Central Bank of Uzbekistan Identifies Age Group with the Highest Risk of Loan Delinquencies

Tashkent, Uzbekistan (UzDaily.com) — The Statistics and Research Department of the Central Bank of the Republic of Uzbekistan has published the results of a study examining factors that influence delays in loan repayments.

According to the analysis, borrowers aged between 20 and 35 most frequently face loan delinquencies. The regulator explains that this trend among younger borrowers is driven by income instability, limited experience in using credit products, and a higher propensity for financial risk.

The study also revealed gender differences in credit behavior. Women were found to be less likely to fall into arrears than men, which experts interpret as evidence of a more cautious and disciplined approach by women to meeting their credit obligations.

As of 1 December 2025, the outstanding volume of loans to individuals in the country amounted to 217.1 trillion soums. The database contains information on more than 8.5 million loan agreements and nearly 5 million borrowers.

An analysis of the loan portfolio by gender and age shows that delinquency rates, defined as the ratio of overdue loans to the total number of loans, vary significantly across demographic groups and decline as borrowers grow older.

In particular, the highest delinquency rates were recorded among men aged 20–25 at 22.8 percent and 25–30 at 20.9 percent. Among women, the highest level of delinquency was observed in the 20–25 age group at 13.3 percent. This age-related pattern is attributed to income volatility, a lack of credit experience, and behavioral factors such as risk-taking, impulsive spending, and procrastination.

From the age of 30 onward, delinquency levels gradually decrease. In the 35–50 age group, the average delinquency rate stabilizes in the range of 14.8 to 15.5 percent, while among borrowers over 55 it falls further to 12.5–13.2 percent.

Gender differences persist across all age categories, with men consistently exhibiting higher delinquency rates than women. The gap is particularly pronounced among borrowers under 30, where it reaches 8–11 percentage points. After the age of 60, delinquency indicators for men and women largely converge, settling at around 12.5–13.1 percent.

The distribution of delinquency cases by age group indicates that the main credit risk is concentrated among borrowers aged 20–35. This segment accounts for the largest share of overdue contracts, at 53.4 percent, including 15.9 percent among those aged 20–25, 18.6 percent in the 25–30 group, and 18.9 percent among borrowers aged 30–35. This points to a high concentration of credit risk within economically active age groups carrying a heavy debt burden.

An analysis of income levels by age shows that men’s incomes tend to rise up to a certain age, while women’s incomes remain largely unchanged until the age of 55. The highest average income among men is observed in the 30–40 age group. For both genders, a noticeable decline in income occurs after 55, likely reflecting reduced economic activity at older ages.

Overall, the study indicates that the income gap between male and female borrowers reaches 65–70 percent, with this disparity narrowing significantly after retirement age.

An assessment of delinquencies by income level shows that delinquency rates decrease as income rises. For both men and women, the highest levels of delinquency are characteristic of low-income groups. At the same time, repayment discipline among men improves more rapidly as income increases, while among women delinquency indicators tend to cluster around the average level.

An analysis of overdue debt by interest rate reveals a positive relationship between interest rates and delinquency levels. The lowest delinquency rate was recorded in the 20–34 percent interest rate range at 13.3 percent, indicating the highest portfolio quality in this segment.

For loans with interest rates below 20 percent, the share of overdue contracts stands at 19.4 percent, exceeding that of the core portfolio segment. In the 34–46 percent range, delinquency rises to 19.8 percent, after which risk dynamics begin to accelerate markedly.

The most significant deterioration in portfolio quality is observed at interest rates above 46 percent. In the 46–54 percent segment, the share of overdue contracts reaches 20.8 percent, rising to 27.5 percent in the 54–62 percent range. The highest delinquency rate, at 33.7 percent, is recorded for loans carrying rates of 62–76 percent, indicating an extremely high concentration of credit risk in this segment.

Overall, the results show that the threshold for a sharp increase in delinquencies lies in the 45–50 percent interest rate range, beyond which portfolio quality deteriorates substantially. This points to the advisability of introducing enhanced oversight of high-interest loans.

Across most interest rate ranges, borrower gender plays a secondary role. On average, the share of overdue debt is slightly higher among men, but the pattern of delinquency in relation to interest rates is virtually identical for both genders.

The analysis also revealed a nonlinear relationship between loan maturity and the level of credit risk. The lowest share of delinquencies is observed for short-term loans with maturities of up to three months, at 4.4 percent.

As loan maturities lengthen, delinquency rates increase, reaching 13.7 percent in the six to twelve month segment and 18.6 percent in the twelve to twenty-four month segment. The highest delinquency level is recorded in the twenty-four to thirty-six month segment at 19.8 percent, indicating the greatest risk concentration among medium-term loans.

For longer maturities, delinquency levels decline. In the thirty-six to sixty month segment the rate stands at 18.9 percent, while for loans with maturities exceeding sixty months it falls to 11.3 percent.

