Submit Search
Upload
Credit Neural Network with Neural Designer
•
1 like
•
391 views
AI-enhanced title
C
crmbasel
Follow
credit neural network
Read less
Read more
Economy & Finance
Slideshow view
Report
Share
Slideshow view
Report
Share
1 of 69
Recommended
Chapter 4 credit assessment with neutral network
Chapter 4 credit assessment with neutral network
Quan Risk
Model-Based Reinforcement Learning @NIPS2017
Model-Based Reinforcement Learning @NIPS2017
mooopan
Introduction to bayesian_networks[1]
Introduction to bayesian_networks[1]
JULIO GONZALEZ SANZ
Bayesian Networks - A Brief Introduction
Bayesian Networks - A Brief Introduction
Adnan Masood
25 introduction reinforcement_learning
25 introduction reinforcement_learning
Andres Mendez-Vazquez
Wilczynski_BNFinder_BOSC2009
Wilczynski_BNFinder_BOSC2009
bosc
Final review nopause
Final review nopause
j4tang
Reinforcement Learning (DLAI D7L2 2017 UPC Deep Learning for Artificial Intel...
Reinforcement Learning (DLAI D7L2 2017 UPC Deep Learning for Artificial Intel...
Universitat Politècnica de Catalunya
Recommended
Chapter 4 credit assessment with neutral network
Chapter 4 credit assessment with neutral network
Quan Risk
Model-Based Reinforcement Learning @NIPS2017
Model-Based Reinforcement Learning @NIPS2017
mooopan
Introduction to bayesian_networks[1]
Introduction to bayesian_networks[1]
JULIO GONZALEZ SANZ
Bayesian Networks - A Brief Introduction
Bayesian Networks - A Brief Introduction
Adnan Masood
25 introduction reinforcement_learning
25 introduction reinforcement_learning
Andres Mendez-Vazquez
Wilczynski_BNFinder_BOSC2009
Wilczynski_BNFinder_BOSC2009
bosc
Final review nopause
Final review nopause
j4tang
Reinforcement Learning (DLAI D7L2 2017 UPC Deep Learning for Artificial Intel...
Reinforcement Learning (DLAI D7L2 2017 UPC Deep Learning for Artificial Intel...
Universitat Politècnica de Catalunya
Graph Gurus Episode 19: Deep Learning Implemented by GSQL on a Native Paralle...
Graph Gurus Episode 19: Deep Learning Implemented by GSQL on a Native Paralle...
TigerGraph
03.2 homogeneous debt portfolios
03.2 homogeneous debt portfolios
crmbasel
Chapter 7 homogeneous debt portfolios
Chapter 7 homogeneous debt portfolios
Quan Risk
Big Data LDN 2018: SHAPING AN AI-DRIVEN FUTURE WITH AUGMENTED INTELLIGENCE FO...
Big Data LDN 2018: SHAPING AN AI-DRIVEN FUTURE WITH AUGMENTED INTELLIGENCE FO...
Matt Stubbs
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET Journal
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET Journal
World models v0.14
World models v0.14
Duane Nielsen
Learning when to give up: theory, practice and perspectives
Learning when to give up: theory, practice and perspectives
Giuseppe (Pino) Di Fabbrizio
09.2 credit scoring
09.2 credit scoring
crmbasel
Robust Immunological Algorithms for High-Dimensional Global Optimization
Robust Immunological Algorithms for High-Dimensional Global Optimization
Mario Pavone
An Empirical Investigation Of The Arbitrage Pricing Theory
An Empirical Investigation Of The Arbitrage Pricing Theory
Akhil Goyal
XGBoostLSS - An extension of XGBoost to probabilistic forecasting, Alexander ...
XGBoostLSS - An extension of XGBoost to probabilistic forecasting, Alexander ...
Erlangen Artificial Intelligence & Machine Learning Meetup
PORTFOLIO DEFENDER
PORTFOLIO DEFENDER
Anuj Gopal
Algorithmic Game Theory for Critical Infrastructure Security and Resilience
Algorithmic Game Theory for Critical Infrastructure Security and Resilience
Linan Huang
Poster for IRMC Florence, and GA EALE, Iceland, June 2017
Poster for IRMC Florence, and GA EALE, Iceland, June 2017
Ivelin Zvezdov
Network Connectivity and Systematic Risk - Monica Billio. December, 15 2014
Network Connectivity and Systematic Risk - Monica Billio. December, 15 2014
SYRTO Project
Neural Nets Deconstructed
Neural Nets Deconstructed
Paul Sterk
Pro max icdm2012-slides
Pro max icdm2012-slides
Laks Lakshmanan
Profit Maximization over Social Networks
Profit Maximization over Social Networks
Wei Lu
A simple framework for contrastive learning of visual representations
A simple framework for contrastive learning of visual representations
Devansh16
13.2 credit linked notes
13.2 credit linked notes
crmbasel
20.2 regulatory credit exposures
20.2 regulatory credit exposures
crmbasel
More Related Content
Similar to Credit Neural Network with Neural Designer
Graph Gurus Episode 19: Deep Learning Implemented by GSQL on a Native Paralle...
