Symposium: EMU and the Euro: Current and future prospects: The Euro and monetary policy transparency

Symposium: EMU and the Euro: Current and future prospects: The Euro and monetary policy transparency

Caporale, Guglielmo Maria

Over the last two years the external value of the euro has attracted considerable attention-not surprisingly, given the potential economic and politic repercussions of its steady decline. By November 2000, the value of the European single currency had fallen by 30 percent against the dollar compared to its value at the date of its launch on 1 January 1999, though it has most recently staged a recovery, stabilizing at a higher level. Various reasons for its weakness have been cited. For instance, it has been pointed out that real growth and interest rates differentials were in favor of the United States, which led to significant capital flows from Europe to the United States. Further, it has been argued that the sharp increase in the issuance of euro-denominated bonds and the relative attractiveness of the U.S. equity markets may account for the depreciation of the euro. Its decline has also been attributed to the monetary– fiscal policy mix adopted in Euroland, which, it is argued, might lead to free-riding behavior of fiscal authorities in any particular country. In essence, the low interest rates prevailing in the euro area could result in overborrowing to finance budget deficits, thereby increasing the risk of fiscal unsustainability [Cohen and Loisel, 2000].

This paper focuses on another possible explanation for the weakness of the euro, namely the lack of transparency of the European Central Bank’s (ECB) monetary policy. Recently, both the IMF’s Economic Outlook [2000] and the Economist [2001] have stressed that, notwithstanding its regular press briefings, there is a widespread perception that the ECB’s communication strategy has so far lacked clarity. As argued by Cuckierman and Meltzer [1986], some “creative” ambiguity may be useful in a few instances, but it must be weighed against the uncertainty that it generates, which translates into higher and more volatile borrowing costs. Faust and Svensson [1999] and Geerats [2000] have shown that transparency can act as an implicit commitment mechanism and help to build an institution’s reputation. Therefore, as Favero et al. [2000] point out, what has worked well in terms of secrecy for the Bundesbank and other central banks (under different circumstances) may not be the right approach for the ECB.

This paper provides some empirical evidence on the extent to which the communication strategy chosen by the ECB has resulted in uncertainty about its monetary policy objectives, thereby weakening the euro. In particular, an increase in this source of risk might be the reason for the observed capital outflows from the euro area to the United States, with their obvious consequences for the exchange rate. Since we are interested in market uncertainty about monetary policy, we focus on short-term interest rates. Although other factors could affect their volatility, undoubtedly the lack of monetary policy transparency is one of the most important ones, as argued by various authors. For instance, Dotsey [1987] shows that secrecy shrinks the feasible information set of private agents on monetary policy operating procedures. Consequently, the interest rate forecasting error and the interest rate conditional variance tend to increase.

Options prices are being increasingly employed to extract market expectations and views about monetary policy. For this purpose, Gros et al. [2000] employ riskneutral probability functions to synopsize the information contained in the prices of short-term futures options. They pay particular attention to the second moment of these distributions, to extract the so-called implied volatility. This can be seen as a measure of the uncertainty in the underlying asset price (money market rates, in this context). However, since estimating risk-neutral probability functions is computationally burdensome, they focus only on a few of the several days (over the period 1999-2000) surrounding the Federal Open Market Committee (FOMC) or the ECB’s Governing Council meetings [Gros et al., 2000]. It is standard practice to use the Black-Scholes [1973] formula for option pricing to obtain a time-varying measure of the implied volatility [Gros et al., 2000]. However, it should be noted that the BlackScholes formula predicts a constant volatility across time-to-maturity and strike prices, which is not consistent with the empirical evidence.

The contribution of this paper is twofold. First, we take an alternative approach to measuring uncertainty, and estimate a Stochastic Volatility model [Ghysel et al., 1996] to obtain a time-varying variance from the unpredictable innovations. This, rather than the implied volatility calculated in the standard way, is used as our measure of monetary policy uncertainty. To be more specific, we focus on the deviations of the spot level of short-term interest rates from policy-determined, “official rates.” These differences, we argue, are more representative of market-determined interest rates. Secondly, we estimate directly the impact of monetary policy uncertainty on the euro-dollar exchange rate. The paper is organized as follows: we begin by reviewing a number of theoretical studies on monetary policy transparency and continue by summarizing the available empirical evidence. We then explain the empirical methodology and describe the empirical findings, specifically the implications for the eurodollar exchange rate.

