintile (Q1): mechanical negative impact of -3.4 percent reversed to a net positive effect of 3 percent.
  - Quintile 2: improved from -2.9 percent to -1.8 percent of annual income.
  - Quintiles 4 and 5: increasingly negative repercussions; Q5 worsened from -3.4 percent to -4.8 percent of annual income.
- Cross-country variation driven by household debt composition, asset structures, and institutional frameworks.
- Countries with net borrower household profiles (Cyprus, Netherlands) gained from inflation-induced reductions in real value of liabilities.
- Behavioral adjustments are crucial determinants of wealth-channel outcomes.

### V.B Wealth Channel — Including Real Assets
#### B.A Mechanical Effect (real assets included)
- Real assets often account for more than ten times the average annual income of households in several economies; in Malta, real assets for some quintiles can exceed 25 times annual income.
- Real assets are substantial for both lowest and highest quintiles in many countries (U-shaped distribution of net wealth across quintiles).
- Mechanical assumption: nominal value of real assets unchanged → complete depreciation by inflation in real terms.
- Mechanical wealth losses including real assets:
  - In some countries and quintiles, welfare losses exceed 200 percent of household income.
  - Average mechanical welfare loss: 94.1 percent of the average annual household income across the full sample.
  - Comparison: when only financial assets are included, mechanical average is -2.2 percent of household average annual income.
- Losses are larger on average for the poorest households:
  - For the lowest quintile: approximately -135 percent of average annual income.
  - Lithuania, first quintile: wealth loss of -271 percent of annual household income.
- Mechanical wealth losses are non-positive across all countries and quintiles, showing a U-shaped pattern across quintiles.
  - Fourth quintile: lowest average wealth loss at -75 percent of average annual household income.

#### B.A.2 Behavioral Effects (real assets included)
- Behavioral outcomes depend on relative growth of real asset values compared to inflation and household responses.
- When real asset valuation growth exceeds inflation, households experienced net wealth gains; otherwise, net wealth erosion.
- Using assumptions and data in Section II.C, average behavioral wealth loss when including real assets is -12.4 percent of the average household's annual income (Table 4 “Behavioral” column).
- Cross-country heterogeneity:
  - Luxembourg and Slovakia: inclusion of real assets reverses sign of wealth-channel effect compared to financial-wealth-only results due to substantial real asset appreciation.
  - Cyprus: household wealth deteriorated markedly; highest income quintile negative impact averaged -93.8 percent of annual income.
  - Finland, France, Italy, Spain: inclusion of real assets emphasizes U-shaped distributional pattern with largest changes for lowest and highest quintiles.
- Caveat: isolating inflation’s sole effect on real asset prices is challenging because house prices and other real asset valuations may be influenced by other market forces; Table A.1 in Annex B suggests interquartile range of house price changes closely aligns with inflation in most analyzed countries.

### VI. Distributional Effects Combining the Three Channels — financial wealth only
- Two combined assessments performed:
  - Combined effects with wealth channel including only financial assets.
  - Combined effects with wealth channel including real assets (discussed separately).
- When financial wealth only is considered:
  - Consumption basket channel: smallest relative impact among three channels.
  - Income and wealth channels: dominant in shaping distributional outcomes.
  - Most cross-country variation arises from the wealth channel; income channel is more constant across quintiles (see Table 2).
- Mechanical vs Behavioral combined effects (financial wealth only):
  - Mechanical effect: welfare losses more pronounced at 12.9 percent of household annual income.
  - Behavioral effects included: welfare losses at 7 percent of household annual income.
- Quintile impact:
  - Highest welfare losses occur in first and second poorest quintiles (Quintiles 1 and 2), illustrating poorest are most impacted by the inflationary shock.
- Cross-country patterns:
  - Estonia: largest well-being losses across most quintiles, with losses diminishing as income quintiles increase.
  - Cyprus and the Netherlands: fluctuate between losses and gains across quintiles.
  - In some countries, first (poorest) quintile experiences welfare gains due to low but positive net worth, while fifth (richest) quintile endures welfare losses.

*IMF Working Paper — Distributional Impacts of Inflation Accounting for Behavioral Effects and Real Assets*

### Annex A also brings the same distributional analysis with the assumptions and results presented in the annex.

### wpiea2026022-source-pdf - Annex A also brings the same distributional analysis with the assumptions and results presented in the annex.

