Independent monetary policies, synchronised outcomes

Espen Henriksen, Finn Kydland, Roman Šustek, 2 October 2013



The recession in the Eurozone has given new life to optimal-currency-area thinking. The argument goes that the disadvantages of a single currency come from the loss of flexibility and ability to use monetary policy to respond to “asymmetric shocks” (Krugman and Obstfeld 2009). The often-unarticulated presumption is that countries with independent monetary policies would make different policy decisions as long as contemporaneous shocks to output and employment were asymmetric.

A similar notion of ‘monetary-policy independence’ underlies the classical ‘trilemma’ of international finance. The trilemma states that it is impossible to have free capital mobility, fixed exchange rates, and independent monetary policy at the same time (e.g. Obstfeld and Taylor 2004). Hence, according to the trilemma, if there are free capital flows, only floating exchange rates permit monetary-policy independence.

The ‘trilemma’ has recently been put into a starker light – it has been rendered obsolete by financial globalisation. Risky assets, such as equities and corporate bonds, move in lockstep across the global economy, regardless of exchange-rate regimes. According to recent research, governments therefore face a dilemma, or an ‘irreconcilable duo’, instead of a trilemma – free capital flows may inevitably mean a loss of monetary-policy independence (Rey 2013).

Empirical regularity: Highly synchronised nominal variables

Empirical inspection of time series for key monetary-policy variables, such as inflation and short-term nominal interest rates, gives reason for caution when using the notion of ‘monetary-policy independence’. As shown in Figure 1, in a sample of the largest industrialised countries, cyclical fluctuations in the nominal variables have been substantially more synchronised across countries than cyclical fluctuations in real activity. The high cross-country correlations of short-term nominal fluctuations hold true through extended periods of time, different exchange-rate regimes, and different degrees of capital controls.

Figure 1. Cross-country comovement of nominal variables vs. cross-country comovement of real GDP, 1960 Q1–2006 Q4

Source: Henriksen, Kydland, and Šustek (2013).
Note: The sample includes the largest industrial economies: Australia, Canada, Germany, Japan, the United Kingdom, and the US for the period 1960 Q1–2006 Q4. In addition, from 1970 Q1 the sample also includes Austria and France, and from 1976 Q1 Switzerland. The correlations are for the cyclical components of the data obtained with a band-pass filter.

An explanation

It seems difficult to argue that we understand how nominal variables are determined in an international environment unless we can account for this striking feature of the data. Though puzzling at first, within a parsimonious analytical framework with no monetary-policy coordination across countries, and regardless of any specific assumptions about capital controls, a transparent mechanism can quantitatively account for this empirical fact (Henriksen, Kydland, and Šustek, 2013). Three assumptions turn out to be key for the mechanism:

  • Monetary policy in each country can be approximated as if it was following a Taylor-type rule,
  • GDP is co-integrated across countries, and
  • There are no systematic arbitrage opportunities between real and nominal domestic assets.

These assumptions have broad support. A large literature argues that the monetary policies of major central banks have been reasonably well-approximated by the so-called ‘Taylor rule’ (e.g. Taylor 1993; Clarida, Gali, and Gertler 2000; Woodford 2003). Furthermore, a number of studies document positive spillovers of total factor productivity – although of different strength – across countries (e.g. Backus, Kehoe, and Kydland 1992; Heathcote and Perri 2002; Rabanal, Rubio-Ramirez, and Tuesta 2011).

The mechanism is found to be quantitatively important and robust for a wide range of plausible parameterisations for both co-integration/spillovers of shocks and Taylor-type rules. The faster are the spillovers, the stronger is the effect. For highly persistent shocks, however, even a small degree of cross-country spillovers is quantitatively sufficient. The mechanism is also found to be robust to modifications of the economic environment that help account for other important features of fluctuations in the valuations of risky assets, and in domestic and international aggregate quantities.

Economic intuition

Monetary policy according to a Taylor-type rule, together with the absence of systematic arbitrage opportunities between real and nominal assets, implies that a country’s current price level and the nominal interest rate depend on the country’s expected output and real returns to capital (i.e. its marginal product) in future periods, appropriately discounted.

If real economic conditions are cointegrated across countries, a persistent domestic shock affects not only current and future output and capital returns in the domestic economy, but also in foreign economies – a shock to economic conditions in the domestic economy is expected to affect economic conditions in the foreign economy in a similar way in the future. Even if the monetary authority responds only to current domestic inflation and output, in competitive financial markets inflation and nominal interest rates depend on discounted expected future output and returns on real assets.