MONETARY POLICY AND TRANSPARENCY

Previous studies on monetary policy transparency are not always in agreement. While a majority emphasize the reasons for secrecy, a few recent ones advocate openness, by showing that it provides an implicit commitment mechanism, and it can help a central bank build its reputation. Research has focused on five different aspects of monetary policy transparency [Geerats, 2000]. These are:

* openness about policy objectives, such as explicit inflation targets (political transparency)

* disclosure of economic data, models, and a central bank’s forecasts (economic transparency)

* information about the monetary policy strategy and internal policy deliberations, for instance through the release of minutes and voting records (procedural transparency)

* communication of policy decisions and announcements (policy transparency)

* openness about the implementation of policy decisions, market interventions and control errors (operational transparency, market transparency)

Political transparency is analyzed by Nolan and Schaling [1996], who show that less uncertainty about the central bank’s preferences is beneficial and reduces the inflation bias.

As for economic transparency, Gersbach [1998] and Cuckierman [1999] focus on the publication of the central bank’s forecasts, which removes an information asymmetry about economic shocks. They show that disclosing the forecasts to the public reduces welfare. The reason is that the central bank loses its ability to stabilize output because supply shocks revealed by the forecasts are incorporated in inflation expectations. Both models assume that there is perfect information about the central bank’s preferences. Tarkka and Mayes [1999] consider economic transparency in the form of mutual uncertainty about expectations. Specifically, the public is assumed to be uncertain about the central bank’s preferences and its assessment of the public’s inflation expectations. The publication of the central bank’s forecasts removes these uncertainties and leads to greater predictability of monetary policy. Geerats [2000] also analyzes the case of asymmetric information between the central bank and the public, where it is assumed that the central bank has additional economic information, and the public does not know the central bank’s preferences. Concerning the first assumption, the argument is that central banks have an informational advantage in the interpretation of the available information; they are likely to have different (and often better) forecasts than the market. Romer and Romer [1996] provide evidence of such asymmetric information. The second assumption is claimed not to be a contentious one, since many central banks have adopted explicit target ranges for inflation. Geerats [2000] shows that transparency enhances the central bank’s ability to build a reputation, reduces the inflation bias, and gives monetary authorities greater flexibility to respond to shocks to the economy. These advantages can be achieved through the publication of the conditional central bank’s forecasts of both inflation and output.

Procedural and policy transparency (in particular the publication of minutes, voting records, and policy directives) are considered in Goodfriend [1986], Buiter [1999] and Issing [1999]. The implications of disclosing individual voting records are formally analyzed by Gersbach and Hahn [2000].

There are several models of operational transparency. In their seminal paper, Cuckierman and Meltzer [1986], who build on the optimal policy models of Kydland and Prescott [1977] and Barro and Gordon [1983], provide a motivation for secrecy. They show that a central bank, through control errors, can mask its true intentions, and then use inflation surprises to increase output during periods when it is most valuable to do so. Because in their model the loss function is linear in output, the central bank is willing to accept arbitrary increases in the variance of employment to achieve tiny reductions in inflation. Consequently, monetary authorities behave in the same way regardless of whether credibility is low or high. Because monetary policy does not depend on the reputation and credibility of the central bank, less secrecy represents a cost to the central bank only to the extent that it reduces its ability to engineer inflation surprises.

Faust and Svensson [1998] extend the Cuckierman and Meltzer analysis in two ways. First, they consider a more realistic loss function for the central bank (quadratic in both arguments: inflation and employment variability). Secondly, they explicitly distinguish between operational transparency and imperfect monetary control. In their model the central bank is tempted to deviate from an announced inflation target due to fluctuations in an idiosyncratic employment target, which represents information exclusive to the central bank and unobservable to the private sector. Increased transparency allows the private sector to infer the central bank’s employment target with greater precision, which makes both the central bank’s reputation and inflation expectations more sensitive to the actions of the monetary authorities. This, in turn, increases the cost of the central bank’s deviating from the announced inflation target, and hence deters it further from attempting to achieve its idiosyncratic employment target. As a result, the variability of both inflation and employment falls, and the inflation bias is reduced. Consequently, increased transparency gives the central bank the incentive to build its reputation, and, generally, it increases social welfare.

