Research Tools and Real-Time Macroeconomic Data

The Macro Financial Modeling (MFM) project exists to foster research efforts that advance our understanding of the linkages between financial markets and the macroeconomy. Consequently, we are building a collection of resources that researchers in this area may find useful.

  • Real-Time Macroeconomic Data: The MFM group cannot disseminate proprietary data; however, the community become an online repository of datasets useful for replicating the research results of others. Our affiliated scholars always encourage authors to share their data as far as licensing permits, and we will post such data on this website in data section below.
  • Software and Code: Below we provide key software and code that may be of interest to researchers.

Research Tools and Software

MFR Suite

The MFR Suite provides a collection of Python modules for conducting analysis in macro-finance.

The MFR Suite provides a collection of Python modules for conducting analysis in macro-finance. In particular, it provides the model solution to the framework developed in Hansen, Khorrami, and Tourre (2018). In addition, it provides two independent modules to compute stationary density and shock elasticities (see Term Structure of Uncertainty in the Macroeconomy and Shock Elasticities and Impulse Responses).

The MFR Program would like to thank project associate Joseph Huang for developing the MFR Suite. We would like to gratefully acknowledge the Macro Financial Modeling project through the generous financial support from the Alfred P. Sloan Foundation and Fidelity Investments and to thank Amy Boonstra, former MFM Executive Director, for her unconditional support. For their feedback, the MFR Program thanks Yu-Ting Chiang (University of Chicago), Jian Li (University of Chicago), Simon Scheidegger (HEC Lausanne), Elisabeth Proehl (University of Amsterdam), and conference participants at the 2nd MMCN, PASC18, University of Zurich, Northwestern University, and participants at the Economic Dynamics Working Group at the University of Chicago. The MFR Program also would like to thank the Research Computing Center at the University of Chicago (RCC) for their guidance on high performance computing, in particular Peter Carbonetto and Hossein Pourreza.

To report issues or suggest improvements for the MFR Suite team, submit feedback here.

Shock Elasticities Toolbox in Dynare

The Shock Elasticities toolbox is based on research by Jaroslav Borovicka, Lars Peter Hansen, and Jose Scheinkman.

Shock Elasticities Toolbox in Dynare

In their paper, Hansen and Borovicka develop new methods for representing the asset pricing implications of stochastic general equilibrium models. The authors provide asset pricing counterparts to impulse response functions and the resulting dynamic value decompositions (DVDs). These methods quantify the exposures of macroeconomic cash flows to shocks over alternative investment horizons and the corresponding prices or investors’ compensations. Hansen and Borovicka extend the continuous-time methods developed in Hansen and Scheinkman (2012) and Borovička et al. (2011) by constructing discrete-time, state-dependent, shock exposure and shock price elasticities as functions of the investment horizon. Their methods are applicable to economic models that are nonlinear, including models with stochastic volatility. See Examining Macroeconomic Models Through the Lens of Asset Pricing by Lars Peter Hansen and Jaroslav Borovicka

 

Smolyak Method Toolbox

The Smolyak Method Toolbox is based on research by Kenneth L. Judd, Lilia Maliar, Serguei Maliar and Rafael Valero.

Smolyak Method Toolbox

In this research, Judd, Maliar, Maliar and Valero show efficient implementation of the Smolyak method that avoids costly evaluations of repeated basis functions used in the conventional Smolyak formula. Also, the conventional Smolyak method is extended to include anisotropic constructions and an adaptive domain. The code is illustrated by simple examples and is portable to other applications. A MATLAB code of the Smolyak method is provided for download.

A Macroeconomic Model with Financial Panics Toolbox

Macroeconomic Model with Financial Panics Toolbox

See “A Macroeconomic Model with Financial Panics,” with Mark Gertler, Nobuhiro Kiyotaki, and Andrea Prestipino. Review of Economic Dynamics, Special Issue: “Frontier of Business Cycle Research”, 2020, forthcoming.

This paper and its associated toolbox incorporates banks and banking panics within a conventional macroeconomic framework to analyze the dynamics of a financial crisis of the kind recently experienced. We are particularly interested in characterizing the sudden and discrete nature of banking panics as well as the circumstances that makes an economy vulnerable to such panics in some instances but not in others. Having a conventional macroeconomic model allows us to study the channels by which the crisis affects real activity both qualitatively and quantitatively. In addition to modeling the financial collapse, the authors also introduce a belief driven credit boom that increases the susceptibility of the economy to a disruptive banking panic.

