سیاست های پولی و مقررات سرمایه در ایالات متحده و اروپا

The Federal Reserve and the European Central Bank aggressively lowered interest rates during the recent crisis. Both actions were at odds with an anti-inflationary policy stance: in August 2007, inflation expectations were high, particularly in the United States. To explain these actions, we model an economy with a leveraged and regulated financial sector. We find optional Taylor rules using simulated GMM, and find rules consistent with a pro-inflationary reaction during financial crises and a standard output-inflation mandate for the central bank. Our results support procyclical regulation not because of adequacy concerns, but instead due to the impact on monetary policy.

Our paper makes two contributions to the literature. One, we construct a model economy with a leveraged and regulated financial sector. It shows financial accelerator characteristics in normal times, that is, in the state of the world in which banks do not face regulatory capital constraints, there is a significant amplification of monetary policy. More importantly, the model shows the absence of an accelerator when the economy is capital constrained. The two-state set-up highlights an important feature of the model; we impose, via capital constraints, that lending and borrowing can be rationed. Two, we illustrate, via simulation, that optimal policy can become precautionary in a world that contains even a small probability of a constrained state. The monetary authority may have to include the presence of a constrained state as a factor when choosing policy during “unconstrained” times. We show the utility of this approach by calibrating the model to the US and Europe. Using separate calibrations to find rules appropriate to each environment, we find results consistent with (1) a stronger anti-inflationary stance for the ECB in normal times, and (2) a much stronger willingness by the Federal Reserve to sacrifice inflation risk for output in times of crisis. We have three stylized interpretations of our results. One, the Federal Reserve has partial control over bank regulation as well as full control over monetary policy. This permits it to exercise regulatory lenience as a part of monetary policy. It can thus set an initially higher capital threshold and lower it in the face of a crisis. This permits some flexibility in times of moderate stress, but also requires the willingness to face inflation risk in times of crisis. Once banks hit the capital threshold, the situation is truly dire. Two, a stronger output orientation by the Fed means that it will potentially respond more quickly when faced with constrained banks and the lack of an accelerator. Three, our results are consistent with procyclical capital regulation. However, we reach this conclusion not due to the need to support a stressed financial sector in times of crisis, but rather because the interaction between monetary policy and bank regulation necessitates it. Our approach is to incorporate a very simple financial friction into a new-Keynesian synthesis model. The constraint is a regulatory capital minimum for the banking sector.1 This implies two states of the world: well-capitalized (unconstrained) and under-capitalized (constrained).2 We mesh this with a standard formulation of monetary policy in which the central bank maintains a mandate on inflation and growth (Fed) or inflation alone (ECB). Our motivation for this is to assess the implications for monetary policy and bank regulation using only relatively small deviations from a synthesis model.3 We wish to explain patterns of monetary policy from a positive perspective without needing to adjust classic, or legislated, views of the role of monetary policy or resort to other types of financial frictions. We impose only a single friction: the inability of banks to write new loans when undercapitalized. Of course, in the presence of a bank lending channel, this implies that monetary policy necessarily changes at the point of capital (in)adequacy. This single friction will generate the stylized patterns of monetary policy for both economies. Our conclusion is, then, that it makes little sense to estimate a single policy rule across both constrained and unconstrained regimes. Thus, we construct empirical estimates of optimal policy in a model with two regimes. We illustrate the motivation for this using a simple diagram (see Fig. 1). Notice that the regulatory threshold produces the nonlinear effect of an increasingly strong monetary policy as banks approach the constraint, due to the fact that leverage is rising in this region. However, once the constraint is reached, monetary policy cannot impact lending, as banks are legally prevented from expanding lending.Our paper makes two contributions to the literature. One, we construct a model economy with a leveraged and regulated financial sector. It shows financial accelerator characteristics in normal times, that is, in the state of the world in which banks do not face regulatory capital constraints, there is a significant amplification of monetary policy. More importantly, the model shows the absence of an accelerator when the economy is capital constrained. The two-state set-up highlights an important feature of the model; we impose, via capital constraints, that lending and borrowing can be rationed. Two, we illustrate, via simulation, that optimal policy can become precautionary in a world that contains even a small probability of a constrained state. The monetary authority may have to include the presence of a constrained state as a factor when choosing policy during “unconstrained” times. We show the utility of this approach by calibrating the model to the US and Europe. Using separate calibrations to find rules appropriate to each environment, we find results consistent with (1) a stronger anti-inflationary stance for the ECB in normal times, and (2) a much stronger willingness by the Federal Reserve to sacrifice inflation risk for output in times of crisis. We have three stylized interpretations of our results. One, the Federal Reserve has partial control over bank regulation as well as full control over monetary policy. This permits it to exercise regulatory lenience as a part of monetary policy. It can thus set an initially higher capital threshold and lower it in the face of a crisis. This permits some flexibility in times of moderate stress, but also requires the willingness to face inflation risk in times of crisis. Once banks hit the capital threshold, the situation is truly dire. Two, a stronger output orientation by the Fed means that it will potentially respond more quickly when faced with constrained banks and the lack of an accelerator. Three, our results are consistent with procyclical capital regulation. However, we reach this conclusion not due to the need to support a stressed financial sector in times of crisis, but rather because the interaction between monetary policy and bank regulation necessitates it. Our approach is to incorporate a very simple financial friction into a new-Keynesian synthesis model. The constraint is a regulatory capital minimum for the banking sector.1 This implies two states of the world: well-capitalized (unconstrained) and under-capitalized (constrained).2 We mesh this with a standard formulation of monetary policy in which the central bank maintains a mandate on inflation and growth (Fed) or inflation alone (ECB). Our motivation for this is to assess the implications for monetary policy and bank regulation using only relatively small deviations from a synthesis model.3 We wish to explain patterns of monetary policy from a positive perspective without needing to adjust classic, or legislated, views of the role of monetary policy or resort to other types of financial frictions. We impose only a single friction: the inability of banks to write new loans when undercapitalized. Of course, in the presence of a bank lending channel, this implies that monetary policy necessarily changes at the point of capital (in)adequacy. This single friction will generate the stylized patterns of monetary policy for both economies. Our conclusion is, then, that it makes little sense to estimate a single policy rule across both constrained and unconstrained regimes. Thus, we construct empirical estimates of optimal policy in a model with two regimes. We illustrate the motivation for this using a simple diagram (see Fig. 1). Notice that the regulatory threshold produces the nonlinear effect of an increasingly strong monetary policy as banks approach the constraint, due to the fact that leverage is rising in this region. However, once the constraint is reached, monetary policy cannot impact lending, as banks are legally prevented from expanding lending.Broadly, the recent crisis has highlighted the fact that first-generation new-Keynesian models are not well-equipped to interpret the role of monetary policy under financial stress. They were based on a couple of classic imperfections, such as nominal rigidities and monopolistic competition, to allow for nontrivial market power and price setting. The goal, of course, was to illustrate how demand shifts could impact output, and thus how monetary policy shifting the demand could have real effects. These constructs permitted an extensive literature that could study the basic role of policy. The models, however, omitted details of market imperfections that are central to the study of macroeconomics. This omission is in part responsible for the fact that consensus Taylor rules cannot describe the path of monetary policy (Rudebusch, 2006). A new round of (second-generation) new-Keynesian models focuses on the implications of other frictions. Because of the current financial crisis, a number of new papers,4 this one included, have turned their attention to the role of financial- and credit-market imperfections by building on work by Bernanke et al. (1999) and Carlstrom and Fuerst (2001). Indeed, there is renewed interest in a real-economy link that passes through the banking sector. The channel through the banking sector is now widely believed to play an important role in the conduct of monetary policy and highlights the importance of banking regulation. According to one hypothesis, the previous “credit crunch” in the United States was at least partly a consequence of banks' eagerness to meet the 1992 deadline for capital adequacy requirements under the 1988 Basel agreement (Bernanke and Lown, 1991). This argument dovetails with the realization that regulation has, to date, been largely a study in individual bank risk management. As such, policy makers have reemphasized the need both to incorporate cyclicality into prudential regulation (Borio et al., 2001) and to approach regulation from what has been called a macroprudential perspective (Borio, 2003, and many others). To provide a potential solution to this question, we turn our attention to the nexus of monetary policy and bank regulation. In particular, we ask how one can evaluate the macroeconomy in an environment where the monetary authority must cope with a regulated banking sector. The potential conflict between central bankers and financial supervisors has been noted before. A range of research has found that capital adequacy requirements, while potentially important for financial stability, can also be procyclical.5 Of course, if monetary policy is intended to promote stable economic growth, its countercyclical bias will run counter to bank regulation. A joint authority now has the advantage (or the problem) that policy decisions must account for both factors (Fig. 2 and Fig. 3).The presence of a regulated banking sector matters for our purposes here because it provides a potential explanation for the apparent inconsistency in monetary policy during crisis periods. We focus on “crises” as defined as shocks to bank capital. In a new-Keynesian model of the economy with no financial frictions, bank capital shocks are irrelevant. Bank lending is typically determined completely by available deposits. Thus, capital shocks have no impact on the real economy. In a world with either leverage or some other type of financial-sector friction, economic shocks and monetary policy are amplified through the financial sector (Bernanke et al., 1999). Of course, even in this world, the implied optimal policy rules cannot explain dramatic reductions in interest rates in the face of inflationary pressures. As noted, in a regulated banking sector, financial sector leverage has legal limits. Thus shocks that impact bank capital pass through to limitations on lending. In most instances, this is similar to the Bernanke et al. (1999) world in which economic shocks are magnified by the financial sector, in this case, through a leverage effect. However, a sufficiently large shock to bank capital changes the nature of the accelerator. When bank capital falls below a regulatory requirement, lending is capped. Thus, the leverage effect vanishes and monetary policy becomes ineffective at stimulating the real economy until bank capitalization rises again above the constraint.6 Our paper proceeds as follows. Section 2 outlines our extension of the Bernanke et al. (1999) financial accelerator. We continue in Section 3 with a discussion of the setup for our simulation exercises and show some results in Section 4. Section 5 provides some discussion and we conclude in Section 6.

We believe that there are likely many reasons for the differences in ECB and Fed policies over the past two years. Among these is the difference in mandate emphasis between the two that lead to stronger anti-inflationary emphasis at the ECB, differences in the timing of impact on the banking sector, and many others. That said, the model presented here is both reasonably consistent with both authorities' actions and is capable of reproducing some stylized features of each. With this framework in place, there are potentially more open questions that lie beyond the scope of this paper. As mentioned, the Federal Reserve has responsibility both for monetary policy and bank regulation of some of the financial system. Indeed, much of the response to the crisis in the United States consisted of extraordinary support to banking institutions; this is largely equivalent to a relaxation of capital requirements. Thus, to what extent does the ability to change regulatory requirements jointly with monetary policy change the optimization problem? We would hypothesize that this is consistent with current patterns