### Behavioral effects focusing on financial assets
- Redistribution from net creditor households to net borrower households is clearly revealed when behavioral effects are incorporated while focusing exclusively on financial assets.
- Average welfare losses across the full sample when restricted to financial assets:
  - Richest income quantile: welfare loss of 10.9 percent of their average annual income.
  - Second income quintile: welfare loss averaging 7.9 percent of their annual income.
- Country-specific patterns:
  - Cyprus and the Netherlands: households across all income quintiles, except for the richest, experience welfare gains due to a significant and positive wealth channel effect that offsets income losses. Most households are net borrowers and the reduction in the real value of their debt more than compensates for purchasing power erosion from inflation. Only net-creditor households in the richest quintile experience welfare losses under financial-asset-only simulations.
  - Italy and Portugal: the well-being effect demonstrates a U-shaped pattern, with the largest impacts at both the lowest and highest ends of the income distribution.
- Interpretation:
  - The wealth channel plays a central role in determining whether households are net winners or losers from inflationary episodes.
  - While the income channel dominated in some contexts, the wealth channel introduced substantial variation across and within countries, underlining the importance of wealth composition.

### Distributional effects when accounting for real assets
- Adding real assets substantially alters welfare outcomes, typically worsening outcomes for poorer households and removing previously observed U-shaped patterns in some countries.
- When real assets are added to the wealth channel and only mechanical effects are considered:
  - Average welfare loss across countries and quintiles is dominated by the wealth channel and declines monotonically as income quintiles increase.
  - On average, the welfare loss exceeds 104 percent of household annual income for the full sample.
  - By quintile (mechanical wealth-channel with real assets, averages reported):
    - Poorest households: averaging -146 percent of their annual income.
    - Fourth (second richest) quintile: averaging -85.5 percent of their annual income.
  - Country extremes:
    - Poorest households in Lithuania: loss of 290.4 percent of their average annual income.
    - Poorest households in Finland: loss of -45.1 percent of their average annual income.
- Accounting for real assets and behavioral effects (main simulation, Column “Behavioral” in Table 6):
  - Full-sample average welfare effect across countries and quintiles: -18.5 percent of annual household income.
  - First income quintile average loss: -26.7 percent of annual household income.
  - Highest income quintile average loss: -13.3 percent of annual household income.
  - Seven countries (e.g., Croatia, Estonia, Luxembourg, and the Netherlands) show a consistent reduction in welfare loss across quintiles; other countries exhibit nonlinear patterns (e.g., Finland and Spain where welfare loss is greater for the richest quintile than for the poorest).
- Cross-country heterogeneity when real assets are included:
  - In 13 sample countries, the wealth and overall welfare effects of inflation remain negative for all quintiles because changes in net financial assets and real-asset values are smaller relative to the inflation level.
    - Example: Italy — average welfare loss across all income quintiles when including real assets is equal to 72 percent of household income, ranging from a loss of 83 percent for the poorest quintile to a loss of 63.5 percent for Quintile 4.
  - In 5 economies, the welfare effect is positive for some or all income quintiles.
    - Ireland, Slovakia, and Slovenia: households in all income quintile(s) experienced net welfare gains during the inflationary spike.
      - Ireland example: average welfare gain across all income quintiles is equal to 37,5 percent of household income, from 28 percent for the second poorest quintile to a gain of more than 53 percent for the richest quintile.
    - Luxembourg and the Netherlands: richest quintile(s) experience welfare gains rather than negative effects.
      - Netherlands example: average welfare gain across all income quintiles is equal to 24 percent of household income, moving from an average loss of 2 percent for the poorest quintile to a gain of 52,5 percent for the richest quintile.
  - Drivers of positive outcomes in some countries include house prices increasing significantly more than inflation during the sample period, owing partly to housing-market characteristics.
- Historical context:
  - Search for real assets has been a common inflation hedge historically; Wolff (1979) found homeowners, middle-aged households, married couples, and Whites benefited during 1969–1975 in the United States, with homeowners with substantial mortgages and low-income households with mortgages among the largest beneficiaries.