As many have argued, at the heart of the Great Recession was a financial crisis featuring a series of bank runs that culminated in the precipitous demise of a number of major financial institutions. Associated with this financial meltdown was a sharp contraction in real economic activity. Accordingly, this paper incorporates banks and banking panics within a conventional macro-economic framework to analyze the dynamics of a financial crisis of the kind recently experienced. Gertler, Kiyotaki, and Prestipino are particularly interested in characterizing the sudden and discrete nature of banking panics as well as the circumstances that makes an economy vulnerable to such panics in some instances but not in others. Having a conventional macroeconomic model allows them to study the channels by which the crisis affects real activity both qualitatively and quantitatively. In addition to modeling the financial collapse, the authors also introduce a belief driven credit boom that increases the susceptibility of the economy to a disruptive banking panic. The authors then show how the model can both qualitatively and quantitatively capture the recent financial crisis along with the spillover to real activity.

On the methodological side, Gertler, Kiyotaki and Prestipino develop and numerically solve a highly nonlinear macroeconomic model. The main source of nonlinearity arises from the feature that in some regions of the state space the banking system is susceptible to sunspot panics (specifically when bank balance sheets are weak), while in others it is immune (when bank balance sheets are strong.). In this way the framework is consistent with the evidence that financial panics tend to be more likely when financial institutions are highly leveraged.

Epsilon Distinguishable Set (EDS) method and Cluster Grid Algorithm (CGA) Toolbox

This toolbox is based on research by Lilia Maliar and Serguei Maliar.

Epsilon Distinguishable Set (EDS) method and Cluster Grid Algorithm (CGA) Toolbox

In their paper, Lilia Maliar and Serguei Maliar introduce epsilon distinguishable set (EDS) method and cluster grid algorithm (CGA) for solving dynamic economic models on the ergodic set and illustrate the application of the proposed methods in the context of one- and multi-sector dynamic stochastic general equilibrium (DSGE) models, Also, these methods are used to construct an accurate nonlinear solution to a medium-scale new Keynesian model with a zero lower bound on the nominal interest rate. The EDS and CGA codes are available for download.

Credit Booms, Banking Crises and Macroprudential Policy Toolbox

This toolbox is based on research by with MFM Working Group Members Mark Gertler and Nobuhiro Kiyotaki, and co-author Andrea Prestipino forthcoming in the Review of Economic Dynamics' special issue on, "Frontier of Business Cycle Research", 2020.

This paper and its associated toolbox extends the authors’ model of banking panics to consider macroprudential policy. Gertler, Kiyotaki, and Prestipino first adapt the model to account for two important features of the data relevant to macroprudential policy design. First, banking crises are usually preceded by credit booms. Second, credit booms often do not result in crises. That is, there are “good booms” as well as “bad booms” in the language of Gorton and Ordonez (2019). Further, a policy-maker cannot be certain at any moment in time whether a credit boom is good or bad. The authors then consider how the optimal macroprudential policy weighs the benefits of preventing a crisis against the costs of stopping a good boom. Within their model, there are two externalities that can motivate regulatory intervention. The first is the standard pecuniary externality, where bankers do not account for the impact of their leverage decisions on asset price behavior. The second, which the authors think is new, is that bankers do not take account of their leverage decisions on the probability of a panic. Gertler, Kiyotaki, and Prestipino then show that counter-cyclical capital buffers are a critical feature of a successful macroprudential policy. As with their earlier paper, the authors set up and solve a highly nonlinear macroeconomic framework. They also compute the welfare gains from various macroprudential policies.

Systemic Risk Measures Toolbox

The Systemic Risk Measures toolbox is based on research by MFM Project Director Andrew Lo, Dimitrios Bisias, Mark Flood, and Stavros Valavanis.

Systemic Risk Measures Toolbox

In their paper, Dimitrios Bisias, Mark Flood, Andrew Lo, and Stavros Valavanis survey over thirty quantitative measures of systemic risk, which were discussed at the 2012 September meeting of the MFM group. MATLAB code and its documentation for implementing these measures can be downloaded here.

Volatility Lab

The Volatility Institute (V-Lab) provides real time measurement, modeling and forecasting of financial volatility, correlations and risk for a wide spectrum of assets.

