## _cr05225 ## Source details **Canonical URL:** [_cr05225](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2005/_cr05225.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2005/_cr05225.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2005/_cr05225.pdf.json) --- ### I. The External Competitiveness of the Dutch Economy — Key findings - Rankings: - WEF growth competitiveness index (2004): Netherlands ranked 12th out of 103 countries. - IMD ranking (2005): Netherlands ranked 13th out of 60 countries. - IMD ranking changes: 4th (2000) → 13th (2003) → 15th (2004). - Cost competitiveness and exchange rates: - Dutch unit labor costs in manufacturing rose 11 percent faster than euro area competitors and 14 percent faster than competitors in the rest of the world (1998–2003). - In 2004, unit labor costs in the Netherlands decreased by 2.1 percent. - Euro appreciation vs. U.S. dollar: appreciated by 18 percent in 2002 and by a further 20 percent in 2003; closed 2003 at $1.2595 per euro. In 2004, euro rose 7.6 percent against the dollar to $1.3554 and rose 2.8 percent against a basket of currencies. - Trade-weighted euro appreciation: 7.2 percent in 2002 and 12.1 percent in 2003. - Unit labor cost annual changes (selected): - The Netherlands (2001–2004) — Compensation per employee: 4.6, 6.5, 3.7, 2.4; Labor productivity: -0.3, 2.2, 0.3, 4.6; Unit labor costs: 4.9, 4.3, 3.5, -2.1. - Euro competitors (2001–2004) — Unit labor costs: 2.7, 1.9, 1.3, -2.0. - Non-euro competitors (2001–2004) — Unit labor costs: 2.3, -4.9, -11.8, -7.1. - All competitors (2001–2004) — Unit labor costs: 2.5, -1.7, -5.6, -4.7. - Export performance: - Real export growth averaged less than 1 percent in 2001–03 while Dutch export market growth averaged above 2 percent annually (WEO data). - Year-average real export growth (goods and services) in 2004: about 8 percent; much of the increase in 2004 driven by sharp growth in re-exports. - CMS decomposition (1992–2003 and 2000–03): - Change in exports (Value in billions of U.S. dollars): 1992–2003: 92; 2000–03: 47. - Average annual growth rate: 1992–2003: 3.8; 2000–03: 12.3. - Decomposition (Value in billions of U.S. dollars): - World Trade Effect: 1992–2003: 22; 2000–03: 22. - Commodity Composition Effect: 1992–2003: -5; 2000–03: -27. - Market Distribution Effect: 1992–2003: -16; 2000–03: 10. - Residual (Competitiveness): 1992–2003: -10; 2000–03: -26. - Interpretation: residual negative — competitiveness worsened in recent years. ### II. Long-Run Household Consumption Equilibrium — Key findings and model results - Stylized consumption facts: - Since beginning of 2001, average quarterly private consumption growth ≈ 1/4 percent; in the second half of the 1990s average ≈ nearly 1 percent. - Average share of disposable income in GDP fell by over 2 percentage points when comparing 2001–04 with 1995–2000, largely reflecting a decline in property income. - Deceleration of private consumption began earlier than deceleration of per capita real disposable income growth; occurred while unemployment was at a historic low and interest rates had declined. - Wealth, confidence, and timing: - Amsterdam stock exchange price index started to fall in 2001:Q2. - Per capita household net worth to income ratio started to decline toward end-2001. - Consumer confidence deteriorated rapidly with stock prices; DNB linked decline to doubts about pension funds solvency, dissatisfaction with public services, and policy/political uncertainty. - Empirical framework and data: - Permanent income/CRRA representative agent model used; Ut = Ct^(1−γ)/(1−γ). - Data: per capita real total private consumption; per capita real household net worth (net worth excludes pensions); household disposable income and wealth converted from annual to quarterly using Lisman filter; sample 1983Q2–2004Q4; series seasonally adjusted with X11. - Unit-root/cointegration: DFGLS test — all series except unemployment rate contained a unit root at 95 percent; Johansen-Juselius test strongly rejected no-cointegration at 95 percent (λmax statistic 25.42 vs corrected 95% critical value 25.83, close). - Estimation: nonlinear dynamic least squares estimator of Phillips and Loretan (1991). - Long-run estimated consumption equation (reported as equation (8)): - 3.82 0.100.210.01int0.01, ttttt Cincomewealthunemployment=− ++−− with ρ=0.35. (coefficients and signs as reported) - Exact long-run coefficients (Appendix values): - Constant: -3.823558325. - Real per capita disposable income coefficient: 0.098896581. - Real household net worth per capita coefficient: 0.211184767. - 3-month real interest rate coefficient: -0.005808650. - Unemployment rate coefficient: -0.010224217. - Long-run equilibrium adjustment parameter (DELTA / ρ): 0.350438388. - Associated t-statistics: -18.13, 2.19, 4.76, -2.62, -4.22, and 3.71, respectively; all variables significant at the 95 percent level; equilibrium correction term strongly significant. - Diagnostics (Appendix 1): - Centered R**2 = 0.995528; R Bar **2 = 0.993988. - Standard Error of Estimate = 0.008565472. - Sum of Squared Residuals = 0.0044754056. - Cumulated Periodogram Test: Maximum Gap = 0.1399 at Frequency 1.8162; rejection limits 1% = 0.2037, 5% = 0.1700, 10% = 0.1525. - Economic magnitudes: - Dutch marginal propensity to consume out of real net wealth ≈ 5¼ cents to the euro (based on sample average wealth-to-consumption ratio). - Comparative examples cited: United States 4 cents to the dollar; United Kingdom 5 cents to the pound; France less than 3 cents to the euro. - Consumption adjusts to equilibrium in about three quarters on average. - Interpretation and policy implications: - Main determinants: income, wealth, interest rates, and uncertainty (unemployment proxy). - Wealth effects are economically important; low long-run income elasticity possibly reflects inclusion of property income in disposable income and use of total consumption. - With consumption caught up to fundamentals by end-2004, pickup likely if growth improves, wealth increases (stocks/housing), and uncertainty recedes. - Recommendations: structural reforms to raise potential growth and consumer confidence; enhance transparency of pension fund rules to reduce uncertainty and raise observed marginal propensities to consume. ### III. House Prices in the Netherlands — boom, model, and empirical findings - Boom overview: - Between 1995 and 2001, average Dutch house prices rose by almost 80 percent in real terms. - Price-to-disposable-income ratio rose by more than 50 percent since 1995 and is some 40 percent above its 30-year average (study period). - Price-rent ratio rose by more than 75 percent since 1995 and is more than 50 percent above its 30-year average. - Concerns about overvaluation are supported by these valuation indicators, especially price-rent. - Supply-side constraints: - Population density: 478 people per square kilometer. - Undeveloped land scarce; strict zoning; supply elasticity regarded as very low (Swank, Kakes, and Tieman, 2002). - Low supply elasticity implies prices may overshoot and deviations can be persistent. - Model approach and data: - Error-correction model (ECM) chosen; focus on demand given very inelastic supply. - Final variable vector Z = (p, y, r): p = real house prices (logs), y = total real disposable income of households (logs), r = real mortgage interest rate. - Quarterly sample 1974:Q1–2004:Q2; house prices from Kadaster; real disposable income from De Nederlandsche Bank; mortgage rate from Vereniging Eigen Huis, deflated by CPI. - Time-series properties and cointegration: - All three series I(1) by ADF tests. - VAR lag order selected: seven lags. - Johansen cointegration test: trace statistic close to critical value; λmax indicates one cointegrating equation at 10 percent level. - Disposable income found weakly exogenous; reduced two-equation system used for parsimony. - Final empirical model (reduced specification) — dynamics reported: - Equation (3) (∆p) and Equation (4) (∆r) presented with lagged dynamics (coefficients in Table 4). - Error-correction term (ECT) interpretation: equilibrium house prices rise with disposable income and fall with increases in interest rates; long-run defined by ECT = 0. - Speed of adjustment and persistence: - Speed-of-adjustment low: house prices adjust slowly — "in the course of about 20 quarters." - Lagged house prices significant at various lags — persistence in short-run dynamics. - Historical comparisons and main finding: - Late 1970s/early 1980s: prices were "way out of line with fundamentals" and later corrected. - 1990s through 2004: actual house prices moved largely in line with disposable income and interest rate developments; no evidence of an upward deviation from fundamentals by sample end. - Gradual decline in interest rates during the 1990s and steady increase in total disposable income justified price increases. - Volatility and downside risks: - Estimated equilibrium house price is "quite volatile": small changes in income or interest rates can have large effects. - Example elasticity: "a one percentage point increase in the real interest rate reduces the long-run equilibrium house price by about 10 percent." - Because actual prices adjust slowly, short-term volatility in fundamentals may not be promptly reflected in market prices. - Current weaknesses in disposable income growth and potential rises in interest rates pose downside risks. - Cointegrating coefficients and selected estimation results (normalized PYR): - P: 1; Y: -1.500278 (std. err. 0.19089); R: 0.09423 (std. err. 0.01439). - Adjustment coefficients (std. err.): D(P): -0.050980 (0.01716); D(Y): 0.007261 (0.00389); D(R): 0.918144 (0.44643). - ECT coefficients (sample 1974:Q1–2003:Q3, included observations: 119): - ECT in ∆p: -0.04576 (Std. Error 0.01259), t-Statistic -3.63433. - ECT in ∆r: 0.92633 (Std. Error 0.43286), t-Statistic 2.14003. - Short-run selected