Transcript Remittances and capital markets

Remittances and Financial
Development
Experts Meeting
Latin American Economic Outlook 2010
Rolando Avendaño Sebastian Nieto Parra
OECD Development Centre
 19 May 2009 – Paris 
Outline: Remittances and Financial Development
Core sections:
–
–
–
Remittances and Access to Banking
Remittances and Capital Markets
Conclusions and Policy Implications
OECD Development Centre
Outline: Remittances and Financial Development
1
Remittances and Access to Banking
- Costs (transaction costs, currency costs,…). Latin America vs other regions
-Money sending institutions (banking and other sending platforms): empirical facts and
strategies
- New instruments: Mobile banking (Distribution networks and Transactions costs)
- Remittances and banking development
- Box: Historical comparison for banking development (Latin European countries 1870-1913)
OECD Development Centre
Outline: Remittances and Financial Development
2
Remittances and capital markets
- International experience: Securitizations/ Diaspora bonds
- Remittances and capital markets: the case of ratings (risk indicators : debt over exports
and volatility of external flows)
Current economic crisis and remittances
- Box: external capital market shocks and remittances (historical experience: Baring
crises and remittances)
OECD Development Centre
1
Remittances and Access to Banking
2
Remittances and Capital Markets
OECD Development Centre
0
Source: World Bank 2008.
OECD Development Centre
LAC
A rgentina
1995
P eru
Mex ic o
V enez uela, R B
Uruguay
90
E c uador
J amaic a
Dom. R epublic
C olombia
G uatemala
10
B olivia
20
Nic aragua
20
E l S alvador
40
C os ta R ic a
60
B raz il
120
B eliz e
2007
G renada
D omes tic C redit P rov ided by B anking
S ec tor
C hile
80
% G DP
100
P anama
P eru
E c uador
1995
V enez uela, R B
Uruguay
A rgentina
Mex ic o
140
G uatemala
C olombia
E l S alvador
C os ta R ic a
Dom. R epublic
B olivia
LAC
J amaic a
B eliz e
Nic aragua
P anama
C hile
G renada
B raz il
% G DP
Differences in bank concentration
across the region
D omes tic C redit to P riv ate S ec tor
100
2007
80
70
60
50
40
30
0
Significant differences in bank concentration
across the region
Source: Thorsten Beck, Asli Demirgüç-Kunt and Ross Levine, (2000), "A New Database on Financial Development and
Structure," World Bank Economic Review 14, 597-605. Updated November 2008.
Note: Bank concentration defiined as a measure of the degree of concentration in the banking industry, calculated as the fraction of assets held by the
three largest commercial banks in each country, averaged over the period 1995-99.
OECD Development Centre
Remittances and Banking Access
A broader range of services:
•Repatriation insurance
•Free money-sending
•Residence Certificates
•Low cost travel plans
•Legal advice
•Home Loans
Strategies:
Bancarización inducida: encouraging bank
account creation in receiving country.
•Remesas cuenta a cuenta: From 4% to more
than 20% in the last four years.
•Remesas finalistas: Specific destination for
remittances (housing, education, etc.)
Two main obstacles: Distribution / Costs
OECD Development Centre
Why mobile technology for remittances?
Distribution and Coverage across regions
100
Mobile technology has a potential in regions with low density of banks and cash machines.