### Key quantitative decomposition and comparisons
- Full-sample average welfare loss reported in main results: -18.5 percent of the average household annual income.
- Decomposition of the average welfare loss across channels (full country and period sample):
  - Consumption basket channel: -0.04 percent of average household annual income.
  - Income channel: -6.1 percent of average household annual income.
  - Wealth effect channel: -12.4 percent of average household annual income.
- Quintile-specific decomposition (full country sample averages for poorest and richest quintiles):
  - Poorest quintile: -0.6 (consumption), -5.4 (income), and -20.8 (wealth) percent of household annual income, totaling -26.7 percent.
  - Highest quintile: +0.3 (consumption), -6.4 (income), and -7.2 (wealth) percent of household annual income.
- Comparison with existing literature:
  - The paper’s estimated average loss of 18.5 percent of household annual income is larger than estimates in prior work; example: Pallotti and others (2024) find, e.g., 8 percent in Italy and 3 percent in France.

### Conclusions and implications emphasized by the analysis
- Incorporating real assets and behavioral effects materially increases estimated welfare losses from inflationary shocks, especially for poorer households.
- Accurately reflecting the assets and liabilities in household balance sheets is critical to understanding the welfare impact of inflation and explaining diverging welfare effects across countries.
- The wealth channel (including real assets and behavioral responses) is the dominant driver of overall welfare impacts during the 2021:Q3–2022:Q2 inflation surge for the 18 European economies studied.

*IMF Working Paper — Distributional Impacts of Inflation Accounting for Behavioral Effects and Real Assets*

### 13.3 percent.

### wpiea2026022-source-pdf - 13.3 percent.

### Welfare effects when real assets are incorporated
- In 13 of the 18 economies, welfare losses remain negative across all income quintiles because increases in net financial and real assets did not keep pace with inflation.
- Italy: average loss of about 72 percent of annual household income.
- Five economies recorded gains for some or all quintiles where house prices rose faster than inflation:
  - Ireland, Slovakia, Slovenia, and the richest households in Luxembourg and the Netherlands.
- The wealth channel becomes the dominant driver of welfare outcomes once real assets are incorporated.
- The quantitative welfare effects of inflation are highly sensitive to:
  - Assumptions about behavioral responses.
  - Types of assets included (financial assets only versus including real assets).
- Omitting real assets and behavioral adjustments risks understating welfare losses for vulnerable groups and oversimplifying redistribution mechanisms, especially those shaped by local housing and credit market conditions.
- Figure 9 highlights that different assumptions yield notably different patterns across income quintiles, challenging the common view that inflation’s distributional effects are predominantly U-shaped and suggesting a more nuanced pattern, particularly for lower-income households.

### Empirical channel findings (selected quantitative highlights)
- Combined effects and channels (as presented in the paper):
  - Consumption channel, Income channel, Fisher effect, and Total effect are shown across quintiles in Figure 8 (Percent of household income).
- Income channel summary (Table 2; Percent of household income):
  - Mean mechanical effect: -10.7
  - Mean behavioral effect: -5.4 (and other tabulated means by quintile and simulation)
  - Example country-specific values (poorest quintile mechanical vs behavioral):
    - Estonia: -19.4 (mechanical) and -12.9 (behavioral)
    - Lithuania: -18.9 and -7.5
    - Latvia: -16.7 and -5.1
    - Malta (richest shown): -5.7 and 0.0 (poorest to richest across quintiles in table)
- Wealth channel (financial assets only) summary (Table 3; Percent of annual household income):
  - Mean mechanical: -3.4
  - Mean behavioral: 3.0
  - Notable extremes:
    - Estonia poorest quintile simulation (mechanical): -22.3
    - Cyprus poorest quintile simulation (mechanical): 21.3
- Wealth channel including real assets (Table 4; Percent of annual household income):
  - Mean simulation mechanical: -134.8 (poorest quintile)
  - Mean behavioral for poorest quintile: -20.8
  - Large negative impacts when real assets included, e.g.:
    - Lithuania poorest quintile simulation: -271.0 (mechanical)
    - Estonia poorest quintile simulation: -219.6 (mechanical)
  - Some positive behavioral entries in middle/rich quintiles for certain countries (e.g., Slovakia mechanical/behavioral patterns show positive behavioral values in some quintiles).
- Total changes in well-being (including only financial assets) (Table 5; Percent of annual household income):
  - Mean simulation for poorest quintile: -14.8 (mechanical mean: -3.0; behavioral mean: -14.8 and -7.9 across columns)
  - Country examples (poorest quintile simulation):
    - Estonia: -42.7 (mechanical), -35.8 (behavioral)
    - Cyprus: 12.6 (mechanical), 14.8 (behavioral)
- Total changes in well-being (including real assets) (Table 6; Percent of annual household income):
  - Mean simulation for poorest quintile: -146.1 (mechanical mean: -26.7; behavioral mean: -109.5 and -22.2 across columns)
  - Country examples (poorest quintile simulation):
    - Lithuania: -290.4 (mechanical)
    - Estonia: -240.0 (mechanical)
    - Finland: -45.1 (mechanical)
  - Median and dispersion statistics reported in tables (St. Dev., Minimum, Maximum) indicate wide heterogeneity across countries and quintiles.