coefficients (Equation 1, ∆p) and t-statistics: - DP(-1): 0.23912 (t-Statistic 2.60983). - DP(-3): 0.20509 (t-Statistic 2.63109). - DP(-4): 0.50196 (t-Statistic 6.30156). - DP(-6): -0.25712 (t-Statistic -2.86109). - DY(-4): -1.14737 (t-Statistic -2.79971). - Goodness of fit: - Equation 1 (∆p): R-squared 0.591295; Adj. R-squared 0.540693; S.E. of regression 0.022736. - Equation 2 (∆r): R-squared 0.280008; Adj. R-squared 0.190866; S.E. of regression 0.557617. ### IV. Fiscal framework, medium-term expenditure ceilings, and the CPB - Fiscal consolidation and spending growth: - Initial consolidation target: "0.7 percent every year, implying rapid consolidation." - Subsequent period (1998–2002): real spending growth "1.5 percent per year" while average real GDP growth was "2.8 percent (ex post)". - Current government term (2004–07): ceilings set to increase by "1.4 percent a year in real terms", originally implying a nominal deficit of "0.5 percent of GDP in 2007". - Maastricht safeguard: "The 2.5 percent threshold represents the desire to maintain a 0.5 percent safety margin with respect to the Maastricht ceiling." - Structure and rules of expenditure ceilings: - Overall ceilings subdivided into three sectoral subceilings: (i) “core” sector; (ii) social security; (iii) health care. - Coverage: "around 85 percent of public expenditures will be subject to the ceilings." Included: unemployment spending. Excluded: infrastructure fund expenditures and local government spending. - Revenue windfalls are to be saved under all circumstances. - Within-ministry windfalls may compensate for setbacks; cross-ministry compensation allowed, but not across the three main sectors (rule reinforced in current government). - Ceilings set in real terms over four-year plans and translated annually into nominal ceilings using forecast domestic demand deflator; after Spring review current-year ceilings fixed in nominal terms. - Net expenditure concept: nontax revenues (other than infrastructure fund receipts) treated as negative expenditures under ceilings, reducing transparency. - Experience and outcomes: - Achievements (1994–2000): fiscal balance moved from "4.2 percent of GDP deficit" to a "2.2 percent surplus"; public expenditure fell from "over 58 percent of GDP to about 46 percent"; government debt reduced from "about 77 to 53 percent of GDP"; real GDP growth averaged "3.5 percent a year between 1994–2000." - Deterioration since 2000: fiscal situation worsened to a "3.2 percent of GDP deficit in 2003", exceeding Maastricht ceiling. - Ceilings generally adhered to; overruns only in 1998 and 2003 of "€200 and €400 million, respectively, or less than 0.1 percent of GDP in both cases." - Drivers of 2001–03 deterioration: - Forecasting errors and pre-set spending policies: ceiling for 2003 assumed real GDP growth of "2.5 percent in 2003"; actual outturn was "0.9 percent decline of GDP"; forecasting error for 2003 accounted for "about 1¼ percentage points of the deterioration of the fiscal balance." - Cyclicality of revenues and tax reductions: tax revenues and social security contributions declined from "40.6 percent of GDP in 2000 to 38.7 percent in 2003," explaining "almost 2 percentage points of the deterioration in the balance." "Some 0.8 percentage point of the fall in the revenue ratio was related to the tax cuts in 2001." - Local government spending and asset sales contributed a "0.6 percent of GDP deficit" in 2003. - Policy lessons and recommendations: - Framework effectiveness depends on quality of macro projections and potential growth estimates. - Build and maintain a fiscal buffer to insure against cyclical swings. - Avoid using cyclical windfalls for permanent tax cuts; authorities discontinued that practice. - Widen coverage of ceilings or similar measures to include excluded expenditures (e.g., local government) to improve predictability. - Treat ceilings as focused on spending (avoid treating nontax revenues as negative expenditures) to enhance transparency. - Post-2003 corrective measures: - Measures in 2004 achieved structural adjustment of "about 1¼ percentage points of GDP", reducing the deficit to "2.3 percent of GDP." - Emergency measures effectively returned focus to balance targeting when SGP compliance threatened. - The CPB (Netherlands Bureau for Economic Policy Analysis): - Role: independent economic forecaster and evaluator; provides medium-term projections used by political parties and government; supplies macro assumptions and potential growth estimates for the medium-term fiscal framework. - Independence: technically under ministry of economic affairs but "formally guaranteed by law" operational freedom; director appointed by government but appointment "appears to be relatively nonpolitical." - Contribution to policy: CPB forecasts underpin coalition agreements and medium-term expenditure plans; government does not publish its own macro projections. - Analytical work: ad hoc analyses on fiscal cost of aging, regulatory effects on growth, disability and health-care reforms. - CPB functions similarly to an Independent Fiscal Council (IFC) though not formally linked to SGP requirements. - Performance budgeting: - Introduced from 2002 to link policy goals, actions, and financing with quantitative targets and annual ministry reports. - Upsides: greater transparency, prioritization, reduction in number of spending items. - Challenges: only about half of policies covered by measurable quantitative targets after four years; targets' adequacy questioned; reporting burdensome and of mixed quality; parliamentary scrutiny limited; incentive structures for managers not linked to outcomes. - Recommendation: proceed gradually and selectively, focusing first on areas where measurement is straightforward. ### V. Financial sector — health check, reforms, and FSSA recommendations - High-level sector developments (2004): - Banking: improved balance sheets and profitability; capital-to-asset ratio increased; liquidity increased; trading income recovered cyclically; growth in derivatives activity (CDSs, CDOs). - Mortgages and housing finance: house price growth moderated in 2004; mortgage lending continued to rise; LTV ratios for new loans averaged about 105 percent (average for all loans considerably lower); increased take-up of interest-only mortgages and initial interest-free periods. - Insurance and pensions: insurance profitability improved; pension fund returns edged down but coverage ratio rose to 124 percent; solvency on market interest rates would imply decline from 129 percent in 2003 to 123 percent in 2004. - Households: household debt continued to increase in 2004 but at a slower rate; NPLs of households doubled from a very low base (preliminary data). - Corporates: liquidity and profitability improved; corporate NPLs declined. - Pension reforms effective January 1, 2006: - (i) solvency or coverage requirements; (ii) introduction of “fair value accounting” to mark to market assets and liabilities frequently; (iii) increased transparency for beneficiaries over anticipated payouts. - Minimum requirements imply a coverage ratio, on average, of 130 percent where full indexation is not guaranteed, compared with coverage ratio of 124 percent at end-2004. - When coverage ratio lower than 130 percent, it must be restored within a period not exceeding 15 years. - End-2004 pension fund assets roughly €505 billion. - Potential macro impacts: higher coverage ratios could affect contributions, savings, consumption, and asset allocation. - Regulatory and supervisory developments: - Integrated supervision scheme and DNB restructuring produced policy and operational benefits; New Financial Supervision Act expected to take effect January 1, 2006 (first three parts) and finalization of market infrastructure part by July 2006. - DNB strengthened crisis management preparedness and planning; participating in euro system stress tests and multinational exercises. - FSSA recommendations and actions (selected): - Recommendation I: Minister of Finance approval required for large cross-participations among top banks/insurers; new Supervisory Act to delegate licensing and related tasks to supervisors by law. - Recommendation II: AFM cooperation internationally embedded in Financial Supervision Act; information exchange with non-Member States allowed under conditions. - Recommendation III: Improve cross-border securities settlement — support Euroclear single settlement platform and DVP; AFM and DNB approved Euroclear Netherlands proposal; Single Settlement Engine/Single Application Platform initiative targeted for 2007. - Recommendation IV: Pensions supervisory arrangements — minimum coverage ratio, recovery plans, valuation on term structure, early intervention to avoid procyclicality; footnote: requirement linked to 97½ percent probability implies coverage ratio ≈ 130 percent when indexation not guaranteed. - Recommendation V: Crisis management framework and deposit guarantee scheme review — DNB bilateral agreements, crisis committee concept, review deposit guarantee scheme (move toward partial ex ante funding and risk-based premia likely). - Recommendation VI: Macroprudential surveillance — DNB established Financial Stability division (May 2004), developing macro-stress testing, models (e.g., PALMNET), and regular Financial Stability reports (June and December). - Recommendation VII: Phase out tax deductibility of mortgage interest (medium term) — measures already introduced: “bijleenregeling” (2004), fiscal incentives from January 1, 2005 to pay off mortgages; since 2001 interest on mortgages for consumption/second homes not deductible; deductibility period capped at 30 years. - Selected financial statistics (selected table excerpts): - Banks: assets (1998–2004) — 1,206,632; 1,367,120; 1,606,591; 1,749,569; 1,762,511; 1,911,343; 2,168,530. - Pension funds: assets (1998–2004) — 353,096; 436,263; 444,947; 466,070; 434,931; 489,409; 538,690. - Household debt to GDP (1998–2004): 0.80; 0.87; 0.92; 0.95; 1.00; 1.08; 1.11. - Real estate prices (annual figures, 1998–2004): 123,957; 143,382; 180,376; 192,255; 201,163; 208,899; 211,771. - Regulatory capital to risk-weighted assets (1998–2004): 11.41; 11.18; 11.25; 11.44; 11.86; 12.05; 12.07. - Pension funds: Returns (annual average): 7.3; 11.6; 2.5; -2.8; -0.1; 12.5; 10.1 (periods 1990-94, 1995-99, 2000, 2001, 2002, 2003, 2004). - Pension contribution rate (% of wages) (1990-94 to 2004): 8.5; 9.0; 9.5; 10.0; 12.4; 15.0; 16.3. - Asset-liability ratio (% of liabilities) (pension funds): 107; 135; 132; 118; 101; 108; 124 (selected periods). - Macroprudential and supervisory emphasis: - Continue development of macro-stress testing and indicators. - Monitor mortgage credit practices (interest-only mortgages, LTVs) and pension fund transition effects on savings and asset allocation. - Prepare for phased implementation of new supervisory frameworks to mitigate procyclical outcomes. *Source: _cr05225 (IMF staff report; content extracted from the provided PDF).