بازار اوراق قرضه مشترک، نرخ تورم مورد انتظار و تعادل نرخ واقعی ارز GBP-USD

Since the end of the fixed rates in 1973 and after the European Monetary System (EMS) sterling dismissal in 1992, the value of the pound has undergone large cyclical fluctuations on average. Of particular interest to policy makers is the understanding of whether such movements are consistent with the lack or not of a correction mechanism to some long-run equilibrium. The purpose of the present study is to understand those dynamics, how the external value of the British sterling (GBP) relative to the US dollar (USD) evolved during the recent floating experiences, and what have been the driving forces. In this paper we assume the real exchange rate to be determined by forces relating to the goods and capital market in a general equilibrium framework. This entails testing the purchasing power parity (PPP) and the uncovered interest parity (UIP) together. In doing so, we model inflation expectations explicitly. Our findings have two important implications, both for monetary policy. First, we show that some of the observed changes in the bilateral real exchange rate cannot be solely attributed to changes in inflation rates, but, also to capital markets. Secondly, we find a weaker behavior of the US bond rate on international markets, possibly explained by the special US dollar status of World reserve currency.

Since the end of the fixed rates in 1973 and after the European Monetary System (EMS) sterling dismissal in 1992, the value of the pound has undergone large cyclical fluctuations on average. Of particular interest to policy makers is the understanding of whether such movements are consistent with the presence or not of a correction mechanism to some long-run equilibrium. The purpose of the present study is primarily to understand those dynamics, how the external value of the British sterling (GBP) relative to the US dollar (USD) evolved during the recent floating experiences, and what have been the driving forces. In this paper we assume the real exchange rate (RER) to be determined by forces relating to the goods and capital market in a general equilibrium framework. Moving from the definition of two well known zero arbitrage conditions, the purchasing power parity (PPP) and the uncovered interest parity (UIP), it is assumed the RER observed deviations not to be exclusively explained by unidirectional inefficiencies on the goods markets (i.e. price stickiness, role of tradables vs. non-tradables goods, non linearities). 1 On the contrary these deviations are expected to involve real factors acting through the current account, as foreign net asset position or fiscal imbalances. 2 As the PPP does not to hold in the short run, the current account equality states that an increase in the domestic demand for goods can be satisfied by boosting imports and hence with a growing deficit on the balance of payment. 3 The latter can be financed by increasing the interest rate so to raise a relative supply of cash balances creating a wedge from one country to the other. 4 Using this approach, in a cointegrated VAR (CVAR) framework ( [Johansen, 1991] and [Johansen, 1994]), we are able to assess whether interest rates, prices and the real exchange rate are consistent with a UIP–PPP long-run equilibrium. Based on previous empirical results, we deepen the evidence in favor of a PPP-UIP joint relation on two main grounds. First, we explore the “credibility” implicit in the special USD status as World reserve currency. This role is understood as a weaker behavior of the US bond yield in the system, possibly because of dollar-denominated assets appetite/safe-heaven effects. In particular, if a “special USD” status/safe heaven effect is to be observed in the data, the US bond yield should not necessarily be a driving force in a long run good-capital market equilibrium (especially to the extent that this equilibrium reflects fundamentals; e.g., relative inflation and real exchange rate). Instead investors’ willingness to hold dollars and dollar denominated assets should exert downward pressure on both the US long term interest rate and the bilateral (£/$) spot nominal exchange rate, bringing them on a “bubble” path. Second, and most importantly, we assess the validity of previous empirical results (e.g., [Johansen and Juselius, 1992], [Jore et al., 1993], [Juselius and MacDonald, 2000], [Juselius and MacDonald, 2004], [Pesaran et al., 2000] and [Sjoo, 1992]) by including inflationary expectations, modeled here as long-run inflation forecast. In particular, with respect to the extant literature, modeling inflation expectations explicitly represents a useful addiction to the analysis as it allows to pay special attention to (i) symmetry and proportionality issues implicit in PPP/UIP testing; (ii) the role of accumulated shocks to nominal long term bonds onto inflation expectations. On this latter point, long term yields contain a premium for expected inflation, in line with the convention causality of Fisher's (1907) decomposition. Such a premium can be used as an important indicator of the credibility of a central bank's commitment to low inflation. The reasons for focusing on the US vs. UK data owe to the fact that (i) the US is Britain's largest single export market, and (ii) the UK and the US are each other's single largest investor.5 As discussed before, those are key features if one shares the idea that goods and capital markets may interact in keeping the exchange rate in line.6 Moreover, when dealing with dollar-based bilateral parities, it is worth noticing that the dollar benefits from an “exceptional” reserve currency status (the “credibility” issue mentioned in earlier); the implication being a feebler interest rate pressure in the US. Agents’ high willingness to hold dollar-denominated assets ([Juselius and MacDonald, 2000] and [MacDonald, 1998]) can possibly dampen the necessary US capital adjustment on international markets. Secondly, as interest rates in the US do not have to raise as much as in the UK to finance a given current account deficit, such a “status” for the USD will result into a weaker pressure over US current account imbalances. Especially the latter result implies that the argument according to which (short run) good markets misalignments may prompt a capital market reaction in the proportion of 1:1 is undermined in this setting. To preview the results in the paper, we find that the special USD status possibly plays a role, as the US bond displays a weaker behavior on international markets, compared to the UK bond rate, and consistent with Juselius and MacDonald (2000). Inference based on a standard cointegration analysis ([Johansen, 1991] and [Johansen, 1994]) shows moreover that – when inflation expectations are explicitly modeled – a combined UIP–PPP relation is found to hold as a long run condition, albeit not strictly mainly because of UK misalignments. Looking to the adjustment to the long run equilibrium, goods market adjustment is found to be very slow, whilst a major adjustment occurs on the capital and on the exchange rate markets. Nonetheless, the US bond rate is found to contribute to PPP-UIP misalignments by “pushing” in the opposite direction; reconciling with a failure of the UIP itself (e.g. [Bekaert et al., 2007], [Macchiarelli, 2011a], [Macchiarelli, 2011b] and [Macchiarelli, 2013]), and possibly being consistent with a dollar appetite/safe heaven effect explanation. Finally, looking at how the accumulated empirical shocks to each variable affect the others, we find that shocks to the UK and the US long term yields respectively increase expected inflation in the US but not in the UK; consistent, in the former case, with a Fisherian (1907) view of nominal rates. Instead, the fact that empirical shocks to the bond rate do not increase expected inflation in the UK is broadly consistent with the idea that the relationship between the UK real rate and expected inflation tend to become very little at long horizons (see Barr & Campbell, 1997). Finally, the accumulated shocks to the RER significantly affect the expected inflation in the UK but not in the US. Those latter findings signal imperfect price/capital adjustments on the international markets, and may account for the widespread finding of rejection of the UIP/PPP conditions when tested separately. Overall, our analysis has two important implications, both for monetary policy. First, we show that some of the observed changes in the real exchange rate can not be solely attributed to changes in inflation rates, but, also to capital markets, overall adjusting to a long run PPP-UIP equilibrium. When expected inflation is taken into account, this relationship is, anyhow, not found to hold strictly mainly because of UK misalignments. Secondly, we find a weaker behavior of the US bond rate on international markets, possibly explained by the special US dollar status of World reserve currency. The reminder of the paper is organized as follows. Section 2 presents the theoretical model. Section 3 goes through the econometric strategy and the empirical results. Section 4 is devoted to summary and conclusions.

