Level 26
Level 28

#### 95 words 0 ignored

Ready to learn
Ready to review

## Ignore words

Check the boxes below to ignore/unignore words, then click save at the bottom. Ignored words will never appear in any learning session.

**Ignore?**

Credit risk

possibility that borrower will not meet the scheduled repayments and default on their loan

lender

who holds the primary risk

risk

includes lost principal and interest, disruption to cash flows, and increased collection costs

collection costs

any cost associated with recovering debt on which the borrower has defaulted on his obligation to pay

accounts receivable

assets, what is owed to you

deferred revenue

cash in advance, money deferred to a future period where you actually earn it

Assets

Accounts receivable, what others owe to you

amortization

payment of an obligation in a series of installments or transfers

depreciation

decrease in value of an asset due to obsolescence or use

Income Statement

Record cash dividends from the investee's retained earnings. Do not recognize stock dividends (Memo entry only).

Balance Sheet

statements of assets and liabilities

Liabilities

amounts owed; the legal claims against a business or household by non-owners; the sources of funds for financial intermediaries

Equity

Stake of Ownership

cash flow statement

reconciliation between income statement and balance sheet

accounts payable

an accounting entry that represents an obligation to pay off a short term debt to creditors

fair value

what is the current market value

historical value

what you paid

Risk Management

The process of making and implementing decisions that will minimize the adverse effects of accidental losses on an organization

Risk identification

Risk management involves three major undertaking:

Components of Risk Management

Risk identification: Identify and Inventory assets, Classify and prioritize assets, and Identify and prioritize threats.

Asset Identification and Inventory

Iterative process; begins with identification of assets, including all elements of an organization's system (people, procedures, data and information, software, hardware, networking)

People, Procedures, and Data Asset Identification

Human resources, documentation, and data information assets are more difficult to identify

Hardware, Software, and Network Asset Identification

What information attributes to track depends on:

Data Classification and Management

Variety of classification schemes used by corporate and military organizations

Classifying and Prioritizing Information Assets

Many organizations have data classification schemes (e.g., confidential, internal, public data)

Information Asset Valuation

Questions help develop criteria for asset valuation

Identifying and Prioritizing Threats

Realistic threats need investigation; unimportant threats are set aside

Vulnerability Identification

Specific avenues threat agents can exploit to attack an information asset are called vulnerabilities

Risk Assessment

Risk assessment evaluates the relative risk for each vulnerability

Likelihood

The probability that a specific vulnerability will be the object of a successful attack

Risk is

Risk Identification Estimate Factors

The resulting ranked list of risk ratings for the three vulnerabilities is as follows:

Asset B has a value of 100 and has two vulnerabilities: vulnerability #2 has a likelihood of 0.5 with a current control that addresses 50% of its risk; vulnerability # 3 has a likelihoo…

Identify Possible Controls

For each threat and associated vulnerabilities that have residual risk, create preliminary list of control ideas

Documenting the Results of Risk Assessment

Final summary comprised in ranked vulnerability risk worksheet

Risk Control Strategies-

Once ranked vulnerability risk worksheet complete, must choose one of five strategies to control each risk:

Defend

Attempts to prevent exploitation of the vulnerability

Transfer

Control approach that attempts to shift risk to other assets, processes, or organizations

Mitigate

Attempts to reduce impact of vulnerability exploitation through planning and preparation

Accept

Doing nothing to protect a vulnerability and accepting the outcome of its exploitation

Terminate

Directs the organization to avoid those business activities that introduce uncontrollable risks

Selecting a Risk Control Strategy

Level of threat and value of asset play major role in selection of strategy

Cost Benefit Analysis (CBA)

Feasibility Studies Cost Benefit Analysis (CBA)

ROSI: Risk Assessment TIPS

Get a hold on what you have

Evaluation, Assessment, and Maintenance of Risk Controls

Selection and implementation of control strategy is not end of process

Quantitative versus Qualitative Risk Control Practices

Performing the previous steps using actual values or estimates is known as quantitative assessment

Benchmarking and Best Practices

An alternative approach to risk management

Recommended Risk Control Practices

Convince budget authorities to spend up to value of asset to protect from identified threat

Main goal of credit assessment process

To approve acceptable loan applications, reject clients that will probably default in the future, and set up loan pricing so that credit losses are covered by collected credit margins

Expected Loss

Expected Loss = PD*EAD*LGD

How are the AR, CAP, and AUC related?

The AR (aka Gini's Coefficient) is obtained from the CAP and AUC.