* ### References.............................................................................................................. ### _cr05225 - References ### I. The External Competitiveness of the Dutch Economy — Key findings - Rankings: - WEF growth competitiveness index (2004): Netherlands ranked 12th out of 103 countries. - IMD ranking (2005): Netherlands ranked 13th out of 60 countries. - IMD ranking changes: 4th (2000) → 13th (2003) → 15th (2004). - Cost competitiveness: - Dutch unit labor costs in manufacturing rose 11 percent faster than euro area competitors and 14 percent faster than competitors in the rest of the world (1998–2003). - In 2004, unit labor costs in the Netherlands decreased by 2.1 percent. - Euro appreciation: appreciated by 18 percent against the U.S. dollar in 2002 and by a further 20 percent in 2003; closed 2003 at $1.2595 per euro. On a trade-weighted basis, euro appreciated by 7.2 percent in 2002 and by 12.1 percent in 2003. In 2004, euro rose 7.6 percent against the dollar to $1.3554 and rose 2.8 percent against a basket of currencies. - Unit labor cost annual changes (selected): - The Netherlands (2001–2004): Compensation per employee: 4.6, 6.5, 3.7, 2.4; Labor productivity: -0.3, 2.2, 0.3, 4.6; Unit labor costs: 4.9, 4.3, 3.5, -2.1. - Euro competitors (2001–2004): Unit labor costs: 2.7, 1.9, 1.3, -2.0. - Non-euro competitors (2001–2004): Unit labor costs: 2.3, -4.9, -11.8, -7.1. - All competitors (2001–2004): Unit labor costs: 2.5, -1.7, -5.6, -4.7. - Export performance: - Real export growth averaged less than 1 percent in 2001–03 while Dutch export market growth averaged above 2 percent annually (WEO data). - Year-average real export growth (goods and services) in 2004: about 8 percent; much of the increase in 2004 driven by sharp growth in re-exports. - Constant Market Share (CMS) results (1992–2003 and 2000–03): - Change in exports (Value in billions of U.S. dollars): 1992–2003: 92; 2000–03: 47. - Average annual growth rate: 1992–2003: 3.8; 2000–03: 12.3. - Decomposition (Value in billions of U.S. dollars): - World Trade Effect: 1992–2003: 22; 2000–03: 22. - Commodity Composition Effect: 1992–2003: -5; 2000–03: -27. - Market Distribution Effect: 1992–2003: -16; 2000–03: 10. - Residual (Competitiveness): 1992–2003: -10; 2000–03: -26. - Interpretation: CMS decomposition indicates competitiveness worsened in recent years (residual negative). ### II. Long-Run Household Consumption Equilibrium — Key findings - Consumption trends: - Since beginning of 2001, average quarterly private consumption growth ≈ 1/4 percent; in the second half of the 1990s average ≈ nearly 1 percent. - Average share of disposable income in GDP fell by over 2 percentage points when comparing 2001–04 with 1995–2000, largely reflecting a decline in property income. - Income and labor market context: - Deceleration of private consumption began earlier than deceleration of per capita real disposable income growth. - The deceleration occurred while unemployment was at a historic low and interest rates had declined. - Wealth and confidence: - Amsterdam stock exchange price index started to fall in 2001:Q2. - Ratio of per capita household net worth to income started to decline toward end-2001, partly influenced by slowing house price increases. - Consumer confidence deteriorated rapidly in tandem with the decline in stock prices. - De Nederlandsche Bank associated decline in consumer confidence with doubts about pension funds solvency, dissatisfaction with public services, and overall policy and political uncertainty. - Empirical model and determinants: - A cointegration model tests determinants suggested by literature: disposable income, net wealth, interest rates, inflation, unemployment (proxy for uncertainty), and demographic variables. - Long-run determinants of Dutch total private consumption identified as: disposable income, net wealth, interest rates, and unemployment. - Short-run results indicate a very small positive role for inflation (as measured by the consumption deflator). - Conclusion on recent consumption reduction: resulted from deceleration of income growth, fall in households’ net wealth, and economic uncertainty. ### III. Document structure and empirical materials (selected) - Sections present in the content unit include: - III. House Prices in the Netherlands (subsections A–G; empirical results and unit root/cointegration tests referenced). - IV. Budgetary Policymaking in the Netherlands (subsections A–F; discussion of Medium-Term Expenditure Framework, role of the CPB, performance budgeting). - V. The Financial Sector in the Netherlands: A Health Check and Progress Report on the FSSA Recommendations (subsections A–B; Financial System Structure and Financial Soundness Indicators tables referenced). - II. Long-Run Household Consumption Equilibrium (cointegration analysis; Appendix II.1 contains detailed econometric results). - I. The External Competitiveness of the Dutch Economy (aggregate and disaggregate trade data; CMS methodology and decomposition). - Figures and tables referenced (titles preserved; data and figure/time spans indicated): - Figures I.1–III.5, IV.1–IV.2, II.1–II.7 (examples: I.1 Real Exchange Rate; III.1 Ratio of House Prices over Disposable Income 1970:Q1–2004:Q2; II.1 Consumption, Income, and Net Worth). - Tables I.1–V.4 (examples: I.1 Unit Labor Costs in the Netherlands and its Competitors; III.1 Unit Root Tests; V.1 Financial System Structure; V.4 Performance of Dutch Pension Funds). - CMS methodology formula and decomposition explicated: global market growth effect, commodity composition effect, market distribution effect, competitiveness (residual) effect. *Source: _cr05225 - References* ### 14. This paper uses the standard permanent income framework to explain ### _cr05225 - 14. This paper uses the standard permanent income framework to explain ### Model framework - Representative agent maximizes: 0{} ttt MaxE VU τ τ τ β + ≥ = ∑ (equation (1) as given). - Intertemporal budget constraint and CRRA utility: Ut = Ct^(1−γ)/(1−γ) (equation (3) as given). - Key variables: - Ct: level of per capita real consumption. - Bt: bonds proxying the stock of financial wealth. - r: rate of return on each unit of that wealth between periods t-1 and t (assumed constant for simplicity). - Yt: per capita real disposable income. - Et: mathematical expectation operator. - β: discount factor; γ: relative risk aversion. - First-order condition (Euler equation) summarized as (5) in the source: - 1 1 . 1 t t t C E C r γ β − + ⎡⎤ ⎛⎞ ⎛⎞ ⎢⎥ = ⎜⎟ ⎜⎟ + ⎝⎠ ⎝⎠ ⎢⎥ ⎣⎦ - Linearized solved-out consumption function reported as (6): 0123 . tttt CAAYABAr=+ + + ### Data and estimation approach - Data: - Per capita real total private consumption used (nondurables series starts only in 1995Q1 and was not used). - Per capita real household wealth proxied by per capita real household net worth (net worth excludes pensions). - Household per capita disposable income and wealth available only at annual frequency; transformed to quarterly using the Lisman filter by the Dutch central bank. - Household net worth and 3-month interest rates deflated by the private consumption deflator. - Unemployment rate from OECD database. - Sample period: 1983Q2–2004Q4. - All series seasonally adjusted using X11. - Unit-root and cointegration testing: - Elliot, Rothenberg, and Stock (1996) DFGLS test: with 95 percent confidence, all series except the unemployment rate contained a unit root; first differences were stationary (Table 1). - Johansen-Juselius test (small-sample bias corrected): strongly rejected no-cointegration at the 95 percent level using λmax or trace statistics (Table 2). The λmax statistic could not reject one cointegrating vector against two; its value (25.42) was very close to the corrected 95 percent confidence value (25.83). - Residuals were white noise. Inclusion of the unemployment rate improved overall fit. - Estimation method: - Equation (6) estimated by the nonlinear dynamic least squares estimator of Phillips and Loretan (1991). - Estimated dynamic form given as equation (7) where Yt includes per capita real disposable income, per capita real net wealth, and the interest rate; Xt includes unemployment rate and inflation (consumption deflator). ### Long-run elasticities and key econometric results - Long-run estimated consumption equation as reported (equation (8) in source): - 3.82 0.100.210.01int0.01, ttttt Cincomewealthunemployment=− ++−− with ρ=0.35. - Associated t-statistics reported: -18.13, 2.19, 4.76, -2.62, -4.22, and 3.71, respectively. - All variables significant at the 95 percent level; equilibrium correction term (ρ) strongly significant. - Signs and magnitudes: - Constant: -3.823558325 (A0 in Appendix 1). - Real per capita disposable income coefficient: 0.098896581 (A1). - Real household net worth per capita coefficient: 0.211184767 (A2). - 3-month real interest rate coefficient: -0.005808650 (A4). - Unemployment rate coefficient: -0.010224217 (A6). - Long-run equilibrium adjustment parameter (DELTA): 0.350438388. - Short-run dynamics: inflation enters short-run consumption dynamics with a positive sign (inflation not significant in long run). - Diagnostic statistics (from Appendix 1): - Centered R**2 = 0.995528; R Bar **2 = 0.993988. - Standard Error of Estimate = 0.008565472. - Sum of Squared Residuals = 0.0044754056. - Cumulated Periodogram Test: Maximum Gap = 0.1399 at Frequency 1.8162; approximate rejection limits 1% = 0.2037, 5% = 0.1700, 10% = 0.1525. - Economic magnitudes and comparisons: - Wealth effects are economically and statistically significant; long-run income elasticity is low (likely reflecting inclusion of property income in real disposable income and use of total consumption). - Based on sample average wealth-to-consumption ratio, the Dutch marginal propensity to consume out of real net wealth is about 5¼ cents to the euro. - Comparisons cited in the source: - United States: 4 cents to the dollar (for gross stock market wealth). - United Kingdom: 5 cents to the pound (for gross stock market wealth). - France: wealth elasticity less than 3 cents to the euro (example of continental Europe). - Dynamics of adjustment: - On average, consumption adjusts to its equilibrium level in about three quarters. - Empirical narrative: long-run estimated consumption caught up with actual consumption around 2000; negative shocks since mid-2001 reduced equilibrium long-run consumption to a trough in 2003Q3; by Q4 2004 actual consumption had caught up with long-run fundamentals. ### Findings and policy implications - Main determinants of total private consumption in the Netherlands: income, wealth, interest rates, and uncertainty (proxied by unemployment rate). - Interpretation: - Large role of real net wealth suggests a liquid and efficient credit market facilitating consumption smoothing and relatively quick adjustment of consumption to persistent changes in wealth. - Uncertainty (unemployment rate) plays a statistically significant role but its economic magnitude is smaller than income or wealth. - Prospects: - With consumption having caught up with fundamentals by end-2004, reasonable prospects for a pick up in consumption exist if growth improves, stock market and house prices boost wealth, and unemployment and other sources of uncertainty subside. - Policy recommendations and structural considerations: - Further structural reforms could raise consumption not only directly via higher potential growth but also indirectly via consumer confidence (consumer confidence sensitive to Dutch stock market developments). - Enhancing transparency of pension fund rules should reduce uncertainty about present value of expected income; reduced uncertainty would increase observed marginal propensity to consume out of income and wealth. *Source: _cr05225 - 14. This paper uses the standard permanent income framework to explain (IMF staff report; source PDF content provided).* ### 27. Between 1995 and 2001, average Dutch house prices rose by almost 80 percent in ### Between 1995 and 2001, average Dutch house prices rose by almost 80 percent in real terms ### Overview of the housing boom and concerns - Between 1995 and 2001, average Dutch house prices rose by almost 80 percent in real terms, raising concerns about sustainability and potential overvaluation. - Observers, including ING, warned that house prices appear out of line with economic fundamentals and suggested a substantial correction could be imminent. - Historical precedent: late 1970s rapid price rise in the Netherlands was followed by a complete reversal in the early 1980s. ### Simple valuation and affordability indicators - Price-to-disposable-income ratio: - Since 1995 this ratio has risen by more than 50 percent in the Netherlands, indicating a sharp decline in housing affordability. - At its current level (as of the study period), the ratio lies some 40 percent above its 30-year average. - Price-rent ratio: - The price-rent ratio has risen by more than 75 percent since 1995. - Its current level is more than 50 percent above the average of the past 30 years. - Interpretation: - On both measures, especially the price-rent ratio, concerns about overvaluation appear well-founded. ### Supply-side context and implications - Structural features of the Netherlands affecting supply: - Population density: 478 people per square kilometer. - Undeveloped land is in short supply and zoning laws tend to be very strict. - Supply elasticity of housing is usually regarded as very low (empirical evidence cited: Swank, Kakes, and Tieman, 2002). - Consequences: - Low supply elasticity can be associated with a tendency for prices to overshoot and greater persistence in price deviations from equilibrium. - Some observers (e.g., Rabobank, 2003) argue the structural supply shortage supports and justifies current high prices. ### Recent developments through 2004: soft landing vs. correction - Macro developments: - After rapid growth, economic conditions deteriorated in 2001 following the global collapse of equity prices and the world economic slowdown. - The Dutch downturn included an unusually sharp contraction of domestic consumption; in 2003 average annual growth turned negative for the first time in 20 years. - House price path: - House price increases moderated substantially during the downturn. - Average prices showed no general decline except at the very high end of the market. - The housing market experienced a softer landing compared with the early 1980s episode. ### Research question and paper structure - Key question: whether the rise in house prices reflects changes in fundamentals, a catch-up toward long-run equilibrium, or an unsustainable deviation implying a future correction. - Paper structure (as presented): - Section B: literature review. - Section C: development of a basic error-correction model of house prices. - Sections D–E: data and econometric issues. - Section F: empirical results. - Section G: concluding remarks. ### Literature synthesis relevant to model specification - Two main approaches in the literature: - Housing as durable consumption: models include real disposable income, interest rates, retail prices, mortgage credit, lagged house prices (e.g., Hendry, 1984; Vladkova Hollar, 2003). - Housing as an asset: comparisons of prices to rents and returns in other asset classes (e.g., Krainer and Wei, 2004). - Supply considerations: - Many studies omit supply when supply elasticity is very low; low supply elasticity increases persistence and reduces speed of mean reversion (Capozza and others, 2002). - Other demand-side drivers: - Inflation can explain a large share of house price variation in cross-country work (Tsatsaronis and Zhu, 2004). - Wealth/stock-market effects: stock prices can lead house prices at a two-three year lag in the Netherlands (Van den End and Kakes, 2002); the Dutch stock market lost about half of its value since 2000, suggesting scope for downward movement in house prices. - Prior empirical assessments for the Netherlands: - Verbruggen and others (2005) find current prices “somewhat” overvalued and sticky downwards. - PricewaterhouseCoopers (2002) conclude 1999–2001 increases likely speculative. - Lopes (2004) finds overvaluation of about 10 percent; IMF (2004) broadly confirms when including population and credit growth. - Catte and others (2004) find weak correlation between house prices and business cycle in the Netherlands. ### Conceptual framework and chosen model - Rationale: - Given very inelastic supply in the Netherlands, the paper focuses on demand factors. - An error-correction model (ECM) is employed to separate short-term dynamics from long-run relationships. - General ECM specification (as given): - ΔZt = μ + Π Zt−1 + Σ (Γi ΔZt−i) + εt, for i = 1..k - Z is an n-vector; Π captures long-run relationships; Γ contains lagged-information; μ is a vector of constants. - Final variable selection (after a General-to-Specific procedure): - Z = (p, y, r) - p denotes real house prices. - y denotes total real disposable income of households. - r denotes the real mortgage interest rate. - Variable choices justified: - Use of total disposable income (not income per household) captures both increase in average household income and rise in number of households; increasing proportion of single-person households important in the Netherlands (Kakes, 2004). - Real mortgage interest rates preferred over nominal rates based on empirical tests and consistency with other studies. ### Data description - Sample frequency and period: - Quarterly data covering 1974:Q1 through 2004:Q2. - Data sources: - House prices: Netherlands’ land registry (Kadaster). - Real disposable income: De Nederlandsche Bank. - Mortgage interest rate: Dutch home owners association (Vereniging Eigen Huis), deflated with CPI from Statistics Netherlands. - Transformations: - House price and income series transformed into logs. - Stylized facts from data plots: - Real house prices show boom-bust in late 1970s/early 1980s, more gradual developments thereafter, and acceleration in the second half of the 1990s with particularly strong increases during 1999–2001. - Growth rates during the latest boom were more moderate than in the late 1970s; since 2001 the growth rate slowed quite abruptly. - Real disposable income: the number of households increases steadily and faster than population growth; real income per household is almost flat. - Real mortgage interest rate: took off around 1976 due to falling inflation, peaked in 1987, then descended during the 1990s. ### Time-series properties and preliminary tests - Integration: - All three series are I(1) (contain a unit root in levels but not in first