Gender analysis shows that across all loan maturities the share of overdue debt is higher among men than among women. The difference is particularly pronounced for loans with maturities of more than sixty months, where delinquency reaches 13.3 percent among men compared with 8.9 percent among women.

Overall, the proportion of borrowers who fall into arrears on online loans is higher than on offline loans. At the same time, age-related differences persist in separate analyses of online and offline lending, with the highest delinquency levels recorded among borrowers aged 18–25. For both types of loans, delinquency rates are consistently higher among men than among women.

An analysis of the number of loan contracts per borrower shows that repayment discipline deteriorates as the number of credit agreements increases. Among men, delinquency rises from 19.1 percent for borrowers with a single contract to 45.3 percent for those with more than ten contracts. Among women, the corresponding increase is from 16.5 percent to 36.6 percent. This indicates a substantial rise in the probability of delinquency as the number of credit obligations grows.

The average share of overdue contracts per borrower also shows an upward trend as the number of contracts increases. Among men, this indicator rises from 14.9 percent for borrowers with two contracts to 27.5 percent for those with more than ten contracts. Among women, it increases from 11.4 percent to 21.4 percent, remaining consistently below male levels.

The proportion of borrowers with 50 percent or more of their contracts overdue also increases with the total number of contracts. Among men, this share ranges from 13 to 29 percent and reaches 29 percent in the segment with more than ten contracts. Among women, it ranges from 9 to 22 percent, with a maximum of 22.2 percent.

An analysis of the portfolio structure shows that 71.3 percent of borrowers have only one active contract, while a further 16.6 percent have two contracts. As a result, around 90 percent of borrowers are concentrated in segments with a limited number of credit obligations. Borrowers with three to four contracts account for about 8 percent of the portfolio, those with five or more contracts for 3 percent, and the segment with ten or more contracts for just 0.7 percent. Women are more likely to have a single contract, at 74.5 percent compared with 68.6 percent among men, while men account for a higher share among borrowers with a large number of loans.

An analysis of overdue debt by type of collateral reveals significant differences in portfolio quality. The highest delinquency levels are observed for loans secured by property rights, at 32.8 percent among men and 25.0 percent among women, as well as for loans secured by insurance, at 21.3 percent and 20.0 percent respectively. Unsecured loans are characterized by moderate delinquency levels, at 15.3 percent among men and 12.2 percent among women, likely reflecting the relatively smaller amounts of such loans.

The highest-quality loan portfolios are formed by loans guaranteed by third parties, with delinquency rates of 7.2 percent among men and 3.9 percent among women, as well as by loans secured by passenger vehicles, at 9.9 percent and 12.0 percent respectively. For loans secured by real estate and by individuals’ salaries, delinquency levels are at a moderate level of around 15–16 percent.

In addition, loans with a co-borrower are associated with lower delinquency rates. For loans without a co-borrower, the share of overdue debt stands at 18.4 percent among men and 12.6 percent among women, while the presence of a co-borrower reduces these figures to 14.8 percent and 8.9 percent respectively.

Where the co-borrower is male, the delinquency rate is 11.0 percent, while when the co-borrower is female it is lower, at 9.9 percent.

The age of the co-borrower has only a minor impact on delinquency levels, averaging around 10.5 percent overall, although the indicator is somewhat higher in older age groups, at 12.1 percent among those aged 66 and above.

The study covered 8.577 million individuals who had received loans. For data-cleaning purposes, borrowers with a zero outstanding loan balance were excluded from the sample. In addition, to ensure consistency for borrowers holding more than one loan, only one loan per borrower was taken into account. Records of loans allegedly issued to individuals under the age of 18 were also removed from the dataset. As a result, the regression analysis was conducted using data on 4.949 million borrowers.

When developing scoring models, commercial banks are advised to apply differentiated risk assessment approaches for men and women. In particular, reducing the level of the risk premium for women could contribute to more equal conditions in the credit market and broaden women’s access to credit resources.

At the same time, the study shows that although the average monthly income of women is 1.6 times lower than that of men, the probability of loan delinquency among women is lower.

It is also advisable to revise age-related parameters in credit risk assessment, specifically by applying a higher risk level to borrowers aged 18–30 compared with those in the 30–60 age group.

In addition, there is a need to increase risk levels or tighten requirements when issuing online loans. Another important conclusion of the study is the necessity of applying differentiated approaches to client communication on debt repayment issues.

In particular, separate experimental studies are required to identify the reasons why borrowers’ gender and age influence their repayment behavior.

It is also recommended to test different formats of delinquency notifications, or nudging, depending on borrowers’ gender and age, and, based on an analysis of their responses, to develop the most effective communication strategies for different customer categories.

Furthermore, borrowers should be provided with more information on the terms and consequences of borrowing prior to loan origination, including through the preparation of dedicated informational video materials. At the same time, special attention should be paid to the accuracy of data entry at the loan issuance stage, which will subsequently facilitate similar studies and improve their quality.

In particular, during this study data were identified on loans allegedly issued to individuals aged 13 to 18, which is most likely attributable to incorrect entry of borrowers’ year of birth.