SlideShare a Scribd company logo
1 of 29
1
Poverty Sensitive Scorecards to
prioritize grants and lending
Maximo Torero
M.Torero@cgiar.org
International Food Policy Research Institute
Manuel Hernandez
M.A.Hernandez@cgiar.org
International Food Policy Research Institute
What we know
• Access to credit and grants helps improving
economic opportunities
• Micro and small enterprises play a central role
in economic development
• Agriculture is essential in reducing poverty
2
What we know
• In underdeveloped markets lending risks are
high because:
– contracts are difficult to enforce
– problem of adverse selection (wrong choices
when the type of the borrower is unknown)
• In developed financial markets a system of
score cards are used to mitigate the problem
of adverse selection by identifying
creditworthiness
3
What is missing
• Riskiness of a borrower or a grantee is not the
only criterion in case of development lending
• If the objective is development the menu of
projects has to be assessed also in terms of
their potential for reducing poverty
• Thus, for the optimal use of funds there might
be a possible trade off between profitability
and poverty impacts
4
What we have done
• We have implemented a two dimensional
score card:
–risk score of the grantees
–poverty score card
• We combine both score cards so that project
selection will not only focus on targeting the
poor but also in assuring sustainability
5
Universe of projects
6
Pass threshold of risk score
7
Pass threshold of poverty score
and risk score
8
Building the risk score card
• Develop a risk scoring algorithm that is suitable for
lending to the SMEs or small Rural projects
• Generate a working database on risk scores based on
data to be collected in the future to improve model
prediction
• Operationalize the risk scoring mechanism through a
simple implementable program in open source
software (spreadsheets)
9
Building the risk score card - Steps
Step1
• Information on similar grants and specific characteristics is
collected
• A proxy of performance (default or more specifically
sustainability for grants) will be used
Step 2
• Estimation of performance as a function of characteristics
• Traditionally is done parametrically, we implemented a semi
parametric model following Klein and Spady (1993) to
capture non linearities
Step 3
• Use the estimated parameter to prioritize projects that apply
and select the ones that assure more sustainability
10
Building the risk score card - Steps
Step1
Yi = past performance of grant (sustainability)
Step 2
Yi = 0 + X i + I Traditional method
Yi = g( Xi ) + i Semi-parametric method
We estimate and g where g is an unknown function
Step 3
Use predict and g to estimate predicted Yi to prioritize projects
that apply and select the ones that assure more sustainability
11
Examples of characteristics
12
Personal characteristics Financial and Entrepreneurial characteristics
Age of the beneficiary Experience in the enterprise for which grant/loan is
being sought
Gender of the beneficiary Experience as an entrepreneur in any enterprise
Marital status of the beneficiary Degree of specialization in the intended enterprise
Education level of the entrepreneur Number of business organizations that entrepreneur is a
part of
Geographical location (nearness to economic centers) Number of certifications and vocational qualifications
that the entrepreneur holds
Borrowing situation (level of indebtedness) Has there been a history of default on the borrowings by
the beneficiary
Asset details of the beneficiary
(Examples: Housing/Bank Balance/Other durable assets)
What has been the amount on which there has been
default
Has there been any criminal proceeding against the
beneficiary in the past
Frequency of borrowing in the last five years from any
source
Does the borrower own land? Amount of repayment in last five years from any source
Has the beneficiary borrowed from informal sources
Is this project form the main source of income
Has the beneficiary worked with the government or has
political affiliation with the ruling
Variable Coeff
Edad 0.0082
Sexo 0.1148
Educacion superior universitaria -0.1358
Casado -0.1115
Anos de la empresa -0.0048
Tamaño empresa (nro. trabajadores) -0.0187
Si vivienda es propia -0.2837
Monto prestamos (en miles soles) 0.0116
Tasa de interes (en %) 0.0016
Plazo del prestamo (en meses) 0.0465