### Historical and institutional modifiers of asset protection
- Homeownership can provide inflation protection but may be undermined by government policies:
  - France after World War I and during world wars: rent control led rents to fall to one-tenth of their real value between 1913 and 1950 (Piketty 2003).
  - Germany between 1913 and 1927: real estate lost one-fifth of its value due to rent regulation and taxation (Albers, Bartels, and Schularick 2022).
- Land was the only asset that appreciated in value historically, indicating a pronounced rural-urban divide in inflation effects.

### Policy implications
- Pronounced inflation spikes produce significant redistributive effects across household groups.
- The poorest quintile bears the largest welfare losses in most economies analyzed.
- Policy measures should:
  - Prioritize social benefit programs to support households facing rising living costs.
  - Account for inflation’s effects on real as well as financial assets.
- Because real assets and behavioral responses substantially shape household welfare, they can also affect aggregate consumption and investment decisions.

### Suggested avenues for further research
- Assume that, absent inflation, growth in nominal household income and wealth equals real GDP growth to isolate real valuation effects beyond inflation.
- Deepen understanding of inflation’s distributional effects by:
  - Incorporating real assets.
  - Accounting for behavioral effects, particularly in emerging markets and developing economies.
- Analyze scenarios of persistently high inflation or hyperinflation to understand adaptation of households and the financial sector under severely distorted relative prices.
- Investigate why real asset values and holdings respond strongly to inflationary pressures.

*IMF WORKING PAPERS Distributional Impacts of Inflation Accounting for Behavioral Effects and Real Assets — INTERNATIONAL MONETARY FUND*

### Annex A. Alternative Scenario with Market Price

### Annex A. Alternative Scenario with Market Price Adjustments Only

### A.1. Alternative assumption and methodology
- Scenario: includes only market price adjustments on financial assets, real assets, and liabilities; excludes behavioral responses on real assets and liability contraction or liquidation.
- Financial assets:
  - Deposits: changes use country-level average change in deposits from the European Central Bank's MIR-MFI Interest Rate Statistics, focusing on "Euro-denominated deposits with an agreed maturity of more than one year from euro area households (percentages per annum, rates on new business)."
  - Other financial assets (stocks): changes based on average returns from European financial markets (Bloomberg). Assumes households can freely reallocate financial assets across integrated European markets.
- Liabilities: account for changes in interest rates at country level between the beginning and the end of the sample period; assume households do not acquire new debt or repay early to adjust to inflationary shocks.
- Real assets: initial stock valuation from Eurosystem HFCFS dataset; house price valuation uses quarterly Housing Price Index from the European Central Bank, change between 2021:Q3 and 2022:Q2, discounted for inflation consistent with the last term of Equation 7 in Section II.B with 훽
గவ଴
௛ assumed equal to 1.

### A.2. Wealth channel and well-being with only financial assets (market-price adjustments)
- Aggregate comparisons with baseline:
  - Average wealth loss in this alternative (financial assets only) is 0.7 percentage points lower than the baseline scenario.
  - Average welfare loss (wealth channel included across countries and quintiles) is -6.4 percent of annual household income under this assumption, compared to -7 percent of annual household income in the baseline.
- Distributional heterogeneity and drivers:
  - In 8 countries (example countries: Ireland, Latvia, the Netherlands) where stock valuations exceed the European average, the average wealth loss across quintiles is lower in the baseline than in this alternative.
  - Baseline scenario generally yields a less negative wealth effect for lower income quintiles (they are typically net creditors with limited debt), while higher quintiles tend to benefit relatively more under the market-price-only alternative.
- Aggregate table means (Table A1 — wealth channel, financial assets, percent of annual household income):
  - Mean values presented in Table A1: -3.4, -1.8, -4.0, -2.3, -0.6, 0.6, 0.7, 2.3, -3.4, 0.2.
- Aggregate table means (Table A2 — total well-being, financial assets only, percent of annual household income):
  - Mean values presented in Table A2: -14.8, -7.7, -14.8, -8.6, -11.3, -5.6, -9.9, -3.9, -13.8, -5.9.
- Notable outcomes:
  - With predominantly negative income and consumption basket effects, total welfare results in an average loss across income quintiles for most countries, with exceptions (Cyprus and the Netherlands).
  - The average welfare effect across countries and quintiles shows an inverted U-shape when only financial assets are considered.

