The Federal Reserve Board eagle logo links to home page

The Puzzling Peso

Carlos Arteta, Steven B. Kamin, and Justin Vitanza**

NOTE: International Finance Discussion Papers are preliminary materials circulated to stimulate discussion and critical comment. References in publications to International Finance Discussion Papers (other than an acknowledgment that the writer has had access to unpublished material) should be cleared with the author or authors. Recent IFDPs are available on the Web at http://www.federalreserve.gov/pubs/ifdp/. This paper can be downloaded without charge from the Social Science Research Network electronic library at http://www.ssrn.com/.

Abstract:

In the past decade, some observers have noted an unusual aspect of the Mexican peso's behavior: During periods when the U.S. dollar has risen (fallen) against other major currencies such as the euro, the peso has risen (fallen) against the dollar. Very few other currencies display this behavior. In this paper, we attempt to explain the unusual pattern of the peso's correlation with the dollar by developing some general empirical models of exchange rate correlations. Based on a study of 29 currencies, we find that most of the cross-country variation in exchange rate correlations with the dollar and the euro can be explained by just a few variables. First, a country's currency is more likely to rise against the dollar as the dollar rises against the euro, the closer it is to the United States and the farther it is from the euro area. In this result, distance likely proxies for the role of economic integration in affecting exchange rate correlations. Second, and perhaps more surprisingly, a country's currency is more likely to exhibit this unusual pattern when its sovereign credit rating is more risky. This may reflect that currencies of riskier countries are less substitutable in investor portfolios than those of better-rated countries. All told, these factors well explain the peso's unusual behavior, as Mexico both is very close to the United States and has a lower credit rating than most industrial economies.

Keywords: Mexico, peso, dollar, exchange rates, interest rate differentials, inflation, output gap, output growth differentials

JEL classification: F30, F31

I.  Introduction

The Mexican peso has been floating since shortly after the country's financial crisis at the end of 1994. Since then, some observers have noted an unusual aspect of the peso's behavior: During periods when the U.S. dollar has risen (fallen) against other major currencies such as the euro, the peso has risen (fallen) against the dollar.1 This pattern implies that when the dollar rises (falls) against other major currencies, the peso rises (falls) against those other currencies by an even greater extent.

Chart 1 illustrates this behavior, plotting the level of the nominal peso/dollar exchange rate against the level of the dollar/euro rate during the period 1997 through mid-2008; the correlation between the two series is 0.56.2 (The dollar/euro exchange rate is represented by the ECU, or European Currency Unit, for dates prior to the euro's inception in 1999; see Appendix for additional detail.) The relationship between the two exchange rates is also apparent in Chart 2, which plots monthly percentage changes; the correlation is 0.18.

Chart 3 puts the correlation between the levels of the peso/dollar and dollar/euro exchange rates in perspective, comparing it with correlations of other nominal exchange rates against the dollar with the dollar/euro rate. Chart 3 makes clear that the peso/dollar exchange rate's response to movements in dollar/euro is unusual. Of the 29 currencies shown, only 6 exhibit a positive correlation in levels; only the Argentine peso and the Venezuelan bolivar - which have not been market-determined in most of the sample period - have exhibited a higher correlation. Chart 4 presents similar information. It plots the response of different countries' exchange rates against the dollar to the dollar/euro exchange rate, estimated using OLS regressions; the diamond for each country is the coefficient on the dollar/euro exchange rate, while the vertical lines represent two-standard-error bands. The Mexican coefficient is clearly positive, significantly different from zero, and very precisely estimated, albeit lower than that of Turkey, Argentina and Venezuela. Chart 5 compares estimates of correlations with estimates of regression coefficients; it suggests that, for the most part, the regression coefficients and the ordinary correlations are providing similar information.

Evidence of the unusual behavior of the peso is reinforced by Chart 6, which presents correlations of monthly percent changes in nominal exchange rates against the dollar with changes in dollar/euro exchange rates. Of the 29 currencies shown, only four exhibit a positive correlation in percent changes, and the Mexican peso's is the highest. Chart 7 presents the analogous regression coefficients based on monthly percent changes in exchange rates. Mexico has the only coefficient that is significantly above zero. Chart 8 again confirms that regression coefficients and ordinary correlations provide similar information.

Because placecountry-regionMexico has long experienced inflation rates that have exceeded rates in the placecountry-regionUnited States, focusing on nominal exchange rates could be misleading. Charts 9-14 replicate the analysis shown in previous charts, but based on bilateral real exchange rates--in which nominal exchange rates are deflated by relative CPIs--rather than nominal exchange rates. They confirm that the real peso/dollar exchange rate has exhibited an unusual positive correlation with the real dollar/euro rate. Using real rather than nominal exchange rates, the correlation between peso/dollar and dollar/euro is one of the highest among the group shown for levels (chart 11), and the highest for percent changes (chart 13). In the analysis based on regression coefficients, the Mexican coefficient is the only one that is significantly greater than zero for percent changes.

Does the peso/dollar exchange rate respond in this unusual manner only to the dollar/euro rate, or does it respond to changes in the dollar's foreign exchange value more generally? The evidence is less robust but appears to support the latter hypothesis. Chart 15 plots correlations of movements in real bilateral exchange rates against the dollar with movements of the Federal Reserve's major currency index, a weighted average of the dollar's value against major industrial-country currencies. It shows that the peso/dollar exchange rate has the highest positive correlation with the dollar major currency index. Chart 16, based on regression analysis, has the peso/dollar rate showing only the third-largest positive response to changes in the dollar index; however, none of the positive coefficients (including Mexico's) is significantly different from zero.

Finally, the unusual correlation between peso/dollar and dollar/euro is not an artifact of outsized movements in these exchange rates during a select, limited time period. Chart 17 plots rolling 90-day correlations of daily levels of peso/dollar and dollar/euro, while Chart 18 plots correlations of daily percent changes in these exchange rates. The results indicate that, while the correlations are volatile and appear to have diminished in recent years, they were strongly positive on balance for long periods of time, and especially through 2003. To provide some perspective on this, Charts 19 and 20 present analogous calculations for the correlation between the Canadian dollar/U.S. dollar exchange rate and dollar/euro. This correlation has been negative for most of the past decade.

Charts 21 through 26 address some of the other exchange rates against the dollar that also exhibit high correlations with the dollar/euro exchange rate. As may be seen, notwithstanding their apparent high correlation in levels, the rolling correlations of daily percent changes for the Venezuelan bolivar, Russian ruble, and Argentine peso provide little evidence of systematic reactions to the dollar/euro exchange rate.

What accounts for the unusual positive correlation of the Mexican peso with the dollar's value against other major currencies such as the euro? The factor that comes most readily to mind is Mexico's proximity to, and thus close integration with, the United States. The United States is the major market for Mexican manufactures, and the manufacturing sector is playing an increasingly important role in overall Mexican economic activity. Possibly, the types of shocks that boost U.S. output, interest rates, and exchange rates relative to the euro area--for example, a shock to U.S. investment spending--might boost Mexican output, interest rates, and exchange rates to an even greater extent. Even so, this cannot be the whole story. Canada is also next door to and highly integrated with the United States, and yet the exchange rate of the Canadian dollar against the U.S. dollar exhibits the more normal negative correlation with the dollar/euro rate.

In the remainder of this paper, we attempt to explain the unusual pattern of the peso's correlation with the dollar. Section II briefly addresses a body of related research. Section III lays out the standard uncovered interest parity relationship between exchange rates and interest rate differentials, and assesses whether correlations among bilateral interest rate differentials can explain correlations among bilateral exchange rates. Section IV drills down a bit further, examining the explanatory power of the factors underlying correlations in interest rate differentials: output and inflation. Section V examines the possible role of a range of measures of trade and financial integration. Section VI concludes.

II.  Previous Related Research

We are not aware of any previous analyses of the unusual behavior of the peso/dollar exchange rate. However, this topic is similar in various respects to an issue that attracted some attention in previous decades: the response of European exchange rates to movements in the deutschemark/dollar rate. (See Frankel, 1985, Giavazzi and Giovannini, 1989, and Galati, 1999.) In particular, it was observed that appreciations of the mark against the dollar tended to be associated with increases in the other European currencies' value against the dollar as well, albeit generally to a less extent; this was described as the "dollar-mark axis" or "dollar-mark polarity". placeCityGalati (1999) found that this pattern could be explained by participation in the ERM, the close trade links between the European countries, and a measure of portfolio bias in international investments.

Unlike in the case of the "dollar-mark axis", however, the focus of this paper is not to explain why a group of currencies move together with respect to other currencies, but to explain why one particular currency--the peso--moves by an outsized amount when its "anchor currency"--the dollar--moves against other major currencies. In this sense, the peso's relation to the dollar is similar to the Swiss franc's relation to the mark in the pre-EMU period; alone among the European currencies, when the mark appreciated against the dollar, the Swiss franc tended to appreciate against the mark. Giavazzi and Giovannini (1989) and Galati (1999) suggest this pattern may have owed to portfolio shifts: a high share of the portfolios of international investors may have been allocated to Switzerland, so that shifts in portfolio allocations that tended to boost the mark against the dollar may have boosted the Swiss franc even more.

It is difficult to believe this portfolio allocation story, by itself, explains the puzzling behavior of the peso/ dollar exchange rate, however. Unlike the case of the deutschemark and the Swiss franc, the dollar and the peso likely offer very different attributes to investors and are placed by them in distinct baskets. The dollar is the world's preeminent reserve currency and offers maximum liquidity and safety; as we will discuss further below, the peso is more likely to be grouped by investors with other emerging market currencies.

The research that comes closest to bearing on the unusual behavior of the Mexican peso is Fratzscher (2008). This paper evaluates the impact of shocks to placecountry-regionU.S. monetary policy and economic performances on the values of different currencies. It finds that shocks tending to lower the value of the dollar (for example, a higher-than-expected employment report) tend to lower the dollar most against the euro and Swiss franc and least against emerging market countries. But most interesting for our purposes is that such shocks would actually boost the value of the dollar against a few currencies: in ascending order, Hong Kong, Argentina, Venezuela, Chile, and most of all, Mexico! Thus, the pattern of correlations we have documented corresponds closely to the pattern of response to shocks documented in Fratzscher (2008).

III.  Exploiting the Uncovered Interest Parity Relationship

Equation (1) presents the standard uncovered interest parity (UIP) relationship:

$\displaystyle i_{t} ^{p_{} } =i_{t} ^{\$ _{} } +e_{t+1} ^{p/\$ } -e_{t} ^{p/\$ } +\varepsilon_{t}$ (1)

$ i_{t} ^{p_{} } $ : peso interest rate

$ i_{t} ^{\$ _{} } $ : dollar interest rate

$ e_{t} ^{p/\$ } $ : log exchange rate, pesos per dollar

Re-arranging terms, the current exchange rate can be expressed as a function of the interest rate differential and the expected future exchange rate:

$\displaystyle e_{t} ^{p/\$ } =i_{t} ^{\$ _{} } -i_{t} ^{p_{} } +e_{t+1} ^{p/\$ } +\varepsilon_{t}$ (2)


If the future exchange rate is expected to revert to some constant equilibrium rate $ \overline{e^{p/\$ } }$, then the exchange rate essentially becomes a function of the interest rate differential alone:

$\displaystyle e_{t} ^{p/\$ } =i_{t} ^{\$ _{} } -i_{t} ^{p_{} } +\overline{e^{p/\$ } }+\varepsilon_{t}$ (3)

It then follows that the correlation of the peso/dollar exchange rate with the dollar/euro exchange rate will reflect the correlation of the peso/dollar interest rate differential with the dollar/euro interest rate differential:3

$\displaystyle corr(e_{t} ^{p/\$ } ,e_{t} ^{\$ /eu} )=corr(i_{t} ^{\$ _{} } -i_{t} ^{p_{} } ,i_{t} ^{eu_{} } -i_{t} ^{\$ } )$ (4)

To what extent do correlations in bilateral interest rate differentials match up with correlations in bilateral exchange rates, and does this relationship help explain the positive correlation between peso/dollar and dollar/euro? Chart 27 presents a scatterplot where each point represents correlations for a single country, computed using monthly data for the period 1997 through mid-2008. The x-axis plots the correlation between that country's interest rate differential with the United States (for Mexico, $ i_{t} ^{\$ _{} } -i_{t} ^{p_{} } $) and the U.S. interest rate differential with the euro area ( $ i_{t} ^{eu_{} } -i_{t} ^{\$ _{} } $). The y-axis plots the correlation between that country's exchange rate against the dollar (for placecountry-regionMexico, $ e_{t} ^{p/\$ } $) and the dollar/euro exchange rate ( $ e_{t} ^{\$ /eu} $). All interest rates are money market rates.4 For the euro area, we use the interbank rate, which is available for the entire sample period.

The scatter plot reveals the expected positive relationship between the two sets of correlations: the correlation of the interest rate differential between a given country and the United States with the interest rate differential between the U.S. and the euro area is positively associated with the correlation between that country's exchange rate against the dollar with the dollar/euro exchange rate. The slope of the regression line is 0.48, and it is significant at the 5 percent level. Note that Mexico is the only country whose correlations of interest rate differentials and exchange rates both exceed zero by a substantial margin. It appears that when U.S. interest rates rise relative to euro rates, Mexican interest rates rise relative to dollar rates--this may explain at least part of the positive response of peso/dollar exchange rates to dollar/euro exchange rates.

Chart 28 repeats this exercise, but with correlations involving 12-month changes in interest rate differentials and in exchange rates. The slope of the regression line is 0.59, and it is again significant at the 5 percent level. Mexico exhibits one of the highest correlations of exchange rates. Although no country had a positive correlation of changes in interest rate differentials, Mexico's correlation is one of the highest in the sample.

