Operational Risk Modeling Analytics



Similar documents
Modeling Individual Claims for Motor Third Party Liability of Insurance Companies in Albania

Chapter 4 - Lecture 1 Probability Density Functions and Cumul. Distribution Functions

Gamma Distribution Fitting

Software for Distributions in R

Properties of Future Lifetime Distributions and Estimation

Package SHELF. February 5, 2016

seven Statistical Analysis with Excel chapter OVERVIEW CHAPTER

Exam C, Fall 2006 PRELIMINARY ANSWER KEY

Determining distribution parameters from quantiles

SUMAN DUVVURU STAT 567 PROJECT REPORT

Distribution (Weibull) Fitting

RISKY LOSS DISTRIBUTIONS AND MODELING THE LOSS RESERVE PAY-OUT TAIL

BNG 202 Biomechanics Lab. Descriptive statistics and probability distributions I

Advanced Excel for Institutional Researchers

Data Modeling & Analysis Techniques. Probability & Statistics. Manfred Huber

Probability and Statistics Prof. Dr. Somesh Kumar Department of Mathematics Indian Institute of Technology, Kharagpur

Course Brochure Chandoo.org

An Internal Model for Operational Risk Computation

5: Magnitude 6: Convert to Polar 7: Convert to Rectangular

Dongfeng Li. Autumn 2010

Lecture 15 Introduction to Survival Analysis

STA 4273H: Statistical Machine Learning

Review of Random Variables

LDA at Work: Deutsche Bank s Approach to Quantifying Operational Risk

Java Modules for Time Series Analysis

Empirical Study of effect of using Weibull. NIFTY index options

Maximum Likelihood Estimation

Lecture 8: More Continuous Random Variables

FITTING INSURANCE CLAIMS TO SKEWED DISTRIBUTIONS: ARE

Survival Analysis of Left Truncated Income Protection Insurance Data. [March 29, 2012]

NOTES ON THE BANK OF ENGLAND OPTION-IMPLIED PROBABILITY DENSITY FUNCTIONS

VISUALIZATION OF DENSITY FUNCTIONS WITH GEOGEBRA

SAS Software to Fit the Generalized Linear Model

Chapter 3 RANDOM VARIATE GENERATION

Descriptive Statistics

Graphs. Exploratory data analysis. Graphs. Standard forms. A graph is a suitable way of representing data if:

Logistic Regression (1/24/13)

Package EstCRM. July 13, 2015

Applied Reliability Page 1 APPLIED RELIABILITY. Techniques for Reliability Analysis

1 Maximum likelihood estimation

Joint Exam 1/P Sample Exam 1

**BEGINNING OF EXAMINATION** The annual number of claims for an insured has probability function: , 0 < q < 1.

Financial Assets Behaving Badly The Case of High Yield Bonds. Chris Kantos Newport Seminar June 2013

MAS108 Probability I

Applying Generalized Pareto Distribution to the Risk Management of Commerce Fire Insurance

Solving Linear Programs in Excel

Modelling operational risk in the insurance industry

Geostatistics Exploratory Analysis

Institute of Actuaries of India Subject CT3 Probability and Mathematical Statistics

THE USE OF STATISTICAL DISTRIBUTIONS TO MODEL CLAIMS IN MOTOR INSURANCE

LOGNORMAL MODEL FOR STOCK PRICES

( ) = 1 x. ! 2x = 2. The region where that joint density is positive is indicated with dotted lines in the graph below. y = x

Quantitative Methods for Finance

Modeling the Claim Duration of Income Protection Insurance Policyholders Using Parametric Mixture Models

The Best of Both Worlds:

A Software Tool for. Automatically Veried Operations on. Intervals and Probability Distributions. Daniel Berleant and Hang Cheng

Data Preparation and Statistical Displays

Supplement to Call Centers with Delay Information: Models and Insights

Module 4: Data Exploration

11 Linear and Quadratic Discriminant Analysis, Logistic Regression, and Partial Least Squares Regression

Overview of Monte Carlo Simulation, Probability Review and Introduction to Matlab

Continuous Random Variables

Homework 2. Page 154: Exercise Page 145: Exercise 8.3 Page 150: Exercise 8.9

Inference on the parameters of the Weibull distribution using records

Why Taking This Course? Course Introduction, Descriptive Statistics and Data Visualization. Learning Goals. GENOME 560, Spring 2012

