SyntheticCreditMarket

The SyntheticCreditMarket class is a container for preprocessed credit market relationship data between financial institutions and borrowers. It organizes initial loan states and parameters for model initialization.

Core Functionality

The class handles:

1. Credit Relationship Data:

   • Bank-firm loan relationships
   • Bank-household loan relationships
   • Initial loan parameters

2. Loan Type Organization:

   • Long-term firm loans
   • Short-term firm loans
   • Consumer loans
   • Payday loans
   • Mortgage loans

3. Loan Parameter Processing:

   • Principal amounts
   • Interest rates
   • Installment calculations
   • Maturity periods

Key Attributes

• country_name: Country identifier for data collection
• year: Reference year for preprocessing
• longterm_loans: Preprocessed firm long-term loan data
• shortterm_loans: Preprocessed firm short-term loan data
• consumption_expansion_loans: Preprocessed consumer loan data
• payday_loans: Preprocessed payday loan data
• mortgage_loans: Preprocessed mortgage loan data

Factory Methods

The class provides a factory method create_from_agents that creates a SyntheticCreditMarket instance by:

1. Processing credit relationship data from various economic agents
2. Matching borrowers with corresponding banks
3. Calculating initial loan parameters
4. Organizing data into loan type categories

Implementation

Module for preprocessing synthetic credit market relationship data.

This module provides a framework for preprocessing and organizing credit relationship data between banks, firms, and households. Key preprocessing includes:

1. Credit Relationship Data:
2. Bank-firm loan relationships
3. Bank-household loan relationships

4. Initial loan parameters

5. Loan Type Organization:

6. Long-term firm loans
7. Short-term firm loans
8. Consumer loans
9. Payday loans

10. Mortgage loans

11. Loan Parameter Processing:

12. Principal amounts
13. Interest rates
14. Installment calculations
15. Maturity periods

Note

This module is NOT used for simulating credit market behavior. It only handles the preprocessing and organization of credit relationship data that will later be used to initialize behavioral models in the simulation package.

SyntheticCreditMarket dataclass

Container for preprocessed credit market relationship data.

This class organizes credit relationship data between financial institutions and borrowers, preprocessing initial loan states and parameters for model initialization. It does NOT implement credit market behavior - it only handles data preprocessing.

The preprocessed data is organized into five loan categories: 1. Long-term Loans: Initial firm long-term borrowing data 2. Short-term Loans: Initial firm short-term credit data 3. Consumer Loans: Initial household consumption borrowing data 4. Payday Loans: Initial household short-term credit data 5. Mortgage Loans: Initial household mortgage data

Each loan category contains: - Principal amounts - Interest rates - Installment schedules - Bank-borrower relationships

Note

This is a data container class. The actual credit market behavior (lending decisions, interest rate setting, etc.) is implemented in the simulation package, which uses this preprocessed data for initialization.

Attributes:

Name	Type	Description
country_name	str	Country identifier for data collection
year	int	Reference year for preprocessing
longterm_loans	LongtermLoans	Preprocessed firm long-term loan data
shortterm_loans	ShorttermLoans	Preprocessed firm short-term loan data
consumption_expansion_loans	ConsumptionExpansionLoans	Preprocessed consumer loan data
payday_loans	PaydayLoans	Preprocessed payday loan data
mortgage_loans	MortgageLoans	Preprocessed mortgage loan data

country_name: str instance-attribute

year: int instance-attribute

longterm_loans: LongtermLoans instance-attribute

shortterm_loans: ShorttermLoans instance-attribute

consumption_expansion_loans: ConsumptionExpansionLoans instance-attribute

payday_loans: PaydayLoans instance-attribute

mortgage_loans: MortgageLoans instance-attribute

__init__(country_name: str, year: int, longterm_loans: LongtermLoans, shortterm_loans: ShorttermLoans, consumption_expansion_loans: ConsumptionExpansionLoans, payday_loans: PaydayLoans, mortgage_loans: MortgageLoans) -> None

create_from_agents(firms: SyntheticFirms, population: SyntheticPopulation, banks: SyntheticBanks, zero_firm_debt: bool, firm_loan_maturity: int, hh_consumption_maturity: int, mortgage_maturity: int) -> SyntheticCreditMarket classmethod

Create a preprocessed credit market data container from agent data.

