Descriptive Analytics Report

Contents
Introduction.................................................................................................................................................1
1.1 Business Background...................................................................................................................1
1.2 Goals/Objective...........................................................................................................................2
2. Methodology:..........................................................................................................................................3
2.2 Measurements...................................................................................................................................4
3. Data Visualizations and Summary Statistics.............................................................................................4
3.1 Results/Findings.................................................................................................................................4
3.2 Discussion........................................................................................................................................13
4.0 Conclusion...........................................................................................................................................15
5.0 Recommendations...............................................................................................................................15
References:................................................................................................................................................17

Introduction

1.1 Business Background

The dataset from a German bank provides insights into how customer demographics (age, job,
sex) and financial data (savings, checking accounts, credit amount, loan duration) affect credit
amounts and loan eligibility.

Overview of the Dataset:

Key variables include:

Demographics: Age, sex, job (assesses socio-economic status)

Financials: Savings, checking accounts, credit amounts, loan duration (reflects financial health
and eligibility)
Loan Purpose: Education, business, car purchase (reveals borrower’s financial needs)

Relevance of the Dataset:

Understanding the factors influencing loan amounts is vital for effective credit risk management.
By analyzing demographic and financial data, banks can improve loan approval processes,
predict creditworthiness, and minimize risks.

Context in the Financial Industry:

Advanced credit scoring models use both demographic and financial data to predict defaults and
determine loan terms. Research by (Chang et al., 2024; Dansana et al., 2023) highlights the role
of factors like job type and education in determining loan eligibility. These models help refine
bank lending strategies and improve decision-making.

Implications for the Dataset:

The dataset allows testing these findings within the context of German banking, improving credit
models, and identifying key customer segments. This supports better financial decisions, risk
management, and portfolio optimization.

1.2 Goals/Objective
1.3

The primary goal of this analysis is to investigate the connections between credit amount and
client attributes such as age, sex, employment, housing, checking and savings accounts, loan
duration, and purpose. The objectives are to find significant credit choice trends and offer helpful
information.

Specific Goals:

Identify Key Determinants of Credit Amount: Investigate which factors, like job type or
housing status, have the strongest influence on credit amounts.

Analyze Demographic Correlations: Understand how demographic factors, such as age and
sex, impact credit allocation, and if any groups are disproportionately affected.
Examine Financial Indicators: Explore the relationship between saving/checking accounts,
loan purpose, and credit amounts to improve loan approval processes.

Assess Loan Duration Impact: Evaluate if longer loan durations lead to higher or lower credit
amounts, aiding in loan optimization.

Support Business Decision-Making: Provide insights that help refine the bank's credit scoring
models and loan products, improving risk management and customer segmentation.

These objectives aim to refine credit decision-making, enhance financial strategies, and ensure
better alignment between customer needs and the bank’s offerings.

2. Methodology:
Data pre-processing ensures the dataset is accurate and ready for analysis. The following steps
were applied:

 Handling Missing Values:

o Numeric Columns: Missing values in variables like Credit Amount, Age, and
Duration were imputed using the median to avoid bias from outliers.
o Categorical Columns: Missing values in categorical variables (e.g., Saving
Accounts, Checking Accounts) were imputed using the mode, preserving the
natural distribution of categories.
 Data Cleaning:
o Duplicate Removal: Duplicate records were removed to ensure data accuracy.
o Standardizing Entries: Inconsistent entries (e.g., "male" vs. "Male") were
standardized for consistency.
 Transformation Techniques:
o Scaling: Continuous variables (Credit Amount, Duration) were scaled to ensure
uniform contribution to the analysis.
o Encoding: Categorical variables like Sex and Job were encoded into factors for
statistical analysis.
 Justification:
Imputation with the median and mode preserves data integrity. Scaling and encoding
 ensure that all variables contribute equally without bias, and data cleaning enhances
accuracy.

The dataset includes 1000 records and 10 variables, with the sample size remaining unchanged
after pre-processing.

2.2 Measurements

The analysis utilized the following measurements:

Credit Amount: Total credit held by the customer.

Age: Customer's age.

Sex: Customer's gender (male or female).

Job: Job type, categorized numerically.

Housing: Housing status (own, rent, free).

Saving Accounts: Customer’s saving account status.

Checking Accounts: Customer’s checking account status.

Purpose: The purpose of the credit (e.g., business, car purchase).

Duration: Loan duration in months.

3. Data Visualizations and Summary Statistics

3.1 Results/Findings
The following visualizations and statistical summaries were generated to analyze the relationship
between variables and the Credit Amount variable:

Box Plot for Sex:

 Figure 1: The box plot shows similar median credit amounts for both genders, but males have a wider
range with more high-end outliers

Scatter Plot for Duration vs. Credit Amount:

 Figure 2: The scatter plot shows that although no clear trend emerges, there are instances of
higher credit amounts for longer loan durations.