differences) as tested by augmented Dickey-Fuller (ADF) tests. - Note on disposable income integration: - Evidence on the order of integration for real disposable income is mixed: - ADF test rejects unit root in first differences at the 5 percent level when only a constant is included. - ADF test fails to reject when a trend is added. - Authors accept disposable income as I(1) because the test statistic is very close to the 5 percent critical value in the trend specification and there is no theoretical reason to suspect higher-order integration. ### VAR/cointegration and model selection steps - VAR setup: - A vector autoregression system with the three variables (p, y, r) over 1974:Q1–2004:Q2. - Lag selection: - Lag order tests (selection criteria) unanimously select an order of seven lags. - Cointegration testing: - Johansen cointegration test applied; results summarized in Table 3. - Trace statistic is too close to the critical value to be conclusive. - λmax value strongly indicates the existence of one cointegrating equation including all three variables at the 10 percent level. - Coefficients in the cointegrating equation are very significant and signs are consistent with economic theory. - Model reduction and exogeneity: - In the initial system the cointegrating vector was significant only in the house price and mortgage rate equations. - Disposable income variable found to be weakly exogenous; the income equation could be dropped without loss of information, leaving a two-equation system. - Two-equation system specification: - Each equation contains a constant, an error correction term, and six lags for each of the three variables (initially). - The initial system comprised 40 coefficients. - Parsimony achieved via Wald tests: - Lagged values of the interest rate were not significant in the short-term dynamics of the house price equation. - Lagged house prices did not appear to be a short-term determinant of the interest rate. - Some coefficients and variables were omitted based on statistical insignificance while preserving explanatory power. *Prepared by David Hofman. (Content up to Section F of the source document.)* ### 56. After the reductions described in the previous section, the final empirical model for ### _cr05225 - 56. After the reductions described in the previous section, the final empirical model for ### Final empirical model for house prices (specification) - Equation (3): - ∑∑ = = −− ∆+∆+++−−=∆ 6 1 6 1 21 )557.3094.0500.1(*046.0002.0 i i itiiti ypryppββ - Equation (4): - ∑∑ = − = − ∆+∆++−++=∆ 6 1 4 6 1 3 )84.3796.1564.10(*087.0034.0 i iti i iti ryyprrββ - Notes: - For presentational reasons, Equations (3) and (4) do not show the coefficients for the lagged variables; these coefficients and their t-statistics are reported in Table 4 (estimation output). ### Cointegration, error-correction term (ECT), and dynamics - The error-correction term (ECT) in Equation (3) implies: - Equilibrium house prices rise with disposable income and fall with increases in interest rates. - Long-run equilibrium defined by ECT = 0. - The ECT coefficient is the “speed-of-adjustment” parameter. - Key dynamic properties: - The low value of the speed-of-adjustment indicates house prices adjust only slowly — "in the course of about 20 quarters" — toward long-run equilibrium. - Persistence in short-run dynamics: lagged house prices are significant at various lags. ### Historical episodes and comparison of actual vs equilibrium prices - Late 1970s / early 1980s boom-bust: - Prices were "way out of line with fundamentals." - In the second half of the 1970s, actual house prices rose with a lag in response to a rise in the equilibrium price driven by increasing disposable income and declining mortgage rates. - Around 1976, the environment changed: inflation fell rapidly after the first oil crisis while nominal interest rates rose; markets initially ignored these changing fundamentals, and house prices continued rising through 1979. - By around 1982, actual house prices fell and realigned with fundamental value. - 1990s and recent period through 2004:Q2: - Analysis does not show an upward deviation from fundamentals. - For most of the 1990s, actual house prices were below their long-run equilibrium, with some catching-up in the last few years of the sample. - Overall, actual house prices moved largely in line with developments in total disposable income and interest rates. - The gradual decline of interest rates during the 1990s had a large positive impact on housing affordability, combined with a steady increase in total disposable income, which justified rapid increases in house prices. - Figure-based comparison: - Figure 5 plots actual real house prices (logs) versus the long-run equilibrium house price derived from the cointegrating equation for 1970:Q1–2004:Q2. ### Volatility of the estimated equilibrium and downside risks - The estimated equilibrium house price is "quite volatile": - Even small changes in income or interest rates have a relatively large impact on the fundamental value of housing. - Example: "a one percentage point increase in the real interest rate reduces the long-run equilibrium house price by about 10 percent." - Because actual house prices adjust slowly, short-term volatility of fundamentals is often not reflected promptly in market prices. - Concluding risk assessment: - No evidence of current overvaluation in the Dutch housing market (as of sample end). - Limited comfort: equilibrium price can change rapidly with income and interest rate developments. - Current weaknesses in disposable income growth pose significant risks; a substantial rise in interest rates ahead could still lead to house price falls. ### Selected estimation results (from Table 4 and related tables) - Cointegration and adjustment (Table 3 / Table 4 excerpts): - Normalized cointegrating coefficients (PYR): - P: 1 - Y: -1.500278 (std. err. 0.19089) - R: 0.09423 (std. err. 0.01439) - Adjustment coefficients (std. err. in parentheses): - D(P): -0.050980 (0.01716) - D(Y): 0.007261 (0.00389) - D(R): 0.918144 (0.44643) - Cointegrating Equation (ECT) coefficients (Table 4, sample 1974:Q1–2003:Q3, included observations: 119): - ECT in Equation 1 (∆p): ECT = -0.04576 (Std. Error 0.01259), t-Statistic -3.63433 - ECT in Equation 2 (∆r): ECT = 0.92633 (Std. Error 0.43286), t-Statistic 2.14003 - Short-run coefficients and selected t-statistics (Equation 1, ∆p): - Constant: 0.0015 (Std. Error 0.0031), t-Statistic 0.48167 - DP(-1): 0.23912 (Std. Error 0.09162), t-Statistic 2.60983 - DP(-3): 0.20509 (Std. Error 0.07795), t-Statistic 2.63109 - DP(-4): 0.50196 (Std. Error 0.07966), t-Statistic 6.30156 - DP(-6): -0.25712 (Std. Error 0.08987), t-Statistic -2.86109 - DY(-4): -1.14737 (Std. Error 0.40982), t-Statistic -2.79971 - Short-run coefficients (Equation 2, ∆r) — selected: - Constant: 0.034300 (Std. Error 0.076700), t-Statistic 0.44722 - DP(-1): -34.99337 (Std. Error 10.53298), t-Statistic -3.32227 - DP(-2): 47.04173 (Std. Error 13.30595), t-Statistic 3.53539 - DP(-4): -35.08956 (Std. Error 10.83516), t-Statistic -3.23849 - DP(-5): 28.83957 (Std. Error 13.11439), t-Statistic 2.19908 - DR(-2): -0.27121 (Std. Error 0.11590), t-Statistic -2.34002 - DR(-4): -0.29388 (Std. Error 0.09394), t-Statistic -3.12852 - Goodness of fit (Table 4): - Equation 1 (∆p): - R-squared 0.591295 - Adj. R-squared 0.540693 - S.E. of regression 0.022736 - Equation 2 (∆r): - R-squared 0.280008 - Adj. R-squared 0.190866 - S.E. of regression 0.557617 *Source: IMF staff calculations, as presented in the provided content.* ### 0.7 percent every year, implying rapid consolidation. In the subsequent period (1998–2002), with the public ### _cr05225 - 0.7 percent every year, implying rapid consolidation. In the subsequent period (1998–2002), with the public ### Fiscal framework and consolidation strategy - Spending growth targets and outcomes: - Initial consolidation: "0.7 percent every year, implying rapid consolidation." - Subsequent period (1998–2002): real spending growth of "1.5 percent per year" while average real GDP growth was "2.8 percent (ex post)". - Current government’s term (2004–07): ceilings set to increase by "1.4 percent a year in real terms", originally implying a nominal deficit of "0.5 percent of GDP in 2007". - Maastricht-related safeguard: - "The 2.5 percent threshold represents the desire to maintain a 0.5 percent safety margin with respect to the Maastricht ceiling." - Conservative budgeting assumptions: - Government uses “cautious” macroeconomic assumptions supplied by the CPB when deciding four-year spending plans. - Implication: surprises during budget execution are likely to be on the upside; approach errs on the side of larger-than-planned adjustment given emphasis on fiscal consolidation. ### Structure and rules of expenditure ceilings - Sectors and subceilings: - Overall spending ceilings subdivided into subceilings for three sectors: (i) a “core” sector; (ii) social security; and (iii) the health care sector. - Subceilings are translated into specific budgets for each ministry; every minister is responsible for meeting his/her ministry’s ceiling. - Rules on windfalls and compensations: - Within a ministry, expenditure windfalls may be used to compensate for expenditure setbacks. - Expenditure windfalls may not be used to fund new spending without further consideration by the cabinet (rule added in 2002). - Compensation may occur between ministries, but not across the three main government sectors. (Footnote: "The rule that bans compensation across the three main sectors has not always been adhered to in the past, but the current government renewed its commitment to the rule when it took office.") - "Revenue windfalls, as mentioned above, are to be saved under all circumstances." - Real-term ceilings and annual nominal translation: - Ceilings are set in real terms in the four-year plan. - Each year, real ceilings