The Volatility Institute (V-Lab) provides real time measurement, modeling and forecasting of financial volatility, correlations and risk for a wide spectrum of assets. V-Lab blends together both classic models as well as some of the latest advances proposed in the financial econometrics literature. The aim of the website is to provide real time evidence on market dynamics for researchers, regulators, and practitioners. View here.

Continuous Time Shock Elasticities Toolbox

See Shock Elasticities and Impulse Responses and Term Structure of Uncertainty in the Macroeconomy.

This toolbox is designed to compute shock elasticities for a flexible number of dimensions. It offers the option to use Pardiso, a high performance computing tool, to solve the linear PDE in shock elasticities computations. This toolbox allows the use to compute stationary density through simulations (in parallel processes by default).

Piecewise Linear Approximations and Filtering for DSGE Models with Occasionally-Binding Constraints Toolbox

Paper and associated software by MFM Working Group Member, Frank Schorfheide (University of Pennsylvania, MFM Working Group Member), Boragan Aruoba (University of Maryland), and Pablo Cuba-Borda (Board of Governors)

Motivation and Goal

Dynamic stochastic general equilibrium (DSGE) models with financial frictions are widely used in central banks, by regulators, and in academia to study the effects of monetary and macroprudential policies and the propagation of shocks in the macro economy. The most recent vintage of these models involves occasionally binding constraints arising from financial frictions and the effective lower bound on nominal interest rates. In order for these models to be useable for a quantitative analysis they need to be solved numerically and their parameters need to be estimated based on historical data.

There exists a small literature on solving DSGE models with occasionally-binding constraints using global projection methods, but, unfortunately, the solution techniques developed in these papers are numerically very difficult to implement, even on models with a relatively small state space. The goal of this project is to provide solution and estimation methods that trade off a bit of accuracy against computational speed and scalability to models with a relatively high-dimensional state space, e.g., the widely used Smets and Wouters (2007) model.

Outcomes

The PIs implemented a piecewise linear continuous (PLC) for a prototypical New Keynesian DSGE model in the public-domain scientific computing language JULIA. They also implemented a conditionally-optimal particle filter that can be used to approximate the likelihood function of this nonlinear model in an efficient manner and to extract latent variables such as a shadow nominal interest rate or an equilibrium real interest rate. The likelihood-function approximation can be used to estimate model parameters in a way that enables the model to track historical time series.

Careful speed and accuracy assessments of the model solution and filtering procedure in the context of a prototypical DSGE model are provided. The solution and filtering procedures have been integrated into a model estimation module. The research has resulted in the working paper “Piecewise-Linear Approximations and Filtering for DSGE Models with Occasionally-Binding Constraints” which is available here. The paper is currently under review for publication in the Review of Economic Dynamics. A revised version as well as carefully documented code are expected to be available by September 15, 2020 and will also be disseminated through the MFM website.

Real-Time Macroeconomic Data

Federal Reserve Bank of Philadelphia Database

Macroeconomic data are frequently revised by statistical agencies; this can complicate quantitative analysis, because it means researchers today may have better information about past conditions than policymakers at the time did. Croushore & Stark (2001) and Orphanides (2001) highlight the importance of using real-time data when analyzing macroeconomic policy and outcomes

The Federal Reserve Bank of Philadelphia maintains a well-documented database of real-time macroeconomic data vintages, available for free on their website.

Bank for International Settlements Database

The Bank for International Settlements (BIS) is an international organization of central banks that aims to increase cooperation and transparency among governments in the conduct of monetary policy. The BIS collects and disseminates a wide variety of international financial data, available here.

Bureau of Economic Analysis Database

The Bureau of Economic Analysis (BEA) tabulates data on U.S. output, personal income, and balance of payments. Their data are available for download here.

Bureau of Labor Statistics Database

The Bureau of Labor Statistics (BLS) collects and analyzes data on U.S. prices, unemployment, and working conditions. Their data are available for download here.

Congressional Budget Office Database

The Congressional Budget Office (CBO) produces independent analyses of budgetary and economic issues to support the Congressional budget process. The agency is strictly nonpartisan and conducts objective analyses. Their data are available here.

European Central Bank Database

The European Central Bank (ECB) is responsible for conducting monetary policy for the euro area—the world’s largest economy after the United States. The ECB maintains a Statistical Data Warehouse on its website, available here.