Constante -0.0560
% total de aciertos 47.8%
Variable Coeff
Edad 1.0000
Sexo 0.1361
Educacion superior universitaria -0.2497
Casado 0.1867
Anos de la empresa 0.0004
Tamaño empresa (nro. trabajadores) 0.0158
Si vivienda es propia 0.0222
Monto prestamos (en miles soles) 0.0090
Tasa de interes (en %) 0.1578
Plazo del prestamo (en meses) 0.0324
% total de aciertos 71.8%
Building the risk score card
Traditional Method: Probit Semi Parametric model: Single
Index
The single index model have better performance in terms of
prediction than the Probit model.
Semi-parametric Model: Single
Index Model
Probability of default: P(y=1|x) = E(y|x) = g(x΄ )
estimated= argmaxβ
where g is aproximated for each i as:
βestimado
Edad promedio beneficiarios 1.000
% Hombres 0.136
% Educación sec. o sup. -0.250
% Casados 0.187
Antigüedad de asociación 0.000
Tamaño de asociación 0.016
Si posee activos 0.022
Monto solicitado/deuda 0.009
Tasa de interés 0.158
Plazo de deuda 0.032
Kernel type (k): second order Gaussian
To estimate the risk score for each project i
with characteristics Xi we use these
βestimated (“weighting factors”) .
First Stage: Results of Risk scoring
0%
20%
40%
60%
80%
100%
1 2 3 4 5 6 7 8 9 1011 12 13 14 15 1617 18 19 20 21 22 2324 25 26 27 28 2930 31 32 33 34 3536 37 38 39
NiveldeRiesgo
Score riesgo Riesgo maximo
67%
Project with the lowest risk
(WWF Guatemala)
Move to
second
stage
• 24 projects passed the minimum risk threshold and they
requested a total value of US$ 4,469,400.
In parenthesis is the number of projects that were submitted and completed their application.
Country # Project Amount (US$)
El Salvador 3 (4) 586,713
Guatemala 6 (11) 968,946
Honduras 10 (14) 1,958,293
Nicaragua 5 (10) 955,448
Total 24 (39) 4,469,400
First Stage: Results of Risk scoring
Building the poverty score card
• Keep projects that meet the threshold of the
risk score card (we initially use 67%)
• Project selection will involve assessing the
poverty reducing potential of the projects
• The projects with the highest poverty score
card will be the ones prioritized
17
Building the poverty score card
18
Information Criteria
Geographical indicators
1. Project located in high poverty
areas
If it falls in an extreme poor or district with poverty
level over 50%.
2. Access to markets If it falls in a district which has a very low accessibility
to main road, i.e. equal or more of an hour to the main
road.
Employment indicators
1. Labor intensive project Number of new employment generated by the project
2.Low skill labor intensive
project
% of low skill labor in the project
3.Potential female employment
intensity of the enterprise
% of female labor in the project
Spillover indicators
1.Effect on supply chain
participants
Number of potential beneficiaries by the project. The
ration will be calculated based on dollars invested.
2.Other spillover effects (for
example provision of some public
good)
Number of potential beneficiaries by the project. The
ration will be calculated based on dollars invested.
Pobreza extrema en
porcentaje:
Baja < 13%
Media 13-31%
Alta > 31
Limite distrital incluido
Acceso al mercado (50hab) en
horas:
Alto <1hora
Medio 1-3
Bajo >3
Departamento de Chiquimula
Municipios Ipala y
Concepcion las minas
Weighting the criteria
• Step 1: Collect for all projects that apply the information
for all the variables
• Step 2: Normalize/Adjust the data by subtracting the
mean to it
• Step 3: Calculate the covariance matrix
• Step 4: Calculate the eigenvectors and eigenvalues of the
covariance matrix
• Step 5: Choose the first and second principal components
(PCPov1 and PCPov2)
• Step 6: Use PCPov 1 and PCPov 2 to rank the projects on
the poverty dimension.
23
24
- 10 proyectos (en Honduras) que aprobaron el corte mínimo de riesgo.
- Se dispone de 1 millón de dólares americanos.
Caracteristicas de los proyectos
Monto # empleos # empleos # empleos # empleos # total empleos no empleos # empleos # empleos Pobreza Acceso
(miles directos indirectos no femeninos de empleos calificados / femeninos / directos / indirectos /
US$) calificados total empleos total empleos inversion inversion
var 1 var 2 var 3 var 4 var 5 var 6 var 7
Proyecto 1 (La Ceiba, Atlantida) 100 100 200 100 150 300 0.33 0.50 1.00 2.00 0.191 6.325
Proyecto 2 (El Paraiso, Copan) 200 220 500 300 500 720 0.42 0.69 1.10 2.50 0.426 7.994