So far, we have referred to nominal variables in our summary of UIP and in our correlation analysis. However, the assumption that the future expected exchange rate is constant makes more sense if the analysis is re-cast in terms of real exchange rates rather than nominal rates. Starting with the nominal UIP equation (1), above, it is straightforward to derive a version of equation (3) that expresses the current real exchange rate as a function of the real interest rate differential and a constant equilibrium real exchange rate:

$\displaystyle rer_{t} ^{p/\$ } =r_{t} ^{\$ _{} } -r_{t} ^{p_{} } +\overline{rer^{p/\$ } }+\varepsilon_{t}$ (5)

$ r_{t} ^{p_{} } $ : real peso interest rate = $ i_{t} ^{p_{} } -(p_{t+1} ^{p_{} } -p_{t} ^{p_{} } )$

$ p_{t} ^{p_{} } $: Mexican price level

$ r_{t} ^{\$ _{} } $ : real dollar interest rate = $ i_{t} ^{\$ _{} } -(p_{t+1} ^{\$ _{} } -p_{t} ^{\$ _{} } )$

$ p_{t} ^{\$ _{} } $: U.S. price level

$ rer_{t} ^{p/\$ } $ : log real exchange rate, pesos per dollar = $ e_{t} ^{\$ /eu} +p_{t} ^{\$ _{} } -p_{t} ^{p_{} } $

Based on equation (5), Charts 29 and 30 repeat the analysis shown in Charts 27 and 28, but showing correlations between real interest rate differentials and real exchange rates. As shown in Chart 29, the relationship between correlations of levels of real interest rate differentials and correlations of levels of real exchange rates is not significant and explains very little of the variation across currencies. However, the relationship among correlations based on 12-month changes, shown in Chart 30, is again statistically significant; the slope of the regression line is .60 with a t-statistic of 2.7. 5

Even so, Mexico is a substantial outlier: Whereas the correlation of its real exchange rate movements against the dollar with dollar movements against the euro (the y-axis) is the highest in the sample, the correlation of Mexico-U.S. real interest rate differentials with U.S.-euro differentials (the x-axis) is negative and unremarkable. This may reflect that our calculations of ex post real interest rates are poor proxies for the ex ante real interest rates that influence exchange rate movements. Alternatively, other factors besides interest rate differentials may be influencing exchange rates.

IV.  Drilling Down Below Interest Rate Differentials

In this section, we drill down a little deeper to assess what factors may explain the pattern of correlations of interest-rate differentials that, in turn, appear to influence the pattern of exchange rate correlations. We start by assuming that interest rates in a given country j are set accordingly to the Taylor-rule type relation shown in equation (6) below:

$\displaystyle i_{t} ^{j_{} } =\overline{i_{t} ^{j_{} } }+\beta(\pi_{t} ^{j_{} } -\overline{\pi_{} ^{j_{} } })+\delta(y^{j} _{t} -\overline{y^{j} _{t} })+\eta^{j} _{t}$ (6)

$ i\overline{_{} ^{j_{} } }$: equilibrium nominal interest rate = $ \overline{r_{} ^{j_{} } +}\overline{\pi_{} ^{j_{} } }$

$ \overline{r_{} ^{j_{} } }$: equilibrium real interest rate

$ \pi_{t} ^{j_{} } $: inflation rate = $ p_{t} ^{j_{} } -p_{t-1} ^{j_{} } $

$ \overline{\pi_{} ^{j_{} } }$: target inflation rate

$ y_{t} ^{j} $: log real output

$ \overline{y_{t} ^{j} }$: log real potential output

Define the inflation and output gaps:

$\displaystyle \pi gap_{t} ^{j_{} } =\pi_{t} ^{j_{} } -\overline{\pi_{} ^{j_{} } } $

$\displaystyle ygap_{t} ^{j_{} } =y_{t} ^{j_{} } -\overline{y_{t} ^{j_{} } } $

Accordingly, the interest rate differential between Mexico and the United States, for example, is expressed as:6

\begin{displaymath}\begin{array}[c]{l} {i_{t} ^{p_{} } -i_{t} ^{\$ _{} } =\overline{i_{} ^{p_{} } }-\overline{i_{} ^{\$ _{} } }+\beta(\pi gap_{t} ^{p_{} } -\pi gap_{t} ^{\$ _{} } )}\\ {+\delta(ygap^{p} _{t} -ygap^{\$ } _{t} )+\eta^{p} _{t} -\eta^{\$ } _{t} } \end{array}\end{displaymath} (7)

With the interest rate differential between the United States and the euro area expressed similarly, it is apparent that the correlation between the Mexico/U.S. and U.S./euro area interest rate differentials--that is, $ corr(i_{t} ^{p_{} } -i_{t} ^{\$ _{} } ,i_{t} ^{\$ _{} } -i_{t} ^{eu} )$ -will depend on the correlations and cross correlations between inflation gap differentials--e.g., correlations of $ (\pi gap_{t} ^{p_{} } -\pi gap_{t} ^{\$ _{} } )$ with $ (\pi gap_{t} ^{\$ _{} } -\pi gap_{t} ^{eu_{} } )$--and output gap differentials--e.g., correlations of $ (ygap^{p} _{t} -ygap^{\$ } _{t} )$ with $ (ygap^{\$ } _{t} -ygap^{eu} _{t} )$.

Previous research supports the view that arguments in the Taylor-rule relation influence exchange rates. Clarida and Waldman (2007) show that positive inflation surprises tend to lead a country's currency to appreciate, and especially so for countries with explicit inflation targets. See also Mark (2005), Engel and West (2006), and Molodtsova and Papell (2008), among others.

To what extent can correlations in inflation gap differentials and output gap differentials empirically explain the cross-country pattern of correlations in interest rate differentials and, ultimately, exchange rates? To answer this question, we depart from the bivariate scatterplot approach utilized above and instead estimate multivariate regressions.

IV.1  Explaining Patterns of Correlations of Interest Rate Differentials

Table 1 presents the results of estimates of equations explaining correlations of interest rate differentials as a function of correlations of output gap differentials and correlations of inflation gap differentials. Output gaps are calculated as the percent difference between industrial production (IP) and a trend measure of IP calculated using an HP filter; we denote them IPgap. Inflation gaps (pgap) are calculated analogously, as 12-month CPI inflation minus an HP filter of inflation.7 The data are analyzed in 12-month changes, indicated by D. Accordingly, correlations of 12-month changes in interest rate differentials for a given country X - Corr[D(i$ - iX), D (ieu - i$)] - are related to correlations of changes in IPgap differentials--Corr[D(IPgap$ - IPgapX), D (IPgapeu - IPgap$)]-and correlations of changes in inflation gap differentials--Corr[D(pgap$ - pgapX), D (pgapeu - pgap$)]. The data are also analyzed in both nominal and real terms.

Two results are worth highlighting. First, correlations of inflation gap differentials are significant and robust explainers of correlations in interest rate differentials. This means that if increases in a country's inflation gap relative to that of the United States are associated with increases in the U.S. inflation gap relative to that of the euro area, it is more likely that increases in a country's interest rate relative to the U.S. rate will be associated with increases in the U.S. interest rate relative to the euro area rate.

Second, and conversely, looking at columns (1) and (3), there appears to be no relationship between correlations in IPgap differentials and correlations in interest rate differentials. It is possible that the output gaps are being mis-measured, or that they are not the best measure of economic slack. To explore this possibility, we estimated another set of regressions, using correlations of the differentials in 12-month percent changes in IP rather than correlations of the differentials in IP gaps. Because these equations are estimated in 12-month changes, with the 12-month percent change in IP denoted Dip, the explanatory variable becomes Corr[D(Dip$ - DipX), D (Dipeu - Dip$)].8 However, as indicated in columns (2) and (4), this did not materially change the results.

How much of the cross-country pattern in correlations of interest rate differentials is explained by the output and inflation correlations? Chart 31 plots the fitted values from the regression in equation (2) in Table 1--based on changes in nominal interest rates--against their actual values; Chart 32 presents plots the fitted and actual values from equation (4), based on changes in real interest rates. The solid lines represent points where the fitted value equals actual; the dashed lines represent the fitted values plus/minus twice the standard error of the regression, a measure of the confidence interval. As can be seen in these charts, this simple regression does a relatively poor job of fitting the nominal interest rate correlations, albeit a somewhat better job of fitting the real interest rate correlations.

IV.2  Explaining Patterns of Correlations of Exchange Rates

Table 2 presents estimation results for equations explaining exchange rate correlations as a function of correlations of IP gap differentials, inflation gap differentials, and interest rate differentials. If output and inflation affected exchange rates exclusively through their effect on interest rates, of course, we would expect them to have little measured effect, once interest rates were added to the equation. The estimation results, however, suggest otherwise. Although the coefficients on correlations of IP gap differentials remain insignificantly different from zero, the coefficients on correlations of IP growth differentials are significantly different from zero. In contrast, the coefficients on correlations of inflation gaps differentials, and especially interest rate differentials, are not consistently significant, particularly in the regressions using real exchange rate changes. Accordingly, correlations of output growth emerge as the single most consistent influence on patterns of exchange rate correlations, and this influence appears to go beyond their effects on interest rates.9

How well do the set of output, inflation, and interest rate correlations explain the cross-country pattern of exchange rate correlations? Chart 33 plots the fitted values from equation (3) in Table 2--based on changes in nominal exchange rates--against their actual values; Chart 34 presents plots the fitted and actual values from equation (6), based on changes in real exchange rates. These models correctly predict Mexico to have the highest exchange rate correlations, both in nominal and real terms, although the predicted values of these correlations are below zero. Moving to the other side of the rankings, the model successfully predicts close-to-negative-one correlations for several European currencies. However, the fit of these models is obviously poor, as evidenced by the wide dispersion of actual correlations for given levels of fitted values.

V.  Other Factors Influencing the Pattern of Exchange Rate Correlations

The evidence summarized in Charts 33 and 34 suggests that, although correlations of output, inflation, and interest rates explain some of the country-country pattern of exchange rate correlations, much of this pattern remains unexplained. In this section, we assess a broad set of additional factors that might help further explain why the correlation of a country's exchange rate against the dollar with the dollar/euro rate might be high or low. Following on work by Fratzscher (2008), we focus on the extent to which a country is integrated with the United States through either trade or finance, along with more general measures of financial integration and maturity. Accordingly, we consider the effect of the following measures:

  1. Log of Distance from the United States, and Log of Distance from the euro area (specifically, Germany): Measured as the Great Circle log distance between country centers .
  2. Trade Share: Measured as the sum of a country's bilateral imports and exports with the United States divided by the country's GDP.
  3. Stock Return Correlation: Measured as the correlation of monthly log changes in a country's stock market index vis-à-vis the S&P 500 index.
  4. U.S. Portfolio Integration: Measured as the sum of a country's claims on and liabilities to the United States divided by the country's GDP.
  5. International Financial Integration: Measured as the sum of a country's total external assets and liabilities divided by the country's GDP.
  6. International Financial Size: Measured as the sum of a country's total external assets and liabilities in absolute dollar terms.
  7. Credit Rating: Based on Moody's and S&P sovereign credit ratings and converted to a numerical system, with the safest rating indicated by 1 and the riskiest rating indicated by 27.

Tables 3 and 4 present estimates of regressions in which measures of the correlation of a country's exchange rate against the dollar with the dollar/euro rate are related to (1) the output, inflation, and interest rate correlations discussed in the previous section, and (2) the additional factors described above. In Table 3, the dependent variable is the correlation of 12-month changes in nominal exchange rates; in Table 4, the dependent variable is based on correlations of changes in real exchange rates. Columns (2) through (8) include the additional factors separately, while Column (9) includes them jointly. Column (10) represents a reduced version of Column (9), where we progressively remove explanatory variables with the smallest t-statistics. Because the distance variables may be well-correlated with other measures of integration between countries, the equation shown in Column (11) represents the outcome of the same exercise, but with the distance variables removed at the outset.

All told, a number of variables are robustly significant determinants of exchange rate correlations. These include, first, the distance variables. The closer a country is to the United States, the higher is its exchange rate correlation--that is, the greater the likelihood that its currency will rise against the dollar when the dollar rises against the euro. Conversely, the closer a country is to the euro area, the lower (more negative) its exchange rate correlation. We interpret these distance variables as proxies for the degree of economic integration between a country and the United States/euro area, and, in fact, they appear to be rather good proxies. Other measures of economic integration--the correlation of IP growth differentials, the correlation of inflation gap differentials, the correlation of interest rate differentials, and the trade share--are only sporadically significant in the regressions, and mainly when the distance variables are not included.

The second robustly significant variable in the regressions shown in Tables 3 and 4 is the credit rating. The coefficient on the credit rating variable is positive and significantly different from zero in all regressions, suggesting that currencies of riskier countries are more likely to rise against the dollar when the dollar rises against the euro. Generally speaking, less developed countries have lower credit ratings. However, the coefficient on credit rating remains positive and significant even after including the country's per capita income as a control variable. Accordingly, it appears to be a country's perceived riskiness that is affecting the exchange rate correlations rather than its level of development per se. It is not clear what accounts for this result. One possibility is that the credit rating variable serves to distinguish the currencies of major financial centers--which may be most substitutable in global investor portfolios--from the currencies of other countries. Accordingly, when a shock renders U.S. investments more attractive, investors may shift their portfolios more out of other low-risk currencies (generally in industrial economies) than out of higher-risk currencies (generally in emerging market economies). Another possibility is that, given the centrality of the United States in the world economy, shocks which boost the U.S. economy and the dollar relative to the euro area are regarded by the market as favorable for reduction of risk around the world.

Finally, measures of a country's financial integration do not appear to have much effect, one way or the other, on exchange rate correlations. The coefficients on Stock Return Correlation, International Financial Integration, and U.S. Portfolio Integration are all insignificantly different from zero, while that on International Financial Size is usually insignificant as well. Perhaps the economies in our sample all have financial systems that are sufficiently linked to world capital markets that variations in integration among them have little bearing on the responsiveness of their currencies to economic conditions.

How much of the cross-country pattern of exchange rate correlations is explained by the augmented models shown in Tables 3 and 4? We replicate the exercise described in Section IV, focusing on the regressions shown in Column (10), which combine high explanatory power with parsimonious specifications. Chart 35 plots the fitted values from Table 3--based on changes in nominal exchange rates--against their actual values; Chart 36 presents plots of the fitted and actual values from Table 4, based on changes in real exchange rates. The charts suggest that the factors considered explain much of the pattern of exchange rate correlations. In particular, Mexico's high exchange rate correlations, both in nominal and real terms, are well-predicted by the models.