Graph Gurus Episode 19: Deep Learning Implemented by GSQL on a Native Paralle...
TigerGraph
03.2 homogeneous debt portfolios
03.2 homogeneous debt portfolios
crmbasel
Chapter 7 homogeneous debt portfolios
Chapter 7 homogeneous debt portfolios
Quan Risk
Big Data LDN 2018: SHAPING AN AI-DRIVEN FUTURE WITH AUGMENTED INTELLIGENCE FO...
Big Data LDN 2018: SHAPING AN AI-DRIVEN FUTURE WITH AUGMENTED INTELLIGENCE FO...
Matt Stubbs
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET Journal
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET Journal
World models v0.14
World models v0.14
Duane Nielsen
Learning when to give up: theory, practice and perspectives
Learning when to give up: theory, practice and perspectives
Giuseppe (Pino) Di Fabbrizio
09.2 credit scoring
09.2 credit scoring
crmbasel
Robust Immunological Algorithms for High-Dimensional Global Optimization
Robust Immunological Algorithms for High-Dimensional Global Optimization
Mario Pavone
An Empirical Investigation Of The Arbitrage Pricing Theory
An Empirical Investigation Of The Arbitrage Pricing Theory
Akhil Goyal
XGBoostLSS - An extension of XGBoost to probabilistic forecasting, Alexander ...
XGBoostLSS - An extension of XGBoost to probabilistic forecasting, Alexander ...
Erlangen Artificial Intelligence & Machine Learning Meetup
PORTFOLIO DEFENDER
PORTFOLIO DEFENDER
Anuj Gopal
Algorithmic Game Theory for Critical Infrastructure Security and Resilience
Algorithmic Game Theory for Critical Infrastructure Security and Resilience
Linan Huang
Poster for IRMC Florence, and GA EALE, Iceland, June 2017
Poster for IRMC Florence, and GA EALE, Iceland, June 2017
Ivelin Zvezdov
Network Connectivity and Systematic Risk - Monica Billio. December, 15 2014
Network Connectivity and Systematic Risk - Monica Billio. December, 15 2014
SYRTO Project
Neural Nets Deconstructed
Neural Nets Deconstructed
Paul Sterk
Pro max icdm2012-slides
Pro max icdm2012-slides
Laks Lakshmanan
Profit Maximization over Social Networks
Profit Maximization over Social Networks
Wei Lu
A simple framework for contrastive learning of visual representations
A simple framework for contrastive learning of visual representations
Devansh16
Similar to Credit Neural Network with Neural Designer
(20)
Graph Gurus Episode 19: Deep Learning Implemented by GSQL on a Native Paralle...
Graph Gurus Episode 19: Deep Learning Implemented by GSQL on a Native Paralle...
03.2 homogeneous debt portfolios
03.2 homogeneous debt portfolios
Chapter 7 homogeneous debt portfolios
Chapter 7 homogeneous debt portfolios
Big Data LDN 2018: SHAPING AN AI-DRIVEN FUTURE WITH AUGMENTED INTELLIGENCE FO...
Big Data LDN 2018: SHAPING AN AI-DRIVEN FUTURE WITH AUGMENTED INTELLIGENCE FO...
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET- Supervised Learning Classification Algorithms Comparison
IRJET- Supervised Learning Classification Algorithms Comparison
World models v0.14
World models v0.14
Learning when to give up: theory, practice and perspectives
Learning when to give up: theory, practice and perspectives
09.2 credit scoring
09.2 credit scoring
Robust Immunological Algorithms for High-Dimensional Global Optimization
Robust Immunological Algorithms for High-Dimensional Global Optimization
An Empirical Investigation Of The Arbitrage Pricing Theory
An Empirical Investigation Of The Arbitrage Pricing Theory
XGBoostLSS - An extension of XGBoost to probabilistic forecasting, Alexander ...
