XXV Meeting of the Central Bank Researchers Network
Digital Event, October 28-30, 2020.
Manuel Ramos-Francia, CEMLA’s General Director made the opening remarks for the XXV (Digital) Meeting of the Central Bank Research Network. He highlighted that a quarter of a century is a milestone to celebrate and reflect upon. In addition, he thanked the keynote speaker, Professor Rajan, for his participation; the Central Bank of Uruguay for co-hosting the event, in particular, Governor Labat and Jorge Ponce; and, all participants that contributed to the meeting. He focused on three broad themes: i) The importance given to research at CEMLA; ii) The economic policy response to the Covid-19 crisis, mainly, on some of the challenges going forward; and iii) He also made remarks on some of the papers presented. (Please note that a written version of his remarks is available on this webpage).
For his part, Professor Rajan explained the economic responses to the COVID-19 crisis, underscoring four key elements. The initial responses should turn into a bridge to the post-crisis world. At least, that was the original intention. First, containing the pandemic. This involves relief to households and businesses. Second, some economies have opened up, and they have seen a recovery. For instance, the third quarter of 2020 had strong growth. That said, an increase in the number of COVID-19 infections led to some doubts about it. Third, there should be a repair and reallocation process. This means repairing balance sheets as well as reallocating assets. This is particularly important in those economies that were unable to provide relief initially. Fourth, due to the virus, growth potential has shrunk. Thus, growth enhancing reforms have to be set in place.
Containment has had different degrees of success, across countries and through time. We find in Australia and New Zealand examples of successful cases. Some countries, bent the curve only initially, like many in Europe. As mentioned, in the 3Q-2020, there was a strong rebound, particularly so in manufacturing, but not in high contact sectors. Thus, without a vaccine, economies will keep operating below pre-pandemic levels.
The uneven recovery has affected global value chains and global demand. The problem now is the presence of second and third waves in some economies. Initially, a potential explanation was that there was better testing. But it was soon seen that this was not the case. Hospitals have been clogged and new lock-ups have been considered.
Outcomes have varied across countries and industries. For example, China has recovered strongly. Its policy responses have entailed less transfers to households and more to firms. Accordingly, while domestic consumption has been weak, its production has recovered fast. Equally important, China has maintained investment. For its part, the U.S. has done well; in particular, in terms of spending. Startups have exploded and unemployment come down once the economy has been opened.
There are various sources of uncertainty: the vaccine, a cure and a test, among others. The vaccine will come next year, at the earliest. There is fiscal uncertainty, which has led to economic uncertainty. There is moratoria on repayments, which can lead to loan losses. Other sources of uncertainty are: financial distress; the U.S. election; and U.S.-China tensions.
In EMEs part of the problem is that fewer jobs can be done at home. For instance, in Bangladesh, only 10% of the workforce can work from home. There are also smaller buffers in business and houses. In addition, there has been a limited relief. Much less fiscal space. A saving grace has been fewer deaths.
Some central banks in EMEs have expanded their balance sheets. This is not QE, rather these central banks have been intermediaries between their private sectors and governments. A few have provided relief to households. That said, the jury is still out on what is the best policy.
In any case, when it comes to fiscal spending, it is clear that being overly cautious can adversely affect growth potential. The right spending is key to minimize any effect. Two groups on which to focus are poor households and SMEs. Growth potential has been much more damaged in EMEs. Thus, you need to repair and reallocate much more, as you have not been able to spend much on relief.
This pandemic has created a solvency problem for corporations. In essence, demand has vanished. Insolvency usually means bad management, for which bankruptcy is the penalty. This is generally not the case this time. Some companies had the wrong capital structure coming into the crisis. Many businesses collapsed not because of bad management, but due to the terrible environment. Thus, a government equity infusion might be sensible, but difficult to justify from the political economy perspective. Many have talked about stimulus. This is not stimulus. This is repair. This is more a supply side problem, rather than a demand side one. Of course, both are related, but it is mainly a supply side situation. One should also prevent contagion to the financial sector, as was pointed out.
