Debt and its velocity as determinants of real economic activity

The reality of the last decades shows us that debt has an important place in modern economies. This thesis focuses on debt in the form of bank loans extended to private economic agents – households and corporates. It explores the causal relationship between debt and real economic activity, expressed in terms of aggregate income, consumption, and corporate default rates. Debt is analyzed not only in absolute terms but also in relation to other factors – namely, its velocity is the object of interest. The income velocity of debt serves as a valuable tool for investigating the relationship between debt and aggregate income, as well as for interpreting how financial shocks transition into the real economy. The transmission of the changes in debt to real economic activity is inherently complex and the research gap especially lies in the absence of in-depth transmission of loan dynamics to certain aspects of the real economy (aggregate income, household consumption and corporate default rate). Therefore, the presented analysis accounts for various relevant variables, such as (but not limited to) distribution of income and wealth, distribution of marginal propensity to consume, financial and non-financial (housing) wealth, policy interest rates, and prudential policies. The first paper discusses the interaction of debt and economic performance and the models making provision for debt. The paper demonstrates how financial shocks affect the income velocity of money under different distributions of wealth/income and marginal propensity to consume across the population. For this purpose, an original agent-based simulation model with a limited loan supply was designed. Proposed model shocks are shocks to loan demand, loan supply, marginal propensity to consume, macro-prudential regulatory ratios, real estate capital gains, repayment ratios, shocks to the structure of loans provided and to the structure of real estate property transactions. It is shown that the more equal the distributions of wealth/income and of the marginal propensity to consume, the higher is the income velocity of money. From financial shocks, the marginal propensity to consume shock and the shock to the structure of new real estate property purchases have the largest impact on velocity. The shock to regulatory ratios has the lowest magnitude. The second paper examines the effects of household indebtedness, income inequality, and asset dynamics on household consumption across 11 developed OECD countries from 1995 to 2021. It evaluates how rising household debt and income inequality may hinder future consumption and analyzes the influence of changes in net, non-financial, and financial assets. The methodology includes a fixed effects model with Driscoll-Kraay standard errors and a panel VAR (PVAR) model with a GMM estimator. Contrary to previous findings, the analysis reveals that increases in household debt and income inequality do not necessarily reduce consumption. Instead, consumption and indebtedness may rise together in the short to medium term. The expected positive impact of financial and net assets on consumption was not proven. While non-financial assets (e.g., housing) boost consumption in the short run via collateral effects, this influence turns negative over longer horizons. Moreover, the anticipated positive relationship between income inequality and household indebtedness was not confirmed. The last paper provides a structured approach for integrating macroeconomic conditions into the modeling of expected credit losses (ECL) under IFRS 9, particularly by adjusting the probability of default (PD). A major part of the study investigates macroeconomic determinants of corporate loan default rates (DR) and their role in the calculation of expected credit loss (ECL) and PD. Using Czech data (2002Q4–2024Q4), this study identifies key DR drivers not only for the aggregate loan portfolio, but also across various loan portfolio subsegments, including economic sectors, loan types, collateral types, and bank sizes. The findings reveal that on the aggregate level, the most important drivers of the corporate default rate are the output gap and the unemployment rate. The relevant macroeconomic drivers of corporate DR significantly vary across subsegments. Finally, the paper demonstrates the incorporation of forecasted economic development into the calculation of PD by constructing a systematic risk factor Z under the Vasicek approach.