2012 Theses Doctoral

Three Essays on the Credit Dimension of Monetary Policy

Martins, Guilherme Batistella

This thesis focus on the credit dimensions of monetary policy. The topic has been an area of active research since the financial crisis of 2008 and 2009, The chapters can be grouped in terms of the questions that motivated them. For the first and the second, it was "Why do Central Banks in emerging market economies intervene in credit markets in response to external shocks?" while for the third the question is more general "Why do Central Banks intervene in credit markets?" In Chapter 1, we describe that, during the financial crisis of 2008-2009, to respond to a sudden stop in capital flows, many central banks in emerging market economies relied on credit policies. We build a quantitative small open economy model to study these credit policies. The main innovation of our setup is the presence of two imperfect credit markets, one domestic and the other international, and of two types of firms. The exporter is assumed to have access to both credit markets, while the wholesale firm can only borrow in the domestic market. During a sudden stop, exporters, faced with higher spreads for international credit lines, repay part of their foreign debt, tap the local market for funds and cause spreads to increase in the domestic market. This increases financing costs for all firms, causes a deterioration of the balance of payments and depresses output. Calibrating the model to match Brazilian data, we assess the effects of two policies implemented by the Central Bank of Brazil: (i) lending to exporters using previously accumulated foreign-exchange reserves and (ii) expanding credit in order to reduce spreads in the domestic market. The model suggests that both policies probably raised GDP, but that the latter may well have decreased welfare. Moreover, had the central bank not been able to use foreign reserves as the source of funding, lending to exporters would also have reduced welfare. In Chapter 2, we expand our focus to the fact that, during the crisis, the emerging markets economies faced a large decline in their terms of trade and an increase in the interest rate they could borrow from abroad. As their counterparts in developed economies, policymarkers intervened in credit markets. A common ground behind the interventions seems to be failures in the banking system. We build a quantitative small open economy model with domestic financial intermediation to study these credit policies. The main innovation of our setup is the presence of a domestic banking system. In this structure, four main channels link external shocks to the financial sector: (1) the profitability of the export sector, (2) asset prices, (3) bank's borrowing cost and (4) the balance sheet position of banks as they hold foreign currency denominated debt. For the calibration we consider, based on Brazilian data, the domestic financial sector has the largest amplification effect in response to an increase in the international interest rate and the corresponding decline in assets price is the main channel. Hence credit interventions are most powerful in response to this type of a shock, reducing by 30% the initial GDP fall. The model is general and appropriate to address several questions. We illustrate that by showing that it can replicate standard business cycle properties and to discuss conventional monetary policy in the context sudden stops, when the domestic banking system is often at the epicenter of the crisis. In Chapter 3, we first note that a number of recent theoretical papers show that margins can affect asset prices. Such results are important, for example, to understand the unconventional polices implemented by the Fed during the great recession of 2007-2010. However, empirical evidence is still scarce. We contribute to fill this gap. We show that an aggregate margin-related factor is able to predict future excess returns of the SP 500 and that stocks with high exposures to the cost of buying on margin pay on average higher returns.