The global mortgage market is an integral component of the financial system, underpinning economic growth and social stability by enabling property ownership to access financial resources. However, this market operates within a highly dynamic environment influenced by macroeconomic factors such as interest rates, GDP growth, inflation, and exchange rate fluctuations (Koju et al., 2020; Kariuki, 2015). These factors directly affect borrowers’ ability to meet mortgage obligations, posing risks to financial institutions and economic stability (Allen & Johnsen, 2008). The debate on the effect of macroeconomic variables on nonperforming loans (NPLs) is yet to be concluded. Klein (2013) and Beck et al. (2015) have on different occasions verified these contradictory observations. Espinoza and Prasad (2010) evaluated 80 (GCC) banks and found that NPLs were positively correlated with increased global financial volatility. Škarica (2014) suggested for a strong correlation between higher NPLs and the economy’s recession, employment status, and inflation rate (IR). Using data from 16 Central, Eastern, and South-Eastern Europe (CESEE) countries between 1998 and 2011 at the national level, Klein (2013) observed that NPLs are negatively correlated with credit growth, unemployment, GDP growth, and inflation.

Similarly, the mortgage markets in both developed and emerging economies are shaped by structural and cyclical factors, with macroeconomic variables often acting as key determinants of market performance (Mian & Sufi, 2014). In developed economies, fluctuations in lending rates and GDP have been linked to significant variations in mortgage default rates (Carvalho et al., 2022; Morgan & Ryan, 2024; Koju et al., 2020). For instance, a study by Gerardi et al. (2018) on the U.S. housing crisis emphasized how sharp declines in GDP growth and rising unemployment rates triggered widespread mortgage defaults. Similarly, Mian and Sufi (2014) found that rising interest rates exacerbated defaults by increasing repayment burdens for adjustable-rate mortgage holders. These findings align with the “Double Trigger Hypothesis,” which posits that mortgage defaults typically result from a combination of negative equity and cash flow constraints (Bergmann, 2020).

Emerging markets, however, present a different narrative, where additional factors such as exchange rate volatility and inflation often play more prominent roles. Güneş and Apaydın (2021), in their study of Turkey’s mortgage market, highlighted how inflation-induced rises in real estate prices and currency depreciation created instability. Limon (2025) in his study in Bangladesh noted that bank-specific factors such as profitability, capital adequacy, and lending practices are important NPL determinants alongside agency costs. This study, however, ignored macroeconomic factors as determinants of NPL and reserved the same for future. Kansheba and Jacob (2020) suggest that macroeconomic volatility, particularly exchange rate fluctuations, was an important determinant of mortgage performance in Tanzania. Tanzania’s mortgage market, though nascent, has experienced significant growth over the past decade. The Banking and Financial Institution (Mortgage Financing) Regulations of 2011 provided a regulatory foundation for the mortgage market, yet financing and NPL risks remain. Kansheba and Jacob (2020) emphasized the need for stronger risk-management practices among lenders to mitigate default risks.

The Tanzania’s economy has experienced moderate GDP growth, averaging 6% annually from 2014 to 2020, driven by sectors such as construction, agriculture, and mining (World Bank, 2022b). However, external shocks, including the COVID-19 pandemic, caused GDP growth to decelerate to 2% in 2020 before rebounding in subsequent years (World Bank, 2022a). Macroeconomic volatility in Tanzania is further compounded by inflation and exchange rate dynamics. IRs have fluctuated between 3% and 5% over the past decade, driven by global commodity prices and domestic supply chain constraints Bank of Tanzania (2022). At the same time, the Tanzanian Shilling has faced consistent depreciation against major currencies, increasing the cost of servicing foreign-denominated debts, including some mortgages. Such conditions make Tanzanian borrowers particularly vulnerable to adverse macroeconomic shocks, aligning with findings from studies in other developing economies (Güneş & Apaydın, 2021).

By 2022, mortgage loans constituted a growing percentage of total loans issued by commercial banks in Tanzania (Bank of Tanzania, 2022). Despite this growth, the market remains susceptible to macroeconomic fluctuations. Interest rates in Tanzania, for example, have remained relatively high at around 16% and volatile, limiting affordability for borrowers and increasing NPL risks (Bank of Tanzania, 2023). This study contributes to policy discourse by providing insights into how macroeconomic variables can enhance the performance of mortgage-offering banks. Since the global mortgage market operates within a complex interplay of macroeconomic variables, in Tanzania, where the mortgage market is still maturing, these dynamics are particularly pronounced. By integrating theoretical perspectives and empirical evidence, this study provides an analysis of the factors shaping performance of mortgage-offering banks, offering lessons for policymakers and financial institutions.

