Financial Accelerator and Endogenous Growth Over Medium-term Business Cycles [Job Market Paper]

This research is motivated by observations of medium-term
evolution in both real business activities and financial sectors, which
nevertheless tends to be ignored in the standard business cycles. The
medium-term cycles are defined as the sum of the high and
medium-frequency (over 8 years) variations. This paper examines whether these medium-term fluctuations reflect some persistent mechanism of economic activity reacting at high-frequency variations and directing to long-run growth. My wavelets analysis with post-war US data indicates that the medium-term cyclical components are important descriptive features for real output, credit to firms, investment, consumption, TFP, human capital, firm’s funded R&D expenditure. These fluctuations are substantially more volatile and persistent than the conventional measures. Then I develop, as a first pass, a quantitative model to provide an integrated explanation of medium-term fluctuations in both real and financial sectors. The model endogenises the movements in productivity and growth rates with devices of human capital, intermediate product variety and firm’s R&D investment, and introduces the firm’s credit constraint collateral on the expected firm value. Intuitively, the high-frequency disturbances would affect the firm value expectations and then credit limit to firms that is related firm’s R&D
investment, thus producing sustained effects on productivity over the
medium term. Incorporating the endogenous growth and financial
accelerator within the dynamic stochastic general equilibrium (DSGE)
framework, my model illustrates that these two mechanisms as well as
their interactions play a central role to prolong and amplify the
high-frequency disturbances that leads to the medium-term oscillations.

National and International Medium-term Economic Fluctuations Since 1870 [Paper Link]

This paper investigates the existence of medium-term cycles in real GDP/real GDP per capita for 30 developed and developing economies since c. 1870, and the synchronisation of these medium-term cycles across countries. Wavelets analysis indicates that the medium-term fluctuations are an important descriptive feature of the national economic growth. Studying the entire period between 1870 and 2016, we find that more country pairs are significantly positive synchronised over the medium-term cycles, compared to the business cycles. The global and group-specific factors for national medium-term cycles are further evaluated using principal component analysis (PCA) and dynamic factor models (DFM), in which countries are classified according to the outcome of a cluster analysis. The results of PCA and DFM suggest that group-specific factors play a much greater role for the domes- tic medium-term cycle, compared to the global factor. Particularly, a common medium-term cycle is found within each of four groups in the sample. Across three sub-periods (1870-1913, 1914-1959 and 1960-2016), we find an up- ward trend of global synchronisations of medium-term cycles over time. For each period, the national medium- term economic fluctuations tend to synchronise within groups rather than globally. The grouping patterns of these medium-term cycles, however, change notably over time; and for each sub-period, a common medium-term cycle can be found within some groups.

An Anatomy of Financial Crisis over Medium-term Fluctuations

This research studies financial crisis in the context of real-financial interaction over medium-term cycles for developed and developing countries. First, we explore the time-varying and country-varying interactions between GDP and credit series over high and medium-term frequencies (2-32 years). Then, the predictability of financial crisis over medium-term window is examined. Our analysis challenges the convectional argument that financial crisis is the credit boom gone wrong. To distinguish good and bad credit booms, we take productivity considerations.

Assessing Low-Income Countries’ Capacity to Repay the IMF (with Nelson Sobrinho) [IMF Working Paper (Forthcoming)]

Preserving low-income countries (LICs)’ ability to service their debt obligations including to the IMF is key for their continued access to concessional financing and development. This paper develops a data-based tool to assess LICs’ capacity to repay the IMF, drawing from a wide array of indicators of debt-carrying capacity. We employ well-known statistical tools, including principal component analysis, factor analysis and unknown component model, to aggregate information from these indicators into a ‘capacity-to-repay composite indicator’, which we interpret as a relative measure of risks to a LIC’s capacity to repay. These composite indicators for each arrangement are further categorised into three groups- high, moderate and low risks of repayment capacity, with K-means method. The tool performed well in validation exercises and could be used to help support capacity to repay assessments in real time.