We use cookies to give you the best experience possible. By continuing we’ll assume you’re on board with our cookie policy

See Pricing

What's Your Topic?

Hire a Professional Writer Now

The input space is limited by 250 symbols

What's Your Deadline?

Choose 3 Hours or More.
2/4 steps

How Many Pages?

3/4 steps

Sign Up and See Pricing

"You must agree to out terms of services and privacy policy"
Get Offer

The Evidence From The ARDL Bounds Tests

Hire a Professional Writer Now

The input space is limited by 250 symbols

Deadline:2 days left
"You must agree to out terms of services and privacy policy"
Write my paper

This survey investigates the relationship between nominal and existent effectual exchange rates. Both short tally and long tally relationships between the two are examined by using Bounds-testing ( ARDL ) attack to cointegration. Consequences of the survey reveal that the nominal and existent exchange rates bear a long tally relationship ( cointegrated ), while the mistake rectification consequences confirm short-term divergences in the relationship.

Don't use plagiarized sources. Get Your Custom Essay on
The Evidence From The ARDL Bounds Tests
Just from $13,9/Page
Get custom paper

The nominal effectual exchange rate ( NEER ) captures the overall fluctuation in the value of a state ‘s currency against the currencies of all trading spouses.

It measures the alterations in overall nominal value of a currency. In contrast, the existent effectual exchange rate ( REER ) measures alterations in existent value of a currency by integrating the monetary value degrees in its building. Thus it is the REER which reflects alterations in external fight of a state. However, it is the nominal exchange rate which is used as a policy tool to better a state ‘s trade balance as the cardinal bank manipulates it.

There are two channels through which nominal depreciation could take to an betterment in the trade balance. One is through doing a state ‘s exports cheaper in footings of foreign currencies, taking to an addition in exports.

The other channel is through doing a state ‘s imports expensive in footings of her domestic currency taking to a diminution in imports. However, it besides has its negative effects as expensive imports normally do lend to domestic ini¬‚ation which over clip spreads into the export sector and this may pass over out favourable effects that nominal depreciation could hold on the exports. Therefore, to look into the impact of exchange rate accommodation, one has to take into history of monetary value alteration as good, both domestic and foreign. Therefore, nominal depreciation could take to betterment in trade balance merely if it leads to existent depreciation. The present survey investigates, whether nominal exchange rate depreciation finally leads to depreciation in existent exchange rate in the short tally every bit good in the long tally or non. To accomplish this nonsubjective, we have used cointegration analysis through the ARDL Bound trial which has an border over the conventional cointegration trials such as Engel-Granger, Johansen, Johansen and Juselius etc

Literature Review

A few surveies in the literature have concentrated on the relation between alterations in the nominal exchange rate and its impact on the existent exchange rate. The early survey was by Vaubel ( 1976 ) who gave new way by demoing that nominal devaluations were effectual to accomplish existent effectual accommodation during 1959-1975. He besides noted that the relationship between nominal and existent effectual exchange rates is clip changing. Later Connolly and Taylor ( 1976, 1979 ), Bruno ( 1978 ) and Edwards ( 1988, 1994 ) conclude that nominal devaluation leads to existent devaluation merely in the short span of clip to medium term.

De Grauwe and Holvoet ( 1978 ) roll up input-output tabular arraies for European Community and conclude that the result is clearly sensitive to the premise of pay indexing while under nothing pay indexation, a 0.70 per cent existent devaluation was led by 1 per cent addition in nominal devaluation, with complete pay indexation, 1 per cent addition in nominal devaluation leads to 0.5 per cent existent alterations in exchange rate. On the contrary, Donovan ( 1981 ), Bautista ( 1981 ) and Morgan and Davis ( 1982 ) claim that the lead impact of nominal devaluation on existent devaluation begins to gnaw in long span of clip. Edwards ( 1988 ) shows that the impact of nominal devaluation on the existent exchange rate erodes over the undermentioned 16 quarters.