Of the explanatory variables in the models, which of them account most for Mexico's unusually high correlation? To address this question, we decompose the fitted values for each country's exchange rate correlation into the respective contributions of the explanatory variables--in practice, this means multiplying the explanatory variables by their coefficients. For ease of interpretation, we combine the constant with the contributions of the two distance variables. Chart 37 presents the estimated contributions the explanatory variables to the nominal exchange rate correlations, while Chart 38 presents the analogous calculations for the real exchange rate correlations. The dashed black lines indicate the fitted values themselves, while the solid grey lines indicate the actual value of the exchange rate correlations.

Focusing on Chart 38, which has fewer variables and is easier to interpret, it is apparent that distance to the United States and the euro area accounts for most of the variation in predicted and actual exchange rate correlations. For Mexico, the combination of the constant and the two distance variables makes a small positive contribution to the exchange rate correlation, whereas for countries close to the euro area, the contribution is large and negative. Accordingly, to the extent that distance proxies for economic integration, Chart 38 lends support to our initial hypothesis that the peso's unusual behavior reflects Mexico's close economic relationship with the United States.

Yet, Canada is almost equally close to the United States in both geographical and economic terms, but it exhibits a negative exchange rate correlation. Chart 38 highlights several factors that differentiate Canada from Mexico. First, although Canada's IP is relatively well correlated with U.S. IP, the correlation of the Canada/U.S. IP-growth differential with the U.S./euro area IP-growth differential is negative (-.55), whereas that of Mexico is positive (.21).10 Second, Canada has a much safer credit rating than Mexico. Accordingly, investors may view Canadian dollars as more substitutable for U.S. dollars in their portfolio (compared with pesos and dollars), or they may view U.S. shocks as having different implications for risk in Canada compared with in Mexico. Finally, Canada's exchange rate correlation is lower than the model prediction, whereas Mexico's is higher, introducing a third, unexplained factor distinguishing the two countries.

VI. Concluding Remarks

This paper is, to our knowledge, the first attempt to systematically document and account for a puzzling feature of the Mexican peso: when the dollar rises (falls) against the euro, the peso tends to rise (fall) against the dollar. We have found strong evidence that this behavior is very unusual. The correlation between changes in the peso/dollar and dollar/euro exchange rates during 1997-2008 has been among the highest of a broad set of currencies, whether measured in nominal or real terms, and is one of only a few currency correlations to exceed zero.

What explains the peso's unusual behavior? Our starting hypothesis was that the Mexican economy is unusually reliant upon the placecountry-regionU.S. economy. Accordingly, shocks that boost U.S. demand relative to euro-area demand will tend to boost Mexican demand even more. Therefore, even as the shocks to U.S. output raise U.S. interest rates relative to euro rates--and thus boost the dollar against the euro--they raise Mexican interest rates relative to U.S. rates--and thus boost the peso against the dollar.

To evaluate this hypothesis, we focused on explaining the cross-country variation in exchange rate correlations. We first estimated a number of simple regression models based on the uncovered interest parity (UIP) condition, which links exchange rates to interest rates, and the Taylor rule, which links interest rates to output and inflation. We showed that correlations of interest rate differentials are significantly related to patterns of exchange rate correlations: countries whose interest rates rise relative to U.S. rates when U.S. rates rise relative to euro rates are also likely to have currencies that rise against the dollar when the dollar rises against the euro. We showed, as well, that across the countries in our sample, correlations of differentials in output growth and inflation are systematically related to correlations in exchange rates.

Even so, the models we estimated based on the UIP condition and the Taylor rule did not explain a great deal of the cross-country variation in exchange rate correlations, nor did they consistently account for Mexico's unusually high and positive correlation. Accordingly, we departed from the simple UIP/Taylor-rule framework and tested the explanatory power of a number of additional variables intended to proxy for countries' economic and financial integration with the United States, the euro area, and the broader global financial system. We found that just a few variables consistently and significantly explained most of the cross-country variation in exchange rate correlations.

First, most of the variation in exchange rate correlations appears to be explained by two variables representing a country's distance from the United States and the euro area, respectively. The closer a country is to the United States, and the farther from the euro area, the more likely a country's currency will rise against the dollar when the dollar rises against the euro. Clearly, distance represents a proxy for a range of economic and financial ties that are too diverse to be captured by just a few economic or financial statistics. Another measure of economic integration, the correlation of a country's industrial production with that of the United States and the euro area, was found to be a significant influence on exchange rate correlations in many, but not all, of the models we estimated.

Second, a country's sovereign credit rating is a robust, statistically significant influence on a country's exchange rate correlation. The safer the credit rating, the more likely that a country's currency will fall against the dollar when the dollar rises against the euro. Our preferred interpretation for this effect is that in the portfolios of international investors, the dollar and the currencies of other highly rated countries are most substitutable with each other. Accordingly, when some shock enhances the attractiveness of U.S. assets, investors are more likely to shift out of the euro and other highly-rated currencies (mainly of industrial countries) than out of lower-rated currencies (mainly of emerging markets).

Our estimated models based on these variables correctly predict Mexico to have a positive correlation of its currency against the dollar with the dollar/euro exchange rate, and among the highest in the sample. This mainly reflects Mexico's proximity to the United States, which has led to considerable integration and, presumably, has led Mexico's economic and financial prospects to be highly dependent upon U.S. economic and financial prospects. Of course, Canada is geographically close to, and economically integrated with, the United States as well. But Canada has a much safer credit rating than does Mexico, and this offsets much of the effect on its exchange rate correlation conferred by its geographical proximity. Finally, given our estimated results, it is no surprise that the currencies of the euro area's neighbors--given their proximity to the euro area and relatively favorable credit ratings--tend to move alongside the euro when the euro moves against the dollar.

References

Banco de Mexico (2003), Annual Report Summary 2002, Mexico City, April.

Clarida, Richard and Daniel Waldman (2007), "Is Bad News About Inflation Good News for the Exchange Rate?" Paper presented at the 8th Jacques Polak Annual Research Conference, International Monetary Fund, November 15-16.

Engel, Charles and Kenneth D. West (2006), "Taylor Rules and the Deutschmark-Dollar Real Exchange Rate," Journal of Money, Credit, and Banking, Vol. 38, No. 5, pp. 1175-1194.

Frankel, Jeffrey A. (1985), "The Implications of Mean-Variance Optimization for Four Questions in International Macroeconomics," NBER Working Paper No. 1617, May.

Fratzscher, Marcel (2008), "U.S. Shocks and Global Exchange Rate Configurations," Economic Policy, April, pp. 363-409.

Galati, Gabriele (1999), "The Dollar-Mark Axis," BIS Working Papers No. 74, August.

Giavazzi, Francesco and Alberto Giovannini (1989), Limiting Exchange Rate Flexibility, MIT Press, placeCityCambridge, StateMA.

Mark, Nelson (2005), "Changing Monetary Policy Rules, Learning, and Real Exchange Rate Dynamics," NBER Working Paper No. 11061, January.

Molodtsova, Tanya and David H. Papell (2008), "Out-of-Sample Exchange Rate Predictability with Taylor Rule Fundamentals," Journal of International Economics, forthcoming.


Data Appendix

Industrial Production

Seasonally adjusted industrial production.

Source: Haver Analytics for Argentina, Chile, Colombia, India, Indonesia, Russia, Taiwan, and Venezuela; CEIC for Thailand; IFS line 66..b/c for Brazil, Canada, Czech Republic, Denmark, Hungary, Israel, Japan, Korea, Mexico, Norway, Poland, Sweden, Turkey, United Kingdom, and United States. IFS line 66ey (Manufacturing production, not seasonally adjusted) was manually seasonally adjusted for Pakistan and South Africa. Missing Pakistan data is filled by Manufacturing Production series in EMERGEPR of Haver Analytics.

Interest Rates

Nominal short-term interest rate.

Source: IFS line 60b (Money Market Rate). For Chile, Hungary, India, Norway, and Sweden, IFS line 60 (Discount Rate) was used. For Israel, IFS line 60c (Treasury Bill Rate) was used. For Argentina, IFS line 60l (Deposit Rate) was used.

Inflation

12-month percent change in seasonally adjusted consumer price index.

Source: Haver Analytics; CEIC for China, Hong Kong, India, Indonesia, Korea, Malaysia, Philippines, Singapore, Taiwan, and Thailand. IFS line 64 for Pakistan, Peru, and South Africa. Seasonally adjustments were done manually when unavailable.

Exchange Rates

Nominal bilateral exchange rate with the United States (end of period).

Source: IFS line ae. Before January 1999, the EU uses United States line ea (the $/ECU rate).

Equity Prices

Major stock market index.

Source: Bloomberg.

GDP

Gross Domestic Product in current USD.

Source: IFS line 99b(.c).

Trade Levels

Imports and Exports to the United States in current USD.

Source: IMF Direction of Trade Statistics.

Trade Share

Sum of imports and exports to the United States as a percent of GDP

Distance to the United States

Great Circle distance to the United States based on longitude and latitude given by CIA Factbook.

Source: Andrew Rose's Webpage.

International Investment Position

Assets and liabilities in USD.

Source: IFS line 79aad and 79lad.

Portfolio Claims

Portfolio investment from the Coordinated Portfolio Investment Survey.

Source: IMF.

Credit Ratings

Average of S&P and Moody's ratings over the sample period.

Source: Bloomberg.

Real Interest Rate

Calculated by subtracting 12-month inflation rate from the nominal interest rate.

Real Exchange Rate

Calculated by scaling nominal exchange rate changes by US and local inflation.

Interest Rate Differential

Calculated by subtracting local interest rate from the US rate.

Inflation Gap

Calculated by subtracting inflation from an HP-filtered trend of inflation.

Industrial Production Gap

Calculated as the percent deviation of industrial production from an HP-filtered trend measure of industrial production.

Chart 1 - Monthly Exchange Rates: Peso/Dollar and Dollar/Euro

Data for Chart 1 immediately follows

Data for Chart 1 - Monthly Exchange Rates: Peso/Dollar and Dollar/Euro

DatesPeso/$$/Euro
Dec-967.8768571.25299
Jan-977.8288571.1895
Feb-977.8023421.15463
Mar-977.9562141.16173
Apr-977.9058681.1349
May-977.9037381.14906
Jun-977.9497571.13002
Jul-977.8679091.08046
Aug-977.7817861.09704
Sep-977.7809291.113
Oct-977.8708361.14277
Nov-978.2715561.1225
Dec-978.1270681.10421
Jan-988.227151.08348
Feb-988.5020531.09
Mar-988.5680681.07618
Apr-988.5016591.1005
May-988.5848151.10376
Jun-988.9200051.0959
Jul-988.8990431.10717
Aug-989.3712141.11423
Sep-9810.219241.17159
Oct-9810.159381.18398
Nov-989.9684741.1517
Dec-989.9067051.16675
Jan-9910.127921.1384
Feb-9910.005711.1018
Mar-999.7323911.0742
Apr-999.4304451.0597
May-999.3954751.0456
Jun-999.5146141.0328
Jul-999.3699051.0694
Aug-999.3978861.0573
Sep-999.3412861.066499
Oct-999.5752251.045
Nov-999.4161251.0097
Dec-999.4271091.0046
Jan-009.4935250.979096
Feb-009.4265250.971402
Mar-009.2885870.955301
Apr-009.3936850.908496
May-009.5058730.930302
Jun-009.8343410.955603
Jul-009.4192250.9243
Aug-009.2724350.890599
Sep-009.3614750.876501
Oct-009.5369520.841701
Nov-009.5081430.868402
Dec-009.467250.930501
Jan-019.7687860.929299
Feb-019.7108160.924804
Mar-019.5989770.883197
Apr-019.3275950.887603
May-019.14750.847997
Jun-019.0880950.847997
Jul-019.1682380.875503
Aug-019.1331870.915801
Sep-019.4252720.9131
Oct-019.3390770.9042
Nov-019.224980.889798
Dec-019.1573750.881298
Jan-029.1636190.8637
Feb-029.1049740.865097
Mar-029.0640240.872402
Apr-029.1649090.900804
May-029.5098860.938703
Jun-029.7670750.997496
Jul-029.7791680.978301
Aug-029.8389090.983304
Sep-0210.070780.985999
Oct-0210.094090.986398
Nov-0210.195180.992704
Dec-0210.225071.0487
Jan-0310.622311.0816
Feb-0310.944681.0782
Mar-0310.905341.0895
Apr-0310.58871.1131
May-0310.252761.182199
Jun-0310.502851.1427
Jul-0310.458081.1318
Aug-0310.783021.0927
Sep-0310.922861.1652
Oct-0311.179641.1622
Nov-0311.149441.1994
Dec-0311.251481.262999
Jan-0410.920321.238399
Feb-0411.031931.241799
Mar-0411.018991.2224
Apr-0411.270071.1947
May-0411.519931.2246
Jun-0411.39261.2155
Jul-0411.467821.2039
Aug-0411.395291.211099
Sep-0411.487021.2409
Oct-0411.403731.2737
Nov-0411.370981.329501
Dec-0411.201161.362101
Jan-0511.262711.314301
Feb-0511.137341.325701
Mar-0511.155231.2964
Apr-0511.112091.295699
May-0510.976441.2331
Jun-0510.819661.2092
Jul-0510.672391.2093
Aug-0510.686241.219799
Sep-0510.785831.2042
Oct-0510.835381.2023
Nov-0510.671511.1769
Dec-0510.626641.1797
Jan-0610.542241.2118
Feb-0610.484181.1875
Mar-0610.749291.2104
Apr-0611.048861.2537
May-0611.090771.286799
Jun-0611.393381.271301
Jul-0610.983011.276701
Aug-0610.873461.285099
Sep-0610.988761.266001
Oct-0610.88541.269599
Nov-0610.913281.32
Dec-0610.854581.317001
Jan-0710.95591.295401
Feb-0710.995071.3211
Mar-0711.114421.331801
Apr-0710.980241.3605
May-0710.82211.345299
Jun-0710.833011.350501
Jul-0710.814561.3707
Aug-0711.04381.370499
Sep-0711.031911.417901
Oct-0710.821421.4447
Nov-0710.881151.4761
Dec-0710.84631.472099
Jan-0810.905691.487
Feb-0810.767891.5167
Mar-0810.732761.5812
Apr-0810.51461.553999
May-0810.438131.550801
Jun-0810.326921.576399

