Palisade
Live Chat
Palisade Risk Conference
Probabilistic Techniques for
Decision Making in Banking and Finance
June 23, 2015 Toronto
2015
2015
  PALISADE

Palisade Risk Conference

Probabilistic Techniques for

Decision Making in Banking and Finance


June 23rd, 2015 • Toronto

Probabilistic Techniques for Decision Making in Banking and Finance

Join Industry Leaders and Software Experts

Palisade invites you to an intensive one-day conference in Toronto, specializing in banking and finance risk. We’ll take an in-depth look at using risk analysis software and probabilistic techniques to define uncertainty in banking and finance. A sampling of topics includes:

  • Quantifing risk of default for a portfolio of bank loans, using a stochastic risk-adjusted return on capital (RAROC)
  • Calculating value at risk (VaR) for an investment portfolio
  • Applying neural networks technology to credit scoring banking, employing “Live Prediction” capabilities
  • Using a statistical application to create a multiple regression model for prediction of corporate loans

This is a “must attend” learning, information sharing, and networking event for professionals dealing with risk. We’ll try to cover as much as possible in one day, with real-world case studies and software training covering innovative approaches to managing risk and uncertainty. We’ll explore some of the sophisticated methods within @RISK and the DecisionTools Suite that make risk modelling and analysis powerful and accessible.

This event it is free to attend and promises to be an invaluable opportunity for broadening any attendee's knowledge of risk modelling, and networking with decision-making professionals in the financial and banking industries.

Testimonials from Previous Conferences

“We saw and discussed methods that directly apply to our problems, and learned to build upon them to come up with solutions.”

Michael Watson, PMP, Senior Staff Integrated Planning, Lockheed Martin

“Excellent conference. If there are numbers that vary, you should be here.”

George Chiang, Independent Cost Analyst, Boeing Company

“Very good conference with good usage of time.”

Mark Mendonca, Program Risk Management, Ernst & Young

“Excellent conference. Very informative. Great topics.”

Laurie Rutherford, Director, Enterprise Risk Management at CenterPoint Energy

Learn from the Experts

The Palisade Risk Syposium will offer two types of sessions to maximize your experience.

Case Studies:

Hear users in banking and finance describe how they apply @RISK and DecisionTools software to their problems.

Software Training:

Here’s your chance to get exposure to the latest in risk and decision analytics software and techniques. Get more from @RISK and each of the products in Palisade’s DecisionTools Suite. Sessions are taught by Palisade’s expert consultants and trainers. Learn direct from the source!

Networking Opportunities – Learn from Each Other

The Palisade Risk Conference will offer plenty of opportunities to mingle and share ideas with other delegates, presenters, and Palisade staff. Lunches, breaks and networking receptions are all included with your registration.

Join a Global Group

The 2015 Palisade Risk Conference is part of a global, annual series of meetings where professionals share ideas and advance the practice of risk and decision analysis. Events in North America include Chicago, New York, San Francisco, Calgary, and our two-day Annual Risk Conference in New Orleans.

See our global schedule of 2015 Conferences here.

Schedule

  Networking Sessions      
  Software Presentations      
   Industry Case Studies

Tuesday 23 June 2015

Schedule is subject to change without notice.

 

8:00 - 9:00
Registration
9:00 - 9:15
WELCOME

Welcome and Introduction
Jaime Weisberg
Palisade Corporation

9:15 - 10:15
SOFTWARE TRAINING

Introduction to @RISK
Dean Cardno
BC Hydro

10:15 - 11:00
SOFTWARE TRAINING

Value at Risk (Var) Model for an Investment Portfolio
Gustavo Vinueza
Palisade Corporation

11:00 - 11:15
Break
11:15 - 12:00
SOFTWARE TRAINING

Quantifying RAROC and Probabilistic Losses on a Credit Portfolio
Gustavo Vinueza
Palisade Corporation

12:00 - 1:00
Lunch
1:00 - 2:00
SOFTWARE TRAINING

Selecting the Right Distribtion
Dean Cardno
BC Hydro

2:00 - 2:30
CASE STUDY

“Downside” Risk: What? Why? How?
Irina Dorogan
Cougar Global Investments

3:00 - 3:30
Break
3:30 - 4:30
SOFTWARE TRAINING

Risk Analysis for Corporate Loans
Gustavo Vinueza
Palisade Corporation

4:30 - 4:45
Q&A
4:45 - 5:00
Closing Remarks


Jaime Weisberg
Palisade Corporation

5:00
Reception

Presenters

  Dean Cardno

Consultant
BC Hydro

Dean Cardno qualified as a Chartered Accountant with Coopers & Lybrand (now PwC) in 1987. After leaving public practice, Dean worked with companies involved in mining, heavy civil construction, and real estate development. For the past 15 years Dean has been a consultant in project evaluation, development, and financing, with an emphasis on economic valuation and project risk identification and management. Dean concentrates his practice in the energy sector, primarily natural gas transmission and distribution, and electricity generation and transmission.