XGBoostLSS - An extension of XGBoost to probabilistic forecasting, Alexander ...
PORTFOLIO DEFENDER
PORTFOLIO DEFENDER
Algorithmic Game Theory for Critical Infrastructure Security and Resilience
Algorithmic Game Theory for Critical Infrastructure Security and Resilience
Poster for IRMC Florence, and GA EALE, Iceland, June 2017
Poster for IRMC Florence, and GA EALE, Iceland, June 2017
Network Connectivity and Systematic Risk - Monica Billio. December, 15 2014
Network Connectivity and Systematic Risk - Monica Billio. December, 15 2014
Neural Nets Deconstructed
Neural Nets Deconstructed
Pro max icdm2012-slides
Pro max icdm2012-slides
Profit Maximization over Social Networks
Profit Maximization over Social Networks
A simple framework for contrastive learning of visual representations
A simple framework for contrastive learning of visual representations
More from crmbasel
13.2 credit linked notes
13.2 credit linked notes
crmbasel
20.2 regulatory credit exposures
20.2 regulatory credit exposures
crmbasel
19.2 regulatory irb validation
19.2 regulatory irb validation
crmbasel
18.2 internal ratings based approach
18.2 internal ratings based approach
crmbasel
17.2 the basel iii framework
17.2 the basel iii framework
crmbasel
16.2 the ifrs 9
16.2 the ifrs 9
crmbasel
15.2 financial tsunami 2008
15.2 financial tsunami 2008
crmbasel
14.2 collateralization debt obligations
14.2 collateralization debt obligations
crmbasel
12.2 cds indices
12.2 cds indices
crmbasel
11.2 credit default swaps
11.2 credit default swaps
crmbasel
10.2 practical issues in credit assessments
10.2 practical issues in credit assessments
crmbasel
08.2 corporate credit analysis
08.2 corporate credit analysis
crmbasel
07.2 credit ratings and fico scores
07.2 credit ratings and fico scores
crmbasel
06.2 credit risk controls
06.2 credit risk controls
crmbasel
05.2 credit quality monitoring
05.2 credit quality monitoring
crmbasel
04.2 heterogeneous debt portfolio
04.2 heterogeneous debt portfolio
crmbasel
02.2 credit products
02.2 credit products
crmbasel
01.2 credit risk factors and measures
01.2 credit risk factors and measures
crmbasel
05.3 credit quality monitoring
05.3 credit quality monitoring
crmbasel
03.3 homogeneous debt portfolios
03.3 homogeneous debt portfolios
crmbasel
More from crmbasel
(20)
13.2 credit linked notes
13.2 credit linked notes
20.2 regulatory credit exposures
20.2 regulatory credit exposures
19.2 regulatory irb validation
19.2 regulatory irb validation
18.2 internal ratings based approach
18.2 internal ratings based approach
17.2 the basel iii framework
17.2 the basel iii framework
16.2 the ifrs 9
16.2 the ifrs 9
15.2 financial tsunami 2008
15.2 financial tsunami 2008
14.2 collateralization debt obligations
14.2 collateralization debt obligations
12.2 cds indices
12.2 cds indices
11.2 credit default swaps
11.2 credit default swaps
10.2 practical issues in credit assessments
10.2 practical issues in credit assessments
08.2 corporate credit analysis
08.2 corporate credit analysis
07.2 credit ratings and fico scores
07.2 credit ratings and fico scores
06.2 credit risk controls
06.2 credit risk controls
05.2 credit quality monitoring
05.2 credit quality monitoring
04.2 heterogeneous debt portfolio
04.2 heterogeneous debt portfolio
02.2 credit products
02.2 credit products
01.2 credit risk factors and measures
01.2 credit risk factors and measures
05.3 credit quality monitoring
05.3 credit quality monitoring
03.3 homogeneous debt portfolios
03.3 homogeneous debt portfolios
Recently uploaded
20240429 Calibre April 2024 Investor Presentation.pdf
20240429 Calibre April 2024 Investor Presentation.pdf
Adnet Communications
Dharavi Russian callg Girls, { 09892124323 } || Call Girl In Mumbai ...
Dharavi Russian callg Girls, { 09892124323 } || Call Girl In Mumbai ...
