مدارک و شواهد ساختار مدت در هموارسازی نرخ بهره و اینرسی سیاست های پولی
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|24826||2002||27 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Monetary Economics, Volume 49, Issue 6, September 2002, Pages 1161–1187
Numerous studies have used quarterly data to estimate monetary policy rules or reaction functions that appear to exhibit a very slow partial adjustment of the policy interest rate. The conventional wisdom asserts that this gradual adjustment reflects a policy inertia or interest rate smoothing behavior by central banks. However, such quarterly monetary policy inertia would imply a large amount of forecastable variation in interest rates at horizons of more than 3 months, which is contradicted by evidence from the term structure of interest rates. The illusion of monetary policy inertia evident in the estimated policy rules likely reflects the persistent shocks that central banks face.
How quickly do central banks adjust monetary policy in response to developments in the economy? A common view among economists is that the short-term policy interest rate in many countries is changed at a very sluggish pace over several quarters. The evidence supporting this view is found in the many monetary policy rules or reaction functions estimated in the literature with quarterly data. These policy rules take the general partial adjustment form , where it is the level of the policy interest rate in quarter t, which is set as a weighted average of the current desired level, , and last quarter's actual value, it−1. Based on historical data, estimates of ρ are often in the range of 0.8, so these empirical rules appear to imply a very slow speed of adjustment of the policy rate to its fundamental determinants. This gradual adjustment of the policy rate over several quarters to its desired level is widely interpreted as evidence of an “interest rate smoothing” or “monetary policy inertia” behavior by central banks. For example, Clarida et al. (2000, pp. 157–158) describe their U.S. estimates of various partial adjustment policy rules: “… the estimate of the smoothing parameter ρ is high in all cases, suggesting considerable interest rate inertia: only between 10 and 30 percent of a change in the [desired interest rate] is reflected in the Funds rate within the quarter of the change. Thus, our estimates confirm the conventional wisdom that the Federal Reserve smooths adjustments in the interest rate”. Some of the many other recent papers with a similar inertial interpretation of monetary policy rules include Woodford (1999), Goodhart (1999), Levin et al. (1999), Amato and Laubach (1999), and Sack (1998). Furthermore, a few researchers have also argued recently that the monetary policy inertia apparently present in the real world may be an optimal behavioral response on the part of central banks. For example, one popular such normative argument contends that the quarterly policy inertia and interest rate smoothing behavior helps the central bank focus the expectations of agents in the economy on its stabilization goals and thereby achieve a better outcome (e.g., Levin et al. 1999; Woodford, 1999; Sack and Wieland, 2000). There is another quite separate literature on “interest rate smoothing”, which, at least superficially, may appear to be consistent with the quarterly interest rate smoothing described above. This earlier literature analyzes changes in policy interest rates on a day-by-day basis. Both in the U.S. (e.g., Goodfriend, 1991; Rudebusch, 1995) and in Europe, Japan, and Australia (e.g., Goodhart, 1997; Lowe and Ellis, 1997), central banks appear to follow a pattern of behavior in which changes in the policy rate are undertaken at discrete intervals and in discrete amounts.1 For example, Rudebusch (1995, p. 264) defines short-term (or weekly) interest rate smoothing as the Fed adjusting interest rates “… in limited amounts… over the course of several weeks with gradual increases or decreases (but not both)…”.2 Many have assumed—including the monetary policy rule papers cited above—that such short-term interest rate smoothing implies the quarterly interest rate smoothing found in the empirical policy rules. However, the earlier short-term interest rate smoothing refers to a partial adjustment over the course of several weeks, while quarterly interest rate smoothing refers to a partial adjustment over the course of several quarters. With such disparate time frames, the two types of partial adjustment are in fact largely independent, so a central bank could conduct either type of smoothing without much of the other. Indeed, an important point in the short-term interest rate smoothing literature is that although central banks smooth interest rates on a week-to-week or month-to-month basis, there is essentially no quarterly interest rate smoothing. This description follows Mankiw and Miron (1986, p. 225), who note that the postwar term structure of interest rates suggests that at a quarterly frequency “… while the Fed might change the short rate in response to new information, it always (rationally) expected to maintain the short rate at its current level”. Goodfriend (1991, p. 10) provides an identical random-walk characterization of the policy rate and argues that changes in the rate set by the Fed “… are essentially unpredictable at forecast horizons longer than a month or two”. Similarly, Rudebusch (1995, p. 264) characterizes the Fed's behavior as, “… beyond a horizon of about a month, there are no planned movements to react to information already known”. Thus, the earlier short-term interest rate smoothing literature rejects any partial adjustment or policy inertia at a quarterly frequency. 3 This paper argues that quarterly interest rate smoothing (or monetary policy inertia) is a very modest phenomenon in practice, which accords with the earlier characterization of monetary policy partial adjustment as involving only a very short-term smoothing of rates. This argument, however, must account for the many estimated policy rules that appear to indicate that a high degree of quarterly interest rate smoothing is present in the real world. This seemingly straightforward descriptive evidence of slow adjustment from the inertial empirical policy rules is summarized in the next section, while Section 3 outlines the related normative arguments for the optimality of inertial behavior in a New Keynesian model of output and inflation. Evidence against the existence of an inertial policy rule is obtained from the behavior of market interest rates at the short-term end of the yield curve. As documented in Section 4, there appears to be very little information generally available in financial markets regarding future interest rate movements beyond the next 1 or 2 months, which is consistent with the results of Mankiw and Miron (1986) and many others. In contrast, Section 5 derives the term structure implications of monetary policy inertia in a New Keynesian model and shows that the large ρ in an inertial rule implies that typically there are predictable future changes in the policy rate, which under rational expectations should be embodied in the term structure. Thus, there is an inconsistency between the term structure implications of quarterly interest rate smoothing and the historical term structure evidence. Furthermore, this inconsistency is robust to a variety of different assumptions about the specification of the model and the policy rule. Assuming financial markets process information efficiently, the term structure evidence implies that the empirical policy rules displaying substantial partial adjustment are misspecified. Section 6 argues that such partial adjustment could be spuriously attributed to a non-inertial central bank, that is, one that displays no quarterly interest rate smoothing. This argument is based on the econometric near-observational equivalence of the partial adjustment rule and a non-inertial rule with serially correlated shocks. That is, significant persistent shocks may explain the illusion of monetary policy inertia, and the conventional empirical partial adjustment rules are misspecified. Furthermore, when monetary policymakers respond to current information—including the persistent shock—interest rate predictability is quite low, which is consistent with the term structure evidence.