Accuracy Ratio (AR)

the probability of good discrimination minus the probability of wrong discrimination; measures the quality of the rating system measured

Area Under Curve (AUC)

the probability of good discrimination plus one half the probability of no discrimination; measures the ratio of good decisions from all attempts

Cumulative Accuracy Profile (CAP)

an arbitrary rating model that produces a rating score (high rating score -> low default probability)

Receiver Operating Characteristic Curve (ROC)

illustrates distributions of rating scores for defaulting and non-defaulting debtors

KS Statistic

the maximum distance of the ROC from the diagonal multiplied by √2

Confusion Matrix

splits the totoal number of applications into actual goods and bads; the actual goods into those that are predicted as good (approved), and those that are predicted as bad (rejected), and the actual bads …

Type 1 and Type 2 Errors in Confusion Matrix

Type 1 Error: When hypothesis is true (and borrower is good), but the system predicts otherwise (that the borrower is bad)

Hosmer-Lemeshow Test

calculates a weighted sum of normalized squared differences between the forecast probabilities and the realized default rates

Binomial Test

only for one grade, one-sided, with forecast probability of default PDs; test overestimates the significance of deviations in the realized default rate from the forecast default rate, so if there is a positive correlation, …

Key Financial Ratios

Operating Performance (ROE, ROA, Operating Margin, Net Profit Margin, Effective tax rate, S…

Through the Cycle Rating (TTC) vs. Point in Time (PIT) Rating

PIT - the rating expresses the actual expected PD conditional on current economic conditions

Internal Analytics Raings

Asset based - emphasis is put on valuation of borrower's assets; unsecured general corporate lending or product lending - emphasis is put on the borrower's ability to generate future cash flows

Examples of Econometric Models

classification trees, linear and multiple discriminant analysis, linear, logit and probit regression

Shadow rating

used when default data are rare, but there are external ratings from respected rating agencies; try to mimic analytical (external) rating systems; based on a database of rating decisions of an external rating agency ra…

Neural networks; rule-based/expert systems; structural models

Neural networks - try to mimic functioning of human brain

Altman's Z-score

a linear scoring function assigns to a company a linear combination of financial ratios and other explanatory variables in order to discriminate between good and bad borrowers

Logit vs. Probit

Tails are heavier for logit; logit is numerically more efficient; logit provides log odds

Weight of Evidence (WoE)

can be used to replace categorical variables with continuous variables; if positive, it increases log odds, negative decreases log odds

Information Value (IV)

measures an overall discriminatory power of the variable; always a positive number

Survival Analysis

Allows us to estimate probability of default in different time horizons at the same time; graphs illustrate that profiles differ, not sufficient to only focus on one year PD, especially if you want to esti…

Hazard Function

gives the rate at which objects that have survived until the time t, exit right right at time t; its derivative is approximately the probability of exit in the time interval, provided the objec…

What is the difference between survival analysis and logistical regression analysis ?

Logistic regression analysis does not capture well the time of the default dynamics, Sometimes we need to predict default in a short time horizon, in other situations, in a long time horizon. To model…

Survival Function vs. Hazard Function

A survival function gives the probability of survival until time t, whereas the hazard function gives the probability of exit in the time interval, provided that the object is still alive at time t.

Markov Chain Models

An alternative way to capture default dynamics; allows to estimate PDs for longer time horizons

Markowitz Portfolio Theory

Models risk vs. expected return; there is a diversification effect if the correlation between two assets with similar return is less than 1

Credit VaR (Value at Risk)

Credit loss of a credit portfolio can be measured in different ways:

Credit VaR Models

CreditMetrics, CreditRisk+, CreditPortfolio View, KMV portfolio Manager, Vasicek Model - Basel II

CreditMetrics

Based on ratings that are assumed to determine the values of the individual debt instruments; rating migration correlations modeled as asset correlations; asset return decomposed into a systematic and idiosyncratic parts; published by JP Morgan

CreditMetrics Bond Valuation

Simulated forward values require calculation of implied forward discount rates for individual ratings

CreditMetrics Recovery Rates

models the recovery rates as deterministic or independent on the rate of defaults but a number of studies show a negative correlation

Merton's Structural Model (Credit Migration)

Default occurs If assets < debt

Merton's Structural Model

The rating or its change is determined by the distance to the default threshold or its change

Default Rate and Recovery Rate Correlations

There is a negative correlation between default rate and recovery rate (negative correlation between PD and LGD)

CreditRisk+

Analytic "actuarial" approach; exposures are adjusted by fixed LGDs; discrete treatment; allows a scale of ratings and PDs - simulating overall mean number of defaults with a Gamma distribution; portfolio can be divided into indepen…

Advantage of CreditRisk+ over CreditMetrics

Calculations can be done analytically, without the Monte Carlo simulation

CreditPortfolio View

Ties rates of default to macroeconomic factors (only difference from CreditMetrics); created by Wilson and McKinsey

KMV Portfolio Manager

Idea of KMV is to use information from stock market in order to deduce information we cannot see - market value of assets (rather than book value)

KMV Portfolio Manager Relationship

There is a 1-to-1 relationship between stock price and its volatility and (latent) asset value and its volatility

Overview of KMV Model

based on Equity data to determine initial asset values and volatilities

PD Callibration

the distance to default is calibrated to PDs based on historical default observations

Why use risk neutral probabilities?

The calibrated PDs are historical, but we need risk neutral in order to value the loans because it is more consistent to use the risk-free rate.

Comparison of 3 Models

Different models can give different results due to the logic of the model as well as different inputs required by the models.

Vasicek's Model

Mathematical model describing the evolution of interest rates; the instantaneous interest rate follows the stochastic differential equation; if you multiply UDR (unexpected default rate) by a constant LGD and EAD, you get a "simple" esti…

Basel II Problems

Procyclicality: capital requirement becomes high in bad times as PD goes up; causes banks to be much more willing to provide new loans at good times than bad times -> deepens crisis during bad times

Basel III Improvements

higher capital requirements