Coordination failures in business cycles

• Ano de defesa: 2017
• Autor(a) principal: Machado, Caio Henrique
• Orientador(a): Guimarães, Bernardo de Vasconcellos
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Palavras-chave em Português: Falhas de coordenação; Estímulos fiscais; Externalidades de demanda; Fricções financeiras; Crises financeiras; Corridas bancárias; Risco moral
• Palavras-chave em Inglês: Coordination failures; Fiscal stimulus; Demand externalities; Financial frictions; Financial crisis; Bank runs; Moral hazard
• Link de acesso: http://hdl.handle.net/10438/18270

Coordination failures are often said to play an important role in business cycles. If agents’ incentives of taking a given action depend on the amount of other agents expected to take the same action, coordination failures can often arise. Firms may not invest because they do not expect others to invest, confirming their initial expectations. Similarly, banks may not lend because they do not expect others to lend. This dissertation analyzes different environments in which crises arise as a result of coordination failures. The first chapter analyzes an economy that is subject to a dynamic coordination problem. Because of aggregate demand externalities, firms’ incentives to increase their production depend on expected demand, which in turn depends on the amount produced by other firms. The problem is dynamic since firms do not take investment decisions at the same time, implying that a firm deciding today is trying to forecast what other firms will decide in the future. This opens the possibility of dynamic coordination traps: firms do not invest today because they do not believe others will invest tomorrow, generating lower incentives for firms to invest at future dates. This chapter focuses on the following questions: In economies subject to dynamic coordination traps, what is the optimal stimulus policies? Should policy makers provide higher incentives to production in times of low economic activity? The answer is that a constant subsidy implements the first-best in an economy where beliefs are endogenously determined. The reason is that, although it is harder to coordinate in times of low economic activity, agents are naturally more optimistic about the future in times of poor economic activity and reasonably good fundamentals. This optimism arise from the fact that in bad times negative shocks do not change the level of economic activity, while positive shocks may end a recession. The second chapter proposes a model to study unusually deep financial crises. Previous empirical work has found that financial crises are very deep and persistent on average, but there is a lot of heterogeneity across different episodes. Some financial crises feature a very distressed financial sector, but little distress on the real sector, while others are real macroeconomic disasters. In light of this evidence, I propose a model in which there is a highly non-linear feedback between the real and the financial sector. Disaster episodes arise from the dynamic interaction of two frictions: coordination frictions and financial frictions. When banks have weak balance sheets they do not intermediate much capital. This causes firms to get trapped in a self-reinforcing regime with low aggregate demand, which ends up provoking further damage to banks’ balance sheets. I use the model as a laboratory to study unusually deep financial crises and the effects of some policies. It is shown that the effects of disasters go far beyond what we observe during those episodes: they imply very low asset prices, economic growth and welfare, even in good times and when their probability is very small. Policies that protect the financial sector from those episodes can be very beneficial. Moreover, higher risk-taking in bad times may improve economic growth, welfare and financial stability. The third chapter studies the policy trade-off of a regulator that wants to avoid coordination failures, but at the same time does not want to generate distortions arising from moral hazard. Banks have investment opportunities with an expected return that depends positively on the amount of other banks undertaking similar investments, opening room for coordination failures. At the same time, banks may risk-shift to projects with smaller expected return but higher volatility. By providing guarantees in case of failures, a regulator can enhance coordination, but that leads banks to switch to worse projects. It is shown that in some states a regulator will provide no guarantees, even if it that means allowing a coordination failure to happen. Moreover, the possibility of risk-shifting reduces the amount of guarantees needed to avoid a coordination failure.