Pooja Nehwal
Dividend Policy and Dividend Decision Theories.pptx
Dividend Policy and Dividend Decision Theories.pptx
anshikagoel52
Quarter 4- Module 3 Principles of Marketing
Quarter 4- Module 3 Principles of Marketing
MaristelaRamos12
Q3 2024 Earnings Conference Call and Webcast Slides
Q3 2024 Earnings Conference Call and Webcast Slides
Marketing847413
Vip B Aizawl Call Girls #9907093804 Contact Number Escorts Service Aizawl
Vip B Aizawl Call Girls #9907093804 Contact Number Escorts Service Aizawl
makika9823
call girls in Nand Nagri (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️
call girls in Nand Nagri (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️
9953056974 Low Rate Call Girls In Saket, Delhi NCR
🔝+919953056974 🔝young Delhi Escort service Pusa Road
🔝+919953056974 🔝young Delhi Escort service Pusa Road
9953056974 Low Rate Call Girls In Saket, Delhi NCR
Instant Issue Debit Cards - High School Spirit
Instant Issue Debit Cards - High School Spirit
egoetzinger
Bladex Earnings Call Presentation 1Q2024
Bladex Earnings Call Presentation 1Q2024
Bladex
Stock Market Brief Deck for 4/24/24 .pdf
Stock Market Brief Deck for 4/24/24 .pdf
Michael Silva
The Economic History of the U.S. Lecture 17.pdf
The Economic History of the U.S. Lecture 17.pdf
Gale Pooley
Monthly Market Risk Update: April 2024 [SlideShare]
Monthly Market Risk Update: April 2024 [SlideShare]
Commonwealth
Monthly Economic Monitoring of Ukraine No 231, April 2024
Monthly Economic Monitoring of Ukraine No 231, April 2024
Інститут економічних досліджень та політичних консультацій
Veritas Interim Report 1 January–31 March 2024
Veritas Interim Report 1 January–31 March 2024
Veritas Eläkevakuutus - Veritas Pensionsförsäkring
Call Girls In Yusuf Sarai Women Seeking Men 9654467111
Call Girls In Yusuf Sarai Women Seeking Men 9654467111
Sapana Sha
Chapter 2.ppt of macroeconomics by mankiw 9th edition
Chapter 2.ppt of macroeconomics by mankiw 9th edition
MuhammadHusnain82237
Interimreport1 January–31 March2024 Elo Mutual Pension Insurance Company
Interimreport1 January–31 March2024 Elo Mutual Pension Insurance Company
Työeläkeyhtiö Elo
Best VIP Call Girls Noida Sector 18 Call Me: 8448380779
Best VIP Call Girls Noida Sector 18 Call Me: 8448380779
Delhi Call girls
Andheri Call Girls In 9825968104 Mumbai Hot Models
Andheri Call Girls In 9825968104 Mumbai Hot Models
hematsharma006
Recently uploaded
(20)
20240429 Calibre April 2024 Investor Presentation.pdf
20240429 Calibre April 2024 Investor Presentation.pdf
Dharavi Russian callg Girls, { 09892124323 } || Call Girl In Mumbai ...
Dharavi Russian callg Girls, { 09892124323 } || Call Girl In Mumbai ...
Dividend Policy and Dividend Decision Theories.pptx
Dividend Policy and Dividend Decision Theories.pptx
Quarter 4- Module 3 Principles of Marketing
Quarter 4- Module 3 Principles of Marketing
Q3 2024 Earnings Conference Call and Webcast Slides
Q3 2024 Earnings Conference Call and Webcast Slides
Vip B Aizawl Call Girls #9907093804 Contact Number Escorts Service Aizawl
Vip B Aizawl Call Girls #9907093804 Contact Number Escorts Service Aizawl
call girls in Nand Nagri (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️
call girls in Nand Nagri (DELHI) 🔝 >༒9953330565🔝 genuine Escort Service 🔝✔️✔️
🔝+919953056974 🔝young Delhi Escort service Pusa Road
🔝+919953056974 🔝young Delhi Escort service Pusa Road
Instant Issue Debit Cards - High School Spirit
Instant Issue Debit Cards - High School Spirit
Bladex Earnings Call Presentation 1Q2024
Bladex Earnings Call Presentation 1Q2024
Stock Market Brief Deck for 4/24/24 .pdf
Stock Market Brief Deck for 4/24/24 .pdf
The Economic History of the U.S. Lecture 17.pdf
The Economic History of the U.S. Lecture 17.pdf
Monthly Market Risk Update: April 2024 [SlideShare]
Monthly Market Risk Update: April 2024 [SlideShare]
Monthly Economic Monitoring of Ukraine No 231, April 2024
Monthly Economic Monitoring of Ukraine No 231, April 2024
Veritas Interim Report 1 January–31 March 2024
Veritas Interim Report 1 January–31 March 2024
Call Girls In Yusuf Sarai Women Seeking Men 9654467111
Call Girls In Yusuf Sarai Women Seeking Men 9654467111
Chapter 2.ppt of macroeconomics by mankiw 9th edition
Chapter 2.ppt of macroeconomics by mankiw 9th edition
Interimreport1 January–31 March2024 Elo Mutual Pension Insurance Company
Interimreport1 January–31 March2024 Elo Mutual Pension Insurance Company
Best VIP Call Girls Noida Sector 18 Call Me: 8448380779
Best VIP Call Girls Noida Sector 18 Call Me: 8448380779
Andheri Call Girls In 9825968104 Mumbai Hot Models
Andheri Call Girls In 9825968104 Mumbai Hot Models
Credit Neural Network with Neural Designer
1.