Source: Wireless Intelligence (2008), Beck, Demirguc-Kunt and Martinez Peria (2005)
OECD Development Centre
OECD
0
OECD
0
Sub-Saharan Africa
20
North Africa
20
Central America
40
South America
40
Sub-Saharan Africa
60
North Africa
60
80
Central America
%
80
Caribbean
%
100
120
South America
120
Geographic bank
branch
Geographic
branch
Geographic
ATMmachines
Geographic cash
Mobilephone
Phone
Mobile
140
Demographic
branch
Demographic bank
branch
Demographic
ATM
Demographic cash
machines
Mobile phone
Caribbean
140
Costs and actors in the remittances business model
Banks
30
Fee for sending USD 200 (%)
Fee for sending USD 200 (%)
Transaction Costs in Banks and Money Transfer Operators
25
20
15
10
5
0
0
1
2
3
4
5
6
Number of banks per corridor
Money Transfer Operators
30
25
20
15
10
5
0
0
5
10
15
20
25
Number of Money Transfer Operators per corridor
Mobile technology is more likely in countries with a weak presence of Money
Transfer Operators; under any banking market structure
Source: OECD Development Centre (2009), based on World Bank (2008)
OECD Development Centre
Different corridors, different mechanisms:
The role of the sending country
Most promising corridors - Japan, France, Canada, The Netherlands
Latin America and Caribbean
Number of Money
Transfer Operators
0
Transaction cost (%) for
USD 200
19.71
Destination
Brazil
Origin
Japan
Dominican Republic
Haiti
Haiti
Jamaica
Peru
The Netherlands
Canada
France
Canada
Japan
2
2
3
4
6
17.14
15.14
11.38
14.02
19.92
Surinam
AVERAGE FOR CORRIDORS WITH
LOW MTO PRESENCE IN LAC
AVERAGE FOR CORRIDORS WITH
HIGH MTO PRESENCE IN LAC
The Netherlands
3
11.23
2.8
15.5
11.6
6.6
US, Spain, UK
Source: World Bank (2008)
OECD Development Centre
From mobile payments to mobile banking
In LAC mobile banking is more about households than firm access
1
• While 9 / 10 people in LAC have a mobile
phone line, 70-80 % have no banking
access.
Share of households with bank account
0.8
Share of small firms with banks loans
0.6
• Mobile banking is feasible with a large
amount of small deposits
0.4
0.2
OECD
Sub-Saharan Africa
North Africa
Central America
South America
Caribbean
0
• Mobile phones are enabling to trace a
history to access loans when no
collateral
• In credit decisions face-to-face contact
remains essential
Source: Beck, Demirguc-Kunt and Martinez Peria (2005)
Mobile banking operations are largely available in LAC as an additional distribution channel
OECD Development Centre
Policy recommendations for mobile payments
Outlook: Will LAC’s bank-lead model fully integrate mobile advantages?
1. International remittances through mobile technology are lead by..
•
..banks with cash out through branches: Spain  Ecuador, US  Colombia
•
..mobile operators for population without bank access: US  El Salvador
2. While mobile remittances are estimated to grow, LAC will only benefit
•
if banks and money transfer operators (financially compliant) make full use of mobile
operators (distributional networks) or
•
if mobile operators are subject to adequate financial regulation and liquidity is available at
telcos agents’ shops
OECD Development Centre
1
Remittances and Access to Banking
2
Remittances and Capital Markets
OECD Development Centre
S &P R ating (L aunc h)
Source: OECD Development Centre (2009), based
on Dealogic
OECD
Development
Centre
P eru Mar06
B raz il F eb06
B raz il J an06
S &P s overeign rating
B raz il Mar09
B raz il Mar08
P eru A ug07
P eru A ug07
B raz il S ep06
C C C5-
P eru Nov05
200
B raz il A ug03
B + 10
B raz il A ug03
400
B raz il J un03
800
B raz il J un02
1000
E l S alv. Mar02
1400
E l S alv. Mar02
Total Yield (bp) at Launch
J amaic a Mar06
Y ield F uture F low S ec uritiz ation (DP R )
Y ield s overeign
B raz il Mar08
B raz il Mar09
P eru A ug07
P eru A ug07
B raz il S ep06
P eru Mar06
J amaic a Mar06
B raz il J an06
B raz il F eb06
B raz il A ug03
P eru Nov05
B raz il A ug03
B raz il J un02
B raz il J un03
E l S alv. Mar02
E l S alv. Mar02
Future Flow Securitization
(Diversified Payment Rights)
Payments on the receivables do not enter before obligations are met
Rating at Launch (Sovereign vs DPRs)
25
1200
A A 20
-
B B 15
B
600
0
D0
Diaspora Bonds: A mechanism
to tap the migrants’ wealth
Households’ income matter but other factors are crucial (patriotism ,… )
500000
50
45
400000
40
35
300000
200000
30
25
20
100000
15
10
5
0
Aruba
B ermuda
Argentina
V enez uela
P anama
B ahamas
C uba
Uruguay
C hile
Ant&B arb.
T rinidad&T ob.
