fostering sme development - biia.com · 2018-06-06 · the last two rows shows the no. of times a...
TRANSCRIPT
Fostering SME Development
SME Access To Finance Workshop
Date: 5th November 2010,
Venue: Jakarata, Indonesia
Sharing of Experience from
“SME Rating Agency of India Ltd
(SMERA)
Sharing of Indian Experience-SMERA
SME segment overview of the country- split of Micro/Small/Medium Ent
Macro economic policies of the government regarding SMEs-priority sector lending status with quotas etc
Study on accounting standards and financial disclosures to assess their impact on rating decisions
Identification of clusters and execution of studies in clusters
Corporate linkages and its impact on SMEs ( OEMs/SEZ etc)
Access to quality of data on SMEs in public domain-ROC/Credit Bureau
Definition of SMEs followed by lending agencies
A need gap analysis-reasons for setting up a CRA for SMEs
Identification of scope of services to be offered by the CRA
SME Rating Agency-An Indian Experience
Source: Fourth All-India Census of Micro, Small & Medium Enterprises with reference year of 2004-05
SME Rating Agency-An Indian Experience
2003 2004 2005
% of NBC
% of Total priority sector
advances % of NBC
% of Total priority sector
advances % of NBC
% of Total priority sector
advances
Public sector Banks 39.7 100 44.7 100 42.5 100
I) Agriculture 15.4 38.9 18.3 40.9 17.2 41.4
II) MSEs 7.8 19.7 11.1 24.8 11.3 26.5
III) others 16.5 41.5 15.3 34.4 14 32.1
Private sector Banks 42.9 100 42.5 100 46.8 100
I) Agriculture 12.7 36 17.1 35.7 15.9 40
II) MSEs 3.9 9.1 13.7 28.6 12 25.2
III) others 26.3 54.9 11.7 35.7 18.9 34.8
Foreign sector Banks 33.4 100 39.5 100 34.3 100
I) Agriculture 18.3 54.7 22.7 57.6 19.4 56.8
II) MSEs 10.3 30.8 12.2 30.8 11.2 32.8
III) others 4.8 14.5 4.6 11.6 3.7 10.4
Sources : 1. RBI Annual Report 2006 2. RBI Report on Trend & Progress of Banking in India 2006-07
SME Rating Agency-An Indian Experience
Bank Lending to SMEs-An Overview
Registered Unregistered
No Finance/Self Finance 87.77 93.08
Finance through Institutional Sources
Finance through Non-Institutional Sources 1.02 2.12
Total 100 100
Source: Fourth All-India Census of Micro, Small & Medium Enterprises with reference year of 2006-07
SME Rating Agency-An Indian Experience
Source: Basic Statistical returns, RBI, Aug 2004
Statistics on Credit Range to SMEs
0
10
20
30
40
50
60
Rs. 25,000
and Less
Above Rs.
25,000 and upto
Rs.2 Lakh
Above Rs. 2
Lakh and upto o Rs.5 Lakh
Above Rs. 5
Lakh and upto Rs.10 Lakh
Above Rs. 10
Lakh and upto Rs.25 Lakh
Above Rs. 25
Lakh and upto Rs.50 Lakh
Above Rs. 50
Lakh and upto Rs.1
Crore
Above Rs. 1
Crore and upto
Rs.4 Crore
Above Rs. 4
Crore and upto
Rs.6 Crore
Above Rs. 6
Crore and upto
Rs.10 Crore
Above Rs. 10
Crore and upto
Rs.25 Crore
Above Rs. 25
Crore
% Number of Accounts
SME Rating Agency-An Indian Experience
SME Rating Agency-An Indian Experience
US 57K
US 57 K to 1.14 Mill
US 1.14 Mill to 22 Mill
US 23 K
US 23K to 454K
US 454K to 1.14 mill
Estimation of the potential market size
Data on type/tenure/size of lending by various lending institutions
Identify market for ratings based on size of units-Bank Loan ratings/Instrument ratings/Debt ratings/NBFC ratings etc.
Divide the market size between regulatory and non-regulatory ratings
Detailed analysis of recruitment/training/marketing/compliance costs
Estimate revenues based on the above
Assess the Financial viability & overall feasibility of the entity with a business plan
Draw a P&L and Balance Sheet of the proposed agency for 5 years
SME Rating Agency-An Indian Experience
Identification of technical partner, promoters & other shareholders
Formation of a core team
Incorporation of the company and other compliances
Identification and appointment of CEO
Identification of independent directors and formation of the board of directors
Formation of internal rating committee/criteria/external rating committee
Appointment of technical consultant for drawing rating process in compliance with IOSCO guidelines
Formulation of policies and guidelines for the company
Finalization of org structure, recruitment and training of staff
Marketing of the concept to lenders-Pre launch
Launch of operations
SME Rating Agency-An Indian Experience
12
National Equity Fund
Integrated Infrastructural Development Scheme
Technology Development & Modernisation Fund Scheme
Technology Up gradation Fund Scheme for Textiles and Jute Industries
Risk Capital for Small Industries
Credit Guarantee Fund Scheme for Small Industries (through separate
Trust)
Credit Linked Capital Subsidy Scheme
Cumulative assistance USD 6 billion to the Indian SME sector since inception!
