Financial Frictions In Production Networks

We study how an economy’s production structure determines the response of aggregate output and employment to sectoral financial shocks. In our framework, economic production is organized in an input-output network in which firms face financial constraints on their working capital. We show how sectoral financial shocks propagate through the network and manifest at the aggregate level through two channels: a fall in total factor productivity and an aggregate labor wedge distortion. The strength of each channel depends on the overall network architecture and the location of shocks. Finally, we calibrate our model to the U.S. input-output tables and use it to quantitatively assess the role of the network multiplier within the context of the recent Financial Crisis and the Great Recession.

Chapter 1 provides a theory of financial frictions as a mechanism for uncertainty shocks to drive aggregate total factor productivity (TFP) fluctuations, and shows how input-output linkages amplify the impact of financial frictions on aggregate economy. Financial frictions distort the allocation of capital and induce capital wedges. These wedges differ across sectors due to different sectoral technology nature and financing demands. Sectors that have more financing needs are disproportionately vulnerable to financial frictions. This variation in distortion across sectors generates inefficiency in capital allocation and reduces aggregate measured TFP. Intermediate input shares amplify the effect of this inefficiency and generate larger TFP fluctuations. Chapter 2 estimates and calibrates the theoretical model in Chapter 1 into US data at 14 sectors, and quantifies the amplification magnitude from input-output linkages. Financial frictions can drive aggregate TFP fluctuations and play a crucial role when uncertainty shocks hit the economy. Adding input-output linkages can further amplify the effects of both TFP and uncertainty shocks. In particular, aggregate output drops an additional 84% under TFP shocks and an additional 40% under uncertainty shocks with input-output linkages. The amplification from input-output linkages is the key for capital misallocation to contributing significantly in generating TFP fluctuations. Furthermore, there are dramatic differences in sectors' sensitivities to a tightening of borrowing constraint. Compared to other sectors, an increase in the dispersion of the return to capital in the Finance sector has the largest impact on aggregate output. Chapter 3 settles the debate between Eeckhout (2004, 2009) and Levy (2009) and offers a simple but neglected explanation for the heavy tail observed in the city size distribution. Using the same data set and statistics reported in Eeckhout (2004), I show that U.S. city sizes are not lognormally distributed and that the upper tail indeed follows a power law. I then show that the aggregate city size distribution is a mixture of lognormal distributions, and that this mixture generates the heavy tail. Finally, I relax the independence assumption and show that city sizes are positively dependent in the data, and that the power law distribution is robust under this positive dependence.

I review the empirical and theoretical literature on trade credit. I identify two main strands of the literature, which focus, respectively, on expensive trade credit to financially constrained firms and on cheap trade credit to customers with high bargaining power. It emerges that trade credit is sometimes an instrument to mitigate financial frictions and ease access to external finance for customers. However, the ability to extend trade credit is also a source of comparative advantage for suppliers and can affect competition in downstream markets. I highlight the consequences of trade credit usage for monetary policy transmission and industrial structure and highlight new avenues of research, which take into account how trade credit affects the stability of production networks.

How does an economy's production structure determine its macroeconomic response to sectoral distortions? We study a static framework in which production is organized in an input-output network and firms' production decisions are distorted. We show how sectoral distortions manifest at the aggregate level via two channels: total factor productivity and the labor wedge. The strength of each channel depends jointly on the input-output structure and the distribution of shocks. Near efficiency, distortions generate zero first-order effects on TFP but non-zero first-order effects on the labor wedge; the latter we show to be determined by the sector's network "centrality." We apply the model to the 2008-09 Financial Crisis and find that the U.S. input-ouput network may have amplified financial distortions by roughly a factor of two relative to a counterfactual economy devoid of intermediate good trade.