Proyecto 3 (Choluteca, Choluteca) 170 150 50 50 50 200 0.25 0.25 0.88 0.29 0.336 2.233
Proyecto 4 (Choluteca, Choluteca) 100 120 400 400 100 520 0.77 0.19 1.20 4.00 0.336 2.233
Proyecto 5 (El Corpus, Choluteca) 180 80 240 150 200 320 0.47 0.63 0.44 1.33 0.541 1.743
Proyecto 6 (Vado Ancho, Paraiso) 120 50 300 200 300 350 0.57 0.86 0.42 2.50 0.477 4.155
Proyecto 7 (Brus Laguna, Gracias a Dios) 300 400 800 1,000 200 1200 0.83 0.17 1.33 2.67 0.527 38.062
Proyecto 8 (Colomoncagua, Intibuca) 120 240 80 200 150 320 0.63 0.47 2.00 0.67 0.445 13.895
Proyecto 9 (San Antonio, Intibuca) 150 130 150 140 100 280 0.50 0.36 0.87 1.00 0.332 23.596
Proyecto 10 (Santa Cruz, Lempira) 250 300 300 400 500 600 0.67 0.83 1.20 1.20 0.496 11.707
Example of poverty scoring
25
Example of poverty scoring (cont.)
Paso 1: Calculo de matriz de covarianzas
Variables Var1 Var2 Var3 Var4 Var5 Var6 Var7
Var1 1 0.649 -0.187 0.293 0.500 0.457 0.674 Se observa una elevada correlacion entre numero total de
Var2 0.649 1 -0.169 0.395 0.522 0.498 0.518 empleos, proporcion de empleos no calificados sobre total
Var3 -0.187 -0.169 1 -0.331 -0.121 0.335 -0.374 de empleos, proporcion de empleos indirectos sobre total
Var4 0.293 0.395 -0.331 1 -0.071 -0.033 0.387 invertido y baja accesibilidad.
Var5 0.500 0.522 -0.121 -0.071 1 -0.023 0.000
Var6 0.457 0.498 0.335 -0.033 -0.023 1 0.256
Var7 0.674 0.518 -0.374 0.387 0.000 0.256 1
Paso 2: Estimacion de valores propios (eigenvalues) de cada componente principal (CP) y seleccion del primer componente
CP1 CP2 CP3 CP4 CP5 CP6 CP7
Eigenvalue 2.95 1.53 1.17 0.67 0.40 0.25 0.03
% Variacion explicada 42.08 21.80 16.74 9.51 5.75 3.63 0.47 El primer componente principal (CP1) explica el
Variacion acumulada 42.08 63.89 80.63 90.14 95.89 99.53 100.00 42% de la varianza observada.
Correlacion de cada componente principal con cada variable
CP1 CP2 CP3 CP4 CP5 CP6 CP7
Var1 0.893 0.150 0.079 -0.165 0.263 -0.264 -0.086 Importancia del numero total de empleos,
Var2 0.875 0.176 0.102 0.233 -0.244 0.270 -0.080 proporcion de empleos no calificados sobre total
Var3 -0.344 0.787 -0.178 0.275 0.369 0.138 -0.003 de empleos y baja accesibilidad en CP1.
Var4 0.498 -0.513 -0.335 0.595 0.096 -0.107 0.040
Var5 0.462 0.174 0.855 0.108 0.067 -0.005 0.092
Var6 0.461 0.714 -0.402 -0.042 -0.286 -0.165 0.069
Var7 0.768 -0.224 -0.343 -0.376 0.206 0.232 0.067
Paso 3: Ranking final de proyectos a partir del primer componente principal (CP1) identificado
Ranking Puntuacion Monto Acumulado
1. Proyecto 7 4.222 300 300 Se seleccionan los proyectos 7 (300,000 dólares), 4 (100,000 dólares), 8 (120,000 dólares),
2. Proyecto 4 1.075 100 400 10 (250,000 dólares) y 2 (200,000 dólares).
3. Proyecto 8 0.514 120 520
4. Proyecto 10 0.481 250 770 Monto total otorgado igual a 970,000 dolares.
5. Proyecto 2 0.024 200 970
6. Proyecto 9 -0.403 150
7. Proyecto 6 -0.825 120
8. Proyecto 5 -1.192 180
9. Proyecto 1 -1.688 100
10. Proyecto 3 -2.207 170
Second Stage: Results on poverty
dimension
• The first two principal components explain 52% of the variability
of the different variables.
• The first principal component (CP1) is highly correlated with the
number of new employment generated, and direct and indirect
employment generated per invested dollar
• The second principal component (CP2) has a high correlation
with the number of female employments, poverty rate, and low
accessibility to markets.
CP1 CP2 CP3 CP4 CP5 CP6 CP7
Eigenvalue 2.219 1.399 1.109 1.054 0.765 0.453 0.001
Variability (%) 31.70 19.98 15.84 15.06 10.94 6.48 0.01
Cumulative % 31.70 51.68 67.52 82.58 93.52 99.99 100.00
Eigenvalues (summary of information):
Eigenvectors (“weights” of the different indicators in the PC):
Indicator CP1 CP2
Number of employment 0.659 -0.023
% Non qualified labor 0.014 0.286
% Female employement -0.139 0.580
Direct employment per $ invested 0.460 -0.317
Indirect employment per $ invested 0.536 0.184
Poverty rate 0.163 0.370
Access to markets 0.141 0.557
Second Stage: Results on poverty
dimension
Selected Projects
-5
-4
-3
-2
-1
0
1
2
3
4
5
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7
CP2(19.98%)
CP1 (31.70 %)
Banhcafé (Honduras)
More poverty
alleviation
Lower poverty
alleviation
WWF (Guatemala)
Lemenigi (Honduras)
Las Segovias (Nicaragua)
COMUCAP (Honduras)
TechnoServe (Honduras)
Helvetas (Nicaragua)
Desarrollo Sostenible (El Salvador)
Aldea Global (Honduras)
Thanks!

