1

The COM fiscal (debt) sustainability analysis

frameworkStéphanie Pamies Sumner

European CommissionECFIN.C2 - Sustainability of Public Finances

ESM workshop on Debt sustainability: current practice and future perspectives

Luxembourg, 11-12 December 2018

Contents

• Main features of the Commission fiscal (debt) sustainability analysis framework

• Focus on special tools• On-going work (forthcoming Fiscal Sustainability

Report)

2

Fiscal risks are still present in some – often large - EU countries

Source: Commission services, IMF (GDD)

020406080100120140160180200

Government debt(% of GDP, 2018)

0

20

40

60

80

100

120

140

160

1960

1964

1968

1972

1976

1980

1984

1988

1992

1996

2000

2004

2008

2012

2016

2020

2024

2028

% of GDP Government debt, Italy

Debt reduction episode Baseline debt projections Debt-to-GDP ratio

-10

10

30

50

70

90

110

1978

1982

1986

1990

1994

1998

2002

2006

2010

2014

2018

2022

2026

% of GDP Government debt, France

Debt reduction episode Baseline debt projections Debt-to-GDP ratio

Main features of the Commission fiscal (debt)

sustainability analysis framework

3

What are the key features of the COM fiscal sustainability framework?

A comprehensiveand multidimensional approach

A consistentframework across all countries (common assumptions and methodologies)

A timelyassessment of fiscal sustainability (updates twice a year)

COM regular DSA components –overview

Risk-classification (S-indicators, debt level &

trajectory, realism of fiscal assumption, uncertainties)

Risk-classification (S-indicators, debt level &

trajectory, realism of fiscal assumption, uncertainties)

Debt profile (currency, holders,

maturity)

Debt profile (currency, holders,

maturity)

Contingent liabilities (incl. linked to the

banking sector)

Contingent liabilities (incl. linked to the

banking sector)Other risk factors (e.g. public assets)Other risk factors (e.g. public assets)

Risk-based approach

Risk-based approach

Benchmarks derived from early-detection model

Banking sector: simulations based on model (SYMBOL)

4

COM framework - core tools used for fiscal sustainability risk classification

• Short-term – liquidity risks: early-detection indicator (S0 indicator) => captures risks from fiscal and macro-financial sides

• Medium-term – solvency risks: DSA and medium-term fiscal gap indicator (S1 indicator)

• Long-term – solvency risks: long-term fiscal gap indicator (S2 indicator)

Overall risk classification by time dimension

COM DSA framework includes• Traditional (deterministic) 10-year public debt

projections • Sensitivity analysis around baseline projections and

alternative policy scenarios (around 20 ‘standard’ and ‘enhanced’ scenarios + additional customised sensitivity tests)

• Stochastic public debt projections (2000 shocks)• Degree of realism of fiscal assumptions• Gross financing needs projections• Forecast accuracy analysis• Public debt profile• Contingent liabilities• Financial markets’ information

Overall DSA risk classification: low/medium/high

5

A synthetic yet granular assessment of fiscal risks

BE DE EE IE ES FR IT CY LV LT LU MT NL AT PT SI SK FIS0 overall index 0.27 0.00 0.20 0.25 0.37 0.24 0.36 0.44 0.21 0.21 0.08 0.05 0.12 0.07 0.41 0.13 0.27 0.07

Overall SHORT-TERM risk category LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW LOW

BE DE EE IE ES FR IT CY LV LT LU MT NL AT PT SI SK FIS1 indicator - Baseline scenario 4.3 -2.0 -4.6 -0.7 5.4 5.5 7.5 -0.6 -1.5 -1.5 -4.4 -5.2 -1.5 -0.5 4.4 1.0 -2.7 0.4

S1 indicator - overall risk assessment HIGH LOW LOW LOW HIGH HIGH HIGH LOW LOW LOW LOW LOW LOW LOW HIGH MEDIUM LOW MEDIUM