The purpose of this analysis was to illustrate the extent to which the real exchange rate (RER) observed deviations could explain bilateral capital movements. This led us to combine the PPP and the UIP hypothesis in a CVAR general equilibrium framework, admitting goods and capital markets to interact in keeping exchange markets in line. There are two main reasons for which one would expect PPP and UIP not to hold separately. First, commodity price movements or speculative activities may lead to significant deviations from PPP/UIP. Secondly, price stickiness out of the equilibrium, together with the presence of limits to arbitrage and related market imperfections, may undermine short-term adjustments. As prices in markets for goods tend to adjust to shocks more slowly than prices in markets for assets, PPP deviations are more likely to persist if compared to UIP deviations ( Pesaran et al., 2000). Combining them both hence allows for a gradual convergence to PPP as it seems that the “PPP relation needs a lift […] to become more stable, and this is what the interest rates do” ( Johansen, 1994). The econometric strategy adopted in our “benchmark” analysis (Section 3.1) gave initial suggestions of the expected adjustments on the side of the exchange markets (RER term) to be very strong. This supports the evidence in favor of the RER acting as error correction mechanism in a PPP-UIP combined relation (see also [Johansen, 1994], [Juselius, 1995], [Juselius, 2006] and [Sjoo, 1992]). Relatively to the existing literature, the current paper innovates on two fundamental grounds. First, we explore a role for the “credibility” of the USD, implicit in its special status as World reserve currency. The literature reports indeed weak forms of PPP to hold for USD bilateral parings (see [Juselius and MacDonald, 2004] and [MacDonald, 1998]) as the result of adopting as numeraire a currency which is “exceptional”. Second, and more importantly, we extend the analysis by allowing for markets expectations. As discussed in Campbell et al. (2007), PPP (and UIP) deviations may be due to people not generally and promptly adjusting to changes in the behavior of prices, so that in our set up expectations are accounted for by means of a dynamic latent factor model. The overall results possibly point to a role for the special status of the USD both when the observed inflation and its long-run expectation enter the CVAR. Using our new estimates for inflation, we find a combined UIP–PPP relation not to hold strictly as a long run condition essentially because of UK misalignments: the cointegrating vector is not found to accommodate symmetry, but only proportionality between the US inflation, the US bond yield and the RER term. Inference based on a typical cointegration analysis moreover shows that the goods market adjustment is very slow, whilst a major adjustment occurs on the capital and on the exchange rate markets. Nonetheless, the US bond rate is found to “push” in the opposite direction, possibly reconciling with a failure of the UIP itself (e.g. [Bekaert et al., 2007], [Macchiarelli, 2011a] and [Macchiarelli, 2013]) and being consistent with a dollar appetite/safe haven explanation. Finally, looking at how accumulated empirical shocks to each variable affect the others, we find that empirical shocks to the US and the UK long term yields respectively increase expected inflation in the US but not in the UK; consistent, in the former case, with a Fisherian (1907) view of nominal rates. Instead, the fact that empirical shocks to the bond rate do not increase expected inflation in the UK is broadly consistent wit the idea that the relationship between the UK real rate and expected inflation tend to become very little at long horizons (see Barr & Campbell, 1997). Finally, the accumulated empirical shocks to the RER significantly affect the expected inflation in the UK but not in the US. Those latter findings signal imperfect price/capital adjustments on the international markets, and may account for the widespread finding of rejection of the UIP/PPP conditions when tested separately.