نتیجه گیری انگلیسی
Empirical monetary policy rules with large estimated coefficients on the lagged policy interest rate, which are very prevalent in the literature, are widely interpreted as indicating a sluggish adjustment of the policy rate to its determinants—on the order of only about 20 percent per quarter. This partial adjustment implies predictable future changes in the policy rate over horizons of several quarters, which does not accord with the lack of information about such changes in financial markets. This paper proposes a resolution of this empirical inconsistency by providing an alternative interpretation of the large lag coefficients in the estimated policy rules. These coefficients reflect serially correlated or persistent special factors or shocks that cause the central bank to deviate from the policy rule. This argument uses indirect term structure evidence to dismiss the interest rate smoothing interpretation of the partial adjustment rule. As noted above, it appears difficult to develop direct evidence against the partial adjustment rule (in the form of non-rejection of the ρ=0 hypothesis). In particular, the uncertainty in modeling the desired policy rate (given the endogeneity of its determinants, the real-time nature of the information set, as well as the small samples available) makes any direct evidence from estimated rules fragile. For example, the rule with partial adjustment and the rule with serially correlated shocks both appear to fit the data as empirical reaction functions. However, they have very different economic interpretations. In the former rule, persistent deviations from an output and inflation response occur because policymakers are slow to react. In the latter rule, these deviations reflect the policymaker's response to other persistent influences. The two types of rules can be distinguished, however, by their very different implications for the term structure. Only the serially correlated shocks rule is consistent with the historical evidence showing that the term structure is largely uninformative about the future course of the policy rate. There may be other possible reconciliations of the policy rule and term structure empirical results. For example, it may be that the rational expectations hypothesis of the term structure cannot be applied and the associated term structure interpretations above are spurious. One way in which this hypothesis may fail is that expectations are not rational, but this would undermine many aspects of any explicitly forward-looking macroeconomic modeling exercise such as the one above. Or term premia for short-term interest rates may be even more volatile than assumed above; however, if rates are driven by volatile term premia, then it seems unlikely that they can communicate the subtle expectations of future monetary policy as required in the literature on optimal monetary policy inertia. It is also possible that there is some intermediate case of partial adjustment, a ρ1 or ρ2 of 0.4, say, along with some serially correlated shocks, that is not strictly rejected by the term structure evidence. However, it should be noted that while real-world discussions of monetary policy sometimes mention the “incrementalism” and “gradualism” of smoothing the policy rate over the next several weeks, there is no acknowledgment of quarterly interest rate smoothing. 35 As the New York Times (July 26, 2000) summarized of recent Congressional testimony: “Alan Greenspan, the Federal Reserve chairman, said today that the central bank's decision about whether to raise interest rates again at its meeting next month would hinge in large part on economic data released in coming weeks.” That is, there was little if any pent-up pressure from the past for further adjustment. In future research, the empirical rules given in Section 6 can be improved as further effort is made in estimating rules without the crutch of partial adjustment. Given the similar estimates above of gπ and gy across rules, it may be that past conclusions about these coefficients, as in Clarida et al. (2000), are robust to the exact formulation of serial correlation in the rule. However, the lagged policy rate, though useful in mopping up residual serial correlation, should not be given a structural partial adjustment interpretation with regard to central bank behavior. In particular, using the partial adjustment rule in a model as a representation of historical policy (as in Levin et al., 1999, and many other studies) may give misleading results, especially about the nature of optimal policy inertia. With regard to optimality, the maintained hypothesis of economics for central banks, as for other agents in the economy, is that the non-inertial policy rule apparently used in practice is optimal, and certainly, the rule can be rationalized as such in particular models as in Table 2. However, it should be stressed that there are many aspects of the monetary policy process still to be modeled, especially imperfect credibility and uncertainty (see Rudebusch, 2001). Also, the absence of partial adjustment does not mean that central banks are not trying to influence long-term interest rates. However, in order to influence the long rate, central banks only must present a clear path for the policy rate that can shape expected future rates. The partial adjustment rule provides one such path, but it is not the only one. As noted by Goodfriend (1991) and Rudebusch (1995), an ex ante constant path, which is approximately what the non-inertial rules deliver, is another obvious choice. Finally, further careful analysis of the empirical policy rule is required in modeling and identifying the shocks. Section 6 provides a simple formulation for adding shocks to a policy rule. A better specification may link persistent shocks in both the rule and the rest of the model. A bout of credit frictions or impediments may lower the equilibrium real rate and provide a persistent negative shock to the policy rule and to the output equation as well (see Rudebusch, 2001). Alternatively, an idiosyncratic inflation scare may provide a shock to the rule and to inflation expectations more broadly.