 

Chart 2 - Month-to-Month Percent Change of Peso/Dollar, Dollar/Euro Exchange Rates

Data for Chart 2 immediately follows

Data for Chart 2 - Month-to-Month Percent Change of Peso/Dollar, Dollar/Euro Exchange Rates

DatePeso/$$/Euro
Jan-97-0.60938-5.06708
Feb-97-0.33868-2.93148
Mar-971.9721280.614916
Apr-97-0.63279-2.30949
May-97-0.026941.247687
Jun-970.582244-1.65701
Jul-97-1.02957-4.38576
Aug-97-1.094621.534532
Sep-97-0.011011.454824
Oct-971.1554892.674753
Nov-975.09119-1.77376
Dec-97-1.7468-1.6294
Jan-981.231463-1.87736
Feb-983.3414080.601765
Mar-980.776466-1.26789
Apr-98-0.775082.259845
May-980.9781140.296229
Jun-983.904447-0.71211
Jul-98-0.234991.028379
Aug-985.305860.637662
Sep-989.0492415.14795
Oct-98-0.585731.057537
Nov-98-1.87912-2.7264
Dec-98-0.619641.306764
Jan-992.232998-2.42984
Feb-99-1.20667-3.21505
Mar-99-2.73163-2.50493
Apr-99-3.10248-1.34985
May-99-0.37083-1.33063
Jun-991.268043-1.22418
Jul-99-1.520913.543777
Aug-990.298633-1.13142
Sep-99-0.602270.87005
Oct-992.504359-2.01591
Nov-99-1.66158-3.37795
Dec-990.116648-0.50511
Jan-000.704525-2.5387
Feb-00-0.70574-0.78586
Mar-00-1.4633-1.65745
Apr-001.131475-4.89952
May-001.1942892.400179
Jun-003.4554242.719645
Jul-00-4.22109-3.27572
Aug-00-1.55841-3.64611
Sep-000.960268-1.58296
Oct-001.874463-3.9703
Nov-00-0.302083.172274
Dec-00-0.430087.150899
Jan-013.18504-0.12917
Feb-01-0.59342-0.48367
Mar-01-1.15169-4.49901
Apr-01-2.82720.498833
May-01-1.93078-4.46216
Jun-01-0.649410
Jul-010.8818443.24374
Aug-01-0.382314.602817
Sep-013.198065-0.29493
Oct-01-0.91451-0.97473
Nov-01-1.22172-1.59274
Dec-01-0.73285-0.95533
Jan-020.068186-1.99687
Feb-02-0.639980.161773
Mar-02-0.449750.844486
Apr-021.113033.255504
May-023.764114.207266
Jun-022.7044346.263279
Jul-020.123816-1.92432
Aug-020.61090.511318
Sep-022.3566220.274108
Oct-020.2315210.040442
Nov-021.001510.639301
Dec-020.293155.640745
Jan-033.8849423.13729
Feb-033.034883-0.3144
Mar-03-0.35951.048099
Apr-03-2.903472.166092
May-03-3.172656.207847
Jun-032.439203-3.34114
Jul-03-0.42623-0.95388
Aug-033.10709-3.45468
Sep-031.2967916.63488
Oct-032.350843-0.25743
Nov-03-0.270063.20084
Dec-030.9151795.302577
Jan-04-2.94327-1.94775
Feb-041.0220060.274562
Mar-04-0.11725-1.56223
Apr-042.278583-2.26599
May-042.2169952.502716
Jun-04-1.10522-0.74316
Jul-040.660246-0.95433
Aug-04-0.632490.598041
Sep-040.8050072.46058
Oct-04-0.725162.64331
Nov-04-0.287144.380971
Dec-04-1.493482.452054
Jan-050.549528-3.50931
Feb-05-1.113120.867406
Mar-050.160616-2.21023
Apr-05-0.38677-0.05403
May-05-1.22072-4.83129
Jun-05-1.42832-1.93823
Jul-05-1.361170.008223
Aug-050.1298130.868253
Sep-050.931941-1.27886
Oct-050.459366-0.15774
Nov-05-1.51231-2.11265
Dec-05-0.420440.237946
Jan-06-0.79432.720975
Feb-06-0.5507-2.00521
Mar-062.5286481.92841
Apr-062.7868643.577308
May-060.3793852.640126
Jun-062.728435-1.20443
Jul-06-3.601810.424758
Aug-06-0.997480.657842
Sep-061.060412-1.48616
Oct-06-0.940550.284263
Nov-060.2560443.969767
Dec-06-0.53789-0.22718
Jan-070.933521-1.64011
Feb-070.3575141.983896
Mar-071.0854360.810001
Apr-07-1.207262.154896
May-07-1.44025-1.11727
Jun-070.100850.386648
Jul-07-0.170331.495708
Aug-072.119802-0.01466
Sep-07-0.107753.458685
Oct-07-1.907951.890102
Nov-070.5519232.173409
Dec-07-0.3203-0.27101
Jan-080.5475671.012201
Feb-08-1.263571.997343
Mar-08-0.326234.252638
Apr-08-2.03259-1.72028
May-08-0.72729-0.20579
Jun-08-1.065461.650648

Chart 3 - Correlation of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Nominal Levels, Jan 1997-Jun 2008

Data for Chart 3 immediately follows

Data for Chart 3 - Correlation of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Nominal Levels, Jan 1997 - Jun 2008*

CountryCorrelation
DN-0.98441
NO-0.96362
UK-0.9565
SD-0.94654
CZ-0.94581
HU-0.90842
PL-0.9051
CA-0.89927
SI-0.77694
KO-0.5933
TH-0.5615
PE-0.47448
IN-0.42621
MA-0.35073
CL-0.3467
JA-0.34115
SF-0.33567
TA-0.23975
IS-0.20981
RU-0.09587
BZ-0.08121
CO-0.08121
ID-0.01112
PH0.021059
PK0.213051
TK0.229169
MX0.56283
AR0.580504
VE0.731434

Chart 4 - Regression Coefficients of Country X's Exchange Rate against Dollar on Dollar/Euro Exchange Rate: Nominal Levels , Jan 1997 - Jun 2008

Data for Chart 4 immediately follows

Data for Chart 4 - Regression Coefficients of Country X's Exchange Rate against Dollar on Dollar/Euro Exchange Rate: Nominal Levels , Jan 1997 - Jun 2008

Country CodeBeta2*sigma
CZ-1.470980.094435
HU-0.999990.067878
DN-0.99630.004814
PL-0.976680.089456
NO-0.957550.040625
SD-0.907690.036462
CA-0.810430.075234
UK-0.656670.032376
KO-0.55460.12459
SF-0.462570.224449
TH-0.433370.118998
SI-0.364710.054164
CL-0.349920.164966
PE-0.263530.087676
IN-0.219580.081283
MA-0.209140.108486
RU-0.200240.610524
JA-0.157440.078005
BZ-0.145120.376077
CO-0.145120.376077
IS-0.12620.107271
TA-0.081710.059812
ID-0.050740.387991
PH0.0084980.209821
PK0.1303760.147831
MX0.3728630.100604
TK0.8049890.84834
AR2.1225720.498575
VE2.471210.496128

Chart 5 - Nominal Levels of Exchange Rates

Data for Chart 5 immediately follows.

Data for Chart 5 - Nominal Levels of Exchange Rates

Country CodeRegression CoefficientCorrelation
CZ-1.47098-0.94581
HU-0.99999-0.90842
DN-0.9963-0.98441
PL-0.97668-0.9051
NO-0.95755-0.96362
SD-0.90769-0.94654
CA-0.81043-0.89927
UK-0.65667-0.9565
KO-0.5546-0.5933
SF-0.46257-0.33567
TH-0.43337-0.5615
SI-0.36471-0.77694
CL-0.34992-0.3467
PE-0.26353-0.47448
IN-0.21958-0.42621
MA-0.20914-0.35073
RU-0.20024-0.09587
JA-0.15744-0.34115
BZ-0.14512-0.08121
CO-0.14512-0.08121
IS-0.1262-0.20981
TA-0.08171-0.23975
ID-0.05074-0.01112
PH0.0084980.021059
PK0.1303760.213051
MX0.3728630.56283
TK0.8049890.229169
AR2.1225720.580504
VE2.471210.731434

Chart 6 - Correlation of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Nominal Monthly Percent Changes, Jan 1997 - Jun 2008

Data for Chart 6 immediately follows.

Data for Chart 6 - Correlation of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Nominal Monthly Percent Changes, Jan 1997 - Jun 2008*

Country CodeCorrelation
DN-0.99201
SD-0.86423
NO-0.81075
CZ-0.8044
HU-0.79823
UK-0.67053
PL-0.58896
SI-0.42054
JA-0.39124
TA-0.26761
TH-0.26346
SF-0.22296
CA-0.21718
TK-0.16739
CL-0.15917
PE-0.15237
IS-0.1408
PH-0.13682
ID-0.12256
IN-0.11773
MA-0.09916
KO-0.07155
BZ-0.05401
CO-0.05401
PK-0.02167
AR0.006261
RU0.096552
VE0.14132
MX0.18481

Chart 7 - Regression Coefficients of Country X's Exchagne Rate against Dollar on Dollar/Euro Exchange Rate: Nominal Monthly Percent Changes, Jan 1997 - June 2008

Data for Chart 7 immediately follows.

Data for Chart 7 - Regression Coefficients of Country X's Exchange Rate against Dollar on Dollar/Euro Exchange Rate: Nominal Monthly Percent Changes, Jan 1997 - Jun 2008

Country CodeBeta2*sigma
CZ-1.086960.137673
DN-1.006580.021953
HU-0.951260.123107
SD-0.943130.094158
NO-0.894340.110746
PL-0.714310.168098
ID-0.625240.868336
UK-0.53710.101913
JA-0.461990.18637
TH-0.433630.2723
SF-0.386860.290076
TK-0.349360.352898
SI-0.28770.106445
TA-0.177210.109425
CA-0.174480.134492
PH-0.163680.203237
CL-0.162250.172594
BZ-0.154880.491054
CO-0.154880.491054
KO-0.146260.349668
IS-0.121560.146583
MA-0.119740.206064
PE-0.079440.088371
IN-0.064440.093212
PK-0.015730.124493
AR0.0166620.456366
MX0.1676860.152927
VE0.3291570.395439
RU0.3478930.615048

Chart 8 - Nominal Percent Changes of Exchange Rates

Data for Chart 8 immediately follows

Data for Chart 8 - Nominal Percent Changes of Exchange Rates

Country CodeRegression CoefficientCorrelation
CZ-1.08696-0.8044
DN-1.00658-0.99201
HU-0.95126-0.79823
SD-0.94313-0.86423
NO-0.89434-0.81075
PL-0.71431-0.58896
ID-0.62524-0.12256
UK-0.5371-0.67053
JA-0.46199-0.39124
TH-0.43363-0.26346
SF-0.38686-0.22296
TK-0.34936-0.16739
SI-0.2877-0.42054
TA-0.17721-0.26761
CA-0.17448-0.21718
PH-0.16368-0.13682
CL-0.16225-0.15917
BZ-0.15488-0.05401
CO-0.15488-0.05401
KO-0.14626-0.07155
IS-0.12156-0.1408
MA-0.11974-0.09916
PE-0.07944-0.15237
IN-0.06444-0.11773
PK-0.01573-0.02167
AR0.0166620.006261
MX0.1676860.18481
VE0.3291570.14132
RU0.3478930.096553

Chart 9 - Real Monthly Exchange Rates: Peso/Dollar and Dollar/Euro

Data for Chart 9 immediately follows

Data for Chart 9 - Real Monthly Exchange Rates: Peso/Dollar and Dollar/Euro

DATESpeso/$$/Euro
Dec-96122.924140.1996
Jan-97120.9778133.3281
Feb-97118.7301129.2218
Mar-97119.1826129.9503
Apr-97118.8051126.7066
May-97117.2262128.6177
Jun-97116.8212126.3233
Jul-97113.2977120.8575
Aug-97111.1673122.7202
Sep-97110.8177124.2976
Oct-97113.6967127.5329
Nov-97114.0274125.3188
Dec-97111.5007123.2577
Jan-98113.9988120.8363
Feb-98115.3249121.6337
Mar-98113.1665120.0776
Apr-98111.8695122.7815
May-98116.1396123.0265
Jun-98116.6687122.1972
Jul-98113.78123.2646
Aug-98125.4844123.9984
Sep-98125.3364130.2572
Oct-98124.2045131.3285
Nov-98119.8366127.6802
Dec-98116.839129.1719
Jan-99118.7023125.9047
Feb-99114.5862121.9546
Mar-99108.7288118.9909
Apr-99105.7741116.8065
May-99110.0498115.2223
Jun-99106.1208113.8767
Jul-99104.2757117.7922
Aug-99103.6558116.4185
Sep-99102.9227116.8619
Oct-99105.5484114.5484
Nov-99101.872110.6821
Dec-99103.1965110.1852
Jan-00102.7814107.3244
Feb-00100.9339106.2454
Mar-0099.32543103.9667
Apr-00100.441498.85302
May-00101.1305101.0823
Jun-00105.417103.692
Jul-0098.67757100.2508
Aug-0096.6426796.83247
Sep-0098.4467895.15352
Oct-00100.381391.29886
Nov-0097.5417994.33478
Dec-0098.66063101.105
Jan-01100.1956100.2115
Feb-01100.468799.60984
Mar-0198.4961595.36761
Apr-0195.2706696.10063
May-0193.27491.71596
Jun-0192.8382291.67891
Jul-0193.9631394.83342
Aug-0192.8665999.23077
Sep-0196.4149898.72054
Oct-0192.8986298.17171
Nov-0193.0827796.73579
Dec-0191.7807596.15613
Jan-0291.7295894.47276
Feb-0291.0485994.5067
Mar-0290.2516395.25216
Apr-0292.9688698.12409
May-0295.38511102.2586
Jun-0298.60312108.742
Jul-0295.27314106.5941
Aug-0297.18721106.9561
Sep-0299.59701107.2308
Oct-0299.37309107.3556
Nov-0299.03784107.9547
Dec-02100.4367114.1692
Jan-03107.3325117.5938
Feb-03108.0524116.9535
Mar-03105.1414118.2958
Apr-03100.9939121.1043
May-03100.4125128.6443
Jun-03100.954124.497
Jul-03101.0015123.1511
Aug-03105.4671118.6779
Sep-03105.3791126.4573
Oct-03106.7161126.3732
Nov-03108.654130.6997
Dec-03107.4738137.3729
Jan-04104.3339134.5659
Feb-04105.6652134.6712
Mar-04106.1171132.6523
Apr-04107.7644129.6252
May-04108.9436132.6714
Jun-04108.5488131.3991
Jul-04108.9029130.1803
Aug-04107.4399131.0585
Sep-04107.4335134.1368
Oct-04108.3253137.379
Nov-04105.7847142.9437
Dec-04105.9199146.6988
Jan-05106.376141.4505
Feb-05104.415142.5922
Mar-05106.1566139.3156
Apr-05104.3185138.7473
May-05101.8789132.5015
Jun-05101.2306130.1401
Jul-0599.59108129.5937
Aug-05102.5377130.2812
Sep-05103.3858127.5022
Oct-05104.0539127.3101
Nov-0599.96473125.0746
Dec-05101.4481125.597
Jan-0698.5707128.4284
Feb-0698.49157126.1776
Mar-06103.009128.5188
Apr-06105.1308132.8703
May-06104.9906136.1844
Jun-06107.4807134.381
Jul-06102.984134.5889
Aug-06103.0464135.1525
Sep-06102.9329133.6433
Oct-0699.29977134.7264
Nov-06102.3179140.1952
Dec-06100.9869139.2946
Jan-07102.5844136.9999
Feb-07102.493139.6004
Mar-07102.7862140.2895
Apr-07101.5648143.2444
May-07100.4779141.2247
Jun-07101.1269141.6422
Jul-07102.1609143.6201
Aug-07102.8419143.8876
Sep-07101.082148.776
Oct-0799.15415151.8559
Nov-07101.7187154.8147
Dec-07101.2154154.124
Jan-08101.0418155.4954
Feb-0899.68632158.9624
Mar-0898.97755165.9139
Apr-0896.43459162.6314
May-0895.53874162.137
Jun-0895.37736163.8184