More recently, Faust and Svensson [1999] endogenize the choice of control and transparency, and show that a minimum degree of transparency can be preferable. The evidence pertaining to the Bundesbank and the U.S. Federal Reserve Board (Fed) appears to be consistent with this view, while the recent experience of high central bank transparency in inflation targeting countries is clearly not. In a model similar to Faust and Svensson [1998] but with a New Keynesian Philips curve, Jensen [2000] finds that greater transparency could actually reduce social welfare.

EARLIER EMPIRICAL STUDIES

As already mentioned, the ECB’s monetary policy, especially when compared with the Fed’s operating strategy, has attracted criticism from the media, market participants, and international organizations. Several empirical studies have tried to demonstrate that faults in a central bank’s communication strategy can cause considerable uncertainty over the future course of monetary policy.

Poole and Rasche [2000] analyze daily data on the Fed funds futures, and find that over the last few years the accuracy of market forecasts of the Fed’s policy decisions has increased. They argue that the improved market understanding of monetary policy is due to the adoption of a better communication strategy, which includes immediate announcements of policy decisions following the Federal Open Market Committee (FOMC) meetings, and confining most policy actions to the time of regularly scheduled FOMC meetings. Two surveys, one from Goldman and Sachs [2000] and another from the Center for European Economic Research (ZEW) in Germany [1999], criticize the ECB for its lack of transparency. Following Soderlind and Svensson [1997], Gros et al. [2000] use the volatility of 3-month money market rates (in both Euroland and the United States) implied by option prices as a measure of monetary policy uncertainty. They show, using cross sections of different strike prices, how to derive the (risk-neutral) probability distribution of option prices on a daily basis. Once this has been estimated, the moments of the distribution (in particular, the variance) can be computed. Volatility levels are normalized to make them comparable, given the different mean levels. The analysis focuses on some crucial dates when monetary policy decisions were expected in the United States and Euroland, between September 1999 and March 2000. The empirical results show that there is no clear ranking: implied volatility is sometimes higher in the United States, and sometimes higher in Euroland. To check the robustness of their results to the dates chosen, the authors also examine the daily implied volatilities estimated by Bloomberg, which are based on the same option prices using the traditional Black-Scholes [1973] formula. The findings confirm that there is no clear ranking: towards the end of the sample, Euroland volatilities seem to be higher, while the opposite is true at the beginning of the sample. In particular, they report that Euroland’s volatility drops after the interest rate hike decisions in November and February, while it is high (and higher than in the United States) before such decisions.

As already mentioned, the Black-Scholes [1973] formula implies a constant volatility, while the evidence based on observed option prices indicates that volatility varies across time-to-maturity and strike prices. Therefore, in contrast to the existing literature [Gros et al., 2000], we measure the uncertainty surrounding expectations on the ECB’s and Fed’s monetary policy by using the variance of the stochastic, or unpredictable component of the spot money market rates. This is modelled adopting a stochastic volatility framework [Harvey et al., 1994], which allows the estimation of a time-varying variance, using an unobserved component model, as explained in the following section.

EMPIRICAL METHODOLOGY

Measuring the implied volatility from option prices is one possible way to quantify the uncertainty in expectations about a financial variable. The cross-sectional dispersion of individual forecasts from surveys summarizes the range of disagreement among individual forecasters. However, it does not give any information regarding each individual’s uncertainty about their own forecast. It is possible for each forecaster to be extremely uncertain about future events but to submit very similar estimates. Under these circumstances, survey measures would fail to capture the amount of existing uncertainty.