Bootstrap Example and Sample Code

ACTUARIAL MODELING FOR INSURANCE CLAIM SEVERITY IN MOTOR COMPREHENSIVE POLICY USING INDUSTRIAL STATISTICAL DISTRIBUTIONS

Duration Analysis. Econometric Analysis. Dr. Keshab Bhattarai. April 4, Hull Univ. Business School

A Primer on Mathematical Statistics and Univariate Distributions; The Normal Distribution; The GLM with the Normal Distribution

Lecture 2 ESTIMATING THE SURVIVAL FUNCTION. One-sample nonparametric methods

Overview of Violations of the Basic Assumptions in the Classical Normal Linear Regression Model

1. A survey of a group s viewing habits over the last year revealed the following

7.1 The Hazard and Survival Functions

2DI36 Statistics. 2DI36 Part II (Chapter 7 of MR)

Package cpm. July 28, 2015

How To Price Garch

ATV - Lifetime Data Analysis

Parametric Survival Models

Advanced Topics in Statistical Process Control

Low-level concentrations of organic and

Lecture 8. Confidence intervals and the central limit theorem

Applied Reliability Page 1 APPLIED RELIABILITY. Techniques for Reliability Analysis

MATH4427 Notebook 2 Spring MATH4427 Notebook Definitions and Examples Performance Measures for Estimators...

UNIT I: RANDOM VARIABLES PART- A -TWO MARKS

Tests for Two Survival Curves Using Cox s Proportional Hazards Model

Petrel TIPS&TRICKS from SCM

Advances in Loss Data Analytics: What We Have Learned at ORX

DYNAMIC PROJECT MANAGEMENT WITH COST AND SCHEDULE RISK

Nonparametric adaptive age replacement with a one-cycle criterion

YASAIw.xla A modified version of an open source add in for Excel to provide additional functions for Monte Carlo simulation.

Finite Mathematics Using Microsoft Excel

NCSS Statistical Software Principal Components Regression. In ordinary least squares, the regression coefficients are estimated using the formula ( )

Transcription:

Operational Risk Modeling Analytics 01 NOV 2011 by Dinesh Chaudhary Pristine (www.edupristine.com) and Bionic Turtle have entered into a partnership to promote practical applications of the concepts related to risk analytics and modelling. Practical and hands on understanding of building excel based models related to operational and credit risk is necessary for any job related to risk management. For this purpose, we would be illustrating step by step model building techniques for risk management. Registrations for Operational Risk are OPEN. Operational Risk Modelling Analytics Fitting distribution to Operational Risk Data Maximum Likelihood Estimation (MLE) in Excel Hi, This is Dinesh from Pristine! I will discuss my favourite area risk modelling with you. Operational Risk Modeling Techniques Operational risk modelling uses Loss Data Analysis (internal or external), Scenario Analysis and data on Business Environment and Internal Controls. Loss data analysis and Scenario Analysis require fitting probability distribution to loss data or scenario data i.e. identifying which distribution best describes the empirical or expert judgement data. There are various methods of fitting distributions to data such as Moments Matching (estimating distribution parameters such that moments such as mean, variance of the data are matched) Quantile/Percentile Matching (estimating distribution parameters such that quantiles like 50 th Quantile i.e. Median, 99% percentile etc. are matched) Probability Weighted Moments Maximum Likelihood Estimation (MLE) Why Maximum Likelihood Estimators (MLE)? MLE method estimates the distribution parameter such that the joint likelihood of observing all empirical data points together gets maximized. Quite unlike moment matching and quantile matching, MLE makes use of all data points instead of only specific moments/quantiles. MLE also allows fitting distribution to truncated and censored data (common feature of operational risk data). Estimating MLE parameter for any distribution with given PDF and CDF functions Start with seed parameter values for the distribution to be fitted to the data Find the probability of observing each data point (using appropriate PDF function)