BE DE EE IE ES FR IT CY LV LT LU MT NL AT PT SI SK FIBaseline no-policy change scenario HIGH LOW LOW LOW HIGH HIGH HIGH MEDIUM LOW LOW LOW LOW LOW LOW HIGH MEDIUM LOW LOW

Debt level (2028) 100.2 37.9 7.7 50.8 109.5 108.2 132.5 66.9 39.4 36.2 11.8 15.0 41.0 55.1 108.5 61.0 34.0 58.6Debt peak year 2017 2017 2017 2017 2028 2028 2028 2018 2017 2017 2017 2017 2017 2017 2017 2017 2017 2017Average Structural Primary Balance (2019-2028) Percenti le rank 48% 25% 70% 33% 69% 76% 30% 15% 73% 54% 45% 19% 48% 39% 22% 55% 49% 57%

Historical SPB scenario HIGH LOW LOW MEDIUM HIGH HIGH HIGH MEDIUM LOW LOW LOW LOW LOW LOW HIGH MEDIUM LOW LOWDebt level (2028) 94.2 41.2 5.4 68.0 108.3 108.6 129.5 85.1 39.4 45.8 3.3 29.9 39.6 57.0 129.3 66.4 49.5 44.3Debt peak year 2017 2017 2017 2028 2028 2028 2017 2018 2017 2028 2017 2017 2017 2017 2028 2017 2017 2017Average Structural Primary Balance (2019-2028) Percenti le rank 37% 28% 68% 63% 68% 75% 27% 28% 73% 67% 30% 35% 45% 42% 50% 62% 72% 33%

Negative shock (-0.5p.p.) on nominal GDP grow th HIGH LOW LOW LOW HIGH HIGH HIGH MEDIUM LOW LOW LOW LOW LOW LOW HIGH MEDIUM LOW MEDIUMDebt level (2028) 105.6 40.6 8.1 53.7 115.2 113.7 140.3 71.8 41.3 38.2 12.7 16.6 43.5 58.5 115.2 64.2 36.0 61.7Debt peak year 2028 2017 2017 2017 2028 2028 2028 2018 2028 2017 2017 2017 2017 2017 2017 2017 2017 2028

Positive shock (+1p.p.) to the short- and long-term interest rates on new ly issued and rolled over debt HIGH LOW LOW LOW HIGH HIGH HIGH MEDIUM LOW LOW LOW LOW LOW LOW HIGH MEDIUM LOW MEDIUM

Debt level (2028) 105.9 40.6 8.0 53.3 115.6 114.5 141.7 69.0 41.5 38.4 12.3 15.8 43.6 57.4 113.3 64.4 35.6 61.3Debt peak year 2028 2017 2017 2017 2028 2028 2028 2018 2028 2017 2017 2017 2017 2017 2017 2017 2017 2017

Negative shock on the PB equal to 50% of the forecasted cumulative change over the tw o forecast years HIGH LOW LOW LOW HIGH HIGH HIGH MEDIUM LOW LOW LOW LOW LOW LOW HIGH MEDIUM LOW MEDIUM

Debt level (2028) 103.7 40.6 9.0 53.3 111.4 114.6 135.9 73.0 43.6 37.8 19.4 26.8 46.0 56.3 111.3 69.6 34.0 62.6Debt peak year 2028 2017 2028 2017 2028 2028 2028 2018 2028 2017 2019 2017 2017 2017 2017 2017 2017 2028

Stochastic projections HIGH LOW LOW MEDIUM HIGH HIGH HIGH MEDIUM MEDIUM MEDIUM LOW LOW LOW LOW MEDIUM LOW LOW LOWProbabili ty of debt in 2022 greater than in 2017 (%) 32% 0% 6% 16.3% 57% 57% 33% 30% 39% 40% 33% 1% 3% 10% 23% 12% 19% 40%Difference of the 10th and 90th percenti le in 2022 (p.p. of GDP) 29.9 15.7 3.2 30.6 18.2 13.6 25.297 44.6 39.5 32.9 22.0 20.0 17.2 27.8 38.3 25.9 28.9 18.8

