Community Research and Development Information Service - CORDIS

ERC

PERSISTDEBT Report Summary

Project ID: 648398
Funded under: H2020-EU.1.1.

Periodic Reporting for period 1 - PERSISTDEBT (Debt and Persistence of Financial Shocks)

Reporting period: 2015-07-01 to 2016-12-31

Summary of the context and overall objectives of the project

In 2007-08, Europe and US were overwhelmed by a financial crisis, followed by a severe, persistent economic recession (some countries are still below their economic activity before the start of the crisis, and the Euro Area recovered their pre-crisis level just last year). Prior to the crisis, there was a debt and asset-price boom. Historical studies show that this is the common pattern: (i) Financial crises are followed by a strong contraction of aggregate output and employment (and credit) and take a longer time to recover (than non-financial recessions); (ii) The best predictor of financial crises is an ex-ante strong credit boom (accompanied by high asset prices).

There is growing finance-macro theory and new policy, but no micro evidence. Therefore, in this project I study: Why are the effects of debt and financial shocks strong and persistent? What are the channels of transmission (households, banks, firms, sovereigns)? As crises are not exogenous, what are the determinants? Can public policy (macroprudential, monetary) alleviate the negative effects? Are there costs or limitations of these policies?

To study these issues, we are constructing several micro datasets that are absolutely new in the literature:
• Comprehensive loan-level data for households, with borrower identifier, matched with household-level info (income, wealth, house variables, consumption, labor…; both administrative data and survey data).
• Security-time-bank level data (i.e., both securities and credit registers): each security a bank has in each period with rating, price, maturity, yield, coupons, whether a market price… and all loans.
• Real effects (of credit channels and policies): matched loan level data with administrative firm-level (and supervisory bank-level) data.

For identification of the above questions, we exploit the loan borrower-lender and security-bank level data, matched with household-firm-bank level info is crucial, including data on randomized house allocation. We can then aggregate up the data to quantify the real effects and exploit different shocks and policy changes.

Bank-level, firm-level, or aggregated macro data cannot identify (or even analyze) the channels and can provide wrong results. We also analyse the biases that aggregate data can have for the above academic and policy questions, and why micro data is better for answering those questions.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

"The specific projects cover three parts:
1. Household credit
2. Securities and credit registers
3. Monetary and prudential policies, can other bank policies help?

On the first set of projects, as I will discuss below, there has been substantial progress on obtaining and cleaning the data. On the second and third, I already have papers that are forthcoming at top academic journals, see below.

In the existing academic literature, there are loan-level research papers (with borrower and lender identifier) for firms, but not for households. There is aggregated data on debt and real effects at county level for households, but this cannot identify credit availability, nor real effects as substantial heterogeneity in a county. Therefore, I am matching several household loan-level datasets with household-level info, in particular three different datasets:

Matched household credit register with household characteristics: We are starting to analyze the loan level data from Spanish credit register, including loan applications, volume, maturity, collateral, credit history and future defaults, with identifier of borrower and lender. We have matched this dataset with The Spanish Survey of Households Finance, there has been 5 waves since 2002, with household detailed information on not only income, wealth, debt…, but also consumption, saving, expectation on house prices, loan rejections. This project is with Olympia Bover and Gabriel Jimenez (both from Bank of Spain). We have the initial results in which we find that financial crises affect more some households via a reduction on the supply of credit, and now we are analyzing whether this mechanism has effects on real outcomes: consumption, employment, health, and other variables such as divorces. We plan to have the first draft this year, in 2017. I also have other projects with Bank of Spain data on the impact of prudential policies (including state-owned banks) on credit and on the real economy (my co-authors are, apart from Gabriel Jimenez, Jesus Saurina and Rafael Repullo).

Moreover, with Niels Johannesen (University of Copenhagen), we have constructed a matched dataset for each Danish household with their main bank, their total credit, and health outcomes, to study in detail how a financial crisis can truly real (real) effects. In Economics we argue that the real effects (compared to for example financial effects) are consumption and other related variables, but health effects stemming from a financial crisis can be argued that are more important (more real) than just a consumption reduction. The data is constructed and we have preliminary results. We believe we will have the first draft this year, 2017.