The theoretical framework guiding this study combines the Double Trigger Hypothesis and Arbitrage Pricing Theory (APT) to provide a comprehensive understanding of how macroeconomic factors affect NPL. The Double Trigger Hypothesis, proposed by Bergmann (2020) asserts that two concurrent factors are necessary for mortgage default: negative equity and financial distress. Negative equity occurs when a borrower owes more on their mortgage than the property is worth, while financial distress refers to the borrower’s inability to meet financial obligations due to job loss, illness, or other financial or economic shocks. According to this theory, these two factors act as “triggers” that initiate the default process. While the Double Trigger Hypothesis emphasizes individual borrower circumstances, it also highlights the influence of broader macroeconomic conditions that can exacerbate financial distress, such as rising unemployment or a downturn in the housing market.

In contrast, the APT, developed by Ross (1976), offers a broader, systemic framework for understanding how economic variables affect asset prices, including housing. The APT suggests that asset prices, such as real estate values, are determined by multiple macroeconomic factors, including GDP growth, inflation, and interest rates. This theory posits that changes in these factors affect mortgage markets by influencing both property values and borrowers’ ability to repay loans. APT is particularly valuable in this context because it captures the cumulative impact of economic conditions on the mortgage market, integrating multiple factors that might simultaneously affect both the supply and demand for housing.

The conceptual framework in Figure 1 integrates the two perspectives discussed earlier, illustrating how macroeconomic variables and individual borrower circumstances interact to influence mortgage defaults. For instance, a sudden increase in interest rates could reduce borrowers’ affordability while simultaneously decreasing property values, thus triggering both factors required for mortgage default according to the Double Trigger Hypothesis. This integrated approach helps to explain the complexities of mortgage market performance under varying macroeconomic conditions. Despite extensive research on macroeconomic factors and NPL, most studies focus on developed economies, with limited attention to emerging markets. While Güneş and Apaydın (2021) and Kansheba and Jacob (2020) provide valuable insights, more research is needed to understand the unique challenges faced by nascent mortgage markets in Africa and Asia. This study, therefore, complements the limited studies in Sub- Saharan African that have focused on macroeconomic determinants of mortgage uptake.

This study utilizes a panel data analysis approach to examine the relationship between macroeconomic factors and mortgage defaults across 10 Tanzanian banks over the period 2014 to 2024. This methodological framework combines cross-sectional data from 10 Tanzanian banks with time-series observations. This approach enables the examination of both temporal and cross-sectional variations, enhancing the precision of estimates compared to purely cross-sectional or time-series analyses (Berrington, 2006). Panel data methods are particularly useful for addressing unobserved heterogeneity and detecting dynamic relationships among variables, making them well suited for analyzing mortgage market performance. By obtaining data from 10 banks, that is, 6 largest banks from the list of 10 banks ranked in Table 2 and 4 other small banks, out of a total of 31 mortgage-offering banks by 2024, the study expounds critical relationship between NPL and macroeconomic variables. This study, however, did not capture the performance of the other 21 banks that though small in size may change some findings of this study.

The study relies exclusively on secondary data sourced from published financial statements found on the websites of the respective banks as summarized in Table 1. The dependent variable, NPL, represents the loans outstanding specific to each bank, as derived from their quarterly and annual financial statements in both percentage and amount in Tanzania Shilling (TZS). NPL quantifies the proportion of loans that borrowers have failed to repay out of the gross loan amount issued for each bank, offering a direct measure of market stability for mortgage-offering banks. The independent variables encompass four critical macroeconomic indicators. The LI, sourced from the Bank of Tanzania (BOT), reflects the cost of borrowing, influencing borrowers’ ability to repay loans. Gross Domestic Product (GDP) growth, obtained from the National Bureau of Statistics (NBS), measures the overall economic performance, which can affect employment and income levels, subsequently impacting loan repayment capacity. The IR, also provided by the NBS, captures price level changes that can strain household finances and influence default rates. Lastly, the Foreign Exchange Rate (LN_ER), sourced from the BOT, indicates the average exchange rate of the Tanzanian shilling against the US dollar, which can impact mortgage repayment for loans tied to foreign currency or influenced by imported goods prices. Although there are other macroeconomic variables such as unemployment and population dynamics that are central in NPL fluctuations, they were not included in the current analysis because they are considered to be highly correlated with GDP and inflation and their inclusion would have limited value addition.