This method processes credit relationship data from various economic agents to prepare: 1. Firm loan data (long and short term) 2. Household loan data (consumer and payday) 3. Mortgage loan data

The preprocessing steps: 1. Extract loan data from firms and households 2. Match with corresponding bank data 3. Calculate initial loan parameters 4. Organize into loan type categories

Parameters:

Name	Type	Description	Default
firms	SyntheticFirms	Firm data container	required
population	SyntheticPopulation	Population data container	required
banks	SyntheticBanks	Bank data container	required
zero_firm_debt	bool	Whether to initialize with zero firm debt	required
firm_loan_maturity	int	Initial maturity for firm loans	required
hh_consumption_maturity	int	Initial maturity for consumer loans	required
mortgage_maturity	int	Initial maturity for mortgages	required

Returns:

Name	Type	Description
SyntheticCreditMarket	SyntheticCreditMarket	Container with preprocessed credit relationship data

Loan Data Classes

The credit market module includes several specialized classes for different types of loans:

LoanData

Base class for preprocessing loan-specific credit data. It provides:

• Principal amounts by bank-borrower pair
• Interest amounts by bank-borrower pair
• Installment amounts by bank-borrower pair

LongtermLoans

Container for preprocessed long-term firm loan data:

• Principal amounts from firm debt data
• Interest amounts using bank long-term rates
• Installment amounts based on maturity

ShorttermLoans

Container for preprocessed short-term firm loan data:

• Principal amounts from firm debt data
• Interest amounts using bank short-term rates
• Installment amounts based on maturity

ConsumptionExpansionLoans

Container for preprocessed consumer loan data:

• Principal amounts from household debt data
• Interest amounts using bank consumer rates
• Installment amounts based on maturity

PaydayLoans

Container for preprocessed payday loan data:

• Principal amounts from household data
• Interest amounts using bank payday rates
• Installment amounts based on maturity

MortgageLoans

Container for preprocessed mortgage loan data:

• Principal amounts from household mortgage data
• Interest amounts using bank mortgage rates
• Installment amounts based on maturity

Implementation

Module for preprocessing loan-specific credit relationship data.

This module provides dataclasses for organizing different types of loan data that will be used to initialize behavioral models. Key preprocessing includes:

1. Loan Parameter Processing:
2. Principal amount calculations
3. Interest rate applications

4. Installment computations

5. Bank-Borrower Relationships:

6. Firm-bank loan mappings
7. Household-bank loan mappings

8. Initial loan state organization

9. Loan Type Specialization:

10. Long-term firm loans
11. Short-term firm loans
12. Consumer loans
13. Payday loans
14. Mortgages

Note

This module is NOT used for simulating loan behavior. It only handles the preprocessing and organization of loan-specific data that will later be used to initialize behavioral models in the simulation package.

LoanData dataclass

Base class for preprocessing loan-specific credit data.

This class provides a framework for organizing loan parameters that will be used to initialize behavioral models. It is NOT used for simulating loan behavior - it only handles data preprocessing.

The preprocessed data includes: - Principal amounts by bank-borrower pair - Interest amounts by bank-borrower pair - Installment amounts by bank-borrower pair

Note

This is a data container class. The actual loan behavior (repayment, default, etc.) is implemented in the simulation package, which uses this preprocessed data for initialization.

Attributes:

Name	Type	Description
principal	ndarray	Initial loan principal amounts
interest	ndarray	Initial loan interest amounts
installments	ndarray	Initial loan installment amounts

principal: np.ndarray instance-attribute

interest: np.ndarray instance-attribute

installments: np.ndarray instance-attribute

__init__(principal: np.ndarray, interest: np.ndarray, installments: np.ndarray) -> None

stack()

Stack loan parameters for preprocessing.

Returns:

Type	Description
	np.ndarray: Stacked array of [principal, interest, installments]

LongtermLoans dataclass

Container for preprocessed long-term firm loan data.

This class organizes initial state data for long-term loans to firms. It processes: - Principal amounts from firm debt data - Interest amounts using bank long-term rates - Installment amounts based on maturity

Note

This is a data container class. The actual loan behavior is implemented in the simulation package.

__init__(principal: np.ndarray, interest: np.ndarray, installments: np.ndarray) -> None

from_agent_data(bank_data: pd.DataFrame, firm_data: pd.DataFrame, firm_loan_maturity: int = 60) -> LongtermLoans classmethod

Create a preprocessed long-term loan data container.