Again in another survey in1994, he concludes that nominal devaluations are translated into existent devaluations in the short to medium tally. Bahmani-Oskooee ( 2001 ) assesses long tally response of trade balance to nominal degenerations and existent depreciation in instance of Middle Eastern states and shows that existent depreciation has favorable long-term consequence on the trade balance of most nonoil exporting In-between Eastern states. Bahmani-Oskooee and Miteza ( 2002 ) use error-correction mold to research the relation between nominal effectual exchange rate and existent effectual exchange rate non merely in short tally but besides for long tally in less developed economic systems including India. They argue that nominal devaluation leads to existent devaluation with important values of variables in the instance for India over 1971-1997 periods.

Bahmani-Oskooee and Gelan, ( 2007 ), have come to the decision that nominal devaluation is associated with existent devaluation in medium to long tally. But in short tally, nominal effectual exchange rate alterations do non take the existent effectual exchange rate alterations except in a few African states. Bahmani-Oskooee and Kandi ( 2007 ) besides investigate the relationship between nominal and existent devaluation in MENA states and conclude that nominal depreciation leads to existent depreciation in all states in the short-run, the short-run-effects last into the long-run, non in all, but in most states. Bahmani-Oskooee and Harvey ( 2007 ), by building quarterly informations of concerned variables over the 1971-2004 periods for less developed states, show important impact of nominal depreciation on existent depreciation for states in the sample. Recently Shahbaz, Muhammad ( 2009 ) utilizing the ARDL technique of cointegration, show that nominal devaluation non merely leads to existent devaluation in longer periods but besides in short span of clip in instance of Pakistan.

Data and Methodological Framework:

Cointegration and error-correction attack is used in this survey to analyze the short-term and long tally relationship between nominal and existent effectual exchange rates. There are many techniques available in economic literature to look into cointegration relationship among macroeconomic variables. For bivariate analysis, Engle-Granger ( 1987 ), and FMOLS process of Phillips and Hansen ( 1990 ) have been outstanding. For multivariate co-integration, the techniques of Johansen ( 1988 ) ; Johansen and Juselius ( 1990 ) ; and Johansen ‘s ( 1995 ) have been popular. In the present survey, car regressive distributed slowdown ( ARDL ) attack to cointegration, developed by Pesaran et Al. ( 2001 ) has been used. This attack, besides known as the ARDL bounds test attack, is preferred over other conventional cointegration trials, as it has several advantages over other conventional trials [ See Emran et Al. ( 2007 ) ]. The ARDL attack ensures estimates that satisfy the little sample belongingss. Further, this attack efficaciously corrects for possible endogeneity of explanatory variables. Yet another advantage of this attack is that it is applicable irrespective of whether the implicit in regressors are strictly I ( 0 ), or strictly I ( 1 ) or a combination of the two. In add-on, both short tally and long tally calculators can be at the same time estimated.

An unrestricted error-correction specification due to Pesaran et Al. ( 2001 ) is adopted here to suit the dynamic accommodation between existent REER and NEER. Formally, this may be written as,

… ( 1 )

Where lnREER and lnNEER are log of trade weighted ( 36-currency bilateral weights ) monthly norm of indices of existent and nominal effectual exchange rates severally, of the Indian rupee with 1993-94 as basal twelvemonth and ( one = 1,2 ) are the long tally multipliers, is the changeless and are the white noise mistake footings.

Through the ARDL theoretical account outlined above we may separate the short tally effects of nominal depreciation from its long tally effects. ARDL is a general dynamic specification theoretical account which uses the slowdown of the dependent variables and the lagged and contemporary values of the independent variables, through which the short tally consequence can be straight estimated, and long tally equilibrium relationship can be indirectly estimated. In the first measure in the ARDL bounds proving we estimate equation ( 1 ) by ordinary Least Square ( OLS ) in order to prove for the being of a long tally relationship among the variables by carry oning an F-test for the joint significance of the coefficients of lagged degrees of the variables, i.e., for equation ( 1 ) the void hypothesis of no cointegration is defined by against the option.