B olivia
C os ta R ica
Dominica
B arbados
G uyana
P uerto R ico
P araguay
B eliz e
J amaica
C olombia
P eru
G renada
B raz il
Nicaragua
Haiti
Dominican R epublic
E cuador
Honduras
G uatemala
S alvador
Mexico
Is rael
S outh Africa
India
Lebanon
S ri Lanka
0
T ota l Ma na g e m e nt a nd P rofre s s iona ls (% of tota l occupa tions ; rhs )
T ota l Ma na g e m e nt a nd P rofe s s iona ls (N b, lhs )
Source: OECD Development Centre (2009), based on OECD stats
OECD Development Centre
Are Remittances
relevant for Rating Agencies?
Motivation
• Remittances flows are an important source of financing for developing countries.
• They act as an instrument to reduce financial vulnerabilities  Lessen the probability of
current account reversals and credit risk.
• World Bank: « Remittances can improve a country’s creditworthiness and enhance its
access to international capital markets.»
• We study:
(i) Role of workers’ remittances in the estimation of sovereign ratings, following the
models Credit Rating Agencies (CRAs).
– Solvency ratio (debt/exports)
– Volatility of external flows
(ii) Shadow ratings for unrated countries (some with high remittances levels)
OECD Development Centre
Why are we interested in ratings?
Linked with financial markets (development, spreads) and ceiling rating
Sub-Sovereign (% Sovereign and Ceiling)
70
60
50
40
30
20
10
0
rated higher
s ame rate
C ountry C eiling rate
Source: Authors based on World Bank and OECD data
rated lower
S overeign rate
Source: OECD Development Centre based on Moody’s, 2009
OECD Development Centre
Period
Cantor and Packer
1996
1995
•
Rowland and Torres
2004
1987-2001
Sutton
2005
2004
•
Mora
2006
1986-2001
•
•
Our model
2009
1993-2006
•
•
Inflation
Fiscal balance
•
•
•
•
•
•
•
External Debt/GDP
External Debt / Exports
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
OECD Development Centre
•
•
•
•
•
•
•
•
•
Dummy European Union
Spread (lag)
•
Dummy default on bonds and bak debt
•
Corrruption Index
Indicator of Economic Development
Default variable (diff. for each model)
Default History
EMBI Dynamic
Volatility External Flows
Ratios short-term bank/Total claims
Openness ((Exports + Imports)/GDP)
Current Account Balance/GDP
Reserves/GDP
Reserves
Dependent
Variable
Debt Service/GDP
External balance
Fiscal Balance/GDP
GDP growth
•
Income per capita
Spread
Institutional rating
Year
Sovereign Rating
Traditional Determinants of Sovereign Ratings
Solvency Ratio (Debt/Exports) excluding remittances is a key variable in traditional models
Independent Variables
•
•
Traditional channel:
Solvency Ratio and remittances
Solvency Ratio (Debt/Exports) including and excluding remittances
2003
2004
2005
2006
2004
2005
2006
2002
2003
2001
2000
1999
1997
1996
1995
1994
tdoverx
tdoverx_wr
tdoverx_wr
El Salvador
250
Mexico
250
1993
2006
2005
2004
2003
2002
0
2001
0
2000
50
1999
100
1998
100
1997
200
1996
150
1995
300
1994
200
1993
400
tdoverx
Colombia
250
1998
Brazil
500
200
200
150
150
100
100
50
tdoverx
tdoverx
tdoverx_wr
OECD Development Centre
2002
2001
2000
1999
1998
1997
1996
1995
1994
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
0
0
1993
50
tdoverx_wr
Remittances are less volatile than other flows:
A source for financial stability
Volatility of external flows 92-07 (% of GDP)
0.3
0.25
0.2
0.15
0.1
0.05
0
R emittanc es
F DI Inward
P ortfolio Inward
O DA
Source: OECD Development Centre, based on IFS, 2009.
OECD Development Centre
E x ports
Capital flows volatility is reduced by remittances:
an externality for sovereign ratings?
Impact of Remittances on Volatility of External
Flows
120
Average 1992-2006
Average 1992-2006
100
0.005
80
% Change
%
N
N
i 1
i j
Varexternal _ flows  ,t   wi2,t  Var ( X i ,t )  2 wi ,t w j ,t Cov( X i ,t , X j ,t )
Source: OECD Development Centre, based on IFS, 2009.