The study focused on 10 out of 31 mortgage-offering commercial banks that collectively represent a significant portion of Tanzania’s mortgage market (BoT, 2023b). These banks were selected because of availability of data on continuous basis since 2014 and the fact that they represent large and small banks in Tanzania with the largest six banks having around 75% of the market share (BoT, 2023a). Six of the banks that offer mortgage products are also among the top 10 largest banks in Tanzania as shown in Table 2. The 4 small mortgage-offering banks that have never featured in the top 10 largest banks as per Table 2 but were included in the study are Amana Bank Limited (AMANA), I&M Tanzania Limited (I & M), Bank of Baroda (BOB), and International Commercial Bank Tanzania Ltd (ICB).

Source. Authors’ construction based on https://www.tanzaniainvest.com/banks, 2025.

The stationarity of variables was examined using the both Pesaran CADF test for cross-sectional dependence and the Fisher-type unit root test for panel-independent variables as per Table 3. Panel stationarity tests were conducted to ensure the reliability of the panel time series analysis in cases of cross-sectional dependence. Since the macroeconomic variables are only time dependent and are constant across panels, the traditional assumption of cross-sectional independence is assumed and the Fisher-type test is utilized. The null hypothesis in the Fisher’s test is that “all series are nonstationary” while the alternative is “at least one series in the panel is stationary.” Both unit root test results as presented in Table 3 revealed a mixed order of integration among the variables, with some being stationary at level and others requiring first differencing. This finding justified the use of the panel ARDL model, which is particularly suited to datasets with mixed integration orders, ensuring the robustness of the subsequent analyses.

b: Probability based on MacKinnon (1996) one-sided p values. Source. Authors’ construction, 2025.

The cross-sectional dependence in Table 4 was found to be significant for all the bank-specific variables thus calling for second generation stationarity test that in this case the Pesaran CADF, that is, cross- sectionally augmented Dickey–Fuller tests was utilized. In this test, the null hypothesis of cross- sectional independence is tested against the alternative of cross-sectional dependence.

The selection of optimal lag lengths for the ARDL model, as shown in Table 5, was determined using the Akaike Information Criterion (AIC). The optimal lag was observed to be 3 given the lowest AIC score of around −32.115, which was also supported by other optimality criteria as reported in Table 5. These lags are critical for accurately capturing both the short- and long-term dynamics of the relationships between macroeconomic variables and NPLs. By carefully balancing model complexity and predictive accuracy, the chosen lags allow for a nuanced analysis of how immediate and delayed effects shape NPL.

LR: sequential modified LR test statistic (each test at 5% level) FPE: Final prediction error AIC: Akaike information criterion HQ: Hannan-Quinn information criterion Source. Authors’ construction, 2025

The Hausman test was employed to select between the Pooled Mean Group (PMG) and the Mean Group (MG) estimator while the AIC was used for optimal lag selection. The Hausman test supported the use of the PMG estimator, confirming long-run homogeneity across banks while accommodating short-run heterogeneity as per Table 6.

The findings of this study highlight the complex and multifaceted relationship between macroeconomic factors and NPLs of mortgage-offering banks in Tanzania, providing insights that align with and, in some cases, diverge from existing literature. Significant long-run effects of LIs, and exchange rate were identified, underscoring the critical role of these variables in shaping mortgage market dynamics. In the long run, GDP growth was found to have a stabilizing effect on NPLs across all banks though its significance varied. This aligns with studies by Louzis et al. (2012) and Kansheba and Jacob (2020), which emphasize the importance of sustained economic growth in reducing default risks. These findings highlight significant variations in how macroeconomic factors influence NPLs across Tanzanian banks, reflecting the diversity in their operational structures, market exposures, and strategic orientations.

Although GDP did not exhibit significant long-run effect on NPL, short-run dynamics suggest for a positive effect especially among larger banks. In this regards, economic expansion enhances borrowers’ income stability, enabling them to meet mortgage obligations more consistently as supported in the previous studies (Koju et al., 2020; Kumar et al., 2023; Christodoulou-Volos, 2025; Bergmann, 2020; Mokas & Nijskens, 2019). However, in this study, this positive relationship varied in the short run, where certain banks exhibited heightened sensitivity to GDP fluctuations, indicating the need for institution-specific strategies to cushion against macroeconomic shocks. These results highlight the need for tailored risk-management strategies that account for the unique characteristics of individual banks and their exposure to specific macroeconomic risks. Overall, the insights derived from this study emphasize the importance of maintaining macroeconomic stability to enhance the performance and resilience of Tanzania’s mortgage market.