The Federal Reserve Board Database

The Federal Reserve Board (FRB) is the central governing body of the Federal Reserve System, which is responsible for conducting monetary policy in the U.S. The FRB collects and releases data on the U.S. Flow of Funds accounts, as well as various interest and exchange rates; their data are available here.

Office of Financial Research Database

The Office of Financial Research (OFR) is a department of the U.S. Treasury created by the Dodd-Frank Act to improve the quality of financial data and facilitate analyses of the financial system. Their data standards page can be found here.

Methods and Findings

Micro-evidence from a System-wide Financial Meltdown: The German Crisis of 1931

Research by MFM Working Group Member, Markus Brunnermeier, Kristian Blickle and Stephan Luck

What determines the stability of a bank during a system-wide bank run? Which banks can withstand deposit outflows successfully when an entire banking system is in distress? The theoretical literature has made great progress on the understanding of the fragility of financial institutions in the light of maturity and liquidity mismatch. However, the empirical understanding of bank runs is—due to the lack of better data—mostly confined to studying the runs on individual institutions.

In this paper, Brunnermeier, Blickle, and Luck use hand-collected detailed monthly bank balance sheet digitized from the newspaper “Deutscher Reichs und Preussischer Staatsanzeiger” to study the determinants of bank stability during a major bank run, the German Banking Crisis during the summer of 1931. Their empirical approach exploits the fact that the German Crisis of 1931 was system-wide with cross-sectional variation in deposit flows as well as bank distress and took place in absence of a deposit insurance scheme. The authors find that interbank deposit flows predict subsequent bank distress early on. In contrast, wholesale depositors are more likely to withdraw from distressed banks at later stages of the run and only after the interbank market has started to collapse. Retail deposits are—despite the absence of deposit insurance—largely stable. Their findings emphasize the heterogeneity in depositor roles, with discipline being best provided through the interbank market.

On the Equivalence of Private and Public Money Journal of Monetary Economics

Research by MFM Working Group Member, Markus Brunnermeier and Dirk Niepelt

When does a swap between private and public money leave the equilibrium allocation and price system unchanged?

To answer this question, this paper sets up a generic model of money and liquidity which identifies sources of seigniorage rents and liquidity bubbles. Brunnermeier and Niepelt derive sufficient conditions for equivalence and apply them in the context of the “Chicago Plan”, cryptocurrencies, the Indian de-monetization experiment, and Central Bank Digital Currency (CBDC). Their results imply that CBDC coupled with central bank pass-through funding need not imply a credit crunch nor undermine financial stability.

The Benchmark Inclusion Subsidy

Research by MFM Working Group Member, Anil Kashyap and co-authors Natalia Kovrijnykh, Jian Li, and Anna Pavlova

In this paper, Kashyap, Kovrijnykh, Li and Pavlova argue that a common practice of evaluating portfolio managers relative to a benchmark has real effects. Benchmarking generates additional, inelastic demand for assets inside the benchmark. This leads to a “benchmark inclusion subsidy:” a firm inside the benchmark values an investment project more than the one outside. The same wedge arises for valuing M&A, spinoffs, and IPOs. This overturns the standard corporate finance result that an investment’s value is independent of the entity considering it.

The authors describe the characteristics that determine the subsidy, quantify its size (which could be large), and identify empirical work supporting their model’s predictions.

Banks’ Risk Exposures

Research by MFM Working Group Members Monika Piazzesi, Martin Schneider and co-author Juliane Begenau

In this paper, Begenau, Piazzesi and Schneider measure the exposure of banks to interest rate and credit risk. The authors combine two pieces of information. First, they exploit the strong factor structure in fixed income instruments to estimate the exposure of individual instruments to interest rate and credit risk. The two factors explain the large majority of variation in returns on these instruments. Second, they use regulatory data from the Call Reports to measure the holdings of individual US banks of these instruments. Finally, they develop a Bayesian estimation approach to extract the exposure of individual banks’ derivative holdings, which consist mostly of swaps.

Associated Code

The codes for the first paper are on the website of the Journal of Monetary Economics. The authors are in the process of updating the data and finalizing the estimation for the “Banks’ Risk Exposures.” This process will be completed soon. The authors have received many requests for their codes that load the individual bank data and perform the estimation. The authors will post all codes with instructions on how to download the interest rate series and regulatory data on their respective websites in the foreseeable future.