CONCLUSIONS

As pointed out in a number of studies [Faust and Svensson, 1998; Geerats, 2000], transparency can help central banks to build their reputations, thereby enabling monetary authorities to conduct their policies more successfully. This paper has addressed the issue of monetary policy uncertainty in both the United States and Euroland, and it has improved upon the existing literature [Gros et al., 2000] in two ways. First, we have estimated a stochastic volatility model to obtain a time-varying measure of uncertainty. This is in contrast to earlier studies relying on the Black-Scholes [1973] formula, which, implausibly, implies constant volatility. Furthermore, we have used policy-adjusted short-term interest rates, which has enabled us to focus on the component of interest rates which is market-driven. Secondly, we have analyzed directly the impact of monetary policy uncertainty on the euro-dollar exchange rate. Our empirical findings (which are consistent with those of Gros et al.) appear to confirm that the ECB’s communication strategy has not been very successful. In other words, the ECB’s actions and motivations are not well understood by the markets, and this lack of transparency (relative to the U.S. Fed) has been perceived as an additional source of risk by investors, with consequent capital outflows and a weakening of the euro relative to the dollar. Such a view has been receiving increasing support among foreign exchange dealers, media analysts and academic economists.

The ECB’s communication strategy undoubtedly plays a crucial role in the pursuit of its policy objectives. As suggested, for instance, by the IMF [1999], and also by Gros et al. [2000], a number of measures could be taken to improve it. For a start, publications of the inflation forecasts (as in inflation targeting countries, such as New Zealand, the UK and Sweden) and of the models used to produce such forecasts could lead to a better understanding of European monetary policy. For instance, reporting the conditional inflation (and output) forecast allows outsiders to monitor and scrutinize it. Hence, the evaluation of monetary policy does not have to be postponed some two years until actual inflation is observed. The more explicit a central bank is about its targeting rule, the easier it is for outsiders to monitor the extent to which this is adhered to, and to learn how to accurately predict the behavior of the central bank. This implies that the incentives for the central bank to build a reputation become stronger. In the last few months promising signals have emerged that the ECB is changing its attitude. Following an announcement made in November 2000 by its president, Mr. Wim Duisenberg, on 21 December the ECB published for the first time its economic projections (rather than forecasts) for the euro area in the form of ranges of prospective economic growth and inflation over a two-year horizon. However, Mr. Duisenberg stressed that they are “produced by staff experts and do not embody the policy judgments of the Governing Council” [Financial Times, 24 November 2000], and that they are simply one of various inputs into the policy-making process which is the responsibility of the ECB’s Governing Council.

Another important issue is whether the two-pillar strategy adopted by the ECB is the most effective one. The first pillar is targeting the M3 monetary aggregate. However, the ECB does not seem to have paid much attention to its first pillar. Svensson [2000] has pointed out that the emphasis on monetary aggregates is misplaced and confuses financial markets. Therefore, a strategy based only on an inflation target would make it easier for financial markets to understand the motives behind the ECB’s decisions. Improved communication could also be achieved by clarifying the meaning of the general goal of keeping inflation below 2 percent in the medium run. This could be done by specifying a target in terms of “core” inflation, defined as the Harmonized Consumer Price Index minus the more volatile components, such as energy and food. This would enable the ECB to ignore temporary deviation from the target driven by changes in the oil price or shifts in indirect taxes. An explicit distinction between direct and second-round inflationary effects has already been adopted in a number of inflation targeting countries, such as Canada and Sweden.

Finally, at present the communication and explanation of the ECB’s operating decisions takes place through a press conference and not through the publication of the minutes of the ECB meetings. The main reason given against their publication is that they would expose disagreements within the Governing Council, which might confuse rather than help outside observers in their attempts to understand the ECB’s policy. However, occasional comments made informally after an official press conference can be at times equally if not more confusing, especially when they offer conflicting views of the policy outlook and policy priorities. To avoid sending confusing signals to the markets, the EBC should indicate clearly which sources are authorized to make major policy pronouncements.

The costs of adopting such measures would be minimal, because none of them would require amending the Amsterdam Treaty defining the ECB’s institutional setup. By contrast, as our analysis appears to confirm, the benefits could be significant. It is to be hoped that the ECB will be responsive to the suggestions being made in many quarters.

NOTES

We are grateful to participants at the 2001 Eastern Economic Association Annual Conference, New York, February 23-25, 2001 for useful comments, and to Fabio Fraschetti for assistance with the data and helpful conversations.

1. Chadha and Nolan [1999] take the same approach to assess whether, in a regime of increased transparency in the UK, there has been a rise in the volatility of the 3-month Sterling Libor. Their results indicate that, contrary to their theoretical prior, higher transparency has been associated with less volatile interest rates in the UK.