There are several types of firms depending on the situation they are facing. First, there are the big firms, those in manufacturing, and the low leveraged ones. They have readily access to finance and, thus, one should not be concerned about them. Second, there are those firms that are economic viable, have some leverage and limited access to financing. These need credit support. Third, there are firms that have a viable model but have too much financial leverage. They can’t have access to equity. They are illiquid. They need debt restructuring and time. Fourth, there are those firms that have too much debt and are insolvent. If their economic business is viable, it makes sense to write down their debt, instead of, for instance, rescheduling their payments. In some cases, there might be the need for equity infusion. Fifth, there are firms that are economically unviable due to coronavirus. There is a need to liquidate them. A fundamental problem is how to tell them apart from the rest. Again, we have to focus on repair and reallocate systems, but we need to make these distinctions. If you need to target credit and guarantees you need to ask, to whom and on what basis.
For many firms, the bridge might be too short. That said, is it still worth to keep these firms alive or should they be allowed to close? There is the possibility that they will eventually close. The problem of course is that support is expensive, even for the U.S. Should we let a firm shut down and then expect another firm to start? Firms that have significant human, organizational and social capital should have special consideration. Yet, these are mainly the big firms. Small firms come and go. That said, big firms can access the financial markets themselves. Medium firms are the ones we need to focus on. Small firms go in and out all the time.
Targeting has to be done much better in EMEs. Initially, general help made sense as the perception was that the pandemic was going to be temporal. Now we know that this is no longer the case. This is one of the reasons why targeting is quite relevant. One possibility is targeting with equity infusions. Debt re-structuring might help. Equity infusion might be politically problematic. A proposal was to allocate equity infusion based on past tax payments. The issue here is that some firms that did pay such taxes might not be viable, even post-pandemic. A basic issue in this context is the expected return that tax payer will have if a firm obtains an equity infusion.
There is also the possibility of quick and cheap restructuring. In addition, we need to make bankruptcy less costly. We should keep an eye on the recapitalization of the financial system. In this context, we should remember that in many EMEs there has been low relief, an excess combined with debt burden and an inability to access finance. If unviable firms continue to operate under some sort of aid, they could impair profitability. In effect, zombie firms reduce business for healthy firms. Given that EMEs did not provide much relief, repair and reallocation are extremely important.
There has been a lot of pressure on regulators to implement some forbearance. We need to keep in mind that forbearance has been a temporal solution. Over indebtedness tends to increase under forbearance. A financial sector meltdown in any EMEs will be, of course, very costly. In addition, there the issue of a doom loop between financial and governments sectors. In sum, one needs to deal with the financial sector.
As a last point, Professor Rajan mentioned the external sector. We are presently in risk-off mode. But, risk-on could start any time. Central banks need to focus on creating repair and reallocation processes as we go forward. Reforms have become much more important. Sometimes reforms can have short-term costs, and there is little will right now, but there might be no other option. There is a huge amount of uncertainty. In the absence of domestic stimulus, EMEs could grab a lifeline in exports. All of this depends on getting the virus under control.
Moving on to the papers, Cantore and Freund explore the interaction of fiscal policy and household inequality. They show that their two-agent New Keynesian model can be extended to match key implications of more general HANK models.1 In their model, one agent faces limited asset market participation and the other one concentrates firm ownership. Their model leads to a more pronounced redistributive and subdued aggregate effect of fiscal stimulus, relative to a conventional model. Lewis and Vázquez-Grande present evidence that the natural rate of interest is affected by permanent and transitory shocks. To estimate their model, they use Bayesian methods featuring loose priors on unobserved drivers. If transitory shocks are considered, the median estimate of the natural interest rate for the U.S. exhibits a less pronounced secular decline. Hence, with samples after the GFC, their estimates are greater than most in the related literature. Ramos-Francia et al., for their part explore the interaction between the fiscal deficits and inflation dynamics, in particular, the role of the money-demand function. To that end, they extend the model of Sargent, Williams, and Zha (2009) incorporating Selden-Latane’s money-demand function. The model exhibits new equilibrium dynamics and policy implications on what have been the best reforms to have and maintain low and stable inflation in Mexico.
D’Acunto et al. document that consumers rely on the price changes they face in their daily lives while grocery shopping to form inflation expectations. The frequency and size of price changes seem to matter more for individuals’ inflation expectations, rather than their expenditure share. Frache et al. explore the effects of inflation and idiosyncratic cost expectations on firms’ price-adjusting decisions in Uruguay. To that end, they use a new survey. Interestingly, firms’ beliefs about an expected increase in their costs positively affect the probability of adjusting prices. Of course, costs are a key ingredient in modern Philips’ curves theoretical models. Araujo and Piazza implement several forecasting exercises with a variety of ML techniques, using the Brazilian CPI and considering multiple horizons, and using a large database. Their results bolster automated procedures. It remains to be seen in more detail whether machine learning approaches can, for instance, outperform surveys.