Chart 10 - Month-to-Month Percent Change of Real Peso/Dollar, Dollar/Euro Exchange Rates

Data for Chart 10 immediately follows

Data for Chart 10 - Month-to-Month Percent Change of Real Peso/Dollar, Dollar/Euro Exchange Rates

DATESpeso/$$/Euro
Dec-96-2.6083-1.40147
Jan-97-1.58326-4.90119
Feb-97-1.85792-3.07986
Mar-970.3811190.563701
Apr-97-0.31675-2.49604
May-97-1.329031.508238
Jun-97-0.34548-1.78391
Jul-97-3.01612-4.32682
Aug-97-1.880411.541293
Sep-97-0.314431.285316
Oct-972.5979262.602903
Nov-970.290926-1.73612
Dec-97-2.21595-1.64471
Jan-982.240465-1.96451
Feb-981.1632680.659903
Mar-98-1.8716-1.27929
Apr-98-1.14612.251746
May-983.8170470.199584
Jun-980.455589-0.67412
Jul-98-2.475980.873507
Aug-9810.286860.595347
Sep-98-0.117975.047437
Oct-98-0.903020.822486
Nov-98-3.51676-2.77799
Dec-98-2.501361.168326
Jan-991.59473-2.52935
Feb-99-3.46756-3.13738
Mar-99-5.11184-2.43015
Apr-99-2.7175-1.8358
May-994.042349-1.35629
Jun-99-3.57017-1.1678
Jul-99-1.738713.438364
Aug-99-0.59448-1.16618
Sep-99-0.707240.380855
Oct-992.551109-1.97972
Nov-99-3.48311-3.37519
Dec-991.300162-0.44894
Jan-00-0.40225-2.59637
Feb-00-1.79751-1.00539
Mar-00-1.59361-2.14471
Apr-001.123535-4.91861
May-000.6861152.255176
Jun-004.2385842.5817
Jul-00-6.39314-3.31863
Aug-00-2.06217-3.40979
Sep-001.866776-1.73388
Oct-001.964998-4.05099
Nov-00-2.828683.325254
Dec-001.1470397.176755
Jan-011.555802-0.88373
Feb-010.272546-0.60035
Mar-01-1.96333-4.25884
Apr-01-3.274740.768628
May-01-2.09577-4.56259
Jun-01-0.46721-0.04039
Jul-011.2116913.440817
Aug-01-1.166994.636919
Sep-013.820956-0.51419
Oct-01-3.64711-0.55594
Nov-010.198228-1.46266
Dec-01-1.39877-0.59922
Jan-02-0.05575-1.75067
Feb-02-0.74240.035932
Mar-02-0.875310.788791
Apr-023.0107293.015081
May-022.5989794.213578
Jun-023.3737046.340213
Jul-02-3.37715-1.97527
Aug-022.0090310.339615
Sep-022.4795480.256789
Oct-02-0.224830.116432
Nov-02-0.337360.558066
Dec-021.4124215.756531
Jan-036.8658772.999617
Feb-030.670674-0.5445
Mar-03-2.694041.14771
Apr-03-3.944742.374133
May-03-0.575646.226019
Jun-030.539241-3.22385
Jul-030.047131-1.08104
Aug-034.421262-3.63229
Sep-03-0.083446.555058
Oct-031.268744-0.0665
Nov-031.8160123.423594
Dec-03-1.086275.105768
Jan-04-2.92157-2.0434
Feb-041.276040.078268
Mar-040.427673-1.49915
Apr-041.552306-2.28195
May-041.0943192.350022
Jun-04-0.36239-0.95905
Jul-040.326154-0.92756
Aug-04-1.34340.674615
Sep-04-0.005912.348818
Oct-040.8300922.41712
Nov-04-2.345384.050617
Dec-040.1278232.62697
Jan-050.430627-3.57763
Feb-05-1.84350.807122
Mar-051.668013-2.29787
Apr-05-1.73151-0.40794
May-05-2.33863-4.50151
Jun-05-0.6363-1.78217
Jul-05-1.61963-0.4199
Aug-052.9587370.530505
Sep-050.827069-2.13308
Oct-050.646229-0.15064
Nov-05-3.92984-1.756
Dec-051.4838880.417706
Jan-06-2.836322.254357
Feb-06-0.08028-1.75259
Mar-064.5866341.855499
Apr-062.0597823.385923
May-06-0.133342.494182
Jun-062.371777-1.32423
Jul-06-4.183810.154702
Aug-060.0606050.418783
Sep-06-0.11007-1.11667
Oct-06-3.529650.810443
Nov-063.0393744.059223
Dec-06-1.30084-0.64242
Jan-071.58191-1.64739
Feb-07-0.08911.898172
Mar-070.2860690.493671
Apr-07-1.188262.106273
May-07-1.07014-1.40996
Jun-070.6459140.295613
Jul-071.0224251.396432
Aug-070.6666060.186214
Sep-07-1.711283.397394
Oct-07-1.90722.070177
Nov-072.5863811.948418
Dec-07-0.49478-0.44613
Jan-08-0.17150.889767
Feb-08-1.341492.229664
Mar-08-0.7114.373011
Apr-08-2.56923-1.97841
May-08-0.92897-0.30401
Jun-08-0.168921.037066

Chart 11 - Correlation of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Real Levels, Jan 1997 - Jun 2008*

Data for Chart 11 immediately follows

Data for Chart 11 - Correlation of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Real Levels, Jan 1997 - Jun 2008*

Country CodeCorrelation
DN-0.98505
NO-0.95509
UK-0.92185
SD-0.92023
IN-0.89001
CZ-0.88757
HU-0.87949
CA-0.87256
PL-0.82311
TK-0.82287
RU-0.76491
SF-0.67085
KO-0.61922
PK-0.59911
TH-0.58946
BZ-0.46959
PE-0.46273
ID-0.46215
CL-0.44348
PH-0.37962
CO-0.37024
SI-0.33341
MA-0.25331
IS-0.02126
VE0.19427
TA0.228745
MX0.295117
AR0.304872
JA0.34258

Chart 12 - Regression Coefficients of Country X's Exchange Rate Against Dollar on Dollar/Euro Exchange Rate: Real Levels, Jan 1997 - Jun 2008

Data for Chart 12 immediately follows

Data for Chart 12 - Regression Coefficients of Country X's Exchange Rate against Dollar on Dollar/Euro Exchange Rate: Real Levels, Jan 1997 - Jun 2008

Country CodeBeta2*sigma
RU-1.603170.212508
HU-1.407040.138696
CZ-1.397930.137245
TK-1.333270.159654
DN-0.940770.009544
PL-0.936390.126474
BZ-0.923050.288612
ID-0.910970.241461
SF-0.91040.154646
NO-0.866430.043772
SD-0.848050.04929
CA-0.758930.080145
CO-0.666880.271571
KO-0.612860.12672
UK-0.563460.035934
TH-0.493610.123094
IN-0.48850.049605
CL-0.450010.155469
PK-0.405250.092377
PH-0.375180.159222
PE-0.232170.078896
SI-0.189080.094424
MA-0.158180.112582
IS-0.014950.10716
TA0.1261840.107485
MX0.1768730.077872
JA0.220540.110391
VE0.2777180.178923
AR1.0133360.446133

Chart 13 - Correlation of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Real Monthly Percent Changes, Jan 1997 - Jun 2008*

Data for Chart 13 immediately follows

Data for Chart 13 - Correlation of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Real Monthly Percent Changes, Jan 1997 - Jun 2008*

Country CodeCorrelation
DN-0.99027
SD-0.86111
NO-0.79883
HU-0.78799
CZ-0.78417
UK-0.66401
PL-0.55647
SI-0.41421
JA-0.3866
TH-0.2715
TA-0.23182
CA-0.21377
SF-0.2126
IS-0.15683
CL-0.15436
PE-0.13977
ID-0.12364
PH-0.12364
IN-0.1043
MA-0.09861
TK-0.0911
KO-0.06687
BZ-0.05568
PK-0.04884
CO-0.04725
AR-0.01994
RU0.060861
VE0.142231
MX0.204521

Chart 14 - Regression Coefficients of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Real Monthly Percent Changes, Jan 1997 - Jun 2008

Data for Chart 14 immediately follows

Data for Chart 14 - Regression Coefficients of Country X's Exchange Rate against Dollar with Dollar/Euro Exchange Rate: Real Monthly Percent Changes, Jan 1997 - Jun 2008

Country CodeBeta2*sigma
CZ-1.05860.143802
DN-1.005890.024267
SD-0.943890.095613
HU-0.903890.121165
NO-0.885270.114431
PL-0.658610.17098
ID-0.601010.829577
UK-0.530170.102501
JA-0.457190.187978
TH-0.441160.269254
SF-0.368280.291064
SI-0.292380.110231
CA-0.174660.136716
TK-0.173850.326023
BZ-0.160870.492067
TA-0.158290.11478
CL-0.155230.173827
PH-0.146730.20457
CO-0.134150.483319
KO-0.131090.338486
IS-0.128520.139698
MA-0.118720.207264
PE-0.07540.093543
IN-0.063080.104709
AR-0.04890.425323
PK-0.035130.126574
RU0.115580.31992
MX0.188020.153973
VE0.3355340.39604

Chart 15 - Correlation of Country X's Exchange Rate against Dollar with Major Currencies Index: Real Monthly Percent Changes, Jan 1997 - Jun 2008*

Data for Chart 15 immediately follows

Data for Chart 15 - Correlation of Country X's Exchange Rate against Dollar with Major Currencies Index: Real Monthly Percent Changes, Jan 1997 - Jun 2008*

Country CodeCorrelation
DN-0.57299
SD-0.54543
UK-0.51593
JA-0.47432
CZ-0.46098
SI-0.43481
NO-0.40421
HU-0.40244
CA-0.34041
TH-0.32202
TA-0.31414
PL-0.29485
MA-0.22148
PH-0.22114
CL-0.19656
ID-0.18106
IN-0.17538
KO-0.1726
IS-0.16091
SF-0.14736
PE-0.13545
PK-0.07926
AR-0.04608
TK-0.02224
VE0.044089
RU0.054881
BZ0.109346
CO0.122895
MX0.139995

Chart 16 - Regression Coefficients of Country X's Exchange Rate against Dollar on Major Currencies Index: Real Monthly Percent Changes, Jan 1997 - Jun 2008

Data for Chart 16 immediately follows

Data for Chart 16 - Regression Coefficients of Country X's Exchange Rate against Dollar on Major Currencies Index: Real Monthly Percent Changes, Jan 1997 - Jun 2008

Country CodeBeta2*sigma
ID-1.400251.304345
CZ-0.994290.328253
SD-0.954980.251675
DN-0.929610.228032
JA-0.89970.286379
TH-0.837810.422428
HU-0.73740.287672
NO-0.715670.277736
UK-0.658340.187462
PL-0.562680.312739
KO-0.545290.533673
SI-0.490380.174174
CA-0.443710.210186
MA-0.429570.324365
PH-0.424490.321042
SF-0.408740.4705
TA-0.345060.178842
CL-0.322150.275589
IS-0.211940.222936
AR-0.182550.678716
IN-0.171930.165522
PE-0.119120.149433
PK-0.093530.20175
TK-0.067820.522753
RU0.1638240.511169
VE0.164320.638558
MX0.2050990.248779
BZ0.5018760.782421
CO0.5536970.766821

Chart 17 - Rolling 90-day Correlation of Daily Exchange Rates: Mexico*

Chart 17 depicts calculations of the correlation between the nominal levels of the peso-dollar exchange rate and the dollar-euro exchange rate, measured over rolling 90-day periods, during 1997-2008. It shows that these calculations, while volatile, generally were positive, although results were more mixed after 2003.