Set up in April 1990 under an Act of Parliament, SIDBI is the Principal Financial
Institution for Promotion, Financing and Development of Small Scale Industries and to
Coordinate the functions of Institutions Serving the Small Industries
Small Industries Development Bank of
India-Joint Promoter
13
• D&B India had data base on 146,000 entities with P&L and B/S in SME Space
Source – D&B database (2003-2006)
• D&B India has compiled detailed reports on over 50,000 Indian entities
since its inception.
• There are over 11,000 active* ratings in D&B India database.
• D&B India receives over 32,000 rating requests on cross border entities
annually.
* Ratings are valid 13 months from the date of rating carried out.
14
Board of SMERA comprises independent directors, nominees of SME centric Banks, SIDBI, D&B.
No single entity has a controlling stake-10 Banks and 1 D&B
Shareholding of Banks – commitment to the SME sector
SIDBI, 34%
D&B India, 34%
ICICI Bank, 11%
SBI, 9.50%
Citigroup, 5%
Indian Bank, 4%
Standard Chartered
Bank, 4.50%
Other PSBs, 28%
15
Request for
Interview and
site visit Rating Request SMERA Correspondents Entity
Questionnaire
Documentation,
audited results
and certified
projections
Site Visit
Assessment
Report
ROC
Information
Rating
Model
Industry
Assessment
and Cluster
data
Rating Analyst
External Committee Review
Third Party Data
– e.g. Litigation
Information
Conducts site
visit and
interviews
Management
Documents Obtained
External Data
SMERA
Database
Entity Rating Final Rating
SMERA started operations with its inauguration by
The Finance Minister of India on Sept 2005
On Sept 2010 SMERA completed 5 years with 9000 MSME
Ratings
Only Independent agency dedicated for SME Ratings
SMERA to play active role on newly created SME Exchange
SME Ratings/Green & Brown field Grading/MFI Grading/
Educational Institutions Rating/Green Rating
Building a suitable SME Rating model & processes for rating of SMEs
Mapping of SMERA model with Banks internal model-Interest rate benefit
Lenders’ resistance to share default data
Defining the process for ratings
High initial marketing and sales cost for brand buildup & awareness
Compiling a robust database of rated entities-Identification of RDBMS
Access to experienced and trained manpower & retention
Lack of adequate experienced members for external rating committee
Acceptability of external ratings by the lenders-mindset & legacy issue
Slow Regulatory approval due to lack of default data-Basel-II
Stricter compliance by regulations for rating agencies-post crisis
Lack of Information availability in the public domain
Submission of un-reliable and insufficient data by the SMEs for rating
High expectations from Ratings-SMEs/Lenders/Govt.
SME Rating Agency-An Indian Experience
19
Variables are selected for modeling after considering so many variable selection/reduction process
Table below shows the different statistical and machine learning variable selection/reduction
process considered and the list of variables selected in each of the process
The Last two rows shows the no. of times a variable had been selected in the selection process
Variables which are selected more no. of times will be considered for further analysis/modeling
Variables which are considered important by the Bank will also be included in the modeling
20
Prob
Level Non- Non- Sensi- Speci- FALSE FALSE
Event Event tivity ficity POS NEG
0.05 734 1260 3208 32 38.1 95.8 28.2 81.4 2.5
0.1 671 2366 2102 95 58 87.6 53 75.8 3.9
0.146 562 3093 1375 204 69.8 73.4 69.2 71 6.2
0.2 442 3589 879 324 77 57.7 80.3 66.5 8.3
0.5 57 4401 67 709 85.2 7.4 98.5 54 13.9
Classification Table
Correct Incorrect Percentages
Event Event Correct
Classification table, also called as confusion matrix is the best way to assess the predictive power of the
model.
The above table is the output of classifying the observations according to whether the predicted
probabilities are above or below the cut-off values in the range (0,1). In our case we had chosen 5 different
cutoff points (0.05,0.1,0.146,0.2,0.5) for demonstration.
The accuracy of the model is measured by its Sensitivity (ability to predict an default event correctly) and
Specificity (ability to predict an non-default event correctly).
Sensitivity – Proportion of default companies that were predicted to be defaults.
Specificity – Proportion of non-default companies that were predicted to be non-defaults.
21
A Receiver Operating Characteristic (ROC)
curve is a graphical representation of the trade
off between the false negative and false positive
rates for every possible cut off.