Credit Neural Network with
Neural Designer This presentation file is prepared in accordance with the text book “Credit Neural Network with Neural Designer” Website : https://sites.google.com/site/quanrisk E-mail : quanrisk@gmail.com Copyright © 2019 CapitaLogic Limited
2.
Declaration Copyright ©
2019 CapitaLogic Limited. All rights reserved. No part of this presentation file may be reproduced, in any form or by any means, without written permission from CapitaLogic Limited. Authored by Dr. LAM Yat-fai (林日辉), Director, CapitaLogic Limited, Adjunct Professor of Finance, City University of Hong Kong, Doctor of Business Administration, CFA, CAIA, CAMS, FRM, PRM. 2Copyright © 2019 CapitaLogic Limited
3.
Outline Data preparation
Classical regressions Monotonic neural network Continuous response network Credit neural network Shadow credit rating LGD of residential mortgages 3Copyright © 2019 CapitaLogic Limited
4.
What is research?
There is a result Theory Which factors cause this result? How increases and/or decreases in these factors impact the result? Testing Does the theory work in real life scenarios? 4Copyright © 2019 CapitaLogic Limited
5.
Financial research Response
variable (y) To be explained and then estimated Explanatory variables (x1, x2, x3, … , xN) Independent Individually impacts the response variable Collectively sufficient to explain the response variable Random noise Impact but immaterial to the response variable 1 2 3 Ny = f x ,x ,x , ... ,x + Random noise 5Copyright © 2019 CapitaLogic Limited
6.
Data preparation Outliers
Relevancy Inter-dependency Randomization Consistency Sampling Copyright © 2019 CapitaLogic Limited 6 Example 1.1
7.
Outliers Smallest 1%
value of the variables Largest 1% value of the variables To exclude outliers Copyright © 2019 CapitaLogic Limited 7 Example 1.2
8.
Relevancy Between the
response variable and an explanatory Quantified by the Spearman’s ρ Smaller t-statistic and larger p-value suggest weaker relevancy To exclude irrelevant explanatory variables Copyright © 2019 CapitaLogic Limited 8 Example 1.3
9.
Inter-dependency Between any
two explanatory variables Quantified by the Spearman’s ρ Larger Spearman’s ρ suggests stronger inter- dependency between two explanatory variables To drop one of the inter-dependent explanatory variables Copyright © 2019 CapitaLogic Limited 9 Example 1.4
10.
Randomization For a
continuous response variable, order the records randomly For a categorical response variable, order the records randomly for each category of the response variable Copyright © 2019 CapitaLogic Limited 10 Example 1.5
11.
Consistency Consistent variables
Cover a similar range Increase in an explanatory variable => Increase in the response variable while holding other explanatory variables fixed More effective and efficient for computer implementation 11Copyright © 2019 CapitaLogic Limited Example 1.6
12.
Consistent transformation 12 Consistent variable Value
of the variable - Average of the variable = Standard deviation of the variable Spearman's ρ between the variable × Sign and the response variable Copyright © 2019 CapitaLogic Limited
13.
Inverse consistent transformation 13 Value
of the variable = Consistent variable × Standard deviation of the variable Spearman's ρ between the variable × Sign and the response variable + Average of the variable Copyright © 2019 CapitaLogic Limited
14.