OECD Development Centre
Chile
Venezuela, RB
Paraguay
Mexico
Nicaragua
Peru
Bolivia
Colombia
Honduras
Ecuador
-20
Dominican Republic
Colombia
Brazil
Guatemala
El Salvador
Peru
Bolivia
Chile
Ecuador
Uruguay
Mexico
Argentina
Costa Rica
0
Honduras
0
Dominican Republic
20
Venezuela, RB
0.001
Paraguay
40
Nicaragua
0.002
Guatemala
60
El Salvador
Volat External Flows (without remitt.)
0.003
Argentina
Average Volat External Flows
Uruguay
0.004
Brazil
0.006
Percent Change on Volatility Excluding
Remittances
Costa Rica
0.007
Our model and Counterfactual Analysis
Predicted model:
Rating i ,t   0   1controlvariables   2 SolvencyRatio _ wri ,t   3VolatExtFlows _ wri ,t   t  i   i ,t
Counterfactual model:
Rating i ,t   0   1controlvariables   2 SolvencyRatioi ,t   3VolatExtFlowsi ,t   t  i   i ,t
Three estimations:
•Observed
•Predicted
•Counterfactual
OECD Development Centre
ˆ
Our model and Counterfactual Analysis
Results: Standard and Poor’s estimations
Observed, Predicted, Counterfactual
Counterfactual – Predicted
(Rating gain: 1=one notch)
0.8
A 16
B B B14
+
0.6
B B B12
B B 10
0.4
B+ 8
B-
0.2
6
Source: OECD Development Centre, based on Avendano, Gaillard and Nieto-Parra (2009)
OECD Development Centre
E l S alv.
G uatemala
Dom. R ep.
C ounterfac tual
E c uador
Mex ic o
C hile
E l S alv.
G uatemala
Dom. R ep.
E c uador
P redic ted
0
A rgentina
O bs erved
A rgentina
Mex ic o
C hile
CCC4
Our model and Counterfactual Analysis
Results: Moody’s estimations
Counterfactual – Predicted
Observed, Predicted, Counterfactual
(Rating gain: 1=one notch)
A 16
6
0.2
CCC4
0.1
OECD Development Centre
E l S alv.
Nic aragua
Honduras
G uatemala
Mex ic o
Source: OECD Development Centre, based on Avendano, Gaillard and Nieto-Parra (2009)
E c uador
C ounterfac tual
Dom. R ep.
P redic ted
A rgentina
O bs erved
0
C hile
C hile
B-
E l S alv.
0.3
Nic aragua
B+ 8
Honduras
0.4
G uatemala
B B 10
E c uador
0.5
Dom. R ep.
B B B12-
A rgentina
0.6
Mex ic o
B B B14+
Rating agencies do not have an unique model
Model for High-Remittance Receptors: Defining a threshold
Significance and sign of the key variables (remittances matter)
Rating i ,t   0  1control var.   2 Solvencyi ,t  3Vol. flowsi ,t   4Threshold  Solvencyi ,t
 5Threshold Vol.Flowsi ,t   6Threshold i ,t   t  i   i ,t
Threshold variable
(remittances/GDP > 5%)
Solvency ratio
(Debt over exports and remittances)
4
0.005
3
0
Interactive of Solvency ratio
(Threshold*Solvency ratio)
0.015
0.01
0.005
0
-0.005
2
-0.005
1
0
-0.01
-0.015
-0.02
-0.025
-0.03
-0.01
-1
-0.015
-2
-0.02
-3
1
S&P’s
Low er bound
2
Moody’s
Upper bound
Fitch 3
coefficient
S&P's
Moody's
Low er bound
Upper bound
Fitch
coefficient
Source: OECD Development Centre, based on Avendano, Gaillard and Nieto-Parra (2009)
OECD Development Centre
S&P's
Low er bound
Moody's
Upper bound
Fitch
coefficient
Rating agencies do not have an unique model
Model for High-Remittance Receptors: Defining a threshold
Indirect effect (premium): Reduce negative impact of volatility ext. flows
Interactive of Volatility External Flows
(Threshold*Vol.External Flows)
Volatility of External Flows
(% of GDP)
100
800
0
600
-100
400
-200
200
-300
0
-400
-200
-500
S&P's
Low er bound
Moody's
Upper bound
S&P's
Fitch
coefficient
Low er bound
Source: OECD Development Centre, based on Avendano, Gaillard and Nieto-Parra (2009)
OECD Development Centre
Moody's
Fitch
Upper bound
coefficient
Shadow ratings for Selected Countries
A1
17
A2
15
A3
Baa1
13
Baa2
Baa3
11
Ba1
Ba29
Ba3
B17
B2
B35
Caa1
Caa23
Caa3
Observed
Predicted
Nicaragua
Observed
Predicted
-1
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
1
Dominican Repubic
A1
17
A2
15
A3
Baa1
13
Baa2
Baa3
11
Ba1
Ba29
Ba3
B17
B2
B35
Caa1
Caa23
Caa3
Observed
Predicted
1
-1
A+
17
A
15
ABBB+
13
BBB
BBB11
BB+
BB
9
BBB+
7
B
B5
CCC+
3
CCC
CCC1
-1
Source: Authors calculation. OECD Development Centre 2009.