More Related Content

Similar to IFPRI's Poverty Scorecard

project-6-bank-loan-case-study.pdf
project-6-bank-loan-case-study.pdfproject-6-bank-loan-case-study.pdf
project-6-bank-loan-case-study.pdfVaibhaviKhedekar1
 
Managing risk from top to bottom by @ericpesik
Managing risk from top to bottom by @ericpesikManaging risk from top to bottom by @ericpesik
Managing risk from top to bottom by @ericpesikEric Pesik
 
Assessment 3 - Project Report _Template (2).pdf
Assessment 3 - Project Report _Template (2).pdfAssessment 3 - Project Report _Template (2).pdf
Assessment 3 - Project Report _Template (2).pdfLankaniPerera
 
Credit Risk Evaluation Model
Credit Risk Evaluation ModelCredit Risk Evaluation Model
Credit Risk Evaluation ModelMihai Enescu
 
Learning our lessons: The effectiveness of alternative livelihood projects in...
Learning our lessons: The effectiveness of alternative livelihood projects in...Learning our lessons: The effectiveness of alternative livelihood projects in...
Learning our lessons: The effectiveness of alternative livelihood projects in...Fundsi88
 
Environmental and social impacts Across supply chains - LCA conference 4 Nov ...
Environmental and social impacts Across supply chains - LCA conference 4 Nov ...Environmental and social impacts Across supply chains - LCA conference 4 Nov ...
Environmental and social impacts Across supply chains - LCA conference 4 Nov ...Factor-X
 
Final Project (Real Estate 3400)1. Individual ProjectThe fi
Final Project (Real Estate 3400)1. Individual ProjectThe fiFinal Project (Real Estate 3400)1. Individual ProjectThe fi
Final Project (Real Estate 3400)1. Individual ProjectThe fiChereCheek752
 
Six cigma AJAL
Six cigma AJALSix cigma AJAL
Six cigma AJALAJAL A J
 
WF-IOT-2014, Seoul, Korea, 06 March 2014
WF-IOT-2014, Seoul, Korea, 06 March 2014WF-IOT-2014, Seoul, Korea, 06 March 2014
WF-IOT-2014, Seoul, Korea, 06 March 2014Charith Perera
 
ABN AMRO_SLP Consulting
ABN AMRO_SLP ConsultingABN AMRO_SLP Consulting
ABN AMRO_SLP ConsultingSem de Moel
 
BackLifeUp - Adapting to change
BackLifeUp - Adapting to changeBackLifeUp - Adapting to change
BackLifeUp - Adapting to changeGuillaume Lauzier
 
Pro bono economics
Pro bono economicsPro bono economics
Pro bono economicsSWF
 
Product development for sanitation microfinance
Product development for sanitation microfinanceProduct development for sanitation microfinance
Product development for sanitation microfinanceTrémolet Consulting
 
APICS DC 2016, Gary S Lynch, The Risk Project, LLC
APICS DC 2016, Gary S Lynch, The Risk Project, LLCAPICS DC 2016, Gary S Lynch, The Risk Project, LLC
APICS DC 2016, Gary S Lynch, The Risk Project, LLCThe Risk Project, LLC
 
Fund Your Innovation - SR&ED Tax Credits and other Government Grants
Fund Your Innovation - SR&ED Tax Credits and other Government GrantsFund Your Innovation - SR&ED Tax Credits and other Government Grants
Fund Your Innovation - SR&ED Tax Credits and other Government GrantsBoast Capital
 

Similar to IFPRI's Poverty Scorecard (20)

Cafe Swot Analysis
Cafe Swot AnalysisCafe Swot Analysis
Cafe Swot Analysis
 
project-6-bank-loan-case-study.pdf
project-6-bank-loan-case-study.pdfproject-6-bank-loan-case-study.pdf
project-6-bank-loan-case-study.pdf
 
Managing risk from top to bottom by @ericpesik
Managing risk from top to bottom by @ericpesikManaging risk from top to bottom by @ericpesik
Managing risk from top to bottom by @ericpesik
 
Assessment 3 - Project Report _Template (2).pdf
Assessment 3 - Project Report _Template (2).pdfAssessment 3 - Project Report _Template (2).pdf
Assessment 3 - Project Report _Template (2).pdf
 
Credit Risk Evaluation Model
Credit Risk Evaluation ModelCredit Risk Evaluation Model
Credit Risk Evaluation Model
 
Learning our lessons: The effectiveness of alternative livelihood projects in...
Learning our lessons: The effectiveness of alternative livelihood projects in...Learning our lessons: The effectiveness of alternative livelihood projects in...
Learning our lessons: The effectiveness of alternative livelihood projects in...
 
Environmental and social impacts Across supply chains - LCA conference 4 Nov ...
Environmental and social impacts Across supply chains - LCA conference 4 Nov ...Environmental and social impacts Across supply chains - LCA conference 4 Nov ...
Environmental and social impacts Across supply chains - LCA conference 4 Nov ...
 