### A.3. Combined well-being effect including real assets (market-price adjustments)
- Aggregate comparisons with baseline:
  - Wealth channel: under alternative assumptions including real assets, average wealth channel effect across the full sample is a wealth loss of 1.2 percent of annual household income, compared to a loss of 12.4 percent in the baseline estimation.
  - Welfare: Table A4 reports an average well-being loss of 7.4 percent of annual household income under the alternative with real assets; this is less negative than the baseline estimation of -18.5 percent of annual household income.
- Distributional patterns and country-specific outcomes:
  - Adjustment in real assets (house prices) can lead to an average wealth gain for households in the poorest income quintile, improving their wealth effects relative to the financial-assets-only scenario.
  - Under the alternative assumptions, the second and third quintiles are the most affected by inflation rather than the poorest quintile.
  - Only four countries (Croatia, France, Lithuania, Slovakia) show a monotonic decrease in welfare losses for higher quintiles in this scenario.
  - In seven countries, the poorest quintile experiences either a lower welfare loss or a higher welfare gain than the richest quintile.
  - Five countries (Ireland, Luxembourg, the Netherlands, Portugal, Slovenia) report positive welfare gains across all income quintiles in response to the inflationary spike under these market-price-only assumptions.
- Aggregate table means (Table A3 — wealth channel including real assets, percent of annual household income):
  - Mean values presented in Table A3: -134.8, 0.2, -98.6, -3.2, -80.4, -2.0, -75.0, -0.1, -81.7, -1.0.
- Aggregate table means (Table A4 — total well-being including real assets, percent of annual household income):
  - Mean values presented in Table A4: -146.1, -5.8, -109.5, -9.5, -91.1, -8.2, -85.5, -6.3, -92.1, -7.1.
- Key implication:
  - Welfare effects of the inflation spike remain negative across scenarios but can be substantially less severe than baseline depending on assumptions about market-price adjustments and inclusion of real assets.
  - Standard narratives (for example, "inflation implies a redistribution from older households (net creditors) to younger households (net borrowers)" or "the poorest are typically the most affected by inflation") must be qualified and analyzed in each country's specific context due to sensitivity to asset and liability assumptions.

*Source: IMF Working Paper — Annex A. Alternative Scenario with Market Price Adjustments Only (2021:Q3–2022:Q2).*

