سیاست های پولی تحت دام نقدینگی: شواهد شبیه سازی برای منطقه یورو

In this paper, we analyze the conduct of monetary policy under a zero nominal interest-rate bound (hereafter ZIB) in a model economy of the euro area, namely that of the Area Wide Model. The aggregate euro-area economy is modeled to have relatively sluggish adjustment properties and a private sector with mainly backward-looking expectations. For a given ZIB benchmark, we consider variations in the monetary-policy reaction function to minimize the macro-economic consequences of such a deflationary regime. We rank the effectiveness of these remedial policies using a number of metrics and relate our results to features and properties of the model economy. J. Japanese Int. Economies20 (3) (2006) 338–363.

In this paper, we examine issues relating to the liquidity trap and the zero nominal interest bound (hereafter ZIB) in the context of the euro area. We do so in two dimensions. First, we examine the historical experience of the liquidity trap in the main economies, including the euro area and constituent members. Second, we perform zero-interest bound simulations on a well-known macro-model of the euro area; this allows us to examine the transmission mechanisms underlying a ZIB deflationary regime and potentially to rank the performance of remedial policies, conditional on the reference model. Svensson (2000) defined a liquidity trap as “…a situation with zero [nominal] interest rates, persistent deflation and persistent deflation expectations”.1 Deflations are thus characterized by chronic diverges between demand and potential output leading to weakening employment and growth and, in turn, embedded expectations of continued deflation.2 Deflations essentially result in the nominal interest rate declining to zero at which point the real interest rate paid by borrowers equals the expected rate of deflation. Thus, borrowers' capacity to honor payments on the principal of their obligations may be strained, as falling prices increase the real value of such liabilities. With nominal interest rates at their lower bound, moreover, a liquidity trap severely limits the traditional role of monetary policy in stabilizing or stimulating the economy. A form of nominal rigidity, which may adversely interact with a prolonged deflation, is the difficulty employers face in reducing nominal wages. If employees do not accept such cuts—due to “money illusion”, union bargaining powers, etc.—deflation would result in increasing real wages, thereby further reducing employment and output. Finally, financial-market frictions may obstruct the reallocation of the resources required to restore balanced economic conditions. Deflations, particularly if unexpected, can lead to large redistributions of purchasing power from borrowers to lenders. To the extent that firms are net borrowers, this wealth reallocation, notably if reinforced by an erosion of firms' net worth due to falling asset prices, can have additional indirect effects. As their balance sheets deteriorate, firms may find it difficult to maintain access to external finance. Unavailability of credit may thus curtail investment and deepen the downturn. Conversely, deflationary shocks increase purchasing power and potentially improve external competitiveness. Although the literature on preventing (and escaping from) such a trap is large and multi-faceted, fundamentally most remedies boil down to proposals to raise inflationary expectations or lower expected interest rates; the general idea being for policy makers to lower interest rates along the term structure and flatten the yield curve. This, sometimes called the “policy duration” effect, implies that policy makers follow a rule that would retain a zero nominal interest rate for a period beyond what an optimal policy, in absence of the lower bound, would follow. Indeed, the analysis of Reifschneider and Williams (2000) and Eggertson and Woodford (2003) suggest that monetary policy makers remain unhindered in conducting monetary policy in a low-inflation regime, because of their ability to manipulate financial-market expectations of future policy through long-term interest rates. Additional remedies have included: announcing a higher target of inflation (Krugman, 1998), systematic expansion of the monetary base (Clouse et al., 2000), taxing money (Goodfriend, 2000), and, in an open-economy setting, engaging in announced and controlled devaluations, (McCallum, 2000 and Svensson, 2001). Finally, a number of authors have suggested a supporting role for fiscal policy, e.g., Eggertson and Woodford (2003). In this paper, we analyze strategies to limit the consequences of a deflationary regime. We first identify, in our macro-model, a demand shock sufficient to ensure a four-year zero interest bound. Given this benchmark, we consider variations in the monetary policy reaction function to undo some of the consequences of this shock. These variations include more aggressive and more forward-looking policy rules, the addition (alone or jointly) of price level targets, non-linear or threshold rules, policies which raise the inflation target once a deflationary regime sets in, and “policy-duration” rules. Using a number of informal metrics we attempt to discriminate between these alternatives. The paper proceeds as follows. In Section 2, we review some historical experiences of deflation from a panel of countries. In Section 3 we describe the macro model used for our exercises (the ECB's Area-Wide Model) and the underlying deterministic simulations—the results of which are presented and discussed in Section 4. Section 5 concludes.

In this paper, we analyzed issues surrounding the liquidity trap. Naturally, macro-models cannot hope to address all the relevant issues. For example, the Japanese experience reflected fragilities in the banking sector, divergence of asset prices from fundamentals (e.g., Hoshi and Kashyap, 2004), and possibly an inefficient fiscal-monetary policy mix etc. Macro-models typically work at a high degree of aggregation with a limited role for credit and balance-sheet effects, confidence indicators, chronic expectational errors and other such structural features. Indeed, the most important component in escaping a zero interest bound regime, may require structural reforms in, say, financial markets, on which macro-models have little comparative advantage. Nevertheless, whatever the cause of deflation, a key starting point in its remedy is the implementation of well-designed monetary policies and the generation of higher inflationary expectations; these are issues on which macro-models are not uninformative. Our exercises suggest that a number of strategies exist for (monetary) policy makers to minimize their exposure to a liquidity trap. Policies which imply more aggressive responses to existing policy targets appear very robust and perhaps the most straightforward of those examined. Greater gains can still be made by introducing price targets and this is intuitive in a deflationary environment; of these, targeting prices jointly rather than alone appear to be superior. Threshold and Reifschneider–Williams rules appear to do equally well, and have similar characteristics of holding rates at the ZIB for a relatively long period. Were the present model more forward-looking or incorporated a more active role for long-term interest rates, both these policy types might yield great policy gains. However, if it is safe to assume a backward-looking economy rather than a forward-looking one, then policies which hold interest rates at the lower bound for a protracted period and rely on the policy duration effect may be considered as ones upon which there are better alternatives. Forecast-based rules appear to perform badly in the essentially backward-looking AWM environment; this too may be overturned in forward-looking models (although in such cases, there is often an inherent problem of indeterminacy for especially long lead horizons). The greatest gain, moreover, involves raising the inflation target (e.g., Krugman, 1998). However, the practical success of this policy is uncertain since policy makers face the difficult task of communicating this strategic change; revisions of monetary policy strategies by central banks are necessarily rare and likely to be motivated by long-run structural and institutional considerations. Indeed, all our exercises assumed that policy makers implement credible policies. Given the long lags between policy makers recognizing a deflationary episode, systematically acting upon it in and successfully communicating their actions, these exercises necessarily underestimate the ease of escaping a liquidity trap. Nonetheless, our simulations suggest that policy makers should devote energies to tailoring their monetary policy strategies to extreme outcomes. Furthermore, modeling the process by which agents learn about deflationary shocks (their nature and duration) should be subjects of ongoing research interest. Given the essentially backward-looking nature of our model economy, sluggish expectational errors have been an underlying feature. Finally, we acknowledge that our analysis is conditional on a single model economy. Since conclusions about monetary-policy conduct are likely to be sensitive to model uncertainty, it would be prudent to consider alternative modeling assumptions. Our results could therefore fruitfully be seen in the context of a similar set of exercises preformed on other central-bank models, see Fujiwara et al. (2006).