Final Project (Real Estate 3400)1. Individual ProjectThe fi
Final Project (Real Estate 3400)1. Individual ProjectThe fiFinal Project (Real Estate 3400)1. Individual ProjectThe fi
Final Project (Real Estate 3400)1. Individual ProjectThe fi
 
Six cigma AJAL
Six cigma AJALSix cigma AJAL
Six cigma AJAL
 
WF-IOT-2014, Seoul, Korea, 06 March 2014
WF-IOT-2014, Seoul, Korea, 06 March 2014WF-IOT-2014, Seoul, Korea, 06 March 2014
WF-IOT-2014, Seoul, Korea, 06 March 2014
 
WM.pptx
WM.pptxWM.pptx
WM.pptx
 
ABN AMRO_SLP Consulting
ABN AMRO_SLP ConsultingABN AMRO_SLP Consulting
ABN AMRO_SLP Consulting
 
BackLifeUp - Adapting to change
BackLifeUp - Adapting to changeBackLifeUp - Adapting to change
BackLifeUp - Adapting to change
 
Dodd-Frank Section 1502: Compliance Costs and Externalites of Greater Informa...
Dodd-Frank Section 1502: Compliance Costs and Externalites of Greater Informa...Dodd-Frank Section 1502: Compliance Costs and Externalites of Greater Informa...
Dodd-Frank Section 1502: Compliance Costs and Externalites of Greater Informa...
 
Pro bono economics
Pro bono economicsPro bono economics
Pro bono economics
 
Product development for sanitation microfinance
Product development for sanitation microfinanceProduct development for sanitation microfinance
Product development for sanitation microfinance
 
APICS DC 2016, Gary S Lynch, The Risk Project, LLC
APICS DC 2016, Gary S Lynch, The Risk Project, LLCAPICS DC 2016, Gary S Lynch, The Risk Project, LLC
APICS DC 2016, Gary S Lynch, The Risk Project, LLC
 
Fund Your Innovation - SR&ED Tax Credits and other Government Grants
Fund Your Innovation - SR&ED Tax Credits and other Government GrantsFund Your Innovation - SR&ED Tax Credits and other Government Grants
Fund Your Innovation - SR&ED Tax Credits and other Government Grants
 
Pmp integration chapter 4
Pmp integration chapter 4Pmp integration chapter 4
Pmp integration chapter 4
 
Cdp2014 ebay
Cdp2014 ebayCdp2014 ebay
Cdp2014 ebay
 

Recently uploaded

How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonetsnaman860154
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitecturePixlogix Infotech
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationSafe Software
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsEnterprise Knowledge
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersThousandEyes
 
Azure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAzure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAndikSusilo4
 
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxFactors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxKatpro Technologies
 
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Alan Dix
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024Scott Keck-Warren
 
Pigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping ElbowsPigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping ElbowsPigging Solutions
 
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024BookNet Canada
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdfhans926745
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking MenDelhi Call girls
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreternaman860154
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure servicePooja Nehwal
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptxHampshireHUG
 
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...shyamraj55
 
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...HostedbyConfluent
 
How to Remove Document Management Hurdles with X-Docs?
How to Remove Document Management Hurdles with X-Docs?How to Remove Document Management Hurdles with X-Docs?
How to Remove Document Management Hurdles with X-Docs?XfilesPro
 

Recently uploaded (20)

How to convert PDF to text with Nanonets
How to convert PDF to text with NanonetsHow to convert PDF to text with Nanonets
How to convert PDF to text with Nanonets
 
Understanding the Laravel MVC Architecture
Understanding the Laravel MVC ArchitectureUnderstanding the Laravel MVC Architecture
Understanding the Laravel MVC Architecture
 
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time AutomationFrom Event to Action: Accelerate Your Decision Making with Real-Time Automation
From Event to Action: Accelerate Your Decision Making with Real-Time Automation
 
IAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI SolutionsIAC 2024 - IA Fast Track to Search Focused AI Solutions
IAC 2024 - IA Fast Track to Search Focused AI Solutions
 
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for PartnersEnhancing Worker Digital Experience: A Hands-on Workshop for Partners
Enhancing Worker Digital Experience: A Hands-on Workshop for Partners
 
Azure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & ApplicationAzure Monitor & Application Insight to monitor Infrastructure & Application
Azure Monitor & Application Insight to monitor Infrastructure & Application
 
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptxFactors to Consider When Choosing Accounts Payable Services Providers.pptx
Factors to Consider When Choosing Accounts Payable Services Providers.pptx
 
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...Swan(sea) Song – personal research during my six years at Swansea ... and bey...
Swan(sea) Song – personal research during my six years at Swansea ... and bey...
 
SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024SQL Database Design For Developers at php[tek] 2024
SQL Database Design For Developers at php[tek] 2024
 
Pigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping ElbowsPigging Solutions Piggable Sweeping Elbows
Pigging Solutions Piggable Sweeping Elbows
 
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
#StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
 
[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf[2024]Digital Global Overview Report 2024 Meltwater.pdf
[2024]Digital Global Overview Report 2024 Meltwater.pdf
 
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men08448380779 Call Girls In Greater Kailash - I Women Seeking Men
08448380779 Call Girls In Greater Kailash - I Women Seeking Men
 
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
Neo4j - How KGs are shaping the future of Generative AI at AWS Summit London ...
 
Presentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreterPresentation on how to chat with PDF using ChatGPT code interpreter
Presentation on how to chat with PDF using ChatGPT code interpreter
 
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure serviceWhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
WhatsApp 9892124323 ✓Call Girls In Kalyan ( Mumbai ) secure service
 
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
04-2024-HHUG-Sales-and-Marketing-Alignment.pptx
 
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
Automating Business Process via MuleSoft Composer | Bangalore MuleSoft Meetup...
 
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
Transforming Data Streams with Kafka Connect: An Introduction to Single Messa...
 
How to Remove Document Management Hurdles with X-Docs?
How to Remove Document Management Hurdles with X-Docs?How to Remove Document Management Hurdles with X-Docs?
How to Remove Document Management Hurdles with X-Docs?
 

IFPRI's Poverty Scorecard

  • 1. 1 Poverty Sensitive Scorecards to prioritize grants and lending Maximo Torero M.Torero@cgiar.org International Food Policy Research Institute Manuel Hernandez M.A.Hernandez@cgiar.org International Food Policy Research Institute
  • 2. What we know • Access to credit and grants helps improving economic opportunities • Micro and small enterprises play a central role in economic development • Agriculture is essential in reducing poverty 2
  • 3. What we know • In underdeveloped markets lending risks are high because: – contracts are difficult to enforce – problem of adverse selection (wrong choices when the type of the borrower is unknown) • In developed financial markets a system of score cards are used to mitigate the problem of adverse selection by identifying creditworthiness 3
  • 4. What is missing • Riskiness of a borrower or a grantee is not the only criterion in case of development lending • If the objective is development the menu of projects has to be assessed also in terms of their potential for reducing poverty • Thus, for the optimal use of funds there might be a possible trade off between profitability and poverty impacts 4
  • 5. What we have done • We have implemented a two dimensional score card: –risk score of the grantees –poverty score card • We combine both score cards so that project selection will not only focus on targeting the poor but also in assuring sustainability 5
  • 7. Pass threshold of risk score 7
  • 8. Pass threshold of poverty score and risk score 8
  • 9. Building the risk score card • Develop a risk scoring algorithm that is suitable for lending to the SMEs or small Rural projects • Generate a working database on risk scores based on data to be collected in the future to improve model prediction • Operationalize the risk scoring mechanism through a simple implementable program in open source software (spreadsheets) 9
  • 10. Building the risk score card - Steps Step1 • Information on similar grants and specific characteristics is collected • A proxy of performance (default or more specifically sustainability for grants) will be used Step 2 • Estimation of performance as a function of characteristics • Traditionally is done parametrically, we implemented a semi parametric model following Klein and Spady (1993) to capture non linearities Step 3 • Use the estimated parameter to prioritize projects that apply and select the ones that assure more sustainability 10
  • 11. Building the risk score card - Steps Step1 Yi = past performance of grant (sustainability) Step 2 Yi = 0 + X i + I Traditional method Yi = g( Xi ) + i Semi-parametric method We estimate and g where g is an unknown function Step 3 Use predict and g to estimate predicted Yi to prioritize projects that apply and select the ones that assure more sustainability 11
  • 12. Examples of characteristics 12 Personal characteristics Financial and Entrepreneurial characteristics Age of the beneficiary Experience in the enterprise for which grant/loan is being sought Gender of the beneficiary Experience as an entrepreneur in any enterprise Marital status of the beneficiary Degree of specialization in the intended enterprise Education level of the entrepreneur Number of business organizations that entrepreneur is a part of Geographical location (nearness to economic centers) Number of certifications and vocational qualifications that the entrepreneur holds Borrowing situation (level of indebtedness) Has there been a history of default on the borrowings by the beneficiary Asset details of the beneficiary (Examples: Housing/Bank Balance/Other durable assets) What has been the amount on which there has been default Has there been any criminal proceeding against the beneficiary in the past Frequency of borrowing in the last five years from any source Does the borrower own land? Amount of repayment in last five years from any source Has the beneficiary borrowed from informal sources Is this project form the main source of income Has the beneficiary worked with the government or has political affiliation with the ruling