Systemic Risk 10 Years Later

Research by MFM Working Group Member, Robert Engle published in Annual Review of Financial Economics (2018)

This paper surveys the research from the Volatility Institute since the financial crisis. This research introduces the statistical measure, SRISK which is a market-based measure of undercapitalization of major banks. This measure has been computed and published on the internet at v-lab since it was introduced in about 2010. However, it has undergone several revisions and modifications, many of which are widely used. This survey gives the fundamental econometrics behind these measures and surveys the relation between them. It also gives the first available measures of statistical confidence intervals around the measure. Finally, this paper reports on an econometric model designed to measure how high levels of SRISK lead to a financial crisis and consequently, when is there a risk. It calculates the probability of a financial crisis from the model.

This research was reported at an MFM meeting and then in a major conference in 2008 which included, Bernanke, Fischer and Trichet and many other influential people who participated in different ways in the financial crisis. It was dedicated to reflection on the 10 years since the crisis.

Measuring the Probability of a Financial Crisis

Research by MFM Working Group Member, Robert Engle, NYU Stern, and co-author Tianyue Ruan, National University of Singapore

This paper develops an econometric model of the process by which undercapitalized banks in an effort to strengthen their balance sheets by deleveraging, actually precipitate a financial crisis. The paper uses historical data on the aggregate SRISK from 23 OECD developed countries to model the causes of a financial crisis. The measure used is a classification of financial stress extracted from OECD reports by Romer and Romer on a scale of 0 to 15 from no financial stress to extremely severe financial stress. Because a large fraction of the observations are zeros, the panel model is estimated using the Tobit estimator. Using country fixed effects and aggregate measures of global financial undercapitalization, the model produces an estimate of the probability that the country is in a financial crisis as a function of its own SRISK and global measures. It also delivers a measure of the capacity of each country to endure undercapitalized financial institutions without having a probability of crisis greater than ½.

The key findings are that high levels of domestic SRISK raise the probability of crisis. However, this probability is also affected by the levels of SRISK globally. When the rest of the world is well capitalized, domestic risk is less dangerous. When the rest of the world is already undercapitalized, the crisis risk rises dramatically.

Read Associated Paper: “How Much SRISK Is Too Much?”

OTC Intermediaries

Research by MFM Working Group Member, Andrea Eisfeldt and co-authors Bernard Herskovic, Sriram Rajan and Emil Siriwardane

The paper is aimed at investigating the role of the network structure of the market for over-the-counter derivatives on average prices and price dispersion during normal times, and on the quantitative effect of dealer exit on prices.

To this end, Eisfeldt, Herskovic, Rajan and Siriwardane build a structural model of the market for trading Credit Default Swaps (CDS), and match its features to the actual CDS market using proprietary data from the Deposit Trust and Clearing Corporation, which the authors access through the Office of Financial Research. They find that the credit exposure of the core dealer sector, and of individual dealers, plays a crucial role in credit pricing and risk sharing. In normal times, the dealer sector has a credit-risk-bearing capacity above that of the overall economy, and dealers thus provide credit protection to the rest of the economy. Prices are set bilaterally as a function of counterparties’ post trade exposures, and since most trades involve the dealer sector, average prices are actually lower than in a Walrasian Market. Importantly, however, this reflects imperfect risk sharing, rather than an overall higher economy-wide risk-bearing capacity.

One of their most interesting findings describes what happens to CDS prices upon the exit of a key dealer. In their model, a key dealer is not simply one who is well-connected. In the data, most dealers are essentially fully connected, and the authors’ model matches this feature. Instead, what determines the importance of a dealer for system-wide pricing is their individual role in determining the net credit insurance provision capacity of the entire dealer sector. Eisfeldt et al. find that in position-level data, there are a large number of essentially neutral dealers, a set of dealers who actually demand insurance from the system, and a smaller number of dealers who provide a very substantial amount of credit protection to the overall market. Removing one of these last type of dealers from the system has a quantitatively large impact on prices, leading to an increase in credit spreads as high as 20%. This is because eliminating one of these dealers can change the overall dealer sector from being a net provider of credit protection to being a net demander of credit insurance, and it is the risk-bearing capacity of the OTC intermediary sector that determines the level of credit risk prices in the economy they study.