With several economists from Bank of England, I am analyzing the effects are of limits on mortgage debt for UK households depending on their income. Bank of England put restrictions on DTI (debt to income) and we have access to all mortgages in the UK at the household level, so we are analyzing this macroprudential tool. We are working on the datasets (household level and bank level datasets) and we plan to have the first draft next year, in 2018. The construction of the datasets can also allow us to answer other related questions, such as credit and household inequality, effects of monetary and prudential policies, and others.

Jose G. Montalvo has been working in the construction of the datasets of the core project of the proposal on the real effect of housing wealth on debt and the impact of debt on consumption, employment, health, etc. The process has been particularly slow and intense because in every step there have been some political hurdles to overcome. The basic data sources for this project are administrative data on the lotteries of public housing in the Basque Country. The reason for this choice is the wealth of information on lotteries’ participants and its demographic and economic conditions in this particular region. More than 250.000 participated in the lotteries of public housing during the period 2001-2008. During 2016 the university signed a General Collaboration Agreement and a Non Disclosure Agreement (NDA) with the Government of the Basque Country to get access to those data.

Some technical issues require special care as the fact that the people that could belong to the control group in a particular lottery (they did not get the house) may be fortunate in a posterior lottery. The basic dataset contains also information on the mortgage (size, interest rate, maturity) but only of about 20% of the beneficiaries of a public house. This is because they decided to ask for the mortgage through the public financial facility. Having only 20% of the characteristics of the mortgages obtained by the beneficiaries could be a potentially limiting feature of the study. For this reason, we contacted the Property Registry to ask them to match our dataset with the information on housing and characteristics of the registered mortgages. We still have to sign the General Collaboration Agreement and the NDA with the Property Registry.

In addition, in order to analyze the impact of debt on durable consumption we contacted the Direccion General de Trafico (Department of Motor Vehicles) to study the possibility of matching our data on public housing lotteries with the information on car owners (considering the history of car owning of individuals in the treated and control groups). At the end of 2016 the Subdirección General de Movilidad y la Subdireccion de Investigacion e Intervención approved the research proposal submitted for their collaboration. We hope to sign the General Collaboration Agreement for the matching and the use of the data during the first quarter of 2017.

Finally, as originally reported in the ERC proposal, we also want to analyze the effect of debt on health and, in particular on mental health. We contacted the Department of Health of the Basque Country to study the possibility of matching the individuals in the control-treated group with their history of hospitalization, drug usage, etc. This data effort is the less developed of all the avenues investigated up to this point. We have contacted with the technicians in charge of maintaining the health dataset. They confirm that the matching would be, in principle, possible and they looked quite excited about the possibility of doing such a large study. But, as in previous efforts to obtain administrative data, we still need to obtain the approval of the political instances.

Jose G. Montalvo has also been working on the empirical specification to apply to the data once the full dataset is constructed. This is relevant for the special characteristics of the data (successive lotteries with individuals who can potentially switch from the control group to the treated group in different periods of time.

I also have a set of projects analyzing securities registers, in addition to credit registers. History shows that credit and asset-price booms are the key ex-ante correlates of financial crises, and there is theory on banks’ credit crunch by investing in securities in crises. Policy restrictions on security-trading by banks: Volcker rule in Dobb-Frank in U.S.; the Liikanen Report in Europe, and the Vickers Report in the UK. Empirical analysis has been elusive due to lack of detailed micro, comprehensive data on activities of banks in security markets. We are working on securities register from Germany and Italy, and soon for the whole Europe, which are comprehensive security-time level info for each bank in conjunction with credit registers (including firm and bank level data), monthly data.