Chart 18 - Rolling 90-day Correlation of Percent Changes in Daily Exchange Rates: Mexico*

Chart 18 depicts calculations of the correlation between the nominal daily percent changes in the peso-dollar exchange rate and the dollar-euro exchange rate, measured over rolling 90-day periods, during 1997-2008. As with the correlations of the levels of these exchange rates presented in Chart 17, it shows that these correlations, while volatile, generally were positive during 1997-2003, but became more predominantly negative 2004-2008.

Chart 19 - Rolling 90-day Correlation of Daily Exchange Rates: Canada*

Chart 19 depicts calculations of the correlation between the nominal levels of the Canadian dollar-U.S. dollar exchange rate and the U.S. dollar-euro exchange rate, measured over rolling 90-day periods, during 1997-2008. It shows that these calculations, while volatile, generally were negative, especially after 2000.

Chart 20 - Rolling 90-day Correlation of Percent Changes in Daily Exchange Rates: Canada*

Chart 20 depicts calculations of the correlation between the nominal daily percent changes in the Canadian dollar-U.S. dollar exchange rate and the U.S. dollar-euro exchange rate, measured over rolling 90-day periods, during 1997-2008. As with the correlations of the levels of these exchange rates presented in Chart 19, it shows that these correlations, while volatile, generally were positive during 1997-2003, but became more predominantly negative 2004-2008.

Chart 21 - Monthly Exchange Rates: Venezuelan Bolivar/Dollar and Dollar/Euro

Data for Chart 21 immediately follows

Data for Chart 21 - Monthly Exchange Rates: Venezuelan Bolivar/Dollar and Dollar/Euro

DATES$/EuroBolivar/$
Dec-961.252990.475157
Jan-971.18950.476682
Feb-971.154630.474465
Mar-971.161730.478486
Apr-971.13490.479512
May-971.149060.483663
Jun-971.130020.485776
Jul-971.080460.492593
Aug-971.097040.495937
Sep-971.1130.496905
Oct-971.142770.498661
Nov-971.12250.500173
Dec-971.104210.50299
Jan-981.083480.507678
Feb-981.090.515425
Mar-981.076180.521682
Apr-981.10050.531259
May-981.103760.537265
Jun-981.09590.543819
Jul-981.107170.558474
Aug-981.114230.571882
Sep-981.171590.583846
Oct-981.183980.57068
Nov-981.15170.569665
Dec-981.166750.565891
Jan-991.13840.569798
Feb-991.10180.577319
Mar-991.07420.580058
Apr-991.05970.587786
May-991.04560.59648
Jun-991.03280.603286
Jul-991.06940.611175
Aug-991.05730.615953
Sep-991.0664990.625411
Oct-991.0450.630749
Nov-991.00970.634802
Dec-991.00460.644282
Jan-000.9790960.652808
Feb-000.9714020.659441
Mar-000.9553010.666825
Apr-000.9084960.672728
May-000.9303020.679996
Jun-000.9556030.680955
Jul-000.92430.685863
Aug-000.8905990.689174
Sep-000.8765010.69039
Oct-000.8417010.692863
Nov-000.8684020.69577
Dec-000.9305010.698845
Jan-010.9292990.700021
Feb-010.9248040.703358
Mar-010.8831970.706061
Apr-010.8876030.710387
May-010.8479970.714863
Jun-010.8479970.717265
Jul-010.8755030.72272
Aug-010.9158010.731974
Sep-010.91310.743463
Oct-010.90420.743224
Nov-010.8897980.745098
Dec-010.8812980.753643
Jan-020.86370.762404
Feb-020.8650970.898514
Mar-020.8724020.922657
Apr-020.9008040.871376
May-020.9387030.985796
Jun-020.9974961.212067
Jul-020.9783011.317375
Aug-020.9833041.379728
Sep-020.9859991.458389
Oct-020.9863981.440502
Nov-020.9927041.358607
Dec-021.04871.328286
Jan-031.08161.714452
Feb-031.07821.651079
Mar-031.08951.6
Apr-031.11311.6
May-031.1821991.6
Jun-031.14271.6
Jul-031.13181.6
Aug-031.09271.6
Sep-031.16521.6
Oct-031.16221.6
Nov-031.19941.6
Dec-031.2629991.599864
Jan-041.2383991.6
Feb-041.2417991.818947
Mar-041.22241.92
Apr-041.19471.92
May-041.22461.91976
Jun-041.21551.92
Jul-041.20391.92
Aug-041.2110991.92
Sep-041.24091.92
Oct-041.27371.91808
Nov-041.3295011.9152
Dec-041.3621011.9152
Jan-051.3143011.9152
Feb-051.3257011.9152
Mar-051.29642.124652
Apr-051.2956992.1446
May-051.23312.1446
Jun-051.20922.1446
Jul-051.20932.1446
Aug-051.2197992.1446
Sep-051.20422.1446
Oct-051.20232.1446
Nov-051.17692.14466
Dec-051.17972.144619
Jan-061.21182.14464
Feb-061.18752.144621
Mar-061.21042.1446
Apr-061.25372.1446
May-061.2867992.144545
Jun-061.2713012.1446
Jul-061.2767012.144565
Aug-061.2850992.1446
Sep-061.2660012.1446
Oct-061.2695992.1446
Nov-061.322.1446
Dec-061.3170012.1446
Jan-071.2954012.1446
Feb-071.32112.1446
Mar-071.3318012.1446
Apr-071.36052.1446
May-071.3452992.1446
Jun-071.3505012.1446
Jul-071.37072.1446
Aug-071.3704992.1446
Sep-071.4179012.1446
Oct-071.44472.1446
Nov-071.47612.1446
Dec-071.4720992.1446
Jan-081.4872.1446
Feb-081.51672.1446
Mar-081.58122.1446
Apr-081.5539992.1446
May-081.5508012.1446
Jun-081.5763992.1446

Chart 22 - Rolling 90-day Correlation of Percent Changes in Daily Exchange Rates: Venezuela*

Chart 22 depicts calculations of the correlation between the nominal daily percent changes in the Venezuelan bolivar-dollar exchange rate and the dollar-euro exchange rate, measured over rolling 90-day periods, during 1997-2008. These correlations generally bounce around a mean of zero.

Chart 23 - Monthly Exchange Rates: Ruble/Dollar and Dollar/Eur

Data for Chart 23 immediately follows

Data for Chart 23 - Monthly Exchange Rates: Ruble/Dollar and Dollar/Euro

DATES$/EuroRuble/$
Dec-961.252995.536
Jan-971.18955.601
Feb-971.154635.654
Mar-971.161735.704
Apr-971.13495.747
May-971.149065.77
Jun-971.130025.78
Jul-971.080465.787
Aug-971.097045.811
Sep-971.1135.846
Oct-971.142775.875
Nov-971.12255.902
Dec-971.104215.941
Jan-981.083485.9951
Feb-981.096.0497
Mar-981.076186.0896
Apr-981.10056.1238
May-981.103766.1485
Jun-981.09596.1797
Jul-981.107176.2163
Aug-981.114236.7495
Sep-981.1715914.5257
Oct-981.1839815.9227
Nov-981.151716.47
Dec-981.1667519.9904
Jan-991.138422.2876
Feb-991.101822.9021
Mar-991.074223.4788
Apr-991.059724.7419
May-991.045624.4552
Jun-991.032824.2856
Jul-991.069424.3048
Aug-991.057324.6965
Sep-991.06649925.4704
Oct-991.04525.7135
Nov-991.009726.3028
Dec-991.004626.7996
Jan-000.97909628.1891
Feb-000.97140228.7276
Mar-000.95530128.4581
Apr-000.90849628.5924
May-000.93030228.3108
Jun-000.95560328.2432
Jul-000.924327.8496
Aug-000.89059927.7356
Sep-000.87650127.7981
Oct-000.84170127.87
Nov-000.86840227.8092
Dec-000.93050127.9663
Jan-010.92929928.3592
Feb-010.92480428.5942
Mar-010.88319728.6769
Apr-010.88760328.8464
May-010.84799729.0183
Jun-010.84799729.1144
Jul-010.87550329.2173
Aug-010.91580129.3452
Sep-010.913129.4288
Oct-010.904229.5344
Nov-010.88979829.7956
Dec-010.88129830.0916
Jan-020.863730.4669
Feb-020.86509730.8007
Mar-020.87240231.059
Apr-020.90080431.1717
May-020.93870331.2476
Jun-020.99749631.4021
Jul-020.97830131.5144
Aug-020.98330431.5563
Sep-020.98599931.6251
Oct-020.98639831.6926
Nov-020.99270431.8083
Dec-021.048731.8371
Jan-031.081631.8152
Feb-031.078231.7018
Mar-031.089531.4538
Apr-031.113131.2122
May-031.18219930.9194
Jun-031.142730.4813
Jul-031.131830.3597
Aug-031.092730.3478
Sep-031.165230.5953
Oct-031.162230.1649
Nov-031.199429.8138
Dec-031.26299929.4391
Jan-041.23839928.9234
Feb-041.24179928.5154
Mar-041.222428.5334
Apr-041.194728.6759
May-041.224628.9867
Jun-041.215529.0315
Jul-041.203929.0817
Aug-041.21109929.213
Sep-041.240929.2218
Oct-041.273729.0776
Nov-041.32950128.5844
Dec-041.36210127.9201
Jan-051.31430127.935
Feb-051.32570127.9736
Mar-051.296427.616
Apr-051.29569927.8205
May-051.233127.9203
Jun-051.209228.5042
Jul-051.209328.6888
Aug-051.21979928.4756
Sep-051.204228.3645
Oct-051.202328.5472
Nov-051.176928.7565
Dec-051.179728.8111
Jan-061.211828.4122
Feb-061.187528.1968
Mar-061.210427.8777
Apr-061.253727.5724
May-061.28679927.0578
Jun-061.27130126.9839
Jul-061.27670126.915
Aug-061.28509926.7653
Sep-061.26600126.7442
Oct-061.26959926.8558
Nov-061.3226.6239
Dec-061.31700126.2865
Jan-071.29540126.4749
Feb-071.321126.3351
Mar-071.33180126.1108
Apr-071.360525.842
May-071.34529925.8183
Jun-071.35050125.9256
Jul-071.370725.5564
Aug-071.37049925.6305
Sep-071.41790125.343
Oct-071.444724.8939
Nov-071.476124.4737
Dec-071.47209924.5659
Jan-081.48724.5011
Feb-081.516724.5347
Mar-081.581223.7607
Apr-081.55399923.5127
May-081.55080123.7294
Jun-081.57639923.6376

Chart 24 - Rolling 90-day Correlation of Percent Changes in Daily Exchange Rates: Russia*

Chart 24 depicts calculations of the correlation between the nominal daily percent changes in the Russian ruble-dollar exchange rate and the dollar-euro exchange rate, measured over rolling 90-day periods, during 1997-2008. These correlations generally bounce around a mean of zero until 2004, when they drop toward -1.

Chart 25 - Monthly Exchange Rates: Argentine Peso/Dollar and Dollar/Euro

Data for Chart 25 immediately follows

Data for Chart 25 - Monthly Exchange Rates: Argentine Peso/Dollar and Dollar/Euro

DATES$/EuroAr Peso/$
Dec-961.252990.99975
Jan-971.18950.999769
Feb-971.154630.999759
Mar-971.161730.999783
Apr-971.13490.999776
May-971.149060.999738
Jun-971.130020.99973
Jul-971.080460.999767
Aug-971.097040.999688
Sep-971.1130.99975
Oct-971.142770.99982
Nov-971.12250.999799
Dec-971.104210.99971
Jan-981.083480.999752
Feb-981.090.999745
Mar-981.076180.999769
Apr-981.10050.999715
May-981.103760.999748
Jun-981.09590.99975
Jul-981.107170.999762
Aug-981.114230.999739
Sep-981.171590.99975
Oct-981.183980.99976
Nov-981.15170.999735
Dec-981.166750.999623
Jan-991.13840.999726
Feb-991.10180.999787
Mar-991.07420.999777
Apr-991.05970.999758
May-991.04560.999777
Jun-991.03280.999758
Jul-991.06940.999765
Aug-991.05730.999763
Sep-991.0664990.999745
Oct-991.0450.999746
Nov-991.00970.999766
Dec-991.00460.999718
Jan-000.9790960.999756
Feb-000.9714020.999438
Mar-000.9553010.999758
Apr-000.9084960.999641
May-000.9303020.999718
Jun-000.9556030.999709
Jul-000.92430.999729
Aug-000.8905990.999656
Sep-000.8765010.999675
Oct-000.8417010.999663
Nov-000.8684020.999589
Dec-000.9305010.999322
Jan-010.9292990.999718
Feb-010.9248040.999738
Mar-010.8831970.999758
Apr-010.8876030.999439
May-010.8479970.99965
Jun-010.8479970.999709
Jul-010.8755030.999341
Aug-010.9158010.999315
Sep-010.91310.999548
Oct-010.90420.99946
Nov-010.8897980.998273
Dec-010.8812980.99937
Jan-020.86371.555141
Feb-020.8650971.99575
Mar-020.8724022.433333
Apr-020.9008042.916364
May-020.9387033.307826
Jun-020.9974963.608
Jul-020.9783013.598623
Aug-020.9833043.615909
Sep-020.9859993.6355
Oct-020.9863983.647826
Nov-020.9927043.522105
Dec-021.04873.478571
Jan-031.08163.252857
Feb-031.07823.159474
Mar-031.08953.061667
Apr-031.11312.893636
May-031.1821992.831905
Jun-031.14272.806071
Jul-031.13182.797841
Aug-031.09272.925
Sep-031.16522.915643
Oct-031.16222.855682
Nov-031.19942.87975
Dec-031.2629992.957386
Jan-041.2383992.891071
Feb-041.2417992.92914
Mar-041.22242.894452
Apr-041.19472.831955
May-041.22462.92119
Jun-041.21552.957986
Jul-041.20392.949555
Aug-041.2110993.012023
Sep-041.24092.994045
Oct-041.27372.966643
Nov-041.3295012.953241
Dec-041.3621012.969809
Jan-051.3143012.943343
Feb-051.3257012.914955
Mar-051.29642.92347
Apr-051.2956992.898624
May-051.23312.8895
Jun-051.20922.880909
Jul-051.20932.867362
Aug-051.2197992.886487
Sep-051.20422.910045
Oct-051.20232.963895
Nov-051.17692.962564
Dec-051.17973.012932
Jan-061.21183.044623
Feb-061.18753.067785
Mar-061.21043.074739
Apr-061.25373.066245
May-061.2867993.054261
Jun-061.2713013.080127
Jul-061.2767013.080857
Aug-061.2850993.077674
Sep-061.2660013.099267
Oct-061.2695993.097705
Nov-061.323.075014
Dec-061.3170013.059365
Jan-071.2954013.083391
Feb-071.32113.101415
Mar-071.3318013.100159
Apr-071.36053.089919
May-071.3452993.07973
Jun-071.3505013.078062
Jul-071.37073.110786
Aug-071.3704993.150935
Sep-071.4179013.14665
Oct-071.44473.159109
Nov-071.47613.134919
Dec-071.4720993.139325
Jan-081.4873.144432
Feb-081.51673.15849
Mar-081.58123.155433
Apr-081.5539993.16525
May-081.5508013.150414
Jun-081.5763993.043086

Chart 26 - Rolling 90-day Correlation of Percent Changes in Daily Exchange Rates: Argentina*

Chart 26 depicts calculations of the correlation between the nominal daily percent changes in the Argentine peso-dollar exchange rate and the dollar-euro exchange rate, measured over rolling 90-day periods, during 1997-2008. These correlations generally bounce around a mean of zero.