The effect of LIs on mortgage defaults, particularly in the short run was not conclusive contrary to previous studies that suggest a positive effect (Zheng & Zhang, 2021; Christodoulou-Volos, 2025; Kumar et al., 2023; Mokas & Nijskens, 2019). Higher interest rates significantly increase repayment burdens, leading to higher default probabilities only for two banks, and for the rest it was not statistically significant. The significant positive relationship between NPL and lending rate is observed only in the long run indicating that higher borrowing costs impose substantial financial burdens on mortgage holders, increasing the likelihood of future rather than immediate defaults. Banks with greater sensitivity to interest rate changes, particularly those that are heavily reliant on variable-rate mortgage products may not realize immediate default, rather face heightened risks under volatile interest rate conditions. These findings not only corroborate earlier studies emphasizing the adverse effects of rising interest rates on mortgage performance but also highlight heterogeneity among banks. Institutions with significant exposure to variable-rate mortgages could be more vulnerable thus adopting fixed-rate products or implementing caps on variable rates to mitigate risks may be proposed.

Foreign exchange rates also play a pivotal role in influencing the performance of select mortgage- offering banks in Tanzania specifically I & M and EXIM. This could be the case for banks with substantial foreign-denominated loans (see I & M and EXIM in Table 10) or customers with income streams tied to exchange rate movements. Depreciation of the Tanzanian shilling can, in theory, adversely affect borrowers with foreign-currency obligations as they need more of local currencies to buy foreign currency for their loan repayment obligations. Christodoulou-Volos, (2025) observed a positive association between appreciation of the local currency and NPL similar to long-run effect observed in Table 9 although in this study, most of the effect was captured in personal characteristics of the borrower. This underscores the interconnectedness of the mortgage market with international financial dynamics and the risks posed by external shocks. Banks with greater foreign exchange exposures must enhance their risk-management frameworks, including currency hedging and diversification strategies.

Inflation, while a notable macroeconomic factor (Kumar et al., 2023), shows a less consistent impact on mortgage defaults. The findings suggest that its influence on NPL could be contingent on the broader economic and institutional contexts, such as wage adjustments and cost-of-living factors alongside (Koju et al., 2020), although these other variables were not part of this study. The finding of this study regarding inflation, however, diverges from the study by Güneş and Apaydın (2021), who found significant inflationary impacts on NPL, suggesting that the role of inflation may depend on other macroeconomic environments, such as wage dynamics and inflation pass-through mechanisms. In Tanzania, inflationary pressures may be partially offset by government interventions (Bank of Tanzania, 2023) or borrower adaptations, reducing its direct influence on mortgage performance thus resulting into insignificant effect similar to observations by Goyal et al. (2023) who evaluated both developed and developing countries and found a significant and positive inflation effect on NPL in low-income countries. Since Tanzania is transiting out of that income brackets toward lower middle income, the inflation effect could be insignificant.

The varying adjustment speeds observed across banks highlight significant differences in their resilience to macroeconomic shocks. According to Yoldas and Senyuz (2018), differences in adjustment toward long-run equilibrium require large shocks to risk premia such as balance sheet adjustment by banks as well as policy interventions that aim to reduce liquidity and credit risk premia. It seems, in these findings, that the studied banks respond differently to the central bank’s intervention policy given their diverse sizes and the nature of interventions that may be selective in favor of against certain banks. For purposes of mortgage in Tanzania, Banks through the Tanzania Mortgage Refinancing Company (TMRC), a credit facility that ensures Bank’s liquidity, can request refinance based on their ability to issues such loans.

Another explanation could be linked to integration of the banking system as investigated by Koutsomanoli-Filippaki and Mamatzakis (2010) in a study of European banks. Their insights highlight that bank integrations have had some positive impact on the speed of adjustment toward long-run equilibrium among banks. Since Bank’s in Tanzania have differing speed of adjustment, these provide some indication that they are not integrated. Similarly, those banks that displayed high speed toward their long-run equilibrium suggested for a competitive environment that could be highly questionable in Tanzania given the notable bank concentration (Sanga, 2022). The heterogeneity of these results emphasizes the necessity for tailored risk-management strategies that account for individual bank characteristics and exposures. Banks with higher sensitivity to interest rates and foreign exchange fluctuations may benefit from implementing hedging strategies and diversifying their mortgage portfolios. Policymakers, on the other hand, are advised to prioritize macroeconomic stabilization, particularly through measures that ensure stable interest and exchange rates, as these have the most profound effects on mortgage market performance.