  • 13. Variable Coeff Edad 0.0082 Sexo 0.1148 Educacion superior universitaria -0.1358 Casado -0.1115 Anos de la empresa -0.0048 Tamaño empresa (nro. trabajadores) -0.0187 Si vivienda es propia -0.2837 Monto prestamos (en miles soles) 0.0116 Tasa de interes (en %) 0.0016 Plazo del prestamo (en meses) 0.0465 Constante -0.0560 % total de aciertos 47.8% Variable Coeff Edad 1.0000 Sexo 0.1361 Educacion superior universitaria -0.2497 Casado 0.1867 Anos de la empresa 0.0004 Tamaño empresa (nro. trabajadores) 0.0158 Si vivienda es propia 0.0222 Monto prestamos (en miles soles) 0.0090 Tasa de interes (en %) 0.1578 Plazo del prestamo (en meses) 0.0324 % total de aciertos 71.8% Building the risk score card Traditional Method: Probit Semi Parametric model: Single Index The single index model have better performance in terms of prediction than the Probit model.
  • 14. Semi-parametric Model: Single Index Model Probability of default: P(y=1|x) = E(y|x) = g(x΄ ) estimated= argmaxβ where g is aproximated for each i as: βestimado Edad promedio beneficiarios 1.000 % Hombres 0.136 % Educación sec. o sup. -0.250 % Casados 0.187 Antigüedad de asociación 0.000 Tamaño de asociación 0.016 Si posee activos 0.022 Monto solicitado/deuda 0.009 Tasa de interés 0.158 Plazo de deuda 0.032 Kernel type (k): second order Gaussian To estimate the risk score for each project i with characteristics Xi we use these βestimated (“weighting factors”) .
  • 15. First Stage: Results of Risk scoring 0% 20% 40% 60% 80% 100% 1 2 3 4 5 6 7 8 9 1011 12 13 14 15 1617 18 19 20 21 22 2324 25 26 27 28 2930 31 32 33 34 3536 37 38 39 NiveldeRiesgo Score riesgo Riesgo maximo 67% Project with the lowest risk (WWF Guatemala) Move to second stage
  • 16. • 24 projects passed the minimum risk threshold and they requested a total value of US$ 4,469,400. In parenthesis is the number of projects that were submitted and completed their application. Country # Project Amount (US$) El Salvador 3 (4) 586,713 Guatemala 6 (11) 968,946 Honduras 10 (14) 1,958,293 Nicaragua 5 (10) 955,448 Total 24 (39) 4,469,400 First Stage: Results of Risk scoring
  • 17. Building the poverty score card • Keep projects that meet the threshold of the risk score card (we initially use 67%) • Project selection will involve assessing the poverty reducing potential of the projects • The projects with the highest poverty score card will be the ones prioritized 17
  • 18. Building the poverty score card 18 Information Criteria Geographical indicators 1. Project located in high poverty areas If it falls in an extreme poor or district with poverty level over 50%. 2. Access to markets If it falls in a district which has a very low accessibility to main road, i.e. equal or more of an hour to the main road. Employment indicators 1. Labor intensive project Number of new employment generated by the project 2.Low skill labor intensive project % of low skill labor in the project 3.Potential female employment intensity of the enterprise % of female labor in the project Spillover indicators 1.Effect on supply chain participants Number of potential beneficiaries by the project. The ration will be calculated based on dollars invested. 2.Other spillover effects (for example provision of some public good) Number of potential beneficiaries by the project. The ration will be calculated based on dollars invested.
  • 19. Pobreza extrema en porcentaje: Baja < 13% Media 13-31% Alta > 31
  • 20. Limite distrital incluido Acceso al mercado (50hab) en horas: Alto <1hora Medio 1-3 Bajo >3
  • 21.
  • 22. Departamento de Chiquimula Municipios Ipala y Concepcion las minas
  • 23. Weighting the criteria • Step 1: Collect for all projects that apply the information for all the variables • Step 2: Normalize/Adjust the data by subtracting the mean to it • Step 3: Calculate the covariance matrix • Step 4: Calculate the eigenvectors and eigenvalues of the covariance matrix • Step 5: Choose the first and second principal components (PCPov1 and PCPov2) • Step 6: Use PCPov 1 and PCPov 2 to rank the projects on the poverty dimension. 23