Sampling For a
continuous explanatory variable Training data set No. of explanatory variables × 15 or 1% of all records, which ever is more Training data set No. of explanatory variables × 5 or 0.5% of all records, which ever is more For each value of a categorical explanatory variable Training data set No. of explanatory variables × 15 or 1% of all records, which ever is more Training data set No. of explanatory variables × 5 or 0.5% of all records, which ever is more Copyright © 2019 CapitaLogic Limited 14
15.
Outline Data preparation
Classical regressions Monotonic neural network Continuous response network Credit neural network Shadow credit rating LGD of residential mortgages 15Copyright © 2019 CapitaLogic Limited
16.
Simple linear regression
Response variable (y) To be explained and then estimated Explanatory variable (x) Sufficient to explain the response variable Random noise Impact but immaterial to the response variable Linear relationship Increase in one unit of the explanatory variable always increases the same level of response variable; or Increase in one unit of the explanatory variable always decreases the same level of response variable 0 1y = a + a x + Random noise 16Copyright © 2019 CapitaLogic Limited
17.
Polynomial regression Response
variable (y) To be explained and then estimated Explanatory variable (x) Sufficient to explain the response variable Random noise Impact but immaterial to the response variable Advantage Increase in the order of polynomial will increase the fitness Disadvantage Increase in the order of polynomial may introduce over fitting 2 3 N 0 1 2 3 Ny = a + a x + a x + a x + ... + a x + Random noise 17Copyright © 2019 CapitaLogic Limited Example 2.1 Example 2.2
18.
Multiple linear regression
Response variable (y) To be explained and then estimated Explanatory variables (x1, x2, x3, … , xN) Independent Individually impacts the response variable Collectively sufficient to explain the response variable Random noise Impact but immaterial to the response variable Normally distributed with constant standard deviation Independent Linear relationship Holding other explanatory variables fixed Increase in one unit of an explanatory variable always increases the same level of response variable; or Increase in one unit of an explanatory variable always decreases the same level of response variable 0 1 1 2 2 3 3 N Ny = a + a x + a x + a x + ... + a x + Random noise 18Copyright © 2019 CapitaLogic Limited Example 2.3
19.
Probit transformation 19
2 Probit - -1 1 τ Probistic = exp - dτ 22π Probistic = Φ Probit 0,1 Probit = Φ Probistic - ,+ Copyright © 2019 CapitaLogic Limited
20.
Probistic regression Response
variable (y) Either 0 or 1 Explanatory variables (x1, x2, x3, … , xN) Independent Consistent Individually impacts the response variable Collectively sufficient to explain the response variable Random noise Impact but immaterial to the response variable 0 1 1 2 2 3 3 N NProbit = a + a x + a x + a x + ... + a x + Random noise Probistic = Φ Probit y = 0 if Probistic < 50% = 1 if Probistic 50% 20Copyright © 2019 CapitaLogic Limited Example 2.4
21.
Probit transformation 21Copyright ©
2019 CapitaLogic Limited
22.
22 Maximum likelihood method
0 1 1 2 2 3 3 N N 1 1 1 1 2 3 1 0 0 U 1 2 0 3 Probit = a + a x + a x + a x + ... + a x Probistic = Φ Probit L = Probistic × Probistic × Probistic × × Probistic × 1 - Probistic × 1 - Probistic × 1 - Probistic × × 1 - Probisti 0 V 1 1 1 1 2 3 1 0 0 U 1 2 0 0 3 V c Mazimize ln(L) = ln Probistic + ln Probistic + ln Probistic + + ln Probistic + ln 1 - Probistic + ln 1 - Probistic + ln 1 - Probistic + + ln 1 - Probistic Copyright © 2019 CapitaLogic Limited
23.
Outline Data preparation
Classical regressions Monotonic neural network Continuous response network Credit neural network Shadow credit rating LGD of residential mortgages 23Copyright © 2019 CapitaLogic Limited
24.
Monotonic neural network
To release the limitations of multiple linear regression WITHOUT Introducing complex mathematics An extremely simplified version of neural network The entry level of artificial intelligence, machine learning and deep learning 24Copyright © 2019 CapitaLogic Limited
25.
Requirements of monotonic neural
network Response variable (y) Consistent To be explained and estimated Explanatory variables (x1, x2, x3, … , xN) Consistent Collectively sufficient to explain the response variable Random noise Impact but immaterial to the response variable 1 2 3 Ny = f x ,x ,x , ... ,x + Random noise 25Copyright © 2019 CapitaLogic Limited
26.
Monotonic: Black Scholes
model Explanatory variable Change Call option value Stock price ↑ ↑ Strike price ↓ Volatility ↑ Risk free rate ↑ Time to maturity ↑ 26Copyright © 2019 CapitaLogic Limited
27.
