1.
Fiscal Crises
Tigran Poghosyan
IMF, Fiscal Affairs Department
ADEMU Workshop
“Fiscal Risk and Public Sector Balance Sheets”
University of Bonn, July 6-7, 2017
This presentation is based on an IMF working paper WP/17/86 authored by Juan Farah-Yacoub, Kerstin Gerling, Paulo Medas, Tigran Poghosyan and Yizhi Xu. The views are those of the authors
and do not necessarily represent the views of the IMF, its Executive Board, or IMF management.

2.
Motivation
 Fiscal sustainability is a key policy objective …
 … but empirical identification of fiscal crises is scarce
 Previous literature
 Focuses mainly on public debt defaults (legal definition) …
 … mostly external …
 … mostly in advanced and emerging economies
 Baldacci et al. (2011) is an exception (economic criteria)
• Also adopted by the EC recently (Sumner & Berti 2017)
 Our contribution
 Wider country and time coverage (all 188 IMF members, 1970-2015)
 Improved identification methodology
 Event study analysis – dynamics of macroeconomic variables around
fiscal crises
1IMF, Fiscal Affairs Department
Our database
436 fiscal crises in total, of which:
 25 in Advanced Markets (AM)
 154 in Emerging Markets (EM)
 171 in Low-Income and Developing
Countries (LIDC)
 86 in Small Developing States (SDS)
Existing databases
 Reinhart & Rogoff (2011): 75 external
and 26 domestic defaults
 Baldacci et al. (2011): 176 fiscal crises
 Laeven & Valencia (2012): 67 debt
defaults
 Bruns & Poghosyan (2016): 201 fiscal
crises
 Sumner & Berti (2017): 88 fiscal crises

3.
Outline
I. New Database on Fiscal Crises
II. Descriptive statistics
III. Empirical Analysis
IV. Concluding remarks
2IMF, Fiscal Affairs Department

4.
I. New Database on Fiscal Crises
3IMF, Fiscal Affairs Department

5.
Red Flags: How To Identify Fiscal Crises?
4IMF, Fiscal Affairs Department
• Any year in which actions of the sovereign reduce the PV of public debt
• Exclude technical defaults
• Examples: Bulgaria 1990; Macedonia 2010
1. Credit event
• Any year under a large IMF financial arrangement …
• … with fiscal adjustment as a program objective
• Examples: Hungary 2008; Ireland 2010
2. Official financing
• Any year with a very high inflation
• Any year with domestic arrears accumulation (new!)
• Examples: Belarus 1999; Russia 2007
3. Implicit default
• Any year with a very high price of market access
• Any year with loss of market access (new!)
• Examples: Ukraine 2008; Romania 2009
4. Market confidence

6.
IMF, Fiscal Affairs Department 5
Criterion Sources
AMs EMs LIDCs SDSs
(1) Credit event Any operation that makes creditors incur material
economic losses on the sovereign debt they hold
(e.g. default, restructuring, or rescheduling)
(i) of substantial size (in percent of GDP p.a.)
(ii) if (i) holds and the defaulted nominal amount
grows by a substantial amount (in percent p.a)
(2) Exceptionally
large official
financing
High-access IMF financial arrangement with fiscal
adjustment objective in place (in percent of quota)
▪ IMF Strategy, Policy, and Review
Department (SPR) as well as Finance
Department (FIN)
(3) (a) High inflation rate (in percent of growth of
annual average CPI p.a.)
≥ 35 ≥ 35 ▪ IMF WEO
(b) Steep increase in domestic arrears (in first
difference of the ratio of 'other account payables
(OAP)' to GDP in percentage points)
▪ OECD and Eurostat
(4) (a) Loss of market access ▪ Guscina, Sheheryar and
Papaioannou (2016)
▪ For gaps: authors' research based
on Gelos, Sahay, and Sandleris
(2004) and rating agencies' reports
▪ For gaps: authors' research based(b) High price of market access (in basis points,
sovereign spreads or CDS spreads)
▪ Reuters Datastream, Bloomberg,
IMF IFS
Thresholds
▪ Bank of Canada (2016)
▪ For gaps: authors' research based
on rating agency reports (esp.
Moody's and S&P), WB IDS and WDI
data, and Cruces and Trebesch (2013)>0.2
≥ 10
≥ 100
≥ 100
≥ 1
Loss of market
confidence
when market access is lost
(after maintaining market
access for a 1/4 of the sample
time and 2 consecutive years
before the loss year)
≥ 1,000 bps
Implicit
domestic
public default
Note: AM = advanced markets, EM = emerging markets, LIDC = low-income and developing countries, SDS = small developing states.