are translated into operational nominal ceilings using expected price developments (specifically, "a forecast for the domestic demand deflator is used"). - In Spring updates, deflator forecast is reviewed and ceilings adjusted; "After the Spring review, the ceilings for the current year remain fixed in nominal terms." - Coverage and exclusions: - Typical annual coverage: "around 85 percent of public expenditures will be subject to the ceilings." - Included under ceilings: unemployment spending. - Main excluded items: expenditures from a special fund for investment in infrastructure and the spending of local governments. - Infrastructure fund specifics: - Fund financed by "revenues from natural gas sales and interest savings resulting from the sale of government properties." - Purpose: exclude volatile revenue sources from general budget and safeguard infrastructure investments (e.g., "highways and high speed rail connections"). - Earmarking disadvantage: separates infrastructure investments from general budget trade-offs and reduces reach of expenditure ceilings. - Local government accounting issues: - Local governments required to balance budgets over medium term but compile accounts on a different basis than central government. - Local governments may book annual depreciation rather than full investment cost, enabling large investments without breaching medium-term balance requirement. - In consolidated general government accounts on an ESA 95 basis, such investments must be booked up front, creating risks for EMU-consistent balances. - Additional risk: local governments’ ability to accumulate assets and sell them at timing of their choosing. - Net expenditure concept anomaly: - Ceilings pertain to a net expenditure concept where nontax revenues (other than those used for infrastructure investments) are treated as negative expenditures under the ceilings. - Consequence: reduces transparency of expenditure ceilings and hampers monitoring of compliance. ### Experience with the framework (performance and weaknesses) - Entrenchment and political ownership: - Framework served coalitions of various compositions and is "deeply entrenched in Dutch fiscal policymaking." - Closely associated with Minister of Finance Gerrit Zalm, involved in original design (1993 Study Group on the Budget Margin) and in office during most application years (except a short break in 2002/03). - Achievements, 1994–2000: - Fiscal balance moved from an initial "4.2 percent of GDP deficit" to a "2.2 percent surplus." - Public expenditure curtailed from "over 58 percent of GDP to about 46 percent." - Government debt reduced from "about 77 to 53 percent of GDP." - Real GDP growth averaged "3.5 percent a year between 1994–2000." - Deterioration since 2000: - Economic cycle turned in 2001; fiscal situation worsened, culminating in a "3.2 percent of GDP deficit in 2003," exceeding the Maastricht deficit ceiling. - Ceilings generally adhered to; overruns only in 1998 and 2003 of "€200 and €400 million, respectively, or less than 0.1 percent of GDP in both cases." - Main factors behind 2001–03 deterioration: - Relative stance of expenditure policy: - Pre-set spending policies, determined under more favorable outlooks, turned out more expansionary relative to actual GDP. - Example: ceiling for 2003 assumed real GDP growth of "2.5 percent in 2003"; actual outturn was "0.9 percent decline of GDP." - Forecasting error for 2003 accounted for "about 1¼ percentage points of the deterioration of the fiscal balance." - Cyclicality of revenues and tax reductions: - Tax revenues and social security contributions declined from "40.6 percent of GDP in 2000 to 38.7 percent in 2003," explaining "almost 2 percentage points of the deterioration in the balance." - "Some 0.8 percentage point of the fall in the revenue ratio was related to the tax cuts in 2001." - Cyclical windfalls had been used to finance permanent tax cuts under prevailing rules at the time; wealth effects amplified the cycle and initially regarded gains later proved cyclical. - Increases in spending not subject to ceilings: - Local governments unexpectedly ran a "0.6 percent of GDP deficit" in 2003 due to spending financed by sale of previously accumulated assets, contributing to breaching the "3 percent of GDP Maastricht deficit ceiling." - Lessons and policy implications from experience: - Framework effectiveness depends critically on quality of macroeconomic projections and accurate estimates of potential growth. - Fixing expenditure levels four years ahead provides precommitment advantages but risks committing to unintended expansionary stances if potential growth is overestimated. - Cutting taxes when revenue windfalls are sizable and distinguishing cyclical vs. structural revenues had drawbacks; authorities have discontinued using revenue windfalls to finance tax cuts. - Need to build and maintain a fiscal buffer as insurance against sharp cyclical swings in revenues. - Exclusion of relevant expenditure items from the framework reduces assurance about fiscal outcomes; widening coverage of ceilings or similar measures would mitigate this risk. - Authorities’ trial with an "internal stability pact" coordinating central and local government fiscal policies is encouraging. - Post-2003 corrective measures: - Government implemented measures in 2004 to bring deficit under the Maastricht ceiling; achieved structural adjustment of "about 1¼ percentage points of GDP", reducing the deficit to "2.3 percent of GDP." - Emergency measures amounted to a temporary return to balance targeting, consistent with framework provisions requiring additional measures when SGP compliance is threatened. ### Role, mandate, and influence of the CPB - Institutional role and evolution: - CPB = Netherlands Bureau for Economic Policy Analysis; established in 1945 as the "Central Planning Bureau." - Evolved into an independent economic forecaster and evaluator of public sector policies; takes an “academic” approach, stating facts and expected effects but refraining from normative judgments. - Widely accepted and respected across political parties and public as an independent and credible source of economic projections and policy analysis. - Independence: - Technically under the ministry of economic affairs but enjoys complete operational freedom "formally guaranteed by law." - Director appointed by government; appointment process "appears to be relatively nonpolitical"; directors typically outlive governments. - Role in political and budgetary process: - Prior to elections: provides comprehensive medium-term macro forecasts used by political parties as basis for platforms; larger parties voluntarily submit platforms to CPB for impact assessment. - During coalition formation: supplies updated medium-term projections forming the basis for the coalition agreement and medium-term expenditure plan; often supplies several growth scenarios, with coalition partners traditionally opting for the most “cautious” scenario. - CPB supplies key inputs for the medium-term fiscal framework such as macroeconomic assumptions and estimate of potential growth; it does not set fiscal targets but assesses consistency between proposed policies and expected fiscal outcomes. - Government relies on CPB forecasts for annual budgets, Stability Program updates, and fiscal projections; government does not make or publish its own macroeconomic projections. - Role helps eliminate politically-motivated forecast biases, though CPB projections are not infallible. - Ad hoc and analytical work: - CPB provides ad hoc analyses on issues like fiscal cost of aging, effects of regulatory environment on growth, and analyses of new policy proposals (e.g., disability scheme overhaul, health care reform). - CPB responds to requests from government and parliament and also acts on its own initiative. *Italic source: Content extracted from the provided IMF PDF chapter/section.* ### 95. While the CPB is a very influential institution, given its central role in the policy ### _cr05225 - 95. While the CPB is a very influential institution, given its central role in the policy ### The CPB and Independent Fiscal Councils (IFCs) - The CPB functions effectively by virtue of its reputation for impartiality and sound economic reasoning; much of its work is published and under continuous public and academic scrutiny. - The CPB’s performance is regularly assessed by outsiders: in 2001 its role in policy formation was reviewed by an ad hoc committee of domestic officials; in 2003 it was assessed by an international independent scientific review committee. - Case made for strengthening national fiscal institutions in EMU member states through establishment of Independent Fiscal Councils (IFCs) (Annett (2004) and Annett and others (2005)). - An IFC would be a national watchdog providing independent assessments of fiscal policies and monitoring compliance with the SGP; a sufficiently credible IFC could impose reputational costs on governments by highlighting deviations from budgets or stability programs. - In practice, the CPB is similar to an IFC: - It provides an independent check and, given its credibility, can bring reputational costs on the government. - Key difference: the CPB’s role is not linked to the EMU requirements or the SGP; the CPB makes independent assessments and does not regard itself as a guardian of Dutch SGP compliance. - Note: in recent periods explicit linkage to the SGP has not been necessary because the Dutch government has been strongly committed to compliance with European rules. ### Performance Budgeting: design and experience - Introduced starting with the 2002 budget to enhance transparency and accountability by explicitly linking: - (i) ultimate policy goals; - (ii) associated policy actions; - (iii) available financing. - Quantitative performance targets are set for ultimate policy goals and throughput variables; after