The authors are currently revising the paper to incorporate the helpful feedback journal referees. They have also presented this work at numerous conferences, including the Laboratory for Aggregate Economics and Finance OTC Conference, the Women in Macroeconomics Conference, the Society for Economic Dynamics Annual Meeting, the Conference on OTC Markets and Their Reform by the SNSF, the NBER Conference on Financial Markets and Regulation, the Chicago Booth Recent Advances in Empirical Asset Pricing conference, the Macro Financial Modeling Annual Meeting, the Rome Junior Finance Conference, the Lubrafin Conference, the Maryland Junior Finance Conference, the NBER 2018 Summer Asset Pricing Meetings, the Junior Workshop in New Empirical Finance at Columbia University, the FMA Conference on Derivatives and Volatility, and the Western Finance Association Annual Meeting.  Similarly, we have presented this work at several seminars, including UCLA Macro Finance lunch, UCLA Finance Brown Bag, Pontificcal Catholic University of Rio de Janeiro (PUCRio), University of Washington, Central Bank of Chile, Federal Reserve Bank of Dallas, Federal Reserve Board, and the Office of Financial Research.

The financing of local government in China: Stimulus loan wanes and shadow banking waxes

Research by MFM Working Group Member, Zhiguo He and co-authors Zhuo Chen and Chun Liu

The upsurge of shadow banking is typically driven by rising financing demand from cer- tain real sectors. In China, the 4 trillion yuan stimulus package in 2009 was behind the rapid growth of shadow banking after 2012, expediting the development of Chinese corporate bond markets in the post-stimulus period. Chinese local governments financed the stimulus through bank loans in 2009 and then resorted to nonbank debt financing after 2012 when faced with rollover pressure from bank debt coming due. Cross-sectionally, using a political-economy-based instrument, the authors show that provinces with greater bank loan growth in 2009 experienced more municipal corporate bond issuance during 2012–2015, together with more shadow banking activities including trust loans and wealth management products. China’s post-stimulus experience exhibits similarities to financial market development during the US National Banking Era.

In this paper, the authors study the unintended, both good and bad, consequences of the 2009 four-trillion stimulus package in China.

Chinese Bond Markets and Interbank Market

Research by MFM Working Group Member, Zhiguo He and his co-author, Marlene Amstad

China’s interbank market has experienced rapid growth since Beijing launched the unprecedented 4-trillion RMB stimulus plan in 2009. Together with the exchange market located in Shanghai and Shenzhen, it hosts the second largest bond market in the world, behind only the United States.

MFM Working Group Member, Zhiguo He and his co-author, Marlene Amstad have contributed a handbook chapter, “Chinese Bond Markets and Interbank Market,” for “The Handbook of China’s Financial System,” edited by Marlene Amsted, Wei Xiong, and Guofeng Sun.

In this chapter, the authors cover a wide range of topics of Chinese bond market during 2008-2019, including its connection to shadow banking, recent trend of regulation, credit rating system, as well as surging corporate bond default. This handbook is forthcoming and to be published by Princeton University Press.

Cross-Country Dynamic Panel Models of Macro-Financial Interactions

Research by MFM Working Group Member, Christopher A. Sims and his co-author, Dake Li

This project aims at developing a multi-country panel structural vector autoregression (SVAR) model of the interaction of monetary policy, credit aggregates, and the macro-economy. The motivation for the work is similar to that which led to Brunnermeier, Palia, Sastry, and Sims (2018), which presented estimates of a moderately large SVAR model of monthly US data. In the wake of the financial crisis of 2007-9, economists have recognized that the empirical macro-economic models up to that time had paid too little attention to the interactions of the financial sector with standard macroeconomic outcomes and policy variables, like unemployment, GDP, inflation, and interest rates. While some work has proceeded to integrate financial variables into the widely used New Keynesian models, the theoretical foundations of these model extensions is fragile, and they have tended to look for one-dimensional measures of financial influence on the economy.

SVAR’s make minimal assumptions on the economic structure, while still separating monetary policy effects from disturbances arising in the private sector. They are appealing when looking for connections between the financial sector and non- financial macro variables because, in this environment where firm a priori knowledge of the nature of these connections is lacking, SVAR’s potentially can give Sim and Li a picture of the causal structure where the role of restrictive assumptions is both transparent and minimal.

Of particular interest in Brunnermeier, Palia, Sastry, and Sims (2018) was the finding that credit aggregates for the most part co-move positively with output and inflation, with any negative connection between credit aggregate growth and subsequent output growth weak, accounting for little variation in either credit or output. This result is controversial, as other researchers, using single-equation approaches that do not account for two-way causality, have claimed to find strong negative relations between credit growth and output growth. One criticism of the Brunnermeier, Palia, Sastry, and Sims (2018) results is that they are based entirely on US data, whereas much other work in the area has looked at multiple countries. The research in this project adapts the SVAR methodology of the US model to a collection of 25-30 countries.