### References

### References

### Complete list

- Adam, Klaus, and Panagiota Tzamourani. 2016. “Distributional consequences of asset price inflation in the Euro Area.” European Economic Review, Vol. 89, pp. 172–92.  
- Albanesi, Stefania. 2007. “Inflation and Inequality.” Journal of Monetary Economics, Vol. 54(4), pp. 1088–114.  
- Albers, Thilo N. H., Charlotte Bartels, and Moritz Schularick. 2022. “The Distribution of Wealth in Germany, 1895–2018.” CESifo Working Paper 9739, Ifo Institute for Economic Research, Munich.  
- Amores, Antonio F., Henrique Basso, Johannes Simeon Bischl, Paola De Agostini, Silvia De Poli, Emanuele Dicarlo, Maria Flevotomou, Maximilian Freier, Sofia Maier, Esteban García-Miralles, Myroslav Pidkuyko, Mattia Ricci, Sara Riscado. 2025. “Inflation, Fiscal Policy, and Inequality: The Impact of the Post-Pandemic Price Surge and Fiscal Measures on European Households.” The Review of Income and Wealth, Vol. 71(1), e12713.  
- Ampudia, Miguel, Dimitris Georgarakos, Jiri Slacalek, Oreste Tristani, Philip Vermeulen, and Giovanni L. Violante. 2018. “Monetary policy and household inequality,” ECB Working Paper 2170, European Central Bank, Frankfurt.  
- Badev, Anton, Thorsten Beck, Ligia Vado, and Simon Walley. 2014. “Housing Finance across Countries. New Data and Analysis.” Policy Research Working Paper 6756, World Bank, Washington, DC.  
- Baez Ramirez, Javier Eduardo, Osman K. Inan, and Metin Nebiler. 2021. “Getting Real? The Uneven Burden of Inflation across Households in Turkey.” Policy Research Working Paper 9869, World Bank, Washington, DC.  
- Balasundharam, Vybhavi, Arika Kayastha, and Marcos Poplawski-Ribeiro. 2023, “Inflation Indexation in Public Finances: A Global Dataset on Current Practices.” IMF Working Paper, No. 23/264 (Washington: International Monetary Fund).  
- Blatnik, Nina, Alina Bobasu, Georgi Krustev and Mika Tujula. 2024. “Introducing the Distributional Wealth Accounts for euro area households.” ECB Economic Bulletin, Issue 5/2024, European Central Bank, Frankfurt.  
- Bolhuis, Marijn A., Judd. N. Cramer, Karl O. Schulz, and Lawrence H. Summers. 2024. “The Cost of Money Is Part of the Cost of Living: New Evidence on the Consumer Sentiment Anomaly.” NBER Working Paper 32163, National Bureau of Economic Research, Cambridge, MA.  
- Bulir, Ales. 2001. “Income Inequality: Does Inflation Matter?” IMF Staff Papers Vol. 48(1), pp. 1–5.  
- Burstein, Ariel T. 2006. “Inflation and output dynamics with state-dependent pricing decisions,” Journal of Monetary Economics, Vol. 53(7), pp. 1235–57.  
- Cardoso, Eliana. 1992. “Inflation and Poverty.” NBER Working Paper 4006, National Bureau of Economic Research, Cambridge, MA.  
- Cardoso, Miguel, Clodomiro Ferreira, José Miguel Leiva, Galo Nuño, Álvaro Ortiz, Tomasa Rodrigo, and Sirenia Vazquez. 2022. “The Heterogeneous Impact of Inflation on Households’ Balance Sheets.” Working Paper 176, Red Nacional de Investigadores en Economía (RedNIE), Madrid.  
- Charalampakis, Evangelos, Bruno Fagandini, Lukas Henkel, and Chiara Osbat. 2022. “The Impact of the Recent Rise in Inflation on Low-Income Households.” ECB Economic Bulletin, Box 4, Issue 7/2022, European Central Bank, Frankfurt.  
- Claeys, Grégory, and Lionel Guetta-Jeanrenaud. 2022. “Who Is Suffering Most from Rising Inflation?” Bruegel (blog), February 1, Bruegel, Brussels. https:// www .bruegel .org/ blog -post/ who -suffering -most -rising -inflation.  
- Coibion, Olivier, Yuriy Gorodnichenko, Lorenz Kueng, and John Silvia. 2017. “Innocent Bystanders? Monetary policy and inequality.” Journal of Monetary Economics, Vol. 88, pp. 70–89.  
- Easterly, William, and Stanley Fischer. 2001. “Inflation and Poor.” Journal of Money, Credit and Banking, Vol 33, No.2, p.160–78.  
- Fagereng, Andreas, Matthieu Gomez, Émilien Gouin-Bonenfant, Martin Holm, Benjamin Moll, and Gisle Natvik. forthcoming. “Asset-Price Redistribution.” Journal of Political Economy, available at https://doi.org/10.1086/736769.  
- Ferreira, Clodomiro, José Miguel Leiva, Galo Nuño Barrau, Alvaro Ortiz, Tomasa Rodrigo and Sirenia Vazquez. 2023. “The heterogeneous impact of inflation on households' balance sheets.” BIS Working Papers, No 1152, Bank of International Settlements, Basel.  