7.
II. Descriptive Statistics
6IMF, Fiscal Affairs Department

8.
Fiscal Crisis Episodes (1970-2015)
IMF, Fiscal Affairs Department 7
Source: Authors’ calculations.
AM = advanced markets, EM = emerging markets, LIDC = low-income and developing countries, SDS = small developing states.
Total AM EM LIDC SDS
With start date within sample period 436 25 154 171 86
Average per country 2.3 0.7 2.2 3.4 2.6
With start and end date within sample period 400 23 143 154 80
Average per country 2.1 0.7 2.0 3.1 2.4
Memorandum items:
Number of countries with no fiscal crisis 37 20 9 1 7

9.
Dynamics of Fiscal Crises
8
0
20
40
60
80
100
120
140
1970 - 1979 1980 - 1989 1990 - 1999 2000 - 2004 2005 - 2009 2010 - 2015
By Criteria
Credit event Exceptional financing Implicit default Market confidence
0
10
20
30
40
50
1970 - 1979 1980 - 1989 1990 - 1999 2000 - 2004 2005 - 2009 2010 - 2015
By Country Groups
Advanced Emerging Low Income Small Developing States
IMF, Fiscal Affairs Department

10.
Overlap between Fiscal, Banking, and Currency Crises
9
Fiscal crises
335
Currency crises
122
Banking crises
82
23
14
47
28
IMF, Fiscal Affairs Department
Source: Laeven and Valencia (2012), authors’ calculations.

11.
III. Empirical Analysis
10IMF, Fiscal Affairs Department

12.
Event Study
 Empirical regularities during the 11-year time window around fiscal
crises (Gourinchas & Obstfeld 2012; Catao & Milesi-Ferretti 2014)
 Does not necessarily imply causality
𝑦𝑖𝑡 = 𝛼𝑖 + ෍
𝑗=−5
5
𝛽𝑗 𝐹𝑡+𝑗 + 𝜀𝑖𝑡
where:
y = variable of interest (e.g., real GDP p.c. growth, public debt-to-GDP ratio)
F = dummy variables taking the value of 1 in period t+j, where t is the fiscal crisis year
 = conditional effect of a crisis over the event window relative to tranquil times
11IMF, Fiscal Affairs Department

13.
What Happens Around Fiscal Crises?
12
-10.0
-5.0
0.0
5.0
10.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
Public debt (% of GDP)
-6.0
-4.0
-2.0
0.0
2.0
4.0
6.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
Real expenditure growth (%)
-10.0
-5.0
0.0
5.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
Primary balance (% of GDP)
-3.0
-2.0
-1.0
0.0
1.0
2.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
Real GDP p.c. growth (%)
IMF, Fiscal Affairs Department
Note: Reported are coefficient estimates with 95% confidence intervals.

14.
Fiscal Crises and Growth: By Country Groups
13
-10.0
-8.0
-6.0
-4.0
-2.0
0.0
2.0
4.0
6.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
AM
-5.0
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
EM
-2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
2.5
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
LIDC
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
SDS
IMF, Fiscal Affairs Department
Note: Reported are coefficient estimates with 95% confidence intervals.
AM = advanced markets, EM = emerging markets, LIDC = low-income and developing countries, SDS = small developing states.