a fiscal year every ministry must submit an annual report discussing effectiveness with reference to quantitative targets. - Practical outcomes and mixed experience: - Upsides: - Transparency of budget memoranda and accessibility to a wider audience enhanced. - Improved prioritization led to a drastic reduction in the overall number of spending items, contributing to manageability. - Downsides and challenges: - Identification of meaningful performance indicators is difficult; in four years the government succeeded in covering only about half of all policies with measurable quantitative targets. - Concerns about adequacy of many targets; targets measure only limited dimensions of policy effectiveness and risk oversimplification. - Risk that easily attainable rather than most relevant targets are selected because public sector actors have little incentive to document failure. - Bureaucracy and extensive paperwork: combined policy evaluations and annual reports by each ministry reportedly span several thousands of pages each year; average quality of evaluations reported below par. - Parliamentarians have shown little appetite for digesting reports; annual parliamentary discussions on past policy performance mostly uneventful. - Managing trade-offs where multiple ministries’ policies interact is difficult; accounting for exogenous factors outside government control complicates assessment. - Incentive structures for public sector managers have not been adjusted: neither success nor failure to meet targets currently has direct consequences for managers. - Recommendation on implementation approach: - Move ahead gradually and selectively. - Focus first on policy areas where measurement is relatively straightforward and quantitative targets make most sense, then expand only gradually to more complex areas. - Recognize that comprehensive coverage of policies may not be attainable for some time. - Illustrative example from source footnote: - Crime policy objective: reduce crime rates by 20-25 percent between 2002 and 2010. - Operational variables: increase size of police force and capacity of penitentiary institutions; quantitative targets and funding set in each year’s budget; annual reports discuss intermediate targets (more police, more prison cells), amounts spent, and progress toward overall objective (less crime). ### Summary and Conclusions on budgetary framework and CPB role - Merits of Dutch framework: - Medium-term focus allows multi-year commitment, lessening role of short-run political considerations and emphasizing structural policy measures. - By targeting real expenditure while allowing automatic stabilizers to play freely on the revenue side, framework normally avoids procyclical policy biases. - Main practical challenge: - Accuracy of four-year macroeconomic projections, including estimate of potential growth; overestimation of potential growth can commit to overly expansionary fiscal policies. - Conservative budgeting helps alleviate this, but risks remain during periods of greater-than-usual uncertainty about economic outlook. - Suggested refinements: - Adopt a cautious approach to ensuring adequate funding for any further tax cuts given strong cyclical components in revenues. - Build and maintain a fiscal buffer to safeguard consolidation objectives and SGP compliance from unexpected swings in revenues. - Extend coverage of the framework to integrate currently excluded items such as local government spending to improve predictability of fiscal outcomes. - Focus spending ceilings exclusively on spending by ending practice of treating nontax revenues as negative expenditures to enhance transparency. - Role of CPB: - The CPB plays a key role in keeping fiscal discipline by publicly assessing consistency between government policy measures and fiscal/economic objectives, fostering accountability. - The CPB is a successful example of an “Independent Fiscal Council,” though it does not prescribe fiscal targets or policy measures and has no particular interest in Dutch SGP compliance. - Performance budgeting recommendation reiterated: - Move ahead in small steps, concentrating on areas where quantitative targets make most sense and expanding gradually into more complex areas. ### The financial sector: recent developments and reforms (high-level findings) - Structural stability since FSSA; minor changes in number of banks and insurance companies; continued decline in number of pension funds as sector consolidates. - Banking sector (2004): - Improved balance sheets and profitability. - Capital-to-asset ratio of deposit-taking institutions continued to increase. - Ratio of trading income to total income followed a cyclical recovery. - Returns on assets and equity increased despite flattening of yield curve and reduced margins, aided by higher transaction volumes and cost reduction efforts. - Liquidity in Dutch banks increased further. - Growth in financial derivatives activity: CDSs and CDOs expanded in line with global trend. - Increase in loans to other financial institutions reflected rise in interbank lending. - Mortgage and housing market: - House price growth moderated in 2004, yet mortgage lending continued to rise partly reflecting refinancing and market view that trough in interest rate cycle had been reached. - Loan-to-value (LTV) ratios for new loans averaged about 105 percent (average for all loans considerably lower). - Market competitive with insurance firms and other financial institutions allowed to sell mortgages; banks reduced margins and offered interest-free initial mortgage periods. - Concerns noted: growing take-up in interest-only mortgages and difficulty tracking risk related to investment of associated funds to cover future principal payments. - Insurance and pensions (2004): - Insurance sector profitability improved; life insurers limited by low interest rates but aided by better stock market performance; nonlife providers benefited from cost reductions and limits on claims. - If low interest rates were protracted, concerns over firms’ solvency could arise—less pertinent to nonlife sector due to low duration gap. - Pension fund returns edged down in 2004 but remained high; asset-liability (coverage) ratio rose to 124 percent. - Pension funds increased contribution rates contributing to rise in coverage ratio. - Note: solvency ratio calculated on market interest rates would imply decline from 129 percent in 2003 to 123 percent in 2004. - Household and corporate sectors: - Household debt continued to increase in 2004 but at a slower rate; authorities indicated nonperforming loans (NPLs) of households doubled (from a very low base, preliminary data). - Corporate liquidity and profitability improved in 2004; corporations reduced debt and refinanced at lower interest rates; corporate NPLs declined. - Regulatory and supervisory developments: - Integrated supervision scheme produced policy and operational benefits, including: - Shared work on Basel II and Solvency II, cross-group information and expertise-sharing. - Improved supervision of large and complex financial institutions, institutional transfer of human capital, and use of risk assessment tools across sectors. - DNB restructured financial crisis management and planning; identified a crisis team with strengthened support; participated in euro system stress test and plans to take part in a multinational “war game” foreseen for Spring 2006. - New Financial Supervision Act expected to become law on January 1, 2006; intended to be “super neutral” implementing all EU legislation while incorporating additional national measures exceeding or extending EU minima. - Authorities continue to strive to reduce administrative burden on firms; Act did not include audit committees for individual banks as suggested by FSSA. - Pension fund institutional changes effective January 1, 2006: - (i) solvency or coverage requirements; - (ii) introduction of “fair value accounting” to mark to market assets and liabilities frequently; - (iii) increased transparency for beneficiaries over anticipated payouts. - Second tier pensions are defined-benefit plans; new regime will guarantee nominal benefits at a minimum and make explicit the extent of guaranteed indexation; higher coverage ratios required as guaranteed indexation increases. - Macroeconomic implications of pension reform: - New minimum requirements imply a coverage ratio, on average, of 130 percent where full indexation is not guaranteed, compared with coverage ratio of 124 percent at end-2004. - When coverage ratio is lower than 130 percent, it needs to be restored, as agreed with the supervisor, within a period not exceeding 15 years. - At end-2004 Dutch pension funds held roughly €505 billion in assets. - Increasing coverage ratios could affect contributions, savings, and consumption behavior; could induce changes in asset allocation (e.g., shift from equities to bonds), with potential macroeconomic impacts that require careful monitoring. *Source: IMF staff report content as provided in the supplied PDF excerpt.