- Fisher, Irving. 1930. “The Theory of Interest, as determined by Impatience to Spend Income and Opportunity to Invest it.” New York: Macmillan Publishers.  
- Furceri, Davide, Prakash Loungani, and Aleksandra Zdzienicka. 2018. “The effects of monetary policy shocks on inequality.” Journal of International Money and Finance, Vol. 85, pp. 168–86.  
- Galli, Rossana, and Rolph van Der Hoeven. 2001. “Is Inflation Bad for Income Inequality. The Importance of the Initial Rate of Inflation.” Employment Paper 2001/29. International Labour Organization.  
- Garcia-Macia, Daniel. 2023. “The Effects of Inflation on Public Finances.” IMF Working Paper, No. 23/93 (Washington: International Monetary Fund).  
- Guren, Adam M., Alisdair McKay, Emi Nakamura, Jón Steinsson. 2021. “Housing Wealth Effects: The Long View.” Review of Economic Studies, Vol. 88(2), pp. 669–707.  
- Hansen, Niels-Jakob, Frederik Toscani, and Jing Zhou. 2023. “Euro Area Inflation after the Pandemic and Energy Shock: Import Prices, Profits and Wages.” IMF Working Paper, No. 23/131 (Washington: International Monetary Fund).  
- Hellebrandt, Tomas, and Paolo Mauro. 2015. “The Future of Worldwide Income Distribution.” PIIE Working Paper Series WP15–7, Peterson Institute for International Economics, Washington, DC.  
- Hu, Ruiyang, Jian Wang, Yibai Yang, Zhijie Zheng. 2024. “Inflation and income inequality in an open-economy growth model with liquidity constraints on R&D,” Journal of International Money and Finance, Vol. 147, 103127.  
- International Monetary Fund (IMF). 2023. “Inflation and Disinflation: What Role for Fiscal Policy.” Fiscal Monitor, Chapter 2, Fiscal Affairs Department, Washington DC, April.  
- Loko, Boileau, Nelie Nembot, and Marcos Poplawski-Ribeiro. 2025. “Private savings and COVID-19 in Sub-Saharan Africa.” Emerging Markets Review, Vol. 69, No. 101363.  
- Mao, Bincheng. 2022. “Inflation Affects Women More Than Men. Civil Society Can Help.” World Economic Forum (blog), October 18. https:// www .weforum .org/ agenda/ 2022/ 10/inflation -crisis -hits -women -harder/.  
- McGregor, Thomas, Suphap hiphat, Nujin, Toscani, Frederik. 2022. Europe’s Consumers are Sitting on 1 Trillion Euros in Pandemic Savings. IMF Blog, February 10. International Monetary Fund, Washington DC. available at: https://www.imf.org/en/Blogs/Articles/2022/02/10/europes-consumers-are-sitting-on-1-trillion- euros-in-pandemic-savings.  
- Mohrle, Sascha, and Timo Wollmershauser. 2021. “On the Distributional Effects of the Current High Inflation Rates.” Ifo Schnelldienst digital, Vol. 16, pp. 1–6.  
- Moura, Mauricio, Marcos Poplawski-Ribeiro, and Caio Piza. 2014. "Are there any distributive effects of land title on labor supply? evidence from Brazil," IZA Journal of Labor & Development, Vol. 3(1), pp. 1–18.  
- Mumtaz, Haroon, and Angeliki Theophilopoulou. 2017. “The impact of monetary policy on inequality in the UK. An empirical analysis.” European Economic Review, Vol. 98(C), pp. 410–23.  
- Pallotti, Filippo, Gonzalo Paz-Pardo, Jiri Slacalek, Oreste Tristani, and Giovanni L. Violante. 2023. “Who Bears the Costs of Inflation? Euro Area Households and the 2021–2022 Shock.” Journal of Monetary Economics, Vol. 148, No.103671.  
- Piketty, Thomas. 2003. “Income Inequality in France, 1901–1998.” Journal of Political Economy, Vol. 111 (5), pp. 1004–42.  
- Süssmuth, Bernd, and Matthias Wieschemeyer. 2022. “Taxation and the Distributional Impact of Inflation: The U.S. Post-War Experience.” Economic Modelling, Vol. 111 (C), No. 105813.  
- Sutherland, Holly, and Francesco Figari. 2013. “EUROMOD: the European Union tax-benefit microsimulation model.” International Journal of Microsimulation, Vol. 6(1), pp. 4–26.  
- Wolff, Edward N. 1979. “The Distributional Effects of the 1969–75 Inflation on Holdings of Household Wealth in the United States.” Review of Income and Wealth, Vol. 25(2), pp. 195–207.  

*Distributional Impacts of Inflation Accounting for Behavioral Effects and Real Assets — Working Paper No. WP/2026/022*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2026/english/wpiea2026022-source-pdf.pdf_