There is previous work on fitting VAR’s (descriptive vector autoregressions that do not attempt to give formal causal interpretations) to panel data, but none, as far as the authors know, that have estimated SVAR’s on panel data, using the approach to identification they are using. Their work, therefore, has two potential useful outputs: the empirical results on international panel data, and the software, which is in the process of becoming an R “package”, i.e. a documented collection of interrelated R code.

The expansion of the data set associated with this research increases the number of parameters being estimated in the non-linear part of the estimation to 198. This has brought out numerical problems in evaluating the likelihood. This has slowed the estimation by requiring frequent re-initializations of the optimization program that maximizes the posterior density. At the same time, the authors have been improving the code to minimize the impact of this numerical issue. The estimation is still underway as this is being written. It has not yet converged tightly, but the estimates look reasonable, make economic sense, and are qualitatively similar to those presented at the MFM conference. The results confirm that

  • their identification methods display a clearly defined monetary policy disturbance;
  • credit aggregates respond in the same direction as GDP following every source of disturbance;
  • credit disturbances are accompanied by positive output growth, followed by a weak but persistent negative output growth, over the span of 5 years or so.

This, in the author’s view, accounts for the finding of negative relations between credit growth and output growth in previous literature.

The model assumes no correlation of disturbances across countries, but the estimated disturbances show some correlation. This is reduced but not eliminated when Sims and Li allow each country to respond to US interest rates. The authors have made some time-consuming, but unsuccessful so far, efforts to expand the modeling of cross-country impacts to account for the estimated cross-correlations. This greatly complicates the code, however, and so far has not eliminated the cross-correlations.

Financing in the Shadow of Banks: China’s Interbank Market

Research by MFM Working Group Members Hui Chen and Zhiguo He

Financing in the Shadow of Banks: China’s Interbank Market

by Hui Chen and Zhiguo He

China’s interbank market has experienced rapid growth since Beijing launched the unprecedented 4-trillion RMB stimulus plan in 2009. Together with the exchange market located in Shanghai and Shenzhen, it hosts the second largest bond market in the world, behind only the United States. Thanks to the generous support of the MFM, we have been actively working on several projects regarding the functioning of China’s bond markets and the fast-growing interbank market, as well as their increasingly important roles in Chinese economy. Here is the list of papers, datasets, and data access guide that we have produced under the MFM project.

1) Zhiguo He has contributed a handbook chapter, “Chinese Bond Markets and Interbank Market,” for “The Handbook of China’s Financial System,” edited by Marlene Amsted, Wei Xiong, and Guofeng Sun, available at https://www.chinafinancialsystem.com/. In this chapter, we cover a wide range of topics of Chinese bond market during 2008–2019, including its connection to shadow banking, recent trend of regulation, credit rating system, as well as surging corporate bond default. This handbook will be published by Princeton University Press, scheduled at the second half of 2020.
2) In “The Financing of Local Government in China: Stimulus Loan Wanes and Shadow Banking Waxes” Journal of Financial Economics 137 (2020), 42-71, Zhiguo He and coauthors study the unintended (both good and bad) consequences of the 2009 four-trillion stimulus package in China.
3) Together with three coauthors resided in China, we have finished a working paper “Pledgeability and Asset Prices: Evidence from Chinese Corporate Bond Market.” By exploiting the unique institutional feature of dual-listed bonds in Chinese bond markets, this paper establishes the causal impact of pledgeability on asset pricing and estimates the pledgeability premium. Moreover, our estimate helps shed light on the degree of financial constraint of those institutions that heavily rely on repo financing. The paper has been presented in numerous conferences, including the NBER Summer Institute (2019, 2020) and AFA (2020), and received the Arthur Warga Award by the Society for Financial Studies. We have submitted this paper for journal publication.
4) We have created a website for the most recent update on Chinese corporate bond market, hosted on Zhiguo He’s personal website: https://voices.uchicago.edu/zhiguohe/data-and-empirical-patterns/chinese-bond-market-data/. On this website, people can find the summary of corporate defaults since the first default occurred in 2014, the up-to-date monthly credit spreads for various ratings groups, as well as the pricing wedge for dual-listed bonds studied in the working paper mentioned in 3).
5) To help facilitate access to the Chinese corporate bond market data, we are preparing a detailed data access guide, including the following:
a. Data sources for information on bond and issuer characteristics;
b. Data sources for daily pricing and transaction information on both the interbank (through the China Foreign Exchange Trading System) and exchange markets (through Wind Information Co.);
c. Data source for repo eligibility and haircuts on the exchange market.
We plan to make our cleaned version of the Chinese corporate bond pricing database used for the paper in 3) available to public after its publication. This dataset carefully addresses various issues, including the aggregation of multiple ratings from different agencies, and adjustment for bond prices for features such as sinking fund and callability. In addition, we will also make a hand-collected dataset on the aggregate interbank market statistics (through the People’s Bank of China Statistics Reports) available to the public.