Histogram for Housing:

Figure 3: The histogram reveals that homeowners typically receive higher credit amounts than renters or
those in free housing.

Bar Chart for Saving Accounts:

Figure 4: The bar chart shows that individuals with 'little' savings account for a large proportion of
the total credit amount.

Box Plot for Job:

Figure 5: The box plot reveals that individuals in mid-level jobs receive higher credit amounts than those
in low or high-level jobs.

Box Plot for Age Group:

Figure 6: The box plot shows a slight increase in credit amounts for older age groups (56-65, 66-
75) compared to younger ones.

Box Plot for Checking Accounts:

 Figure 7: The box plot indicates that customers with 'rich' checking accounts generally receive higher
credit amounts than those with 'little' or 'moderate' checking accounts.

Bar Chart for Purpose

Figure 8:
3.2 Discussion
Gender and Credit Amount:

As seen in Figure 1, men generally receive higher and more varied credit amounts compared to
women. While the median credit amounts for both genders are similar (2,500 USD for males and
2,000 USD for females), males show a wider range, with outliers reaching up to 15,000 USD
compared to 5,000 USD for females. This suggests that gender bias could exist in lending
practices, with males potentially benefiting from higher credit amounts due to differences in
credit scoring or loan approval processes (Fernandez Vidal & Caire, 2024; Karl, 2025).

Loan Duration and Credit Amount:

The data in Figure 2 reveals a positive correlation between loan duration and credit amount.
Customers with longer loan durations (36+ months) generally receive higher credit amounts.
However, some individuals with shorter loan terms still obtain more significant credit amounts,
likely due to higher income levels or strong credit histories (Nichols, 2024). This indicates that
income and creditworthiness are essential in determining loan amounts, aside from the loan
duration.

Housing and Credit Amount:

As shown in Figure 3, homeowners tend to receive higher credit amounts, with about 50%
receiving between 3,000 and 5,000 USD and some exceeding 10,000 USD. Renters and
individuals with free housing generally have lower credit limits. This aligns with the notion that
homeownership is closely associated with financial stability, increasing eligibility for more
significant credit amounts (Federal Reserve Bank of San Francisco, 2022).

Saving Accounts and Credit Amount:

In Figure 4, the distribution of credit amounts based on saving accounts shows that individuals
with “little” savings receive most of the total credit, accounting for 90%. Those with higher
savings tend to obtain smaller amounts, highlighting how financial inclusion efforts may
prioritize lending to those with fewer assets, providing them with more significant loans
(Greystar, 2025).

Job Type and Credit Amount:

As depicted in Figure 5, individuals in higher job categories (Job 3) generally receive more
significant credit amounts than those in lower job categories (Job 0). This reflects how job
stability is crucial in determining credit access, as financial stability linked to job type impacts
loan eligibility (Crine, 2024).

Age Group and Credit Amount:

As shown in Figure 6, older individuals (56-75 years) generally receive more significant credit
amounts, likely due to their longer credit histories and higher financial stability. This is
consistent with Schwahn's (2025) findings, suggesting that older age groups typically have better
credit scores and higher credit eligibility due to more established financial histories.

Checking Accounts and Credit Amount:

As indicated in Figure 7, individuals with “rich” checking accounts tend to receive the highest
credit amounts, with some outliers reaching up to 15,000 USD. Those with “little” or “moderate"
checking accounts typically have lower credit amounts. This underscores the importance of
checking account balances in determining credit eligibility, with higher balances indicating more
excellent financial stability (Huang et al., 2024).

Purpose vs. Credit Amount:

As shown in Figure 8, analyzing credit amounts for loan purposes reveals notable trends.
Business loans, making up 10% of total loans, typically have higher credit amounts, reaching up
to USD 15,000. Auto loans, accounting for 40%, typically range between USD 2,000 and 7,000,
while education loans (25% of total loans) usually fall below USD 5,000. According to Dori
Zinn (2024), lenders adjust credit limits based on loan purpose, with business loans often
requiring more significant amounts due to their investment nature, and personal loans like car or
education loans involving smaller sums due to perceived lower risk.
4.0 Conclusion

Credit amounts and demographic and financial factors, including gender, loan term, housing,
savings, job type, age, and bank accounts, were correlated in this study. The results show that
longer loan durations are linked to more significant credit amounts, men tend to receive more
varied and more substantial credit amounts, and homeowners often receive higher credit limits
than renters. While those with less savings frequently get larger loans, older people benefit from
established credit records. The goals stated in Section 1.2 were accomplished, offering valuable
information about how these factors affect loan availability, enhancing lending procedures, and
guaranteeing more equitable loan distribution among clientele.