The observations in this study also confirm the importance of macroeconomic stability in supporting a robust mortgage market performance. Specifically, GDP growth was shown to have a stabilizing but insignificant influence on NPLs, suggesting though inconclusively that sustained economic expansion reduces the likelihood of mortgage defaults by increasing borrowers’ income and repayment capacity

(Goyal et al., 2023). Conversely, volatility in interest rates and exchange rates emerged as key risk factors, emphasizing the need for financial institutions to develop adaptive strategies to mitigate their effects on borrowers. Given that rising interest rates heighten repayment burdens and exchange rate depreciation poses additional risks to borrowers with foreign-currency mortgages, policy interventions to stabilize these variables are paramount.

This study also highlights the necessity of bank-specific risk-management strategies. The varying adjustment speeds observed across different banks point to differences in their ability to respond to macroeconomic shocks. Banks with more agile risk-management frameworks may be better equipped to absorb external shocks, underscoring the need for enhanced internal controls, diversified loan portfolios, and flexible lending practices (Sanga, 2022). The findings suggest that financial institutions should not adopt a one-size-fits-all approach but instead tailor their strategies based on their unique operational contexts and market exposures.

Furthermore, the limited impact of inflation on mortgage defaults in the Tanzanian context as observed in Tables 9 and 10 (no significant long-run effect of inflation on NPL in Table 9 while short-run dynamics in Table 10 show significant effect of inflation for CRDB bank only), calls for a nuanced understanding of inflationary dynamics and their indirect effects on borrowers. While inflation may not have shown a direct correlation with NPLs, other factors such as income growth (Sanga, 2022), government interventions, and the overall economic environment may play compensatory roles in moderating its influence on mortgage performance. Generally macroeconomic performance is associated with NPL whether directly or indirectly. For example, government policies that foster economic growth will ultimately reduce NPL, while contractionary policies are likely to exacerbate it (Goyal et al., 2023). Financial institutions must therefore consider the broader economic context, including fiscal and monetary policies, when assessing risks related to inflation.

Overall, the findings reinforce the need for tailored risk-management strategies at both the institutional and policy levels. Policymakers should prioritize macroeconomic stabilization, focusing on interest rate and exchange rate stability to safeguard the mortgage market. Observations in Table 9 suggest that all banks tend to have significantly lower NPLs in response to higher exchange rate with the exception of the two major banks whose NPL escalates in response to higher exchange rate. This observation ties the major banks to foreign exchange risk on their loan portfolio. Simultaneously, these major banks may adopt proactive measures, such as diversifying their portfolios, enhancing credit assessment processes, and promoting financial literacy among their respective borrowers. These interventions will help mitigate risks, improve resilience, and foster a stable mortgage market capable of supporting sustainable economic growth.

In light of these findings, several policy recommendations are evident. First, policymakers must prioritize the stabilization of macroeconomic variables such as interest rates and exchange rates, as these have a direct impact on mortgage market stability. Stabilization measures, including monetary policies aimed at controlling inflation and managing exchange rate fluctuations, are essential for reducing default risks. In addition, financial institutions should strengthen their credit risk assessment models, incorporate stress-testing procedures, and promote financial literacy programs to help borrowers better manage their obligations.

Although the government has increased financial literacy initiatives in collaboration with financial institutions, higher learning institutional as well as the private sector (Mmari, 2023), there is still a gap of financial literacy education in line with mortgage products. There is a need to specifically invest in financial literacy programs to educate potential and existing mortgage borrowers on managing their finances, understanding the implications of interest rates, and preparing for economic fluctuations. Providing advisory services to borrowers can help them make informed decisions regarding their mortgage commitments and manage their debt more effectively.

Lastly, the findings highlight the need for continued research on the dynamics of mortgage markets, particularly in emerging economies like Tanzania. Expanding the scope of research to include more diverse financial institutions and considering alternative performance indicators will provide a more comprehensive understanding of the complex interplay between macroeconomic factors and mortgage default rates. By enhancing the resilience of the mortgage market, Tanzania can foster more sustainable economic growth, improve financial inclusion, and ensure greater stability in its financial system.