Chart 27 - Interest Rate Differential vs. Exchange Rate

Data for Chart 27 immediately follows

Data for Chart 27 - Interest Rate Differential vs. Exchange Rate

Country CodeInterest RateExchange Rate
AR-0.509130.580504
BZ-0.04422-0.08121
CA-0.74109-0.89927
CL0.112995-0.3467
CO-0.04422-0.08121
CZ0.003825-0.94581
DN-0.96688-0.98441
HU-0.01755-0.90842
ID0.205432-0.01112
IN-0.56205-0.42621
IS-0.11819-0.20981
JA-0.87873-0.34115
KO-0.10353-0.5933
MA-0.44036-0.35073
MX0.3669420.56283
NO-0.68449-0.96362
PE0.297753-0.47448
PH0.0043610.021059
PK0.0250040.213051
PL0.11689-0.9051
RU0.14238-0.09587
SD-0.89657-0.94654
SF-0.32438-0.33567
SI-0.47357-0.77694
TA-0.33476-0.23975
TH0.014962-0.5615
TK0.1038810.229396
UK-0.76094-0.9565
VE-0.503850.731434

Chart 28 - Interest Rate Differential vs. Exchange Rate in 12 Month Changes

Data for Chart 28 immediately follows

Data for Chart 28 - Interest Rate Differential vs. Exchange Rate in 12 month Changes

Country CodeInterest RateExchange Rate
AR-0.335590.203766
BZ-0.18247-0.12031
CA-0.56437-0.51851
CL-0.2766-0.31606
CO-0.18247-0.12031
CZ-0.63765-0.87562
DN-0.94621-0.98933
HU-0.60374-0.88144
ID-0.16568-0.17476
IN-0.83915-0.46905
IS-0.40516-0.40713
JA-0.84549-0.38656
KO-0.3609-0.24101
MA-0.66752-0.25779
MX-0.038960.334957
NO-0.64904-0.87743
PE-0.04163-0.47952
PH-0.28248-0.36709
PK-0.21003-0.62581
PL-0.42723-0.68348
RU-0.17366-0.15709
SD-0.82408-0.92514
SF-0.56229-0.4964
SI-0.44464-0.56519
TA-0.87727-0.27085
TH-0.26668-0.44748
TK-0.02501-0.60195
UK-0.77889-0.69977
VE-0.490070.300904

Chart 29 - Real Interest Rate Differential vs. Exchange Rate

Data for Chart 29 immediately follows

Data for Chart 29 - Real Interest Rate Differential vs. Exchange Rate

Country CodeInterest RateExchange Rate
AR0.305024-0.05959
BZ-0.469560.258488
CA-0.87268-0.63046
CL-0.443360.168816
CO-0.37021-0.27613
CZ-0.88755-0.13153
DN-0.98508-0.89084
HU-0.87951-0.54858
ID-0.461960.024101
IN-0.88991-0.51375
IS-0.02116-0.2991
JA0.343445-0.78557
KO-0.619030.03686
MA-0.25278-0.66572
MX0.295998-0.10322
NO-0.95509-0.67966
PE-0.462220.249202
PH-0.37928-0.03906
PK-0.598860.123793
PL-0.82326-0.28214
RU-0.76492-0.13736
SD-0.92021-0.78361
SF-0.67082-0.10009
SI-0.332830.052212
TA0.229455-0.19263
TH-0.589140.026461
TK-0.82282-0.19845
UK-0.92176-0.79449
VE0.195662-0.18557

Chart 30 - Real Interest Rate Differential vs. Exchange Rate in 12 Month Changes

Data for Chart 30 immediately follows

Data for Chart 30 - Real Interest Rate Differential vs. Exchange Rate in 12-Month Changes

Country CodeInterest RateExchange Rate
AR0.135546-0.11491
BZ-0.16101-0.04211
CA-0.52947-0.46622
CL-0.28069-0.02741
CO-0.03584-0.14819
CZ-0.86255-0.39835
DN-0.98895-0.8011
HU-0.86871-0.4929
ID-0.22021-0.02827
IN-0.55029-0.33433
IS-0.40612-0.1
JA-0.37616-0.69764
KO-0.25455-0.19256
MA-0.25492-0.67697
MX0.720256-0.20829
NO-0.87465-0.51549
PE-0.43890.019967
PH-0.35602-0.10217
PK-0.615610.158586
PL-0.56275-0.47276
RU-0.18619-0.01345
SD-0.93252-0.67559
SF-0.52638-0.24387
SI-0.557850.197833
TA-0.26181-0.47457
TH-0.46906-0.11509
TK-0.43261-0.16821
UK-0.69421-0.81065
VE0.3105330.080602

Chart 31 - Correlations of 12-Month Changes in Nominal Interest Rates: Actual vs. Fitted*

Data for Chart 31 immediately follows

Data for Chart 31 - Correlations of 12-Month Changes in Nominal Interest Rates: Actual vs. Fitted

Country CodeActualFitted
AR-0.33559-0.23226
BZ-0.18247-0.30435
CA-0.56437-0.41895
CL-0.2766-0.42336
CO-0.18247-0.57237
CZ-0.63765-0.65206
DN-0.94621-0.67803
HU-0.60374-0.64728
ID-0.16568-0.3766
IN-0.83915-0.48914
IS-0.40516-0.35356
JA-0.84549-0.55802
KO-0.3609-0.44355
MA-0.66752-0.5333
MX-0.03896-0.32392
NO-0.64904-0.33388
PE-0.04163-0.35468
PH-0.28248-0.44707
PK-0.21003-0.36054
PL-0.42723-0.50708
RU-0.17366-0.2882
SD-0.82408-0.64091
SF-0.56229-0.30439
SI-0.44464-0.35113
TA-0.87727-0.41808
TH-0.26668-0.38755
TK-0.02501-0.58465
UK-0.77889-0.7953
VE-0.49007-0.32479

Chart 32 - Correlations of 12-Month Changes in Real Interest Rates: Actual vs. Fitted*

Data for Chart 32 immediately follows

Data for Chart 32 - Correlations of 12-Month Changes in Real Interest Rates: Actual vs. Fitted

Country CodeActualFitted
AR-0.114910.052055
BZ-0.04211-0.03645
CA-0.46622-0.15919
CL-0.02741-0.18235
CO-0.14819-0.42559
CZ-0.39835-0.51666
DN-0.8011-0.59876
HU-0.4929-0.52285
ID-0.02827-0.17131
IN-0.33433-0.32116
IS-0.1-0.11245
JA-0.69764-0.46632
KO-0.19256-0.32116
MA-0.67697-0.44586
MX-0.20829-0.13835
NO-0.51549-0.09567
PE0.019967-0.12986
PH-0.10217-0.25678
PK0.158586-0.11646
PL-0.47276-0.32802
RU-0.01345-0.03197
SD-0.67559-0.5773
SF-0.24387-0.01059
SI0.197833-0.15316
TA-0.47457-0.27636
TH-0.11509-0.2334
TK-0.16821-0.49332
UK-0.81065-0.7575
VE0.080602-0.03726

Chart 33 - Correlations of 12-Month Changes in Nominal Exchange Rates: Actual vs. Fitted*

Data for Chart 33 immediately follows

Data for Chart 33 - Correlations of 12-Month Changes in Nominal Exchange Rates: Actual vs. Fitted

Country CodeActualFitted
AR0.203766-0.12498
BZ-0.12031-0.19377
CA-0.51851-0.50797
CL-0.31606-0.40604
CO-0.12031-0.51896
CZ-0.87562-0.78247
DN-0.98933-0.83345
HU-0.88144-0.74691
ID-0.17476-0.23076
IN-0.46905-0.57903
IS-0.40713-0.30964
JA-0.38656-0.59521
KO-0.24101-0.2867
MA-0.25779-0.49111
MX0.334957-0.06423
NO-0.87743-0.33688
PE-0.47952-0.18372
PH-0.36709-0.37519
PK-0.62581-0.27217
PL-0.68348-0.5134
RU-0.15709-0.14263
SD-0.92514-0.70407
SF-0.4964-0.34123
SI-0.56519-0.25011
TA-0.27085-0.41421
TH-0.44748-0.20191
TK-0.60195-0.41273
UK-0.69977-0.94593
VE0.300904-0.35004

Chart 34 - Correlations of 12-Month Changes in Real Exchange Rates: Actual vs. Fitted*

Data for Chart 34 immediately follows

Data for Chart 34 - Correlations of 12-Month Changes in Real Exchange Rates: Actual vs. Fitted

Country CodeActualFitted
AR0.135546-0.12977
BZ-0.16101-0.22551
CA-0.52947-0.56329
CL-0.28069-0.41216
CO-0.03584-0.54219
CZ-0.86255-0.75188
DN-0.98895-0.78389
HU-0.86871-0.74136
ID-0.22021-0.2299
IN-0.55029-0.46404
IS-0.40612-0.28329
JA-0.37616-0.52997
KO-0.25455-0.22333
MA-0.25492-0.46198
MX0.720256-0.11999
NO-0.87465-0.34543
PE-0.4389-0.21583
PH-0.35602-0.3646
PK-0.61561-0.23615
PL-0.56275-0.56486
RU-0.18619-0.15641
SD-0.93252-0.63508
SF-0.52638-0.33956
SI-0.55785-0.09847
TA-0.26181-0.28925
TH-0.46906-0.15742
TK-0.43261-0.44469
UK-0.69421-0.93942
VE0.310533-0.28194

Chart 35 - Correlations of 12-Month Changes in Nominal Exchange Rates: Actual vs. Fitted*

Data for Chart 35 immediately follows

Data for Chart 35 - Correlations of 12-Month Changes in Nominal Exchange Rates: Actual vs. Fitted

Country CodeActualFitted
AR0.2037660.172038
BZ-0.12031-0.05696
CA-0.51851-0.27087
CL-0.31606-0.276
CO-0.12031-0.1554
CZ-0.87562-0.95517
DN-0.98933-0.95282
HU-0.88144-0.8222
ID-0.17476-0.23247
IN-0.46905-0.58084
IS-0.40713-0.49122
JA-0.38656-0.37673
KO-0.24101-0.28048
MA-0.25779-0.40803
MX0.3349570.098775
NO-0.87743-0.76981
PE-0.47952-0.09136
PH-0.36709-0.3839
PK-0.62581-0.38721
PL-0.68348-0.82619
RU-0.15709-0.29284
SD-0.92514-0.76641
SF-0.4964-0.4022
SI-0.56519-0.50956
TA-0.27085-0.44665
TH-0.44748-0.45842
TK-0.60195-0.37087
UK-0.69977-0.89803
VE0.3009040.076402

Chart 36 - Correlations of 12-Month Changes in Real Exchange Rates: Actual vs. Fitted*

Data for Chart 36 immediately follows

Data for Chart 36 - Correlations of 12-Month Changes in Real Exchange Rates: Actual vs. Fitted

Country CodeActualFitted
AR0.1355460.050651
BZ-0.16101-0.13038
CA-0.52947-0.44026
CL-0.28069-0.57037
CO-0.03584-0.52657
CZ-0.86255-0.90506
DN-0.98895-1.05391
HU-0.86871-0.83848
ID-0.22021-0.15287
IN-0.55029-0.47459
IS-0.40612-0.40926
JA-0.37616-0.58138
KO-0.25455-0.37428
MA-0.25492-0.33382
MX0.7202560.038637
NO-0.87465-0.67602
PE-0.4389-0.24701
PH-0.35602-0.38956
PK-0.61561-0.02391
PL-0.56275-0.66273
RU-0.18619-0.11771
SD-0.93252-0.67452
SF-0.52638-0.38095
SI-0.55785-0.6084
TA-0.261810
TH-0.46906-0.25539
TK-0.43261-0.41988
UK-0.69421-0.62707
VE0.310533-0.05953

Chart 37 - Contribution of Explanatory Variables to Predict Nominal Exchange Rate CorrelationsData for Chart 37 immediately follows

Data for Chart 37 - Contribution of Explanatory Variables to Predict Nominal Exchange Rate Correlations (Based on Equation in Table 3, Column 10)