15.
Fiscal Crises and Growth: By Criteria
14IMF, Fiscal Affairs Department
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
2.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
1. Credit event
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
2. Official financing
-12.0
-10.0
-8.0
-6.0
-4.0
-2.0
0.0
2.0
4.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
3. Implicit default
Note: Reported are coefficient estimates with 95% confidence intervals.
-6.0
-5.0
-4.0
-3.0
-2.0
-1.0
0.0
1.0
2.0
3.0
4.0
5.0
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
4. Market confidence

16.
Event Study – Twin Crises
 Empirical specification:
𝑦𝑖𝑡 = 𝛼𝑖 + ෍
𝑗=−5
5
𝛽𝑗 𝐹𝑡+𝑗 + ෍
𝑗=−5
5
𝛾𝑗 𝐹𝐵𝑡+𝑗 + ෍
𝑗=−5
5
𝛿𝑗 𝐹𝐶𝑡+𝑗 + 𝜀𝑖𝑡
where:
y = macro variable of interest (e.g., growth, public debt ratio)
F = fiscal crisis dummy
FB = twin fiscal-banking crisis dummy
FC = twin fiscal-currency crisis dummy
 = effect of a fiscal crisis
 = additional effect of a twin fiscal-banking crisis
 = additional effect of a twin fiscal-currency crisis
15IMF, Fiscal Affairs Department

17.
Twin Fiscal-Banking Crises
16IMF, Fiscal Affairs Department
-30
-20
-10
0
10
20
30
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
-6
-5
-4
-3
-2
-1
0
1
2
3
4
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
Real GDP p.c. growth (%)
Note: Reported are coefficient estimates with 95% confidence intervals.
Public debt (% of GDP)

18.
Twin Fiscal-Currency Crises
17IMF, Fiscal Affairs Department
-30
-20
-10
0
10
20
30
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
-6
-5
-4
-3
-2
-1
0
1
2
3
4
t-5 t-4 t-3 t-2 t-1 t t+1 t+2 t+3 t+4 t+5
Public debt (% of GDP)
Note: Reported are coefficient estimates with 95% confidence intervals.
Real GDP p.c. growth (%)

19.
IV. Concluding Remarks
18IMF, Fiscal Affairs Department

20.
Conclusions
 A new IMF database on fiscal crises
 Updated annually
(https://www.imf.org/en/Publications/WP/Issues/2017/04/03/Fiscal-Crises-44795)
 Large sample (including LIDCs, over 4½ decades) …
 … but some challenges remain (variability of quality/available data across criteria,
time, and country groups)
 Helps understanding implications of fiscal crises
 Crises are disruptive: economic recession, indebtedness rises …
 … but effects can vary across country groups and crisis criteria
 Twin crises can be even more disruptive
 Can be used to conduct further research
 e.g., early warning exercises, causal effects of fiscal crises
19IMF, Fiscal Affairs Department

21.
Thank you!
20IMF, Fiscal Affairs Department

22.
Background Slides
21IMF, Fiscal Affairs Department

23.
Types of Fiscal Crises per Country Groups
IMF, Fiscal Affairs Department 22
AM EM LIDC SDS
Credit event 0 85 141 71
Official financing 11 40 29 6
Implicit default 13 18 9 7
Market confidence 7 25 4 3
Source: Authors’ calculations.

24.
Percent of Cases with Negative Real GDP Per
Capita Growth
IMF, Fiscal Affairs Department 23
Source: Authors’ calculations.
Total crisis
episodes
t-3 t-2 t-1 t t+1 t+2
Total 436 24.8 25.5 28.2 40.6 33.9 29.8
AM 25 12.0 28.0 28.0 48.0 60.0 48.0
EM 154 22.1 20.8 28.6 49.4 40.3 29.2
LIDC 171 29.8 28.1 28.1 33.9 29.2 29.8
SDS 86 23.3 27.9 27.9 36.0 24.4 25.6