Although shocks may be asymmetric and current output growth may vary across countries, future output and returns to capital will converge to similar paths in expectation, leading to similar responses of current inflation and nominal interest rates. Hence, even though countries may be hit by asymmetric shocks and current output may be uncorrelated – and despite fully-independent monetary policies which take only current domestic economic conditions into account – nominal interest rates and inflation will be highly synchronised across countries.

Asset-price fluctuations

As pointed out by Rey (2013), a broad class of macroeconomic models – including those that successfully account for macroeconomic quantities and underlie the ‘trilemma’ – cannot account for several features of international and domestic asset-price fluctuations. A feature of nominal variables that is of particular relevance in this regard is the high volatility of nominal exchange rates in combination with smooth price levels (the so-called Mussa 1986 puzzle). Recent studies (e.g. Atkeson and Kehoe 2009, Cochrane 2011) argue that exchange-rate and interest-rate fluctuations over the business cycle are structurally related to cyclical distortions in standard asset-market Euler equations that occur due to various ‘frictions’, such as limited participation, time-varying risk aversion, or time-varying uncertainty.

The effects of such distortions can be captured by ‘wedges’ in Euler equations in the model behind the mechanism described above. Such extensions, while making the model consistent with asset-pricing facts, do not overturn the basic implications of the mechanism for the high cross-country correlations of price levels and nominal interest rates.

Optimal currency areas and capital controls

The theory of optimal currency areas may be right. However, empirical facts and theoretical underpinnings show that the basic criterion should not be whether contemporaneous shocks are symmetric or not, but instead whether output and returns on real assets co-move in the medium- to long-run. If that is the case, the monetary-policy decisions central banks would hypothetically have made individually would likely appear very similar. The costs of a common currency in terms of a lack of individual adjustment would hence be small.

Capital controls might be called for, but it is not obvious that individual countries’ monetary-policy outcomes would be much different in their presence. The fact is that monetary policies have appeared highly synchronised across industrialised countries both in periods with tight and loose capital controls. Rey (2013) is right to point out that an analytical framework for international economic policy analysis should also be able to account for fluctuations in domestic and international asset prices. A parsimonious analytical framework that can account for why short-term nominal interest rates and price levels are highly synchronised across countries can also account for asset-pricing facts when appropriately extended. When discussing the ‘dilemma’ of international finance and potential capital controls, therefore, a crucial feature of the discussion should be whether output and returns to capital in the countries in question are cointegrated or not.


Atkeson, A and P J Kehoe (2009), “On the need for a new approach to analyzing monetary policy”, in D Acemoglu, K Rogoff, and M Woodford (eds.), NBER Macroeconomics Annual 2008 23: 389–425.

Backus, D K, P J Kehoe, and F E Kydland (1992), “International real business cycles”, Journal of Political Economy 100: 745–75.

Clarida, R, J Gali, and M Gertler (2000), “Monetary policy rules and macroeconomics stability: Evidence and some theory”, Quarterly Journal of Economics 115: 147–80.

Cochrane, J H (2011), “Presidential address: Discount rates”, Journal of Finance 66(4): 1047–1108.

Heathcote, J and F Perri (2002), “Financial autarky and international business cycles”, Journal of Monetary Economics 49: 601–27.

Henriksen, E, F Kydland, and R Šustek (2013), “Globally correlated nominal fluctuations”, Journal of Monetary Economics 60(6), September.

Krugman, P R and M Obstfeld (2009), “Optimum currency areas and the European experience”, in International economics: theory & policy, Boston, MA: Pearson, Addison-Wesley: 565–592.

Mussa, M (1986), “Nominal exchange-rate regimes and the behavior of real exchange rates: Evidence and implications”, Carnegie-Rochester Conference Series on Public Policy 25: 117–213.

Obstfeld, M and A Taylor (2004), Global capital markets: integration, crisis and growth, Cambridge University Press.

Rabanal, P, J F Rubio-Ramirez, and V Tuesta (2011), “Cointegrated TFP processes and international business cycles”, Journal of Monetary Economics 58(2), 156–171.

Rey, H (2013), “Dilemma not Trilemma: The global financial cycle and monetary policy independence”, Paper presented at the Jackson Hole Symposium, August.

Taylor, J B (1993), “Discretion versus policy rules in practice”, Carnegie-Rochester Conference Series on Public Policy 39: 195–204.

Woodford, M (2003), Interest and prices: Foundations of a theory of monetary policy, Princeton University Press.

Topics: Exchange rates, Monetary policy
Tags: capital controls, Central Banks, EMU, exchange-rate policy, inflation, monetary policy

Assistant Professor, University of California Davis
Jeff Henley Professor of Economics, University of California Santa Barbara; Carnegie Mellon University; LAEF; and NBER
Lecturer, University of Southampton; Researcher, LSE—Centre for Macroeconomics