Mourad et al.’s paper is entitled Bank loan forbearance: evidence from a million restructured loans, characterizes when bad bank loans are renegotiated, under deteriorating debtors’ financial conditions for Brazil. They find, for instance, that the greater the difficulty to seize the collateral, the larger the probability of forbearance.
Téllez-Leona et al. Several strands of literature seek to understand the interbank interest rate spread. Their paper explores its determinants in terms of bank network centrality measures, for the Mexican case. Eguren-Martin et al. research characterizes the distribution of capital flows for a group of EMEs. They focus on tail events, finding the strongest reactions in the left one. This observation is compatible with the presence of herd-like behavior among exiting investors. Pozo and Rojas examine the relationship between bank competition and risk-taking, for the Peruvian banking sector. An issue is how competition and financial stability relate to each other, and perhaps more importantly why.
León et al. use a supervised machine learning method to represent the payment of financial institutions using transactions large-value payment system data of Colombia. They use a parameterized neural network to represent its payment patterns, accounting for over a hundred features. The representation is then used to test payment patterns of the same institution in terms of its characteristic patterns. Divakaruni and Zimmerman argue that recent technological innovations have improved the efficiency of Bitcoin as a mean of payment. For instance, they assess: The Lightning Network, a mean of netting payments in the blockchain, among other features. They find an association between the adoption of the Lightning Network and a reduction in blockchain congestion as one of their main results. Elosegui and Pinto explore some of the key constraints that EMEs face in moving towards a cashless economy. The authors use a model of payment methods that includes key determinants of the adoption of electronic ones. They examine the underlying factors of the decisions to adopt and use a credit card in Argentina, emphasizing the role of network effects and informality in such decisions.
Landaberry et al. propose to use a systemic risk measure for a network that features the interbank network, the banks-firms bipartite network and the intra-firm exposures network, in Uruguay. It is one of the first papers in which the intrafirm exposures are estimated and are used for the computation of a systemic risk measure.
Gomez-Cram and Grotteria provide evidence on the price discovery process during FOMC days and insights on investors’ expectations formation and information rigidity. For example, they document that when the Chairman discusses changes between the current and the previous policy statement, price volatility and trading volume spike, and prices move, on average, in the same direction as they did around the statement release. García-Hiernaux et al. use a price convergence test based on a model that measures the speed of price convergence across countries. By implementing this test to the EMU price indices from 2001 to 2011, they find empirical evidence of different price level patterns and the lack of price level convergence in the long run for most countries. Pedersen explores a survey made to the Chilean Financial Traders, which reveals significant heterogeneity in how they make their forecasts. There are also differences in how the traders understand questions regarding the future monetary policy rate. These results can help understand heterogeneity in forecasters’ inflation updates.
Haas and Fajardo explore how personal experiences in the labor market can affect credit behavior. Fernández estimate the impact of a minimum wage increase in Mexico on formal employment by exploiting regional heterogeneity in the implementation of this policy. One of his main findings is that formal employment in municipalities next to the U.S. border is 0.6% lower due to the minimum wage increase. Morales et al. assess the effect of the Covid-19 pandemic and the lockdown of some economic sectors on the Colombian labor market.
Carvalho et al. document that the existence of recent loans of a firm with private domestic banks increases the probability that a loan will be granted by a foreign bank. Carrasco and Florian discuss the role of FX interventions as an instrument for EMEs in response to external shocks originated from various sources. Tang and Zhang propose a model for which the value of a firm has a bubbly component. The model predicts an overshooting of firms’ entries after an expansionary bubble shock.
André et al. analyze the interaction between monetary and fiscal policy in Mexico. To that end, they use a model of a small open economy. They show that the risk premium channel transmits fiscal shocks into the monetary block. The monetary shocks affect the fiscal block mostly through the interest rate affecting the debt service. Mello and Ponce find evidence of a positive correlation between the budget deficit to GDP and inflation expectations of price setters in Uruguay. This implies an interdependence between fiscal and monetary policies. Lankester and Sandoval study the interdependence between monetary and fiscal policy in Costa Rica during the period 1991-2019. They use three different methods. Their results are suggestive of some degree of fiscal dominance.