The asymptotic distributions of the F-statistics are non-standard. Two sets of asymptotic critical values are provided by Pesaran et Al. ( 2001 ). The first set assumes that all variables are I ( 0 ) while the 2nd set assumes that all variables are I ( 1 ). If the computed F-statistic is greater than the upper edge critical value, void hypothesis of no cointegration is rejected irrespective whether the series are I ( 0 ) or I ( 1 ) and we may reason that there exits steady province equilibrium between the variables. Alternatively if the computed F-statistic is less than the lower edge critical value, void hypothesis of no cointegration is non rejected irrespective whether the series are I ( 0 ) or I ( 1 ). But a conclusive illation can non be made without cognizing the integrating order of the series, if the computed F-statistic prevarications between the upper and lower critical values. In the 2nd measure, one time cointegration is established, the conditional ARDL long tally theoretical account for can be estimated as follow ;

… ( 2 )

The above specification is besides based on the premise that the mistake footings are serially uncorrelated. It is hence of import that the slowdown order ( P ) of the implicit in VAR is selected suitably. There is delicate balance between taking P sufficiently big to extenuate the residuary consecutive correlativity job and, at the same clip, sufficiently little so that the conditional ECM is non unduly over-parameterized, peculiarly in position of the limited clip series informations. Therefore, the hardiness of consequences are determined by the appropriate slowdown length sing the consecutive autocorrelation job.

Finally, the short-run dynamic parametric quantities by gauging an mistake theoretical account associated with the long tally estimations can be estimated in the undermentioned specification. … ( 3 )

Where and are the short-run dynamic coefficients and is the velocity of accommodation towards long run term equilibrium.

Although probe of a co-integration relationship utilizing the ARDL attack does non ask proving for a unit root, as Ouattara ( 2004 ) argues, in the presence of I ( 2 ) variables in the relationship might render the F-statistics of Pesaran et Al. ( 2001 ) shut-in. This is on history of the fact that bound trial is based on the premise of variables being I ( 0 ) or I ( 1 ). Therefore, the execution of unit root trials for ARDL attack might still be necessary in order to guarantee that none of the variables are integrated of order two [ I ( 2 ) ] or beyond. To that terminal we apply different unit root trials to both existent effectual exchange rate and nominal exchange rate. The criterion augmented Dickey Fuller ( ADF ) and Phillips Perron ( PP ) unit root trials have been criticized for its low power in separating between unit root and a close unit root procedure ( Campbell and Perron, 1991 ; DeJong et Al. 1992 ). Therefore, we have besides performed KPSS and DF-GLS unit root trials, as both trials are more powerful and dependable for little sample informations sets as in our instance.

Empirical consequences:

To follow the motion of the REER and NEER over the clip, a graphical presentation of logs of variables is given in figure 1.

Figure 1: Tendencies of REER and NEER

From the graph it is apparent that the series move together over clip. The unit root trials consequences are summarized in table 1. It is apparent from the tabular array that there is no unit root in the existent exchange rate. As the ADF and DF-GLS trials with and without tendency for REER are important at 5 percent degree of significance, we may reject the void hypothesis of a unit root. As for the nominal effectual exchange rate, all unit root tests unequivocally show that it is I ( 1 ) for both instances whether tendency is included or non. Therefore from the unit root trial statistics, it is ensured that none of the variable is integrated of order two [ I ( 2 ) ] or beyond.

Table 1: Unit of measurement Root Statisticss





Without tendency

With tendency

Without tendency

With tendency

Without tendency

With tendency















Critical Valuess

1 %







5 %







10 %







Note: the stars *, **, *** denotes the significance degree at 10, 5 and 1 % severally. Consequences of unit root trials of NEER at first difference show that it is stationary at 1 % degree of significance.

In order to implement the ARDL trial, we have to first find the appropriate slowdowns as the consequences are really sensitive to dawdle length. To guarantee comparison of consequences for different slowdown lengths, all appraisals were computed over the same sample period. Lag lengths have been chosen based on the Akaike ‘s and Schwarz ‘s Bayesian Information Criteria, denoted severally AIC and SBC. We have considered seven slowdowns harmonizing to AIC information standards as tabulated in table 2.