Debt sustainability analysis - overall risk assessment HIGH LOW LOW LOW HIGH HIGH HIGH MEDIUM LOW LOW LOW LOW LOW LOW HIGH MEDIUM LOW MEDIUM

Overall MEDIUM-TERM risk category HIGH LOW LOW LOW HIGH HIGH HIGH MEDIUM LOW LOW LOW LOW LOW LOW HIGH MEDIUM LOW MEDIUM

BE DE EE IE ES FR IT CY LV LT LU MT NL AT PT SI SK FIS2 indicator - Baseline scenario 4.3 1.6 0.6 3.1 2.4 1.0 1.8 -1.3 1.1 0.7 8.5 2.8 3.0 2.9 0.7 6.1 2.4 2.9

Overall LONG-TERM risk category MEDIUM LOW LOW MEDIUM MEDIUM LOW LOW LOW LOW LOW HIGH MEDIUM MEDIUM MEDIUM LOW HIGH MEDIUM MEDIUM

Heat m ap for short-term risks in the EU countries

Heat m ap for m edium-term risks in the EU countriesS1 indicator in the EU countries

Sovereign-debt sustainability risks in the EU countries

Heat map for long-term risks in the EU countries

Benchmark values for debt level Debt trajectory Realism of fiscal

assumptions

Probability of debt stabilisation Size of uncertaintiesCritical thresholds for

S-indicators

Across scenarios

A transparent decision tree for the DSA risk classification

6

Assessment criteria used for DSA overall assessment

Focus on special tools and scenarios

7

Short-term (liquidity) risks: the S0 indicator

• An early-detection indicator of fiscal distress based on realized data

• Building on past episodes of fiscal distress in advanced economies, and on the behaviour of a large set of variables ahead of these events

• S0 indicator: composite indicator based on 25 fiscal and macro-financial variables – including from the EU Macroeconomic Imbalance Procedure

• Signal detection approach => critical thresholds derived by minimising the share of missed crises & false alarms

S0 indicator (Spring forecast 2018) Variables considered, thresholds and signalling power

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

DE

MT FI AT

LU

NL SE SI CZ

DK

RO EE

LT

LV

HR FR PL IE BG BE

SK IT ES

UK

HU PT CY

2009 2017

S0 threshold

In-sample, close to 80% of fiscal distress events correctly predicted –performing well (see IMF - Cerovic et al., 2018)

Variables with highest signalling power include (macro-financial) private credit flows, current account, (fiscal) GFN, cycl. adjusted balance

S0 indicator: a clear and comprehensive mapping of short-term vulnerabilities

Variables safety thresholdsignaling

power

Balance, % GDP > -9.61 0.07Primary balance, % GDP > 0.23 0.13Cyclically adjusted balance, % GDP > -2.50 0.23Stabilizing primary balance, % GDP < 2.34 0.08Gross debt, % GDP < 68.44 0.12Change in gross debt, % GDP < 8.06 0.12Short-term debt gen. gov., % GDP < 13.20 0.20Net debt, % GDP < 59.51 0.20Gross financing need, % GDP < 15.95 0.26Interest rate-growth rate differential < 4.80 0.08Change in expenditure of gen. government, % GDP < 1.90 0.11Change in final consumption expend. of gen. government, % GDP < 0.61 0.07Fiscal index < 0.36 0.28L1.net international investment position, % GDP > -19.80 0.29L1.net savings of households, % GDP > 2.61 0.33L1.private sector debt, % GDP < 164.70 0.18L1.private sector credit flow, % GDP < 11.70 0.37L1.short-term debt, non-financial corporations, % GDP < 15.40 0.20L1.short-term debt, households, % GDP < 2.90 0.21L1.construction, % value added < 7.46 0.22L1.current account, 3-year backward MA, % GDP > -2.50 0.34L1.change (3 years) of real eff. exchange rate, based on exports deflator, ref 37 countries < 9.67 0.11L1.change (3 years) in nominal unit labour costs < 7.00 0.18Yield curve > 0.59 0.37Real GDP growth > -0.67 0.10GDP per capita in PPP, % of US level > 72.70 0.22Financial-competitiveness index < 0.49 0.55Overall index < 0.46 0.55