  • 24. 24 - 10 proyectos (en Honduras) que aprobaron el corte mínimo de riesgo. - Se dispone de 1 millón de dólares americanos. Caracteristicas de los proyectos Monto # empleos # empleos # empleos # empleos # total empleos no empleos # empleos # empleos Pobreza Acceso (miles directos indirectos no femeninos de empleos calificados / femeninos / directos / indirectos / US$) calificados total empleos total empleos inversion inversion var 1 var 2 var 3 var 4 var 5 var 6 var 7 Proyecto 1 (La Ceiba, Atlantida) 100 100 200 100 150 300 0.33 0.50 1.00 2.00 0.191 6.325 Proyecto 2 (El Paraiso, Copan) 200 220 500 300 500 720 0.42 0.69 1.10 2.50 0.426 7.994 Proyecto 3 (Choluteca, Choluteca) 170 150 50 50 50 200 0.25 0.25 0.88 0.29 0.336 2.233 Proyecto 4 (Choluteca, Choluteca) 100 120 400 400 100 520 0.77 0.19 1.20 4.00 0.336 2.233 Proyecto 5 (El Corpus, Choluteca) 180 80 240 150 200 320 0.47 0.63 0.44 1.33 0.541 1.743 Proyecto 6 (Vado Ancho, Paraiso) 120 50 300 200 300 350 0.57 0.86 0.42 2.50 0.477 4.155 Proyecto 7 (Brus Laguna, Gracias a Dios) 300 400 800 1,000 200 1200 0.83 0.17 1.33 2.67 0.527 38.062 Proyecto 8 (Colomoncagua, Intibuca) 120 240 80 200 150 320 0.63 0.47 2.00 0.67 0.445 13.895 Proyecto 9 (San Antonio, Intibuca) 150 130 150 140 100 280 0.50 0.36 0.87 1.00 0.332 23.596 Proyecto 10 (Santa Cruz, Lempira) 250 300 300 400 500 600 0.67 0.83 1.20 1.20 0.496 11.707 Example of poverty scoring
  • 25. 25 Example of poverty scoring (cont.) Paso 1: Calculo de matriz de covarianzas Variables Var1 Var2 Var3 Var4 Var5 Var6 Var7 Var1 1 0.649 -0.187 0.293 0.500 0.457 0.674 Se observa una elevada correlacion entre numero total de Var2 0.649 1 -0.169 0.395 0.522 0.498 0.518 empleos, proporcion de empleos no calificados sobre total Var3 -0.187 -0.169 1 -0.331 -0.121 0.335 -0.374 de empleos, proporcion de empleos indirectos sobre total Var4 0.293 0.395 -0.331 1 -0.071 -0.033 0.387 invertido y baja accesibilidad. Var5 0.500 0.522 -0.121 -0.071 1 -0.023 0.000 Var6 0.457 0.498 0.335 -0.033 -0.023 1 0.256 Var7 0.674 0.518 -0.374 0.387 0.000 0.256 1 Paso 2: Estimacion de valores propios (eigenvalues) de cada componente principal (CP) y seleccion del primer componente CP1 CP2 CP3 CP4 CP5 CP6 CP7 Eigenvalue 2.95 1.53 1.17 0.67 0.40 0.25 0.03 % Variacion explicada 42.08 21.80 16.74 9.51 5.75 3.63 0.47 El primer componente principal (CP1) explica el Variacion acumulada 42.08 63.89 80.63 90.14 95.89 99.53 100.00 42% de la varianza observada. Correlacion de cada componente principal con cada variable CP1 CP2 CP3 CP4 CP5 CP6 CP7 Var1 0.893 0.150 0.079 -0.165 0.263 -0.264 -0.086 Importancia del numero total de empleos, Var2 0.875 0.176 0.102 0.233 -0.244 0.270 -0.080 proporcion de empleos no calificados sobre total Var3 -0.344 0.787 -0.178 0.275 0.369 0.138 -0.003 de empleos y baja accesibilidad en CP1. Var4 0.498 -0.513 -0.335 0.595 0.096 -0.107 0.040 Var5 0.462 0.174 0.855 0.108 0.067 -0.005 0.092 Var6 0.461 0.714 -0.402 -0.042 -0.286 -0.165 0.069 Var7 0.768 -0.224 -0.343 -0.376 0.206 0.232 0.067 Paso 3: Ranking final de proyectos a partir del primer componente principal (CP1) identificado Ranking Puntuacion Monto Acumulado 1. Proyecto 7 4.222 300 300 Se seleccionan los proyectos 7 (300,000 dólares), 4 (100,000 dólares), 8 (120,000 dólares), 2. Proyecto 4 1.075 100 400 10 (250,000 dólares) y 2 (200,000 dólares). 3. Proyecto 8 0.514 120 520 4. Proyecto 10 0.481 250 770 Monto total otorgado igual a 970,000 dolares. 5. Proyecto 2 0.024 200 970 6. Proyecto 9 -0.403 150 7. Proyecto 6 -0.825 120 8. Proyecto 5 -1.192 180 9. Proyecto 1 -1.688 100 10. Proyecto 3 -2.207 170
  • 26. Second Stage: Results on poverty dimension • The first two principal components explain 52% of the variability of the different variables. • The first principal component (CP1) is highly correlated with the number of new employment generated, and direct and indirect employment generated per invested dollar • The second principal component (CP2) has a high correlation with the number of female employments, poverty rate, and low accessibility to markets.
  • 27. CP1 CP2 CP3 CP4 CP5 CP6 CP7 Eigenvalue 2.219 1.399 1.109 1.054 0.765 0.453 0.001 Variability (%) 31.70 19.98 15.84 15.06 10.94 6.48 0.01 Cumulative % 31.70 51.68 67.52 82.58 93.52 99.99 100.00 Eigenvalues (summary of information): Eigenvectors (“weights” of the different indicators in the PC): Indicator CP1 CP2 Number of employment 0.659 -0.023 % Non qualified labor 0.014 0.286 % Female employement -0.139 0.580 Direct employment per $ invested 0.460 -0.317 Indirect employment per $ invested 0.536 0.184 Poverty rate 0.163 0.370 Access to markets 0.141 0.557 Second Stage: Results on poverty dimension
  • 28. Selected Projects -5 -4 -3 -2 -1 0 1 2 3 4 5 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 CP2(19.98%) CP1 (31.70 %) Banhcafé (Honduras) More poverty alleviation Lower poverty alleviation WWF (Guatemala) Lemenigi (Honduras) Las Segovias (Nicaragua) COMUCAP (Honduras) TechnoServe (Honduras) Helvetas (Nicaragua) Desarrollo Sostenible (El Salvador) Aldea Global (Honduras)

Editor's Notes

  1. Riskiness of a borrower (for loans in terms of chances of default on repayment and for grants in terms of efficient and adequate use of funds) is not the only criterion in case of development lending.