This dissertation includes three chapters. The first chapter studies the impact of financial shocks and financial frictions on business cycle dynamics in China's economy. The second and third chapters focus on the driving force of structural change and its impact on aggregate fluctuations using an input-output network approach. In the second chapter, I study two questions: (i) How has the U.S. production network structure changed from 1970 to 2017? (ii) What impact does that have on aggregate fluctuations? This paper shows that a few industries, like Finance and Insurance and Professional Services, have become much more central input suppliers over time, while others, like Paper Manufacturing, have become far less important. Therefore, the third chapter considers the driving force behind such structural change. In particular, I study the question of what determines the size of an industry in a production network. China has been one of the world's fastest-growing economies over the past several decades and emerged quickly from the global financial crisis of 2008. Chapter 1, titled DO FINANCIAL SHOCKS DRIVE REAL BUSINESS FLUCTUATIONS IN CHINA, investigates to what extent financial shocks can shape business cycle fluctuations in China. Specifically, I document the cyclical properties of China's macroeconomy and financial market and show the procyclicality of dividend payout and the countercyclicality of debt repurchases with real GDP. To account for these features, I use the real business cycle model incorporating debt and equity financing developed by Jermann and Quadrini (2012) to study how the dynamics of macroeconomic and financial variables are affected by financial shocks in China. This paper finds that financial shocks contribute significantly to business cycle fluctuations in China and can account for over 60% of the variations in the growth rate of output, investment, hours worked, and debt repurchases. Hulton's Theorem states that the impact of an industry-specific shock on the aggregate economy is entirely captured by the size of this industry, regardless of its position in the production network. Chapter 2, titled THE IMPORTANCE OF INPUT-OUTPUT NETWORK STRUCTURE IN THE US ECONOMY, proposes the idea that the network structure in isolation plays an essential role in shaping GDP growth and growth volatility. First, I introduce a new measure of network structure named centrality dispersion and document that the U.S. production network has become sparsely connected from 1970 to 2017, where many industries relied on a few central input suppliers for production. Such changes are associated with slower GDP growth and higher volatility. To account for this evidence, I embed input-output linkages into a multisector real business cycle model and provide a nonlinear characterization of the impact of network structure quantified using centrality dispersion on the macroeconomy. Finally, I study model-implied relationships between production network structure, GDP growth, and growth volatility. The calibrated model accounts for approximately one-quarter of the variation in real GDP growth and 40% of GDP volatility, as observed in the data. Chapter 3, titled THE NETWORK ORIGIN OF INDUSTRY SIZE VARIATIONS, quantifies the origin of industry size variations using the features of a production network. In the analysis, I perform an exact variance decomposition of industry total sales into the supplier, buyer, and final demand components. The findings suggest that matching with many buyers in the network, especially many large buyers is essential in understanding industry size variations. More importantly, these buyer characteristics have become increasingly important in contributing to industry size variations over the 1967-2012 period. Finally, I provide new empirical evidence related to the decomposition results. The evidence reveals a strengthening negative correlation between industry size and the concentration of customer networks in the long run.

In this paper, we show how to combine data on input-output tables and recent insights from the theory of production networks in order to inform policy. We first describe the information contained in input-output tables compiled by statistical agencies, and show how to derive relevant statistics of production networks. We then discuss the implications of production networks for policy intervention in a series of domains, such as fiscal policy, industrial policy, or, finance. Finally, we present a quantitative exercise applied to French data in order to illustrate that production networks shape the overall impact of competition policy on the economy.

We study the impact of bank credit on firm productivity. We exploit a matched firm-bank database covering all the credit relationships of Italian corporations, together with a natural experiment, to measure idiosyncratic supply-side shocks to credit availability and to estimate a production model augmented with financial frictions. We find that a contraction in credit supply causes a reduction of firm TFP growth and also harms IT-adoption, innovation, exporting, and adoption of superior management practices, while a credit expansion has limited impact. Quantitatively, the credit contraction between 2007 and 2009 accounts for about a quarter of observed the decline in TFP.

In our daily life, almost every family owns a portfolio of assets. This portfolio could contain real assets such as a car, or a house, as well as financial assets such as stocks, bonds or futures. Portfolio theory deals with how to form a satisfied portfolio among an enormous number of assets. Originally proposed by H. Markowtiz in 1952, the mean-variance methodology for portfolio optimization has been central to the research activities in this area and has served as a basis for the development of modem financial theory during the past four decades. Follow-on work with this approach has born much fruit for this field of study. Among all those research fruits, the most important is the capital asset pricing model (CAPM) proposed by Sharpe in 1964. This model greatly simplifies the input for portfolio selection and makes the mean-variance methodology into a practical application. Consequently, lots of models were proposed to price the capital assets. In this book, some of the most important progresses in portfolio theory are surveyed and a few new models for portfolio selection are presented. Models for asset pricing are illustrated and the empirical tests of CAPM for China's stock markets are made. The first chapter surveys ideas and principles of modeling the investment decision process of economic agents. It starts with the Markowitz criteria of formulating return and risk as mean and variance and then looks into other related criteria which are based on probability assumptions on future prices of securities.