In "Securities Trading by Banks and Credit Supply: Micro-Evidence", (with Puriya Abbassi, Raj Iyer and Francesc Tous), Journal of Financial Economics (121 (3), 569-594 September 2016), we analyse securities trading by banks during the crisis and the associated spillovers to the supply of credit. We use a proprietary dataset that has the investments of banks at the security level for 2005-2012 in conjunction with the credit register from Germany. We find that – during the crisis – banks with higher trading expertise (trading banks) increase their investments in securities, especially in those that had a larger price drop, with the strongest impact in low-rated and long-term securities. Moreover, trading banks reduce their credit supply, and the credit crunch is binding at the firm level. All of the effects are more pronounced for trading banks with higher capital levels. Finally, banks use central bank liquidity and government subsidies like public recapitalization and implicit guarantees mainly to support trading of securities. Overall, our results suggest an externality arising from fire sales in securities markets on credit supply via the trading behaviour of banks. Moreover, we have new projects with the same data, for example one to understand whether the larger presence of non-bank financial intermediaries are changing the liquidity in financial markets, especially in crisis times, and how this can affect the real economy, and also whether banks manipulate and mask their risks to their supervisors (central banks). These papers are with Puriya Abassi from Bundesbank and Rajkamal Iyer from Imperial College London. I also work in these projects with several PhD students from UPF (Vladimir Manaev, Paul Soto, Dimitry Khametshin. I also have work in progress with German data with Stefan Gissler and Joachim Voth on the effects of the German banking crisis on the 1930s.

With Italian co-authors (Andrea Polo and Enrico Sette), we analyze securities register in Italy and analyze monetary policy. The potency of the bank lending channel of monetary policy may be limited if banks rebalance their portfolios towards securities, e.g. to pursue risk-shifting or liquidity hoarding. To test for the bank lending and risk-taking (reach-for-yield) channels, we therefore analyze banks’ securities trading, in addition to credit supply, in turn allowing us to also study the empirical relevance of key financial frictions. For identification, since the creation of the euro, we exploit the security and credit application registers owned by the central bank of Italy. In crisis times, we find that, with softer monetary policy, less capitalized banks prefer buying securities rather than increasing credit supply (not due to lack of good loan applications). Moreover, more –not less– capitalized banks reach-for-yield, which is inconsistent with the risk-shifting hypothesis. Results suggest that the main drivers at work are access to liquidity and risk-bearing capacity, and not regulatory capital arbitrage. Finally, in pre-crisis times, when financial frictions are limited, less capitalized banks do not expand securities holdings over credit supply. This paper is work in progress and we will have the first draft on March 2017. We are also working on new projects, for example on the impact of negative interest rates.

I have other work in progress based on the above questions using data from Australia, several American countries including Mexico and US, Uganda, and several European countries, including Romania, Norway, France and Turkey. I have two forthcoming papers based on my research on banks (prudential policies and international spillovers), which are "Macroprudential Policy, Countercyclical Bank Capital Buffers and Credit Supply: Evidence from the Spanish Dynamic Provisioning Experiments", (with Gabriel Jiménez, Steven Ongena and Jesús Saurina). Forthcoming at Journal of Political Economy; and “Capital Flows and the International Credit Channel”, (with Yusuf Soner Baskaya, Julian di Giovanni, Sebnem Kalemli-Özcan and Mehmet Fatih Uluk). Accepted at the Journal of International Economics."

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

We provide specific answers to understand the determinants and consequences of financial crises, and how public policy (e.g. prudential or monetary policy) can reduce the likelihood of crises, and conditioning on crises, the negative effects.

By exploiting randomized house allocation via lotteries: once financial crises or strong financial shocks arrive, we can analyse whether there is a (mortgage) debt overhang with negative effects on household-level consumption and employment, or, whether households with houses and debt try to work more as some other finance theories suggest. There are important results on health that we can identify, and as I was arguing above these could be key effects of strong financial shocks. Note that heterogeneity at the county level and non randomized data may provide misleading results, with misleading conclusions for public policy. We can also analyse whether supply or demand of mortgages and consumer lending is more important for the results, including LTV and DTI policy effects. Note that limits on household debt to income can have inequality effects, which we are investigating.

History shows that credit and asset-price booms are the key ex-ante correlates of financial crises, and there is theory on banks’ credit crunch by investing in securities in crises. Policy restrictions on security-trading by banks: Volcker rule in Dobb-Frank in U.S.; the Liikanen Report in Europe, and the Vickers Report in the UK. Therefore, I also have a set of projects analyzing securities registers, in addition to credit registers. We have already shown results that show an externality arising from fire sales in securities markets on credit supply to the real sector via the trading behaviour of banks. Moreover, we have new projects with the same data on whether banks manipulate and mask their risks to their supervisors (central banks). All these results are important on how to regulate and supervise banks.