2. Grier and Perry [2000] use GARCH models to estimate time-varying conditional variances of inflation and output growth. In a GARCH model, the conditional volatility follows a deterministic process, whereas in a stochastic volatility model, it follows a stochastic process. Recently, Panigiritzoglou et al. [2000] have applied GARCH models to policy-adjusted interest rates without obtaining significant coefficient estimates. We found equally unsatisfactory results (not reported) when applying the same framework to our data series.

3. Depending on the information set used, one can have a basic filter or a smoothing filter. The basic filter refers to an estimate of ht based on the information available up to time t, and the smoothing filter refers to an estimate of h, based on all the available information in the sample up to time T. In this paper we used the smoothing filter.

4. We first take the anti-log of the stochastic volatility unobserved component, h,, and then divide it by the absolute levels of the policy-adjusted rates.

5. The focus is on the dates of central bank meetings producing official interest rate variations. For the Federal Reserve, these dates are: 30 June 1999, 24 August 1999, 16 November 1999, 2 February 2000, 21 March 2000, and 19 May 2000. For the ECB, they are: 9 April 1999, 5 November 1999, 4 February 2000, 17 March 2000, 28/April 2000, 9 June 2000, 1/September 2000, and 6 October 2000. 6. See footnote 4.

7. The observed volatility pattern, for both 1 and 3-month rates, might reflect, to some extent, changes in the relative degree of policy activism in Euroland and the United States.

8. See footnote 5.

REFERENCES

Abramovitz, M. and Stegun, N. C. Handbook of Mathematical Functions. New York: Dover Publications Inc., 1970.

Barro, R. J. and Gordon, D. B. A Positive Theory of Monetary Policy in a Natural Rate Model. Journal of Political Economy, August 1983, 589-610.

Breidt, F. J. and Carriquiry, A. L. Improved Quasi-Maximum Likelihood Estimation for Stochastic Volatility Models, in Modelling and Prediction: Honouring Seymour Geisser, edited by C. Jack, W. 0. Lee, and A. Zellner. New York: Springer-Verlag, 1996, 228-47.

Black, F. and Scholes, M. The Pricing of Options and Corporate Liabilities. Journal of Political Economy, May/June 1973, 637-54.

Buiter, W. H. Alice in Euroland. Journal of Common Market Studies, June 1999, 181-209

Chadha, J. and Nolan, C. Inflation Targeting, Transparency and Interest Rate Volatility: Ditching “Monetary Mistique.” Working paper No. 9921, Department of Applied Economics, University of Cambridge, 1999.

Cohen, D. and Loisel, 0. Why Was the Euro Weak? Markets and Policies. CEPR discussion paper No. 2633, CEPR, 2000.

Cuckierman, AL Accountability, Credibility, Transparency and Stabilisation Policy in the Eurosystem. Mimeo, Tel-Aviv University, 1999.

Cuckierman, AL and Meltzer, A. H. A Theory of Ambiguity, Credibility, and Inflation under Discretion and Asymmetric Information. Econometrica, September 1986, 1099-1128.

Doornik, JA and Hansen, H. An Omnibus Test of Univariate and Multivariate Normality. Discussion paper, Nuffield College, Oxford, 1994.

Dornbusch, R. and Fischer, S. Exchange Rate Economics: Where Do We Stand? Brookings Papers on Economic Activity, 1980, 143-205.

Dotsey, M. Monetary Policy, Secrecy and Federal Funds Rate Behavior. Journal of Monetary Economics, November 1987, 463-74.

Economist. The Terrible Twos Begin. 6 January 2001, 67-70.

Faust, J. and Svensson, L.E. Transparency and Credibility: Monetary Policy with Unobservable Goals. NBER working paper No. 6452, 1998.

____.The Equilibrium Degree of Transparency and Control in Monetary Policy. NBER working paper No. 7152, 1999.

Favero, C. et al. One Money, Many Countries. Monitoring the European Central Bank 2, London: Centre for Economic Policy Research, 2000.

Geerats, P. Why Adopt Transparency? The Publication of Central Bank Forecasts. Working Paper, University of California, 2000.

Gersbach, H. On the Negative Social Value of Central Bank’s Transparency. Mimeo, University of Heidelberg, 1998.