The accuracy of the model (i.e. the ability of
the model to correctly classify default and non-
default cases) is measured by the area under the
ROC curve.
An area of 1 represents a perfect model, while
an area of .5 represents a worthless test.
The closer the curve follows the left-hand
border and then the top border of the ROC
space, the more accurate the model; the true
positive rate is high and the false positive rate is
low.
Area under the Curve (AUC) for the model is
78.2
22
Prediction Success for Learn Data
Actual Total Percent 0 1
Class Cases Correct N=3004 N=2230
0 4,468 64.167 2,867 1,601
1 766 82.115 137 629
Actual Total Percent 0 1
Class Cases Correct N=3131 N=2103
0 4,468 65.958 2,947 1,521
1 766 75.979 184 582
Prediction Success for Test Data
0
20
40
60
80
100
0 20 40 60 80 100
% C
lass
% Population
0
20
40
60
80
100
Gains Chart showing the Predictive Power
23
0
20
40
60
80
100
0 20 40 60 80 100
Pct.
Cla
ss 1
Pct. Population
Actual Predicted Predicted Total Percent
Class 0 1 Cases Correct
0 771 3,697 4468 17.26
1 19 747 766 97.52
Actual Predicted Predicted Total Percent
Class 0 1 Cases Correct
0 2,702 1,766 4468 60.47
1 161 605 766 78.98
Gains Chart showing the Predictive Power
Prediction Success @ 0.5 Cut-off Prediction Success @ 0.6 cut-off
24
Classification @ 0.05 Cut-off Classification @ 0.10 Cut-off
Classification @ 0.146 Cut-off Classification @ 0.50 Cut-off
SME Rating Agency-An Indian Experience
SMEs are vulnerable
Information on SMEs is scarce and un-reliable
SMEs are geographically/sectorally dispersed-
difficult to assess the risk
SMEs are transaction intensive
Limited 7%
Partnership 23%
Private Limited 43%
Proprietorship 27%
SME Rating Agency-An Indian Experience
0
200
400
600
800
1000
1200
1400
1600
SME Rating Agency-An Indian Experience
0
500
1000
1500
2000
2500
3000
3500
1 2 3 4 5 6 7 8
SME Rating Agency-An Indian Experience
Provided SME ratings which are: • Centralized
• Comprehensive
• Transparent
• Reliable
• Unbiased and independent - acceptable across the board
Facilitated Sound Credit Decisions: • Whom not to give credit, and more importantly,
• Whom to give credit
• Comfort to lenders and auditors
Provided Tools For Effective Risk Management • Arriving at risk weighted credit pricing • The evaluation of risks when changing the terms of credit agreements
Helped SMEs
• Gain Access to Bank Finance in an easy and cost-effective manner
• Better rated units were able to market their unit • Self improvement tool
SME Rating Agency-An Indian Experience
Statistical
Modeling
Expert
Rules
Skilled
Business
Analysts
Industry
Trends
HYBRID RATING METHODOLOGY
Financial factors
based on
statistical
modeling
Qualitative factors
based on due
diligence and site
visit
SME Rating Agency-An Indian Experience
Rating Factor Schema
Financial Parameters Non-Financial Parameters
Solvency Ratios
Liquidity Ratios
Activity Ratios
Profitability Ratios
Management Quality
Location Advantage
Marketing Network
Legal Issues
Industry and Macro-Economic Assessment
Trend Analysis
An exhaustive list of qualitative and quantitative factors considered for rating.
Each financial parameter is benchmarked within its industry-size peer group
SME Rating Agency-An Indian Experience
Employee Count
Marketing Network
Management Shareholding
Age
Legal
Competitive forces
Industry Trade & Credit Terms
Industry growth
Each non-financial parameter is characterized by objective options which is scored
Management Risk
Industry Risk
Business Risk
SME Rating Agency-An Indian Experience
Customized Rating Scale
Size Indicator reflects the
tangible net worth of a
company.
Composite Appraisal Indicator
reflects a company’s health,
stability and overall condition.
Ensures peer-to-peer comparison of an SME within its own size category (Size Indicator).
Banks’ Internal Risk Rating Models Mostly financial parameters Industry benchmarking – nil or limited Mostly generalization of MSMEs across geographic boundaries
Banks are involved as financiers Backward Looking
Enterprise Rating by SMERA
Rating beyond financials Robust industry benchmarking and also linked to size Consideration to parameters specific to a geographic location Ratings are neutral; credit enhancement measures are not reckoned Looking beyond
SME Rating Agency-An Indian Experience
SME Rating Agency-An Indian Experience
Parag Patki
Chief Executive Officer
SME Rating Agency of India Ltd
Tel. : +91 22 6714 1101
Fax : +91 22 6714 1142
Cell : +91 99303 95736
Email : [email protected]