5.0 Recommendations

Review and Adjust Credit Scoring Models

Due to gender biases in credit ratings, men may be given more variable credit amounts. Financial
institutions should assess and modify their models to foster equity, putting in place gender-
neutral procedures to guarantee that both sexes have equal access to credit (Bermudez Vera et
al., 2025).

Offer Flexible Loan Duration Options

Since loan duration and credit amount are positively correlated, flexible loan terms may improve
customer payback management and expand their eligibility for larger loans. By offering flexible
solutions, financial institutions can satisfy consumer demands for economic flexibility and
increase customer satisfaction and loyalty (GSB, 2024).

Develop Targeted Credit Products for Renters

Renters should be the target of credit products created by financial firms because homeowners
typically receive more significant loan amounts. To increase credit availability and advance
economic inclusion, these products must consider the financial situation of renters and provide
advantageous conditions (Greystar, 2025).

Integrate Savings and Checking Account Data into Credit Evaluations

People with more significant savings and extensive checking accounts tend to have higher credit
amounts, indicating greater financial security. By integrating these characteristics into credit
evaluations, institutions can produce more accurate profiles and make more inclusive lending
decisions, particularly for individuals with low credit histories (Huang et al., 2024).

Create Credit Products Tailored for Older Demographics

Older people typically qualify for more significant credit amounts since they have longer-
established credit histories. Financial institutions ought to provide customized loan products for
this group to reflect their financial soundness and improve access to credit for a significant and
expanding portion of the population (GDS Link, 2023).
References:
1. Chang V, Xu QA, Akinloye SH, Benson V and Hall K (2024) ‘Prediction of bank credit
worthiness through credit risk analysis: an explainable machine learning study’, Annals
of Operations Research, 8:1-25.
2. Crine M (2024) Job Creation Tax Credits: What You Need to Know, Smith Howard
website, accessed 7 April 2025. https://www.smith-howard.com/job-creation-tax-credits-
what-you-need-to-know/
3. Dansana D, Patro SG, Mishra BK, Prasad V, Razak A and Wodajo AW (2024)
‘Analyzing the impact of loan features on bank loan prediction using Random Forest
algorithm’, Engineering Reports, (2):e12707.
4. Fernandez Vidal M and Caire D (2024) Gender-Intentional Credit Scoring, CGAP
website, accessed 7 April 2025. https://www.cgap.org/research/publication/gender-
intentional-credit-scoring
5. FRBSF (Federal Reserve Bank of San Francisco) (2022) Housing and Access to Credit
Are Two Sides of the Financial Inclusion Coin, FRBSF website, accessed 7 April 2025.
https://www.frbsf.org/research-and-insights/blog/community-development/2022/05/24/
housing-and-access-to-credit-two-sides-of-financial-inclusion/
6. GDS Link (2023) How to Make Digital Banking More Accessible for Senior Citizens,
GDS Link website, accessed 7 April 2025.
7. Greystar (2025) Build Credit with Rent Payments, Greystar website, accessed 7 April
2025. https://www.greystar.com/blog/build-credit-with-rent-payments
8. GSB Bank (The Guilford Savings Bank) (2024) A Deep Dive into Term Loans: Flexible
Financing Options for Every Business, GSB Bank website, accessed 7 April 2025.
https://mygsb.bank/news/a-deep-dive-into-term-loans-flexible-financing-options-for-
every-business/
9. Huang Z, Xu Z, Wang X and Xu Z (2024) ‘The analysis of credit governance in the
digital economy development under artificial neural networks’, Heliyon, 10(20).
10. Karl S (2025) Average Credit Scores by Gender, Investopedia website, accessed 7 April
2025. https://www.investopedia.com/average-credit-scores-by-gender-5214525
11. Nichols C (2024) How a Loan’s Maturity and Amortization Impact Credit, SouthState
Correspondent Division website, accessed 7 April 2025.
 https://southstatecorrespondent.com/banker-to-banker/loan-profitability/how-a-loans-
maturity-and-amortization-impact-credit/
12. Schwahn L (2025) What Is the Average Credit Score by Age, and What Is a Good Score
for My Age?, NerdWallet website, accessed 7 April 2025.
https://www.nerdwallet.com/article/finance/what-is-the-average-credit-score-by-age-and-
what-is-a-good-score-for-my-age
13. Vera IM, Restrepo JM and Manotas-Duque DF (2025) ‘Data Mining for the Adjustment
of Credit Scoring Models in Solidarity Economy Entities: A Methodology for Addressing
Class Imbalances’, Risks, 13(2):20.
14. Zinn D (2024) Does your loan purpose matter? Yes — here’s why, Bankrate website,
accessed 7 April 2025. https://www.bankrate.com/loans/personal-loans/does-loan-
purpose-matter/