This method: 1. Extracts firm debt data 2. Matches with bank rate data 3. Calculates initial parameters

Parameters:

Name	Type	Description	Default
bank_data	DataFrame	Bank data with rates	required
firm_data	DataFrame	Firm data with debt	required
firm_loan_maturity	int	Initial maturity. Defaults to 60.	60

Returns:

Name	Type	Description
LongtermLoans	LongtermLoans	Container with preprocessed loan data

ShorttermLoans dataclass

Container for preprocessed short-term firm loan data.

This class organizes initial state data for short-term loans to firms. It processes: - Principal amounts from firm debt data - Interest amounts using bank short-term rates - Installment amounts based on maturity

Note

This is a data container class. The actual loan behavior is implemented in the simulation package.

__init__(principal: np.ndarray, interest: np.ndarray, installments: np.ndarray) -> None

from_agent_data(bank_data: pd.DataFrame, firm_data: pd.DataFrame, firm_loan_maturity: int = 60) -> ShorttermLoans classmethod

Create a preprocessed short-term loan data container.

This method: 1. Extracts firm debt data 2. Matches with bank rate data 3. Calculates initial parameters

Parameters:

Name	Type	Description	Default
bank_data	DataFrame	Bank data with rates	required
firm_data	DataFrame	Firm data with debt	required
firm_loan_maturity	int	Initial maturity. Defaults to 60.	60

Returns:

Name	Type	Description
ShorttermLoans	ShorttermLoans	Container with preprocessed loan data

ConsumptionExpansionLoans dataclass

Container for preprocessed consumer loan data.

This class organizes initial state data for household consumption loans. It processes: - Principal amounts from household debt data - Interest amounts using bank consumer rates - Installment amounts based on maturity

Note

This is a data container class. The actual loan behavior is implemented in the simulation package.

__init__(principal: np.ndarray, interest: np.ndarray, installments: np.ndarray) -> None

from_agent_data(bank_data: pd.DataFrame, household_data: pd.DataFrame, consumption_loan_maturity: int = 1) -> ConsumptionExpansionLoans classmethod

Create a preprocessed consumer loan data container.

This method: 1. Extracts household debt data 2. Matches with bank rate data 3. Calculates initial parameters

Parameters:

Name	Type	Description	Default
bank_data	DataFrame	Bank data with rates	required
household_data	DataFrame	Household data with debt	required
consumption_loan_maturity	int	Initial maturity. Defaults to 1.	1

Returns:

Name	Type	Description
ConsumptionExpansionLoans	ConsumptionExpansionLoans	Container with preprocessed loan data

PaydayLoans dataclass

Container for preprocessed payday loan data.

This class organizes initial state data for household payday loans. It processes: - Principal amounts from household data - Interest amounts using bank payday rates - Installment amounts based on maturity

Note

This is a data container class. The actual loan behavior is implemented in the simulation package.

__init__(principal: np.ndarray, interest: np.ndarray, installments: np.ndarray) -> None

from_agent_data(bank_data: pd.DataFrame, household_data: pd.DataFrame, payday_loan_maturity: int = 1) -> PaydayLoans classmethod

Create a preprocessed payday loan data container.

This method: 1. Extracts household data 2. Matches with bank rate data 3. Calculates initial parameters

Parameters:

Name	Type	Description	Default
bank_data	DataFrame	Bank data with rates	required
household_data	DataFrame	Household data	required
payday_loan_maturity	int	Initial maturity. Defaults to 1.	1

Returns:

Name	Type	Description
PaydayLoans	PaydayLoans	Container with preprocessed loan data

MortgageLoans dataclass

Container for preprocessed mortgage loan data.

This class organizes initial state data for household mortgages. It processes: - Principal amounts from household mortgage data - Interest amounts using bank mortgage rates - Installment amounts based on maturity

Note

This is a data container class. The actual loan behavior is implemented in the simulation package.

__init__(principal: np.ndarray, interest: np.ndarray, installments: np.ndarray) -> None

from_agent_data(bank_data: pd.DataFrame, household_data: pd.DataFrame, mortgage_maturity: int = 120) classmethod

Create a preprocessed mortgage loan data container.

This method: 1. Extracts household mortgage data 2. Matches with bank rate data 3. Calculates initial parameters

Parameters:

Name	Type	Description	Default
bank_data	DataFrame	Bank data with rates	required
household_data	DataFrame	Household data with mortgages	required
mortgage_maturity	int	Initial maturity. Defaults to 120.	120

Returns:

Name	Type	Description
MortgageLoans		Container with preprocessed loan data