8

Alternative fiscal policy scenarios: importance for fiscal surveillance

• SGP scenario• SCP / DBP scenario • Others (historical SPB

scenario, fiscal reaction function scenario)

• DSA: assessed around usual set of criteria: projected debt level / path, and plausibility of the underlying fiscal effort

• Results used for Member States (EA) SCP (DBP) assessment

70.0

75.0

80.0

85.0

90.0

95.0

100.0

105.0

110.0

115.0

120.0

2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028

Gross public debt as % of GDP - BE

Baseline no-policy change scenarioNo-policy change scenario without ageing costsStability and Growth Pact (SGP) scenarioStability and Convergence Programme (SCP) scenario

0%

2%

4%

6%

8%

10%

12%

Less -4 -3 -2 -1 0 1 2 3 4 5 6 More

2019 SPB for EA: percentile rank of 44%

AVG 03-17 SPB for EA: percentile rank of 49%

AVG 19-28 SPB under SGP for EA: percentile rank of 25%

3-year avg SPB greather than 0.7% of GDP represents 44% of the sample

Stochastic projections: a probabilistic assessment

30.0

40.0

50.0

60.0

70.0

80.0

90.0

100.0

110.0

2015 2016 2017 2018 2019 2020 2021 2022

(% of GDP) Stochastic debt projections 2018-2022, IE

p10_p20 p20_p40

p40_p60 p60_p80 p80_p90

p50 baseline scenario

20

40

60

80

100

120

140

IT PT BE CY ES FR UK HR AT SI HU IE DE FI EA

% GDP

Debt ratio, 2017 Median debt ratio, 2022Non-incresing debt cap, 2022

• Stochastic projections results are used to inform the overall DSA risk classification

• Also use to derive ‘non-increasing debt caps’

BE DE EE IE ES FR IT CY LV LT LU MT NL AT PT SI SK FI

Stochastic projections HIGH LOW LOW MEDIUM HIGH HIGH HIGH MEDIUM MEDIUM MEDIUM LOW LOW LOW LOW MEDIUM LOW LOW LOWProbability of debt in 2022 greater than in 2017 (%) 32% 0% 6% 16.3% 57% 57% 33% 30% 39% 40% 33% 1% 3% 10% 23% 12% 19% 40%Difference of the 10th and 90th percentile in 2022 (p.p. of GDP) 29.9 15.7 3.2 30.6 18.2 13.6 25.297 44.6 39.5 32.9 22.0 20.0 17.2 27.8 38.3 25.9 28.9 18.8

Debt sustainability analysis - overall risk assessment HIGH LOW LOW LOW HIGH HIGH HIGH MEDIUM LOW LOW LOW LOW LOW LOW HIGH MEDIUM LOW MEDIUM

Sovereign-debt sustainability risks in the EU countries

9

Beyond (EDP) debt

Tentative categorization of gross financing liabilities of general government

Capturing risks associated to tail events: the Symbol model

• Additional ‘modules’ to assess risks related to contingent liabilities • Based on Eurostat specific reporting (‘Six-pack’)• Heat map on potential triggers for contingent liabilities from

banking sector • Micro-simulation model (Symbol – JRC-FISMA) providing

estimates of the potential impact of banking losses on public finances (implicit contingent liabilities)

bank recap. at 8% bank recap. at 10.5%10.2 0.3 83.9 44.8 -4.2 39.7 0.11% 0.57%

Government's contingent liability risks from banking sector - CY (2016)

Private sector credit flow (%

GDP):

Bank loans-to-deposits ratio

(p.p.):

Share of non-performing loans (%):

Change in share of non-performing

loans (p.p):

NPL coverage ratio

Change in nominal house price index:

Probability of gov't cont. liabilities (>3% of GDP) linked to banking losses and recap needs (SYMBOL):

Upper thresholds from signals' approach, lower thresholds at 80% of signals' approach thresholds; Symbol results: specific thresholds’ calibration