Central banks have massively expanded their balance sheet since 2008, with main monetary rates around zero. However, the large injection of liquidity to banks may not have reached the real sector by means of expanded supply of credit. The potency of the bank lending channel of monetary policy may be limited if banks rebalance their portfolio towards securities holdings, e.g. to pursue liquidity hoarding or risk-shifting, as opposed to lending. For instance, in the words of Jeremy Stein (2013), Governor of the Federal Reserve Board: “A credit crunch may arise as other financial intermediaries (e.g., banks) withdraw capital from lending, so as to exploit the now-more-attractive returns to buying up fire-sold assets. Ultimately, it is the risk of this credit contraction, and its implications for economic activity more broadly, that may be the most compelling basis for regulatory intervention.”

To understand how monetary policy works via banks, including its possible limitations, and to test for the bank lending and risk-taking channels of monetary policy, it is thus crucial to analyse both the supply of bank credit to the real sector and securities trading by banks. Securities holdings by banks are a sizable fraction of their balance sheets, around 20% of assets in the US and Europe (e.g. in Germany and Italy), and several recent policy initiatives aim at limiting security trading by banks (Volker Rule in Dobb-Frank in the US, Likaanen Report in EU and Vickers’ report in the UK). A portfolio rebalance towards securities in crises may be the consequence of a credit demand problem, with few lending opportunities and with risky, highly leveraged borrowers. At the same time, the low level of bank capitalization can contribute to the impairment of the transmission of monetary policy to credit supply: banks, especially less capitalized ones, may e.g. decide to hoard liquid securities rather than issue relatively illiquid loans to SMEs.

Monetary policy may also have unintended consequences in terms of financial stability, e.g. Draghi (2015) argues that: “Our monetary policy measures are necessary to achieve our primary objective of maintaining price stability. But we are nevertheless aware that they may have unintended side effects on the financial system.” Low interest rates have been suggested as a driver of reach-for-yield, which may have contributed to amplify the credit cycle leading to the 2008 financial crisis, consistently with a risk-taking channel of monetary policy. Loan level data support this view prior to the crisis. Yet, in crisis times, risk-shifting incentives may be stronger, notably for less capitalized banks, which, in the face of the huge expansion of central banks’ balance sheets, may have reached-for-yield more easily and quickly by adjusting their securities holdings. Previous evidence, in part because of lack of data, has exclusively focused on analysing GIIPS sovereign debt, which indeed became risky during the crisis, and on the sovereign-bank nexus; however, other GIIPS non-sovereign securities and loans to firms may be riskier, offering higher yields, and are quantitatively more important. On the other hand, increasing credit supply to firms with higher yield and risk, which are more financially constrained, especially in crisis times, may be a desired monetary policy outcome. Therefore, it is crucial to analyse the impact of monetary policy via banks on the real economy with both securities and credit, and we have the first paper analysing it, with substantial implications for the design of policy, as shown in the results summarized in question 2.

Each central bank has a webpage for accessing and work with the data, for example the one on Germany that we published in the Journal of Financial Economics is on https://www.bundesbank.de/Navigation/EN/Bundesbank/Research/RDSC/rdsc.html?nsc=true&https=1; the one on Spain that it is forthcoming at the Journal of Political Economy will be in the journal website once it is on print: see http://www.journals.uchicago.edu/journals/jpe/forthcoming. For each paper we will put all the different web-addresses where the data can be accessed. I have also organized a conference (June 2016, Barcelona) on the bank credit and securities bringing leading researchers in this area, I am planning to repeat this year, as the feedback from all the researchers enhance the project and allows stronger public dissemination.
Though it was not on my proposal nor in my presentation in Brussels for the ERC grant, at this moment I am planning to write a book to a wide audience on central banks’ (monetary and prudential) policies, in great part based on my ERC research. I am discussing this possibility with Franklin Allen from Imperial College London to write the book together. We think we will write the book in 2017-19.
Record Number: 198516 / Last updated on: 2017-05-19