1 The model matches empirical intertemporal marginal propensities to consume that shape the private sector’s dynamic response to policy changes.
Wednesday, October 28
|
Opening Remarks: Manuel Ramos-Francia |
Keynote Lecture: “EMs and the Covid Crisis: Time to Repair” |
Session 1: Macroeconomics / Monetary Policy I |
Workers, Capitalists, and the Government: Fiscal Policy and Income (Re) Distribution |
Discussant: Juan Pablo Medina, Universidad Adolfo Ibañez |
Measuring the Natural Rate of Interest: A Note on Transitory Shocks |
Discussant: Cristiano Cantore, Bank of England |
Inflation Dynamics under Fiscal Deficit Regime-Switching in Mexico |
Discussant: Francisco Vazquez-Grande, Federal Reserve Board |
Session 2: Inflation and Inflation Expectations |
Exposure to Grocery Prices and Inflation Expectations |
Discussant: Fernando Borraz, Banco Central del Uruguay |
Belief-Dependent Pricing Decisions |
Discussant: Wagner Piazza Gaglianone, Banco Central do Brasil |
Machine Learning Methods for Inflation Forecasting in Brazil: New Contenders Versus Classical Models |
Discussant: Serafín Martínez-Jaramillo, CEMLA/Banco de México |
First day ends |
Thursday, October 29
|
Invited Paper: Bank loan forbearance: evidence from a million restructured loans, Winner of the Central Bank Award Rodrigo Gomez 2019 |
Session 3: Financial Stability |
Which network measures relate to the interest rate spread in the Mexican secured and unsecured interbank markets? |
Discussant: Thiago Christiano Silva, Banco Central do Brasil |
Capital Flows-at-risk: Push, Pull and the Role of Policy |
Discussant: Matias Ossandon Busch, CEMLA |
Bank Competition and Risk-Taking |
Discussant: Carola Müller, CEMLA |
Session 4: Payment Systems |
Pattern Recognition of Financial Institutions’ Payment Behavior |
Discussant: Pedro Elosegui, Banco Central de la República Argentina |
Ride the Lightning: Turning Bitcoin into Money |
Discussant: Carlos León, Banco de la República |
Towards a Cashless Economy: The case of Argentina |
Discussant: Peter Zimmerman, Federal Reserve Bank of Cleveland |
Second day ends |
Friday, October 30
|
Invited Paper: Credit Risk and its Systemic Effects |
Session 5: Macroeconomics / Monetary Policy II |
Real-time Price Discovery via Verbal Communication: Method and Application to Fedspeak |
Discussant: Julio Cacho, Inscription Capital |
Eurozone Prices: A Tale of Convergence and Divergence |
Discussant: Michael Pedersen, Banco de Chile |
Surveying the survey. What can we learn about the effect of monetary policy on inflation expectations? |
Discussant: Rodrigo Lluberas, Banco Central del Uruguay |
Parallel Session 5: Labor Economics |
Personal Experiences in the Labor Market and Household Credit Behavior |
Discussant: Leonardo Fabio Morales-Zurita, Banco de la República |
The Impact of Minimum Wage on Low Wage Formal Employment |
Discussant: David Kaplan, Inter-American Development Bank (TBC) |
Effects of the Covid-19 Pandemic on the Colombian Labor Market: Disentangling the Effect of Sector-Specific Mobility Restrictions |
Discussant: León Fernandez Bujanda, Banco de México |
Session 6: Finance |
When in Rome: Lending to Small and Medium Enterprises by Foreign and Domestic banks |
Discussant: Jorge Moreno, Universidad Autónoma de Nuevo León |
External Shocks and FX Interventions in Emerging Economies |
Discussant: Haozhou Tang, Banco de México |
Bubbly Firm Dynamics and Aggregate Fluctuations |
Discussant: Roberto Gómez Cram, London Business School |
Parallel Session 6: 2020 Joint Research Program |
Policy mix in a Small Open Emerging Economy with Commodity Prices |
Discussant: Carlos Urrutia, ITAM |
Fiscal Policy and Inflation Expectations |
Discussant: Santiago García-Verdú, Banco de México / CEMLA |
Fiscal or Monetary Dominance: The Case of Costa Rica |
Discussant: Carlos Urrutia, ITAM |
Closing Remarks |
Third day and meeting end |