Assuming all data points are independent, joint probability of observing all data points together is the product of their probability density functions If two events A and B are independent Joint probability of A and B happening together is P(A) x P(B) Overall probability of observing all loss data amounts together would be all PDF multiplied together. We would like to maximize the joint likelihood to observe all data points. This is achieved either Maximizing P(A)*P(B) or Maximizing log(p(a)) + log(p(b)) or Minimizing [log(p(a)) + log(p(b))] Use an optimization algorithm (like Excel Solver) to maximize joint log-likelihood by changing parameter values Excel Modeling for Estimating Parameters for Gamma Distribution using MLE Let s take a small operational loss severity data set comprising of only 20 loss data points. Illustration shows estimation of parameters of Gamma distribution using MLE: Step-1: Seed values of parameters Gamma distribution has two parameters

Shape/Alpha which controls the shape of the distribution and impacts skewness and kurtosis of the distribution and Scale/Beta which controls the dispersion (variance) of the distribution For Gamma distribution, shape and scale parameters have to be positive, so we start with positive seed values. Step-2: PDF of each loss data point Usually, CDF and PDF functions in Excel have Dist suffixed to the distribution name. For instance, NORMDIST function gives CDF of normal distribution if cumulative argument is TRUE and gives PDF if cumulative argument is entered as FALSE. Similarly, GAMMADIST gives CDF (cumulative = TRUE ) and PDF (cumulative = FALSE) of Gamma distributed random variable. Step-3: Calculation of Joint Probability Joint probability of observing all data points together is calculated as product of PDF of individual data point. We can either Maximise product of all PDF or Maximize the sum of logs of individual PDF (maximising log likelihood) or Minimise the sum of logs of individual PDF (minimizing log likelihood). We decide to minimize the negative logarithm. Column C we calculate the logs of PDF and in C7 we take negative sum of all log PDF. We then use Excel solver to minimize cell C7 by changing parameter values. Step-4: Invoke Excel Solver to minimize negative log likelihood function Excel solver can be used to minimize the log likelihood function on (Cell C7), by changing parameter values (B3 and B4), subject to the constraints that parameters are positive (B3, B4 > = small positive value)

Practical considerations In practice and as allowed in regulatory guidelines, Banks may not collect data on small operational losses below a threshold (called truncation of loss data). MLE parameter estimates need to be corrected for this to avoid bias in estimates. Say if the loss collection threshold is USD 10000 Find conditional PDF, conditional on the fact that losses above 10000 are only being collected To find conditional PDF, we divide PDF by the cumulative probability that losses are above the threshold (as we are capturing only data points above the threshold) Equivalently we divide by (1 cumulative probability that losses are less than the threshold).

Notice that we are now minimizing cell E7, sum of conditional PDF. Conditional probability of each data point is calculated as PDF(Loss Amount)/(1-CDF(Loss threshold)) GAMMADIST(Loss amount, Param-1, Param-2, Cumulative = False)/ (1- GAMMADIST(Loss Threshold, Param-1, Param-2, Cumulative = True)). Another noteworthy point is that after considering loss threshold, shape parameter has changed. Therefore, not adjusting for loss threshold would lead to biased parameter estimates. Templates to download I have created a template for you, where the subheadings are given and you have to link the model to get the cash numbers! You can download the same from here. You can go through the case and fill in the yellow boxes. I also recommend that you try to create this structure on your own (so that you get a hang of what information is to be recorded). Also you can download this filled template and check, if the information you recorded, matches mine or not!

Next Steps Excel functions can be used to model operational loss data. Similar techniques can be used for other common distributions fitted to operational loss data, use: Weibull function for Weibull distribution Expondist function for exponential distribution Normdist function for lognormal distribution after taking Natural Log of all data points Also in the next few tutorials we would find the methodology of fitting distributions to data and the practical problems that we face in the same. By the way, we have launched the course on Operational Risk Modeling, which covers all these practical concepts through video lectures. If you want to join the same, you can join by clicking here. You may also refer to a video that David Harper published on Loss Distribution Approach (LDA) hereor you may watch below. To reach the author, you can send an email to dinesh.chaudhary@asymmetrix.co.in. The case has been drafted for discussion purpose. It has been written by Pristine (www.edupristine.com) and would be discussed by experts from Pristine & Bionic Turtle. There would be a step by step analysis and financial model building to come to a conclusion on the decision. Pristine is an authorized training provider for reputed organizations like CFA Institute (USA), PRMIA (USA), GARP (USA) and has provided trainings for reputed organizations like HSBC, Bank of America, JP Morgan, NUS, IIM Calcutta, etc.