Monotonic: Merton’s model Explanatory
variable Change Credit quality Value of equity ↑ ↑ Value of liabilities ↓ Volatility of equity ↑ 27Copyright © 2019 CapitaLogic Limited
28.
Not monotonic x1 x2
y + ↑ ↑ - ↓ 1 2y = x x + Random noise 28Copyright © 2019 CapitaLogic Limited
29.
Implementation of a
neural network Variables from theory and/or experience A response variable A set of explanatory variables Prepare samples Training : Testing = 3 : 1 Set up the neural network Train the neural network with training data set Calculate values with the neural network Assess the in sample accuracy with the training data set Assess the out sample accuracy with the testing data set Use the neural network to conduct estimation Assess the monotonicity with scenario analysis to verify the theory 29Copyright © 2019 CapitaLogic Limited
30.
Network structure Explanatory variables Response variable Neurons 30 y Copyright ©
2019 CapitaLogic Limited Example 3.1
31.
Optimization For each
neuron k Response variable Sum of squared error Find a set of as and bs to minimize the SSE 31 k k k k k k 0 1 1 2 2 3 3 N N k k k N 0 1 2 3 N 2 n =Φ a + a x + a x + a x + ... + a x y est. = Φ b + b n + b n + b n + ... + b n SSE = y - y est. Copyright © 2019 CapitaLogic Limited
32.
No. of as
and bs No. of nodes N explanatory variables N neurons One response variables Each neuron N + 1 as Response variable N + 1 bs Total number (N + 1)2 Including irrelevant or dependent explanatory variable will waste a lot of computing power 32Copyright © 2019 CapitaLogic Limited
33.
Advantages of monotonic neural
network Higher predictive power Minimum structural assumption Consistency Simple network structure Single neuron layer No. of neurons = No. of explanatory variables Moderate computing power Robust to irrelevant and/or dependent explanatory variables at a cost of computing power Can be easily applied to most financial analysis Particularly suitable for marginally decreasing response variable For example, PD 33Copyright © 2019 CapitaLogic Limited
34.
Disadvantages of monotonic neural
network Rely on theory and/or experience to identify explanatory variables May incorporate the effect of random noise during training No straight forward mathematical formulation More samples 34Copyright © 2019 CapitaLogic Limited
35.
Outline Data preparation
Classical regressions Monotonic neural network Continuous response network Credit neural network Shadow credit rating LGD of residential mortgages 35Copyright © 2019 CapitaLogic Limited
36.
Variables and samples
Response variable y Explanatory variables x1, x2, x3 Sufficient no. of samples 36Copyright © 2019 CapitaLogic Limited Example 4.1
37.
Create a neural
network 37Copyright © 2019 CapitaLogic Limited
38.
Import training data 38Copyright
© 2019 CapitaLogic Limited Example 4.1
39.
Train the neural
network 39Copyright © 2019 CapitaLogic Limited Example 4.2
40.
Conduct estimation 40Copyright ©
2019 CapitaLogic Limited Example 4.3
41.
Testing Estimate y
with the training and testing data sets Compare with the historical response variable Calculate the error 41 y est. Absoulte percentage error = - 1 × 100% y Copyright © 2019 CapitaLogic Limited
42.
Accuracy matrix 42 % error
< Count Percentage 50% 184 92% 30% 176 88% 10% 126 63% 5% 58 29% 3% 36 18% 1% 10 5% Total 200 The larger the better Copyright © 2019 CapitaLogic Limited
43.
Estimation Given a
set of explanatory variables without response variable Use the neural network to estimate the ys 43Copyright © 2019 CapitaLogic Limited Example 4.4
44.
Monotonicity analysis Baseline
scenarios All explanatory variables set to The medians The averages The maximums The minimums While fixing other explanatory variables Vary one explanatory variable from the minimum to the maximum Conduct estimation Plot response variable vs explanatory variable Repeat for other explanatory variables 44Copyright © 2019 CapitaLogic Limited Example 4.6
45.
45Copyright © 2019
CapitaLogic Limited Example 4.7
46.
Exception Violation of
monotonicity The theory and/or experience need to be reviewed Inter-dependency among explanatory variables Too much random noise Response variable insensitive to an explanatory variable The explanatory variable may be irrelevant Remove the explanatory and re-build the neural network 46Copyright © 2019 CapitaLogic Limited
47.