What Happened: Financial Factors in the Great Recession

Research by Simon Gilchrist, New York University, and Mark Gertler, NYU

“What Happened: Financial Factors in the Great Recession” in the Journal of Economic Perspectives – Volume 32, Number 3, Summer 2018 —Pages 3–30 – view paper online here

At least since the Great Depression, major economic calamities have altered the course of research in macroeconomics. The recent global financial crisis is no exception. At the onset of the crisis, the workhorse macroeconomic models assumed frictionless financial markets. These frameworks were thus not able to anticipate the crisis, nor to analyze how the disruption of credit markets changed what initially appeared like a mild downturn into the Great Recession. Since that time, an explosion of both theoretical and empirical research has investigated how the financial crisis emerged and how it was transmitted to the real sector. The goal of this paper is to describe what we have learned from this new research and how it can be used to understand what happened during the Great Recession. In the process, Gertler and Gilchrist also present some new empirical work.

This paper is organized into three main parts. The authors begin with an informal description of the basic theory and concepts, including new developments. This work emphasizes the role of borrower balance sheets in constraining access to credit when capital markets are imperfect. Much of the pre-crisis research focused on constraints facing nonfinancial firms. The events of the Great Recession, however, necessitated shifting more attention to balance sheet constraints facing households and banks. In addition, the crisis brought into sharp relief the need to capture the nonlinear dimension of the financial collapse, prompting a new wave of research.

The next section describes the main events of the financial crisis through the lens of the theory. To tell the story, Gertler and Gilchrist also make use of the new wave of empirical research that has sharpened our insights into how the crisis unfolded. In this regard, the literature has been somewhat balkanized with some work focusing on household balance sheets and others emphasizing banks. The authors argue that a complete description of the Great Recession must take account of the financial distress facing both households and banks and, as the crisis unfolded, nonfinancial firms as well.

Gertler and Gilchrist then present some new evidence on the role of the household balance sheet channel versus the disruption of banking. The authors examine a panel of quarterly state-level data on house prices, mortgage debt, and employment along with a measure of banking distress. Then exploiting both panel data and time series methods, we analyze the contribution of the house price decline, versus the banking distress indicator, to the overall decline in employment during the Great Recession. Gertler and Gilchrist confirm a common finding in the literature that the household balance sheet channel is important for regional variation in employment. However, they also find that the disruption in banking was central to the overall employment contraction.

Macroeconomic Models for Monetary Policy: A Critical Review from a Finance Perspective

Research by Winston Dou, The Wharton School, University of Pennsylvania, Andrew W. Lo Massachusetts Institute of Technology (MIT) - Sloan School of Management; Santa Fe Institute, Ameya Muley, MIT and Harald Uhlig, University of Chicago - Department of Economics

We provide a critical review of macroeconomic models used for monetary policy at central banks from a finance perspective. We review the history of monetary policy modeling, survey the core monetary models used by major central banks, and construct an illustrative model for those readers who are unfamiliar with the literature. Within this framework, we highlight several important limitations of current models and methods, including the fact that local-linearization approximations omit important nonlinear dynamics, yielding biased impulse-response analysis and parameter estimates. We also propose new features for the next generation of macrofinancial policy models, including: a substantial role for a financial sector, the government balance sheet and unconventional monetary policies; heterogeneity, reallocation, and redistribution effects; the macroeconomic impact of large nonlinear risk-premium dynamics; time-varying uncertainty; financial sector and systemic risks; imperfect product market and markups; and further advances in solution, estimation, and evaluation methods for dynamic quantitative structural models.

View Research Paper