Evolution and Determinants of Loan Performance in the U.S. Mortgage Market : A Comparative Study of 2007 and 2019
Srishti Gupta
In the complex landscape of the U.S. mortgage market, the performance of loans is a critical barometer of economic stability and health.
Figure 1
In 2019, the U.S. mortgage landscape reflected a stabilized market with generally low default rates, as indicated by the predominance of green across the states. However, Texas stands out with notably higher default rates, shown in Yellow. This outlier suggests that Texas encountered unique financial stressors or market conditions that diverged from the national trend, potentially due to localized economic conditions or housing market dynamics. Post-2007, the mortgage sector underwent rigorous regulatory reforms aimed at reinforcing lending standards and promoting loan integrity.
Figure 2 (for reference only : Bubble chart depicts default rates by state)
Figure 3
This bubble chart contrasts default rates and loan volumes across U.S. states, revealing significant variations. The size of each bubble indicates the loan volume, while its color represents the default rate, ranging from green for lower rates to red for higher rates. Most states cluster around lower default rates, suggesting stability, but a few with large, redder bubbles signal higher defaults and possible financial risk. Notably, a red bubble highlights a state with a high default rate despite a substantial loan volume, suggesting disproportionate financial distress. This visualization is instrumental for stakeholders in the mortgage market to pinpoint where to focus risk mitigation efforts, ensuring the sector’s health and stability.
Diving Further …
Figure 4
MI_PCT and OLTV Scatter Plot: The scatter plot suggests a generally proportional relationship between MI_PCT and OLTV. This implies that loans with higher loan-to-value ratios typically require more mortgage insurance, a standard practice to mitigate lender risk on high-LTV loans.
Figure 5
Correlation Matrix: Reveals the relationships between various mortgage characteristics. Notable are the strong positive correlations between related financial ratios (like OLTV and OCLTV) and a potential negative correlation between credit scores and default flags. This suggests that as credit scores decrease, the likelihood of default increases, which aligns with standard credit risk assessments.
Figure 6
Loan Age vs. Origination Rate Scatter Plot: The distribution indicates that newer loans (lower loan age) in 2019 generally had lower origination rates than in 2007, reflective of the broader interest rate environment, which can fluctuate based on economic conditions and monetary policy.
(Connecting the dots:
The MI_PCT and OLTV correlation suggest cautious lending practices—higher LTVs, which imply riskier loans, usually have corresponding mortgage insurance.
The correlation-matrix adds depth to the risk profile by quantifying the strength of associations between all pairs of variables. For instance, higher DTI and lower credit scores being correlated with default flags could indicate financial stress among borrowers.
The Loan Age versus Origination Rate plot across the years reflects the macroeconomic conditions influencing mortgage interest rates, such as Federal Reserve policies and market demand.
The linkage emphasizes the interplay of risk factors (e.g., credit scores, loan-to-value ratios, debt-to-income ratios) that lenders must balance against economic conditions (interest rates, market stability) crucial for understanding market trends, borrower behaviors, and the impact of external economic factors on the mortgage industry.)
Figure 7
The comparison of borrower credit scores between 2007 and 2019 shows a shift towards higher scores. In 2007, scores were generally lower, with most below the median. By 2019, the median score increased, and the distribution became tighter, indicating higher average credit scores. This suggests stricter lending standards post-financial crisis and improved borrower creditworthiness over time.
Figure 8
Histogram displays the frequency distribution of borrower’s credit scores for 2007 and 2019. Suggesting an overall improvement in credit scores from 2007 to 2019, with more borrowers having higher credit scores in the latter year. This is evident from the rightward shift of the distribution’s peak towards higher credit scores. This shift could indicate a recovery from the financial crisis with improved borrower quality, reflecting tightened credit policies and potentially better financial health among consumers.
Figure 9
Bar chart illustrates clearly that refinance loans dominate, reflecting a prevalent trend of borrowers perhaps looking to take advantage of favorable interest rates or to consolidate debt during this pre-crisis period. The number of loans for purchases is significant but less than refinances, which could be due to market conditions at that time. Cash-out refinances are the least common, indicating a more cautious approach to borrowing against equity before the financial crisis hit.
Boxplot and histogram indicate an overall improvement in borrower credit scores, suggesting tighter lending criteria and possibly a healthier economic climate for borrowers. In scatter plots, correlation matrix, and the bar chart together, a story unfolds about the mortgage market’s health and the risk factors affecting loan performance. Factors like credit scores, loan-to-value ratios, and debt-to-income ratios seem to have consistent influence over the years, and their impact on loan defaults remains substantial. Hence, the broader mortgage market may have stabilized, individual borrower characteristics and localized economic factors continued to play critical roles in loan performance.
Word Count:(787)
Footnotes
The years 2007 and 2019 stand out as significant periods in the recent history of mortgage lending. In 2007, the U.S. witnessed an escalation in loan defaults that prefaced the global financial crisis, marking a pivotal moment for entities such as Fannie Mae, which experienced unprecedented losses. By contrast, 2019 is notable as the period just before the COVID-19 pandemic, representing a different economic climate with its own unique challenges and outcomes.
The focus of this project is on the evolution and determinants of loan performance and eventually deducing that Fannie Mae had stabilized its mortgage market, reflecting lower default rates compared to 2007, which is consistent with historical trends following the regulatory changes post-financial crisis.
Conclusion further indicates that though broader mortgage market may have stabilized, individual borrower characteristics and localized economic factors continued to play critical roles in loan performance. Highlighting the need for ongoing scrutiny of loan characteristics and market conditions, both individual and macroeconomic factors intertwine to shape the performance of loans.
Figure Appendix
## Warning: package 'usmap' was built under R version 4.3.3## Warning: package 'gridExtra' was built under R version 4.3.3##
## Attaching package: 'gridExtra'## The following object is masked from 'package:dplyr':
##
## combine## Warning: package 'plotly' was built under R version 4.3.3##
## Attaching package: 'plotly'## The following object is masked from 'package:ggplot2':
##
## last_plot## The following object is masked from 'package:stats':
##
## filter## The following object is masked from 'package:graphics':
##
## layoutFigure 1: : U.S. mortgage landscape : The Outlier, echo
Figure 2: : Default rates by state, echo
Figure 3: : Default rates by state, echo
## `geom_smooth()` using formula = 'y ~ x'
## `geom_smooth()` using formula = 'y ~ x'
Figure 4: MI_PCT and OLTV
## `geom_smooth()` using formula = 'y ~ x'
## `geom_smooth()` using formula = 'y ~ x'## Warning: package 'corrplot' was built under R version 4.3.3## corrplot 0.92 loadedFigure 5: Correlation Matrix
## `geom_smooth()` using formula = 'y ~ x'
## `geom_smooth()` using formula = 'y ~ x'
Figure 6: Loan Age vs. Origination Rate
## Warning: Removed 10509 rows containing non-finite outside the scale range
## (`stat_boxplot()`).## Warning: Removed 10859 rows containing missing values or values outside the scale range
## (`geom_point()`).
Figure 7: Comparing borrower credit scores
## Warning: The `size` argument of `element_line()` is deprecated as of ggplot2 3.4.0.
## ℹ Please use the `linewidth` argument instead.
## This warning is displayed once every 8 hours.
## Call `lifecycle::last_lifecycle_warnings()` to see where this warning was
## generated.
Figure 8: The frequency distribution of borrower’s credit scores
Figure 9: : The distribution of loan purposes in the fourth Quarter of 2007