Simple PD and LGD Estimation in MS Excel

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

Step by Step process to Estimate PD and LGD for Loan Portfolio in Microsoft

Excel

To estimate the Probability of Default (PD) and Loss Given Default (LGD) for a loan portfolio
in Microsoft Excel, we can use the “Logistic Regression” model for PD estimation and the
“Linear Regression” model for LGD estimation. These models are widely used in risk
management and can be implemented using Excel’s built-in functions and tools.
Data Required
1. Loan Data:
- Loan ID/ Customer ID
- Borrower Name
- Loan Amount
- Borrower Financial Metrics (e.g., income, debt-to-income ratio, credit score)
- Loan Characteristics (e.g., interest rate, term)

2. Default Data:
- Default Status (Make it binary i.e. 1 for default and 0 for no default)
3. Recovery Data:
- Recovery Amount (for defaulted loans, for a given period or till date, whatever is required)
Step-by-Step Process to be done in Excel
1. Logistic Regression for PD
Step 1: Prepare the Data
Create a worksheet named `Loan Data` with the following columns:
- `A`: Loan ID/Customer ID
- `B`: Borrower Name
- `C`: Loan Amount
- `D`: Borrower Financial Metrics (e.g., Credit Score)
- `E`: Loan Characteristics (e.g., Interest Rate)
- `F`: Default Status (1 for default, 0 for no default)

Step 2: Calculate the Logistic Regression Coefficients


- Use Excel’s “Data Analysis Tool pack” to run logistic regression (if you don’t have this add-
in, you can enable it from Excel options).
Step 3: Logistic Regression Model
The logistic regression formula is:
PD= (1/ (1+ e^ -(β0 + β1.X1+ β2. X2+ β3. X3-------+ βn. Xn)))
Where:
β0 is the intercept
β1, β2, β3….. βn are the coefficients for the predictor variables X 1, X2, X3… Xn

Calculate Logistic Regression Coefficients:

 Ensure you have the Data Analysis Toolpak enabled in Excel (File -> Options -> Add-Ins ->
Manage Excel Add-ins -> Go -> Check Data Analysis Toolpak).
 Use the Data Analysis Toolpak to perform logistic regression:
 Go to Data -> Data Analysis -> Regression.
 Set your dependent variable (Y Range) to the Default Status column.
 Set your independent variables (X Range) to the columns containing Borrower Financial
Metrics and Loan Characteristics.
 Click OK to run the regression analysis.
 Note the coefficients (𝛽0, 1, 𝛽2, etc.) from the output.

Step 4: Implement the Formula in Excel


1. Calculate the linear predictor:
- In cell `G2`, enter:
= β0+ β1*D2+ β2*E2
- Replace `BETA_0`, `BETA_1`, `BETA_2` with the actual coefficients obtained from the
regression analysis.

2. Calculate the PD:


- In cell `H2`, enter:
=1 / (1 + EXP(-G2))
- Drag the formula down to apply it to all rows in your data set.
2. Linear Regression for LGD
Step 1: Prepare the Data
Create another worksheet named `Recovery Data` with the following columns:
- `A`: Loan ID/Customer ID
- `B`: Recovery Amount
- `C`: Loan Amount
- `D`: Financial Metrics
- `E`: Loan Characteristics

Step 2: Calculate the LGD


- LGD is calculated as:
LGD = 1 – (Recovery Amount / Loan Amount)
1. In cell `F2`, enter:
=1 - (B2 / C2)
2. Drag the formula down to apply it to all rows.
Step 3: Linear Regression Model
- Use Excel’s “Data Analysis Toolpack” to run linear regression with LGD as the dependent
variable and financial metrics, loan characteristics as independent variables.

Step 4: Implement the Formula in Excel

1. Calculate the predicted LGD using the regression coefficients:


- In cell `G2`, enter:
=BETA_0 + BETA_1 * D2 + BETA_2 * E2
2. Drag the formula down to apply it to all rows.
Assumptions
1. Data Quality: The data provided is clean, with no missing values or outliers that could skew
the results.
2. Linear Relationships: For linear regression, assume that relationships between independent
variables and LGD are linear.
3. Logistic Relationships: For logistic regression, assume the relationship between
independent variables and the log-odds of the default is linear.

You might also like