Outline Data preparation
Classical regressions Monotonic neural network Continuous response network Credit neural network Shadow credit rating LGD of residential mortgages 47Copyright © 2019 CapitaLogic Limited
48.
Merton’s corporate default
model Market’s view of credit quality can be derived from observable x1 = Market value of equity x2 = Book value of liabilities x3 = Volatility of equity 48Copyright © 2019 CapitaLogic Limited
49.
Create a neural
network 49Copyright © 2019 CapitaLogic Limited Example 5.1 Example 5.2 Example 5.3
50.
Variables and samples
Response variable Coded PD of the listed companies 0 for survival and 1 for default Explanatory variables x1, x2, x3 Sufficient no. of samples 50Copyright © 2019 CapitaLogic Limited
51.
Testing Conduct estimation
with the training and testing data sets on the coded PD Use the neural network to estimate a PD If PD < 50%, then a bad borrower If PD > 50%, then a good borrower Compare with the historical response variable 51Copyright © 2019 CapitaLogic Limited
52.
Accuracy matrix 52 Match ?
Count Percentage Yes 180 90% No 20 10% Total 200 The more Yes the better Copyright © 2019 CapitaLogic Limited
53.
Estimation Given a
set of explanatory variables without the coded PD Use the neural network to estimate the PDs If PD < 50%, then a bad borrower If PD > 50%, then a good borrower 53Copyright © 2019 CapitaLogic Limited Example 5.4
54.
Outline Data preparation
Classical regressions Monotonic neural network Continuous response network Credit neural network Shadow credit rating LGD of residential mortgages 54Copyright © 2019 CapitaLogic Limited
55.
Merton’s corporate default
model Market’s view of credit quality can be derived from observable x1 = Market value of equity x2 = Book value of liabilities x3 = Volatility of equity 55Copyright © 2019 CapitaLogic Limited
56.
Create a neural
network 56Copyright © 2019 CapitaLogic Limited Example 6.1 Example 6.2 Example 6.3
57.
Shadow credit rating
The idea of using credit ratings from major credit agencies to derive a relationship between credit rating and explanatory variables Assume that the credit ratings are largely accurate 57Copyright © 2019 CapitaLogic Limited
58.
Variables and samples
Response variable Credit rating Explanatory variables x1, x2, x3 Sufficient no. of samples 58Copyright © 2019 CapitaLogic Limited
59.
Testing Estimate the
probabilities of credit ratings with the training and testing data sets Select the credit rating with the highest probability Map the credit rating to the rank Compare with the historical response variable 59Copyright © 2019 CapitaLogic Limited
60.
Accuracy matrix 60 Variation Count
Percentage 0 152 76% 1 42 21% 2 4 2% 3 2 1% Total 200 The more 0 variation the better Copyright © 2019 CapitaLogic Limited
61.
Estimation Given a
set of explanatory variables without credit rating Use the neural network to estimate the probabilities of credit ratings Select the credit rating with the highest probability 61Copyright © 2019 CapitaLogic Limited Example 6.4
62.
Outline Data preparation
Classical regressions Monotonic neural network Continuous response network Credit neural network Shadow credit rating LGD of residential mortgages 62Copyright © 2019 CapitaLogic Limited
63.
LGD of collateralized
lending Factor impacting the LGD Outstanding loan amount Current value of collateral Drift of collateral value Volatility of collateral value Explanatory variables x1 = Loan to value ratio x2 = Drift of collateral value x3 = Volatility of collateral value 63Copyright © 2019 CapitaLogic Limited
64.
Create a neural
network 64Copyright © 2019 CapitaLogic Limited Example 7.1 Example 7.2 Example 7.3
65.
Variables and samples
Response variable Credit rating Explanatory variables x1, x2, x3 Sufficient no. of samples 65Copyright © 2019 CapitaLogic Limited
66.
Testing Estimate the
LGD with the training and testing data sets Compare with the historical response variable 66Copyright © 2019 CapitaLogic Limited
67.
Estimation Given a
set of explanatory variables without LGD Use the neural network to estimate the LGDs 67Copyright © 2019 CapitaLogic Limited Example 7.4
68.
Deep learning Many
explanatory variables Many layers of neurons Several response variables Can handle very complex relationships Non-monotonic relationships Periodic relationships Require huge computing power 68Copyright © 2019 CapitaLogic Limited
69.
Deep learning neural
network Explanatory variables Response variables Layers of neurons 69 y Copyright © 2019 CapitaLogic Limited