10

Grasping the size of the long-term fiscal sustainability challenge: long-

term projections and S2 indicator

1.71.10.5

2.3 2.5

0.8

2.7

0.6

2.52.7

0.8

2.9

22

24

26

28

30

32

Bas

elin

e

TF

P ri

sk

AW

G r

isk

TF

P_A

WG

ris

k

HL

E

Polic

y re

vers

al

Bas

elin

e

TF

P ri

sk

AW

G r

isk

TF

P_A

WG

ris

k

HL

E

Polic

y re

vers

al

EU EA

Cos

t of

agei

ng a

s %

of

GD

P

2016 Change 2016-2070 Additional change compared to the baseline scenario

Long-term fiscal gap indicator (S2 indicator) and sub-components, by country (Spring forecast 2018)

Source: Ageing Report 2018, forthcoming Commission FSR

-4

-2

0

2

4

6

8

10

HR

CY

DK EE

LT

PT FR SE LV

BG

DE IT ES

SK PL MT

AT FI

NL IE UK

CZ

BE

HU

RO SI

LU

Gap to debt-stabilizing primary balance Cost of ageing S2 indicator

Medium risk

High risk

Projected change in cost of ageing, baseline and risk scenarios, 2016-2070

On-going work (forthcoming Fiscal

Sustainability Report)

11

On-going work

• Enriching the short-term analysis (definition and measure of financing needs, markets’ information)

• Complementing the set of sensitivity tests (e.g. financial assumptions, uncertainties over the long-term)

• Long-term fiscal sustainability assessment: better accounting for risks due to medium to high debt levels

• Government assets and net debt / net worth

Defining and measuring government financing needs

• GFN is an important, complementary, variable to assess fiscal risks

• Yet, lack of harmonisation of GFN definition and estimation

Note: 1) GFN = budgetary deficit + debt securities amortizations. 2) The size of the bubble reflects government interest rate yields.Source: ECB, Commission services

AT BE

CY

DE

ES

FI FR

IE

IT

LT

LV NL

PT

SI

SK

UK EL

-5

0

5

10

15

20

25

30

0 50 100 150 200

GFN

to

GD

P ra

tio

in 2

018

Debt to GDP ratio in 2018

12

Taking into account markets’ expectations to project interest rates?

• Uncertainties regarding future developments of interest rates

• Different options: conventional assumption based on historical patterns, using financial markets’ expectations, assumption on (i – g)?

-10

-5

0

5

10

15

20

25

30

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

interpercentile range (p10-p90) min-max range

EA average average distressed

EL

EELT

EL

Evolution of long-term interest rates, across the euro area countries, %

EU government debt projection, under the baseline and alternative scenario and the standard sensitivity test, %

Long-term fiscal sustainability assessment: accounting for risks

due to medium to high debt levels

• For countries where debt burdens are significant, meeting the intertemporal budget constraint may not be sufficient to secure fiscal sustainability

• Therefore, the long-term fiscal sustainability assessment, currently based on the S2 indicator, needs being complemented

13

Government assets, net debt and net worth

• Moving towards a balance sheet approach has gained popularity but…

• Many measurement and conceptual issues for fiscal sustainability

fixed assets

inventories valuablesnatural

resources

contracts, leases & licences

goodwill & marketing

assetsBE X XBGCZ X X X X X XDK XDE XEE X XIEEL X XESFR X X X X X XHRIT X XCYLV X X XLT X XLU XHU X XMTNL X XAT XPL X XPT X XROSI X XSK X XFI X X XSE X X XUK X X X X

Source: Eurostat

produced non-produced

Data availability for non-financial assets in national accounts (general government; 2015, 2016 or 2017)

0%

50%

100%

150%

200%

EE LV LT CZ SI SE NL SK FI PL HU PT AT FR DK IT BE DE UK LU

Dwellings and other buildings and structures (% GDP; 2015)

National accounts (Eurostat) ECFIN-KPMG

larger discrepancy between estimates

Thank you for your attention

https://ec.europa.eu/info/publications/debt-sustainability-

monitor-2017_en