* ### 122. This section elaborates on the actions taken with regard to the FSSA ### _cr05225 - 122. This section elaborates on the actions taken with regard to the FSSA ### Recommendation I: Role of the minister of finance where autonomy is not fully delegated - Starting in 2004, the approval of the minister of finance has only been needed if one of the five largest banks or insurers takes a participation in each other. - The requirement follows from the minister’s political accountability for the well-functioning of the financial system and the budgetary consequences of a potential systemic crisis. - In the new Supervisory Act mentioned, this task will be delegated to the supervisors directly by law, with the licensing of financial institutions done solely by the supervisory authorities. ### Recommendation II: AFM power to cooperate internationally without “domestic interest” - Cooperation of the AFM with supervisory authorities in other Member States is part of the new Financial Supervision Act. - For non-Member States, the Act stipulates that under certain conditions, information exchange may take place (this is not compulsory). ### Recommendation III: Improve cross-border securities settlement arrangements - Improvements: - Supporting Euroclear's single settlement platform. - Promoting the use of delivery of securities against payment (DVP) in settlements. - Limitations: - DVP will not be used in all cases and there will always be some free of payments transactions (FOP). - Approvals and support: - The AFM and the DNB have approved a proposal from Euroclear Netherlands to admit investment firms as participants. - They actively support the Euroclear initiative for building a Single Settlement Engine/Single Application Platform for cross-border DVP in 2007. - Legislative timing: - The first three parts of the new Financial Supervision Act come into effect at the start of 2006. - The fourth part on market infrastructure is expected to be finalized by July 2006. ### Recommendation IV: Pensions supervisory arrangements—flexibility and anti-procyclicality - Overall objective: - Improve transparency and balance the timeframe for making up shortfalls to avoid procyclical effects without unduly prolonging adjustment. - Key elements of the Financial Supervision Act: - Minimum coverage ratio: - Present value of assets must equal at least 105 percent of the present value of liabilities. - If not satisfied, the fund should immediately submit a recovery plan to the DNB for approval, explaining how the minimum requirement will be restored within one year. - When a large number of pension funds experience a sharp fall in capital, the DNB may allow a longer period. - If the coverage ratio is lower than the required amount, but more than 105 percent, the pension fund must immediately inform the DNB and submit a recovery plan explaining how the requirement will be restored as quickly as possible but in any event within 15 years. - The fund must implement the recovery plan and report annually to the supervisors on whether the measures have been implemented and whether the intended effects have occurred. - Fundamental changes: - In the event of a fundamental change in a fund’s volume of investments, or the composition of its investments or commitments, the situation should be reviewed and, if necessary, a new recovery plan drawn up. - Departures from a recovery plan are permitted only if the DNB approves a new recovery plan, in which case a new recovery term will also be observed. - Indexation and recovery plans: - Applying resources to fund indexation (in full or in part) rather than to increase the coverage ratio influences the pace at which the fund meets solvency requirements. - Indexation will be allowed only if it fits in the recovery plan and is stipulated in advance. - Valuation of liabilities: - The present value of liabilities will be calculated on the basis of a term structure of interest rates (instead of a fixed discount rate as currently done). - For a three-year transitional period, the DNB proposes to allow the use of discount rates aligned as closely as possible with the maturity characteristics of the institution’s liabilities rather than the full term structure of interest rates. - Risk management and supervisory intervention: - Every pension fund is responsible for sound financial risk management and proper capital funding of its liabilities. - Information provided to the DNB for regulatory purposes must give a transparent and reliable view of the financial position and risks of the fund. - When it appears that a pension fund will be unable to meet solvency requirements in the future (continuity analysis), the DNB can intervene. - Early intervention can be seen as a vehicle for preventing procyclical effects. - Footnote summary (as presented): - Pension funds will be required to achieve a coverage ratio implied by a 97½ percent probability of meeting their obligations. - In cases where indexation is not guaranteed, the latter implies a coverage ratio, on average, of 130 percent. - Under the increased transparency associated with the new framework, pension funds that intend to provide full indexation will have to clearly state that the funds available for that purpose are those that exceed the 130 percent required to provide only partial indexation. ### Recommendation V: Crisis management framework and deposit guarantee scheme review - DNB actions: - Entered bilateral discussions on signing another crisis management agreement. - The agreement envisages a crisis management committee as a mechanism for consultation in crises, without prejudice to the formal responsibilities of each institution, and defines the information items which should be available within the signatory financial institutions. - Critically reviewed its crisis management preparedness and decided to keep its approach pragmatic. - Identified areas for improvement: internal coordination, communication, and documentation. - Reestablished a bank-wide team at board/division level to review and improve procedures (including emergency liquidity assistance). - EU-level actions: - A new memorandum of understanding (MoU) for financial crisis situations between the banking supervisors, central banks, and finance ministries of the EU was agreed upon and signed in May 2005. - The Financial Services Committee is reviewing current national crisis management arrangements in the EU to ensure compatibility with the MoU. - Deposit guarantee scheme review: - Authorities are reviewing the Dutch deposit guarantee scheme, which is not ex ante capitalized, and considering the introduction of prefunding and risk-based premia. - To minimize the burden on banks in circumstances in which the banking system is highly concentrated, it is likely that changes to the scheme will result in partial (rather than full) ex ante funding and risk-sharing. ### Recommendation VI: Macro prudential surveillance development - DNB measures: - Established the division of Financial Stability in May 2004 with the main task of identifying risks and vulnerabilities to financial stability. - Published several papers building up an analytical financial stability framework. - Building on the FSAP stress test methodology, developing a framework for macro-stress testing as a regular tool for assessing financial stability. - Developing models and indicators to support analyses, such as the pension model PALMNET and a set of macro prudential indicators from a joint DNB effort. - Financial Stability reports are published regularly in June and December (an internal version includes policy recommendations). ### Recommendation VII: Phase out tax deductibility of mortgage interest (medium term) - Background and measures: - Tax deductibility of mortgage interest payments had already been limited to some degree. - In 2004, the “bijleenregeling” was introduced: the capital gain made on a sold house is assumed to be used for the purchase of the new house and that amount is no longer eligible for mortgage interest deduction. - As of January 1, 2005, a fiscal incentive to pay off mortgage loans completely, or to a large extent, was introduced. - Footnote summary (as presented): - Since 2001, interest on mortgage loans used for consumption or the purchase of a second home have not been deductible. - The period for the deductibility of mortgage interest has been capped at 30 years. ### Selected statistics and indicators (as presented in tables) - Table 1. Netherlands: Financial System Structure (selected figures) - Banks: assets 1,206,632; 1,367,120; 1,606,591; 1,749,569; 1,762,511; 1,911,343; 2,168,530 (years 1998–2004). - State-owned banks: assets 56,667; 61,220; 61,892; 67,799; 73,575; 83,970; 88,598. - Branches of foreign banks: assets 24,129; 24,551; 33,884; 29,732; 28,396; 27,673; 31,481. - Pension funds: assets 353,096; 436,263; 444,947; 466,070; 434,931; 489,409; 538,690. - Banks: deposits 560,392; 637,106; 734,462; 817,582; 838,441; 909,748; 999,711. - Table 2. Netherlands: Encouraged Financial Soundness Indicators (selected figures) - Corporate sector: - Total debt to equity: 0.69; 0.69; 0.70; 0.69; 0.68; 0.66; 0.52 (1998–2004). - Return on equity: 0.18; 0.18; 0.16; 0.14; 0.14; 0.14; 0.16. - Earnings to interest and principal expenses: 2.37; 2.4; 1.84; 1.46; 1.49; 1.93; 5.16. - Deposit-taking institutions: - Capital to assets: 0.04; 0.05; 0.05; 0.06; 0.06; 0.06; 0.07. - Financial derivatives to capital: 71; 87; 84; 94; 99; 99; 109. - Customer deposits to total (non-interbank) loans: 0.82; 0.77; 0.76; 0.80; 0.79; 0.83; 0.84. - Households: - Household debt to GDP: 0.80; 0.87; 0.92; 0.95; 1.00; 1.08; 1.11. - Household debt service and principal payments to income: 0.10; 0.11; 0.12; 0.11; 0.11; 0.12; 0.12. - Real estate prices (annual figures): 123,957; 143,382; 180,376; 192,255; 201,163; 208,899; 211,771 (1998–2004). - Table 3. Netherlands: Financial Sector Indicators (selected figures, in percent) - Regulatory capital to risk-weighted assets: 11.41; 11.18; 11.25; 11.44; 11.86; 12.05; 12.07 (1998–2004). - Regulatory Tier I capital to risk-weighted assets: 8.65; 8.57; 8.76; 8.63; 9.13; 9.55; 9.91. - Return on assets: 0.44; 0.57; 0.60; 0.42; 0.35; 0.46; 0.54. - Return on equity: 10.89; 14.29; 14.74; 10.76; 9.17; 12.27; 13.81. - Table 4. Netherlands: Performance of Dutch Pension Funds (annual averages and ratios) - Returns (annual average): 7.3; 11.6; 2.5; -2.8; -0.1; 12.5; 10.1 (periods 1990-94, 1995-99, 2000, 2001, 2002, 2003, 2004). - Contribution rate (% of wages): 8.5; 9.0; 9.5; 10.0; 12.4; 15.0; 16.3. - Asset-liability ratio (% of liabilities): 107; 135; 132; 118; 101; 108; 124. *Source: National authorities, as presented in the section.* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2005/_cr05225.pdf_