  Irina Dorogan

Senior Research Analyst
Cougar Global Investments

Irina is a Senior Research Analyst for Cougar Global Investments. She has been with the company for almost 5 years and is responsible for the investment research, including the researching of the Global ETF and Index Universe. She operates the firm’s asset allocation models and contributes to asset allocation decisions. She has a degree in Economics from the Academy of Studies of Moldova and holds the Chartered Investment Manager designation. She is also a registered Portfolio Manager-Advising Representative.

  Gustavo Vinueza

Consultant and Trainer
Palisade Corporation

Gustavo Vinueza is a Systems Engineer from University in Cuenca, Ecuador. He also earned an MBA from Torcuato Di Tella University in Argentina and a MS in Finance from Adolfo Ibáñez University in Chile. His main topics of interest include financial and operational modeling, including scientific and academic research into business practice as well as data mining relative matters. He has 16 years of experience and he’s been a consultant for companies in several industries: finance & banking, telecommunications, insurance and IT related services.

His experience includes managing project portfolios both operational and IT related, cost reduction programs, public purchases bidding, operational controls, capacity analysis and audit processes and software development projects, besides technological infrastructure implementation. He has also earned diplomas in Project Management, Usability, Business Process Management and Business Analytics.

Abstracts

Download zip file of presentations from the Palisade Risk Conference in Toronto
(Individual presentations are also linked below.)

  Introduction to @RISK

Dean Cardno
BC Hydro

» Download the presentation

This introduction to @RISK will walk you through a risk analysis using various example models. Key features of @RISK will be highlighted. You will experience the intuitive interface of @RISK as you define distributions, correlations, and other model components. During simulation you will be able to see all charts, thumbnails, and reports update in real time. View results with a variety of graphing and reporting options. There’s so much to see, we’ll cover as much as time permits.

  Value at Risk (Var) Model for an Investment Portfolio

Gustavo Vinueza
Palisade Corporation

The VaR of a portfolio at a future point in time is usually considered as the loss in value of a portfolio at a certain point in time given a certain confidence level. Thus, considering the price change of a portfolio of financial assets and their correlation, it is possible to simulate the VaR of the portfolio at different levels of confidence intervals. This methodology performed with Monte Carlo simulation allows overcoming the inherent disadvantages of traditional parametric methods of calculating VaR; i.e., the possibility of incorporating nonlinear assets in the portfolio and the possibility of considering non-normal returns.

The objective of the model consists in determining the value at risk (VaR) of the portfolio under conditions of asset correlations. Moreover, the model also contains an automatically updated feature in order to forecast the predicted value of the portfolio versus its actual performance. This allows the establishment of a daily management system for subsequently controlling stress tests and systems for assembling and self -calibrating the model thru back testing.

Through a forecasting methodology via time series, this model stochastically predicts correlated prices of the instruments that make up the portfolio. The time series methods and Monte Carlo simulation work together as complementary methodologies in order to forecast the value of the investment portfolio and therefore the calculation of VaR.

  Quantifying RAROC and Probabilistic Losses on a Credit Portfolio

Gustavo Vinueza
Palisade Corporation

The model shown here quantifies the risk of default for a portfolio of bank loans based on a probabilistic Monte Carlo simulation method. By individually calculating probabilistic losses we can calculate the sum of the total portfolio, the total amount of expected losses of the portfolio and the amount of unexpected losses at a certain confidence level. Uncertainties are input as default probabilities which are calculated using a methodology of historical information. This in turn is based on delinquency information provided. Based on this, transition matrices are calculated to evaluate portfolio quality along time.

Therefore, the model is able to calculate the value at risk and equity reserve to be allocated to loan portfolios based on their expected and unexpected losses by a methodology suggested by Basel II.

The model for calculating the loss corresponds to the model called System Internal Credit Risk Rating. This is a method that aims to calculate, as output variable, the expected and unexpected losses (at a confidence level) of the bank's loan portfolio.

The three components of the multiplicative model to estimate loss are: Exposure at default (EAD), which consists of the principal balance at the time the portfolio is being evaluated; Loss Given Default (LGD) which is the ratio of the uncovered operation that defaulted and Probability of default (PD) generated as a binomial distribution with two alternative stochastic probabilities: 0 (no default) and 1 (default), where the factor of default (P) is calculated according to transition matrices.