Country CodeGDP Per CapitaCredit RatingConstant + Contributions of Distance VariableFitted ValueActual Value
CZ0.9532410.289336-2.19775-0.95517-0.87562
DN1.1401290.056872-2.14982-0.95282-0.98933
UK1.1180120.042654-2.0587-0.89803-0.69977
PL0.9289120.347097-2.10219-0.82619-0.68348
HU0.9344420.317417-2.07406-0.8222-0.88144
NO1.1644580.044076-1.97835-0.76981-0.87743
SD1.1312830.077488-1.97518-0.76641-0.92514
IN0.6777080.481812-1.74036-0.58084-0.46905
SI1.1113770.054206-1.67515-0.50957-0.56519
IS1.0903660.284715-1.8663-0.49122-0.40713
TH0.8382330.393839-1.69049-0.45842-0.44748
TA1.0594020.1484-1.65445-0.44665-0.27085
MA0.9167480.33359-1.65837-0.40803-0.25779
SF0.8846780.40699-1.69386-0.4022-0.4964
PK0.6954750.684597-1.76728-0.38721-0.62581
PH0.763260.496919-1.64408-0.3839-0.36709
JA1.1622460.084775-1.62375-0.37673-0.38656
TK0.9233830.606753-1.90101-0.37087-0.60195
RU0.8271740.595023-1.71504-0.29284-0.15709
KO1.0284390.337144-1.64606-0.28048-0.24101
CL0.9410770.311374-1.52845-0.276-0.31606
CA1.1135890.073045-1.45751-0.27087-0.51851
ID0.7401660.666647-1.63929-0.23247-0.17476
CO0.8393390.488033-1.48277-0.1554-0.12031
PE0.847080.555568-1.49401-0.09136-0.47952
BZ0.9101130.588625-1.5557-0.05696-0.12031
VE 0.9388650.638922-1.501380.0764020.300904
MX0.9609820.437381-1.299590.0987750.334957
AR0.9897340.726718-1.544410.1720380.203766

Chart 38 - Contribution of Explanatory Variables to Predict Real Exchange Rate Correlations (Based on Equation in Table 4, Column 10)

Data for Chart 38 immediately follows

Data for Chart 38 - Contribution of Explanatory Variables to Predict Real Exchange Rate Correlations

Country CodeIP GrowthCredit RatingConstant + Contributions of Distance VariableFitted Value>Actual Value
CZ-0.24870.181127-0.87715-0.94472-0.86255
DN-0.142740.035602-0.8206-0.92774-0.98895
UK-0.204210.026702-0.73686-0.91436-0.69421
HU-0.213920.198706-0.80001-0.81523-0.86871
PL-0.153310.217286-0.80968-0.74571-0.56275
NO-0.057190.027592-0.68385-0.71344-0.87465
SD-0.049950.048508-0.68856-0.69-0.93252
IS-0.092580.178235-0.70123-0.61558-0.40612
SI0.0207440.033934-0.64259-0.58791-0.55785
SF-0.156720.25478-0.64656-0.54851-0.52638
IN-0.099520.301619-0.66323-0.46113-0.55029
JA-0.003380.05307-0.50729-0.4576-0.37616
TA0.0685870.0929-0.57436-0.41287-0.26181
CL-0.191780.194923-0.38512-0.38197-0.28069
TK-0.040380.379833-0.71143-0.37198-0.43261
MA0.0426940.20883-0.62146-0.36994-0.25492
PH-0.096410.311076-0.58443-0.36976-0.35602
PK-0.11010.428564-0.66067-0.34221-0.61561
TH0.0804510.246547-0.63265-0.30566-0.46906
KO0.0815340.211055-0.53451-0.24192-0.25455
CA-0.233240.045727-0.05129-0.2388-0.52947
ID-0.034190.417327-0.61069-0.22756-0.22021
RU-0.037990.37249-0.54796-0.21346-0.18619
BZ-0.090710.368485-0.36856-0.09079-0.16101
CO-0.164130.305513-0.21697-0.07559-0.03584
VE -0.169840.399971-0.2309-0.000770.310533
AR-0.029810.454932-0.415820.0092960.135546
PE-0.052690.347791-0.281360.013741-0.4389
MX0.0910860.2738050.145580.510470.720256

Table 1. Cross-Country Regressions for Correlations of Interest Rate Differentials - Dependent Variable: Correlation (interest rate differential between country X and US, interest rate differential between US and Euro Area)

 

Nominal Interest Rate Changes1 – (1)

Nominal Interest Rate Changes1 – (2)

Real Interest Rate Changes2 – (3)

Real Interest Rate Changes2 – (4)

corr(Δ πgap differentials)3

0.50, (2.7)

0.51, (2.8)

0.79, (5.3)

0.80, (5.2)

corr(Δ IPgap differentials)4

0.19, (1.0)

 

0.23, (1.5)

 

corr(Δ IP growth differentials)5

 

0.25, (1.2)

 

0.22, (1.3)

Adjusted R2

.25

.25

.50

.49

t-statistics in parentheses, n=29
1.  corr(Δi$ -Δix, Δieu -Δi$)
2.  corr(Δr$ -Δrx, Δreu -Δr$)
3.  corr[Δ(πgap$ - πgapx), Δ(πgapeu - πgap$)]
4.  corr[Δ(ipgap$ - ipgapx), Δ(ipgapeu - ipgap$)]
5.  corr[Δ(Δip$ - Δipx), Δ(Δipeu - Δip$)]

Table 2. Cross-Country Regressions for Correlations of Exchange Rates � Dependent Variable: Correlation (exchange rate of country x again dollar, exchange rate of dollar against euro)

 

Nominal Exchange Rate Changes1 - (1)

Nominal Exchange Rate Changes1 - (2)

Nominal Exchange Rate Changes1 - (3)

Real Exchange Rate Changes2 - (4)

Real Exchange Rate Changes2 - (5)

Real Exchange Rate Changes2 - (6)

corr(Δ πgap differentials)3

 

0.49, (2.0)

0.54, (2.3)

 

0.33, (0.9)

0.42, (1.3)

corr(Δ IPgap differentials)4

 

0.26, (1.2)

 

 

0.21, (0.8)

 

corr(Δ IP growth differentials)5

 

 

0.51, (2.2)

 

 

0.59, (2.2)

corr(Δ interest differentials)6

0.59, (2.8)

0.34, (1.5)

0.28, (1.3)

0.60, (2.7)

0.36, (1.1)

0.27, (0.9)

Adjusted R2

.31

.27

.36

.18

.16

.28

t-statistics in parentheses, n=29
1.  corr(Δex/$, Δe$/eu)
2.  corr(Δex/$, Δe$/eu)
3.  corr[Δ(πgap$ - πgapx), Δ(πgapeu - πgap$)]
4.  corr[Δ(ipgap$ - ipgapx), Δ(ipgapeu - ipgap$)]
5.  corr[Δ(Δip$ - Δipx), Δ(Δipeu - Δip$)]
6.  corr(Δi$ - Δix, Δieu- Δi$), Δ(Δipeu - Δip$) or corr(Δr$ - Δrx, Δreu- Δr$)

Table 3. Cross-Country Regressions for Correlations of Nominal Exchange Rates with Additional Regressors � Dependent Variable: Correlation (12-month change in nominal exchange rate of country x against dollar, 12-month change in nominal exchange rate of dollar against euro)

 

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

corr(Δ πgap differentials)

0.54, (2.3)

0.15, (0.8)

0.49, (2.1)

0.49, (2.1)

0.63, (2.5)

0.45, (1.8)

0.63, (2.5)

0.35, (1.6)

-0.02, (-0.1)

 

0.50, (2.4)

corr(Δ IPgap differentials)

0.51, (2.2)

0.20, (0.9)

0.42, (1.6)

0.51, (2.2)

0.58, (2.4)

0.46, (1.8)

0.49, (1.9)

0.49, (2.3)

0.30, (1.5)

 

0.48, (2.4)

corr(Δ interest differential)

0.28, (1.3)

0.08, (0.5)

0.36, (1.5)

0.23, (1.1)

0.28, (1.3)

0.24, (0.9)

0.39, (1.6)

-0.07, (-0.3)

-0.45, (-2.1)

 

 

Distance from the United States

 

-0.22, (-2.7)

 

 

 

 

 

 

-0.26, (-2.7)

-0.18, (-2.8)

 

Distance from the Euro Area

 

0.20, (3.7)

 

 

 

 

 

 

0.21, (3.8)

0.21, (6.4)

 

Trade Share

 

 

0.00, (1.2)

 

 

 

 

 

-0.00, (-0.2)

 

0.01, (2.2)

Correlation of Stock Market Returns

 

 

 

-0.44, (-1.5)

 

 

 

 

0.12, (0.5)

 

 

U.S. Portfolio Integration

 

 

 

 

0.00, (1.0)

 

 

 

0.00, (1.5)

 

 

International Financial Integration

 

 

 

 

 

-0.03, (-0.9)

 

 

-0.04, (-1.3)

 

 

International Financial Size

 

 

 

 

 

 

0.00, (1.2)

 

-0.00, (-0.8)

 

0.00, (2.3)

Credit Rating

 

 

 

 

 

 

 

0.05, (2.6)

0.06, (3.8)

0.04, (4.0)

0.04, (3.7)

GDP per Capita

 

 

 

 

 

 

 

0.10, (1.6)

0.11, (2.1)

0.11, (2.6)

 

Adjusted R2

.36

.63

.37

.39

.36

.37

.36

.46

.78

.77

.57

T-statistics in parentheses, n=28 in regressions (3), (6), (7), (10), and (11), n=29 in all others

Table 4. Cross-Country Regressions for Correlations of Real Exchange Rates with Additional Regressors � Dependent Variable: Correlation (12-month change in nominal exchange rate of country x against dollar, 12-month change in real exchange rate of dollar against euro)

 

(1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10)

(11)

corr(Δ πgap differentials)

0.42, (1.3)

-0.01, (-0.1)

0.32, (0.9)

0.46, (1.4)

0.50, (1.4)

0.31, (0.9)

0.49, (1.4)

0.28, (0.9)

-0.11, (-0.4)

 

 

corr(Δ IPgap differentials)

0.59, (2.2)

0.36, (1.5)

0.51, (1.8)

0.60, (2.2)

0.65, (2.3)

0.56, (2.0)

0.61, (2.1)

0.53, (2.2)

0.32, (1.3)

0.43, (2.3)

 

corr(Δ interest differential)

0.27, (0.9)

0.24, (1.0)

0.37, (1.2)

0.15, (0.5)

0.28, (0.9)

0.17, (0.5)

0.34, (1.0)

-0.02, (-0.1)

-0.03, (-0.1)

 

 

Distance from the United States

 

-0.36, (-4.1)

 

 

 

 

 

 

-0.31, (-2.8)

-0.37, (-5.2)

 

Distance from the Euro Area

 

0.18, (3.0)

 

 

 

 

 

 

0.17, (2.5)

0.15, (3.5)

 

Trade Share

 

 

0.01, (1.5)

 

 

 

 

 

0.00, (0.1)

 

0.01, (2.3)

Correlation of Stock Market Returns

 

 

 

-0.35, (-1.0)

 

 

 

 

0.09, (0.3)

 

 

U.S. Portfolio Integration

 

 

 

 

0.00, (0.8)

 

 

 

0.00, (0.8)

 

 

International Financial Integration

 

 

 

 

 

-0.05, (-1.4)

 

 

-0.04, (-1.1)

 

 

International Financial Size

 

 

 

 

 

 

0.00, (0.9)

 

-0.00, (-0.3)

 

 

Credit Rating

 

 

 

 

 

 

 

0.06, (2.7)

0.05, (2.7)

0.03, (3.4)

0.07, (4.3)

GDP per Capita

 

 

 

 

 

 

 

0.12, (1.6)

0.10, (1.5)

 

0.15, (2.1)

Adjusted R2

.28

.63

.31

.27

.27

.30

.27

.40

.72

.75

.47

T-statistics in parentheses, n=28 in regressions (3), (6), (7), (10), and (11), n=29 in all others

Footnotes

**  The authors are Senior Economist, Deputy Director, and Research Assistant, respectively, in the Division of International Finance, Board of Governors of the Federal Reserve System, Washington, D.C. 20551 U.S.A. The views in this paper are solely the responsibility of the authors and should not be interpreted as reflecting the views of the Board of Governors of the Federal Reserve System or of any other person associated with the Federal Reserve System. We thank Marcel Fratzscher and participants in the International Finance Division Workshop for helpful comments. Geoffrey Keim and Matthew Nespoli provided excellent research assistance. Return to text

1.  See Banco de Mexico (2003). Return to text

2.  Throughout this paper, we focus on the period from 1997 onwards in order to avoid complications associated with the aftermath of the Tequila crisis in 1995-96. Return to text

3.  This assumes that expected future values of the peso/dollar and dollar/euro exchange rates remain constant, and that the error terms in the UIP equations for peso/dollar and dollar/euro are uncorrelated. Return to text

4.  Except for placecountry-regionChile, placecountry-regionHungary, and placecountry-regionIsrael, for which we use the discount rate. Return to text

5.  We should note that the uncovered interest parity relationship described in equation (1) is an equilibrium condition, and it does not indicate whether causality runs from interest rate differentials to exchange rates or vice-versa. Accordingly, it is possible that the relationships shown in Charts 27-30 actually depict the impact of exchange rate correlations on interest rate correlations. We do not place a lot of weight on this "reverse causality" scenario, however, and it leaves unresolved what led to the pattern of exchange rate correlations in the first place. Return to text

6.  We assume for simplicity that the coefficients on the inflation and output gap terms are the same across countries. This is almost certainly a substantial simplification. Return to text

7.  For the placecountry-regionUnited States, core CPI inflation, which excludes prices of food and energy, is used instead of total CPI inflation. Return to text

8.  To provide an example, consider the 2002.07 observation for Mexico. To compute the first argument in the correlation, Mexican IP growth between 2001.07 and 2002.07 is subtracted from U.S. IP growth over the same period. From this calculation is then subtracted the difference between U.S. and Mexico IP growth during the preceding year, of 2000.07 to 2001.07. The analogous computations are then made for the second argument of the correlation, the change in IP growth differential between U.S. and euro area. Return to text

9.  This is not implausible, as exchange rates ultimately should be influenced by expected rates of return, and output growth may be a more robust indicator of such returns than short-term interest rates alone. Additionally, the uncovered interest parity relationship may be more applicable to long rates than the short rates used in our research (mainly reflecting data availability). To the extent that output and inflation affect long rates as well as short rates, they may influence exchange rates even if short rates are held constant. Return to text

10.  The correlation of the 12-month rate of IP growth in Canada and the United States is .70, compared with .88 for Mexico and the United States. Return to text

This version is optimized for use by screen readers. Descriptions for all mathematical expressions are provided in LaTex format. A printable pdf version is available. Return to text

Home | Economic research and data | Publications and education resources
Accessibility | Contact us
Last update: September 11, 2009