A+
17
A
15
ABBB+
13
BBB
BBB11
BB+
BB
9
BB7
B+
B
5
BCCC+
3
CCC
CCC1
-1
Guatemala
Observed
Predicted
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Predicted
Chile
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
17
A+
A
15
ABBB+
13
BBB
BBB11
BB+
BB
9
BBB+
7
B
B5
CCC+
3
CCC
CCC1
-1
Observed
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
A+
17
A
15
ABBB+
13
BBB
BBB11
BB+
BB
9
BBB+
7
B
B5
CCC+
3
CCC
CCC1
-1
Honduras
Fitch
Honduras
Observed
Predicted
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
Moody’s
S&P
Conclusions and Policy Lessons
• This paper also provides “shadow” ratings for countries which are not rated by the three main
CRAs, in particular some Central American and Caribbean countries, where relative remittances
flows are higher.
Information and Portfolio Allocation. Public-Private Partnership
• The impact of including remittances in the solvency ratio (debt over exports) and the volatility of
external flows is relatively modest.
Remittance flows cannot, as a whole, enable low and medium income countries
to improve their creditworthiness significantly. Remittances should be
accompanied by policy measures related to the solvency of the country and the
stability of external flows.
OECD Development Centre
Conclusions and Policy Lessons
• Specification for high-remittance receptors: The direct effect on ratings (the dummy variable
remittances over GDP is not significant) is reduced. There may be an indirect, positive impact on
ratings through a premium (captured with the interactive dummy and the variable volatility of
external flows).
Results consistent with previous research (e.g. Roubini and Manasse, 2005) showing that
there is not a unique model to rate countries and not all variables have the same impact for
a sovereign rating.
OECD Development Centre
Role of remittances in rating agencies’ methodology
•
“ (…) Fitch will nonetheless take account of the volatility and potential vulnerability of such receipts – such as
remittances – to domestic and external shocks. "
Fitch Ratings, “Sovereign Rating Methodology", October 12, 2007
•
“Remittances, equivalent to more than a tenth of domestic output and a major driver of consumption, are expected to
drop 5 to 10 percent this year as a slowing global economy puts pressure on wages of Filipino workers abroad.”
“Moody's: Slowing remittances hurt RP”, Manila Bulletin, February 14, 2009
•
“The revision of the outlook on Mexico to negative reflects our assessment of the deterioration in its fiscal and external
positions.....Our projections are for Mexico's external debt (net of liquid assets) to reach more than 40% of current
account receipts by 2010 versus the 'BBB' median of 28%". Standard and Poor’s, 12 May 2009.
•
Standard and Poor's lowered El Salvador's credit rating to "BB" from "BB+": "The weak performance in 2009 is due to
falling consumption, investments, and exports as a result of a significant pass-through from the global recession.
Remittances from the United States fell by 8 percent in the first two months of the year..."
OECD Development Centre
Do Remittances Prevent Financial Disturbances?
Lessons from the Gold Standard Period
Financial crises and remittances before Word War I
Source: Based on data from Esteves and Khoudour-Castéras (2010).
OECD Development Centre
Remittances and Financial
Development
Thank you!
Rolando Avendaño
Sebastian Nieto Parra
OECD Development Centre
 19 May 2009 – Paris 