Gersbach, H. and Hahn, V. Should Individual Voting Records of Central Bankers Be Published? Mimeo, University of Heidelberg, 2000.

Ghysel, S. E., Harvey, A. and Renault, E. Stochastic Volatility, in Handbook of Statistics, edited by G. S. Maddala and C.R. Rao. New York: Elsevier, 1996, 14.

Goodfriend, M. Monetary Mistique: Secrecy and Central Banking. Journal of Monetary Economics, January 1986, 63-92.

Grier, IL B. and Perry, M. The Effects of Real and Nominal Uncertainty on Inflation and Output Growth: Some GARCH-M Evidence. Journal ofApplied Econometrics, January/February 2000, 45-58. Gros, D., Davanne, O., Emerson, M., Tabellini, G., and Thysegen, N. Quo Vadis Euro? The Cost of

Muddling Through. 2″ Report of the CEPS Macroeconomic Policy Group, Brussels: Centre for European Policy Studies, 2000.

Hamilton, J. D. The Daily Market for Federal Funds. Journal of Political Economy, February 1996, 26-56. Harvey, A., Ruiz, E., and Shephard Multivariate Stochastic Variance Models. Review of Economic Studies, April 1994, 247-64.

IMF. Code of Good Practices on Transparency in Monetary and in Financial Policies: Declaration of Principles. Interim Committee, IMF, 26 September 1999.

_____.Why is the Euro So Undervalued? World Economic Outlook, Box 1.1 in Chapter 1: Prospects and Policy Challenges, IMF, 2000.

Isard, P. Exchange Rate Determination: A Survey of Popular Views and Recent Models. Princeton Studies in International Finance, 1978.

Issing, 0. The Eurosystem: Transparent and Accountable, or “Willem in Euroland”. Journal of Common Market Studies, September 1999, 503-19.

Jensen, H. Optimal Degrees of Transparency in Monetary Policymaking. Mimeo, University of Copenhagen,2000.

Kydland, F. E. and Prescott, E. C. Rules Rather than Discretion: The Inconsistency of Optimal Plans. Journal of Political Economy, June 1977, 473-91.

Nolan, C. and Schaling, E. Monetary Policy Uncertainty and Central Bank Accountability. Working paper No. 54, Bank of England, 1996.

Norman, P. Europe: Duisenberg Sets Out Rules for ECBs Projections. Financial Times, 24 November 2000, 13.

Panigiritzoglou, N., Proudman, J., and Spicer, J. Persistence and Volatility in Short-Term Interest Rates. Bank of England Working paper, Bank of England, 2000.

Phillips, P. C. B. and Hansen, B. H. Statistical Inference in Instrumental Variables Regression with I(1) Processes. Review of Economic Studies, January 1990, 99-125.

Poole, W. and Rasche, R. H. Perfecting the Market’s Knowledge of Monetary Policy. Working paper, 2000-010A, Federal Reserve Bank of St. Louis, 2000.

Quiros, G. B. and Mendizabal, H. R. The Daily Market for Funds in Europe: Has Something Changed with the EMU? Working paper, European Central Bank, 2000.

Romer, D.H. and Romer, C.D. Federal Reserve Private Information and the Behavior of Interest Rates. NBER working paper No. 5692, 1996.

Soderlind, P. and Svensson, L. E. 0. New Techniques to Extract Market Expectations from Financial Instrument. Journal of Monetary Economics, October 1997, 383-430.

Svensson, L. E. 0. The First Year of the Eurosystem: Inflation Targeting or Not? Working Paper #681. Institute for International Economic Studies, 2000.

Tarkka, J. and Mayes, D. The Value of Publishing Official Central Bank Forecasts. Discussion paper No. 22/99, Bank of Finland. 1999.

Guglielmo Maria Caporale

South Bank University, London

and

Andrea Cipollini

South Bank University, London

Guglielmo Maria Caporale: Centre for Monetary and Financial Economics, South Bank University, 103 Borough Road, London SE1 0AA, UK. E-mail: g.m.caporale@sbu.ac.uk

Copyright Eastern Economic Association Winter 2002

Provided by ProQuest Information and Learning Company. All rights Reserved