This model also allows for profitability analysis by various factors according to bank’s convenience; either by region, product, or any other categorization. This is performed through a stochastic RAROC (Risk Adjusted Return on Capital) by simulating the eventual behavior of losses and net interest margins based on the portfolio profile. In turn, profitability analysis allows to strategic and tactic allocation of resources, pricing strategies, category evaluation.

RAROC ("Risk Adjusted Return On Capital") is a common methods used by financial institutions and insurers to measure the profitability of its loan portfolio and exposure limit of its clients and creditors taking into account a certain probability of loss. It not only measures profitability, but it also ponders the level of risk assumed. The RAROC method was originally designed back in the 70s by the American bank Bankers Trust and since then is used as a measure of risk management and how cost benefit analysis of each business unit. Therefore, using the RAROC method (or similar systems developed by the entities themselves) the bank is able to measure the credit risk while maximizing return.

  Selecting the Right Distribtion

Dean Cardno
BC Hydro

» Download the presentation

This session covers the choice of the appropriate distribution in @RISK. A variety of approaches are presented and compared, including pragmatic, theoretical and data-driven methods. The use of distributions to treat a variety of risk modeling situations is discussed, and some more advanced features such as bootstrapping, batch fitting, and the live fit function, will be covered as time permits.

  “Downside” Risk: What? Why? How?

Irina Dorogan
Cougar Global Investments

Irina will talk about the importance of managing risk (in particular, the downside risk) and the reasons why downside risk may be a better risk measure than standard deviation. She will briefly talk about Cougar Global investment process, which has a strict discipline of managing downside risk and will demonstrate how the firm applies Palisade software in its decision making process.

  Risk Analysis for Corporate Loans

Gustavo Vinueza
Palisade Corporation

This model compares two approaches for corporate credit risk analysis: Neural networks and traditional multiple regression via Altmann’s Z-scoring.

Neural networks are capable of learning complex relationships in data. By mimicking the functions of the human brain, they can discern patterns in the data and extrapolate predictions when they provide new data. In this model the fundamental concepts on which neural networks theory is based are introduced, specifying the generalities of categorical and numerical networks. In this model, Palisade’s NeuralTools is used to solve the classic problem of "credit scoring" banking, which allows prediction decide on the granting of a corporate credit facility based on scoring data into different categories of applicants for a loan. This score is in turn based on information from categorical and numerical variables that describe the financial health of borrowers. Information on financial indicators of customers versus quality ("rating") of them will be compared.

Prediction software uses advanced NeuralTools neural networks to reveal the structure of the statistical analysis data set. In this case, the software will learn about the structure of financial indicators that predict the behavior of a loan portfolio quality. The network “learns” the existing data and then uses the knowledge for discovering patterns to make predictions from new raw data. All this happens "behind the scenes" so that the user does not have to be involved in setting the parameters for each algorithm. The software can perform both numeric and categorical predictions.

This software also has a very powerful functionality called "Live Prediction". When new data are entered into an analysis, LivePrediction predictions update instantly. This will save time and will be particularly valuable when combined with Palisade's Evolver ™. By using Evolver, the bank may define a desired target prediction – e.g., timely payments on a bank loan with a certainty of 90%. The optimizer can vary input variables as the loan amount as NeuralTools continually updates its predictions of repayment. Users can adjust the optimizer to continually change the model until the desired prediction is reached.

The second method of complementary prediction refers to the multiple regression model called Altmann’s modified Z-Score. This model will be validated and calibrated versus a traditional scoring model for multiple determinants (model z) to determine ownership and convenience of being able to use in the context of corporate loans. There is no guarantee that it can generate good solutions. This will be performed using Palisade’s StatTools, which is a statistical application that, among other models, allows developing applications for multiple regression and logistic regression.

Toronto

Information Coming Soon

Palisade Corporation
798 Cascadilla Street
Ithaca, NY 14850-3239
800 432 RISK (US/Can)
+1 607 277 8000
+1 607 277 8001 fax
sales@palisade.com
Palisade EMEA & India
+44 1895 425050
sales@palisade-europe.com
Palisade Asia-Pacific
+61 2 9252 5922
salesAP@palisade.com
Palisade アジア・
パシフィック東京事務所
+81 3 5456 5287 tel
sales.jp@palisade.com
www.palisade.com/jp/
Palisade Latinoamérica
+1 607 277 8000 x318
+54-1152528795  Argentina
+56-25813492     Chile
+507-8365675     Panamá
+52 55 5350 2852     México
+511-7086781     Perú
+57-15085187     Colombia
servicioalcliente@palisade.com
ventas@palisade.com
www.palisade-lta.com
Palisade Brasil
+55 (21) 3958 1443
+1 607 277 8000 x318 tel
vendas@palisade.com
www.palisade-br.com