Table 2. Statisticss for Choosing the Lag Order




















529.346 #








Note: # shows optimum slowdown selected based on the AIC and SBC standards.

After make up one’s minding the optimum slowdown order for equation 1, the consequences are reported in table 3. The deliberate F-statistics for joint significance is 7.456 without tendency which is above the upper edge critical value ( Pesaran et al., 2001 ) at 5 % degree of significance. This indicates that the existent and nominal effectual exchange rates in India are co-integrated. The presence of cointegration between nominal and existent effectual exchange rate, farther supports that alterations in monetary value degrees do offset alterations in the nominal exchange rate, which implies that PPP hypothesis holds in India.

Table 3. ARDL Bound Testing for Cointegration Analysis

Computed F-statistics ( FREER ( REER/NEER ) 7.456*

Critical values

1 % degree

5 % degree

10 % degree

I ( 0 )

I ( 1 )

I ( 0 )

I ( 1 )

I ( 0 )

I ( 1 )







Notes: Critical values are from Pesaran et Al. ( 2001, table CI ( three ) -Case III, P.300 ) * denotes important at 5 % degree of significance.

Therefore there is cointegration relationship between existent and nominal effectual exchange rates in India.

We further probe into the short tally and the long tally kineticss. The consequences are obtained for the short-term kineticss and presented in table 4, by the mistake rectification representation of the ARDL ( 1, 6 ) specification as given in equation 3. From the consequences we see that at least two slowdown coefficients are extremely important implying that nominal depreciation/devaluation leads to existent depreciation in the short tally in India. Further the ECM coefficient estimated in the theoretical account shows how quickly/ easy variables return to their equilibrium values. The ECM coefficients should be statistically important with a negative mark. The mistake rectification term ECMt-1, which measures the velocity of accommodation to reconstruct equilibrium, has negative mark and is statistically important at 5 per centum degree, guaranting that long tally equilibrium can be attained. The coefficient of ECMt-1 is equal to -0.069 for short tally theoretical account and implies divergence from the long-run in existent depreciation is corrected by merely about 7 % over each month through nominal depreciation at 5 % degree of significance.

Table 4. Mistake Correction Representation for the Selected ARDL Model.

ARDL ( 1, 6 ) selected based on AIC, Dependent Variable a?†LREER.


Slowdown order














[ 20.228 ] *

[ 0.622 ]

[ 0.448 ]

[ 1.589 ]

[ 0.527 ]

[ 2.740 ] *



[ -0.083 ]

Electronic countermeasures


[ -3.198 ]

Notes: Numbers in [ ] are the t-value ; R2= 0.73849 Adjusted-R2=0.72705 ; F-state= 73.8248 [.000 ] ; DW-statistic=1.9025 ; * show the important slowdown footings

As Bahmani-Oskooee and Ardalani ( 2006 ) have shown, a negative and important coefficient obtained for ECMt-1 would be an alternate manner of back uping cointegration among the variables. Furthermore Banerjee et Al. ( 1993 ) and Banerjee et Al. ( 1998 ) argued that a extremely important mistake rectification term is a farther cogent evidence of the being of stable long tally relationship. They have besides argued that proving the significance of ECMt-1, which is supposed to transport a negative coefficient, is comparatively more efficient manner of set uping co-integration.

Therefore our consequences favour that nominal depreciation/devaluation leads to existent depreciation in the long tally. The long tally inactive parametric quantities are besides estimated utilizing ARDL specification. The estimations reported in table 5 are besides statistically important ( t-statistics ) at 1 % degree of significance, and have the same positive mark as in the instance of short tally. Thus consequences of short tally and long tally put together show that in India depreciation/devaluation of NEER leads to similar REER in long tally but deviates in the short-run.

Table 5: Estimated Long Run Coefficients utilizing the ARDL Approach

ARDL ( 1, 6 ) selected based on AIC, Dependent Variable LREER.



Standard Error













* denotes important at 1 % degree of significance.


In this paper the relationship between nominal and existent effectual exchange rate is examined in the Indian context. The Bounds-testing attack ( ARDL ) of cointegration is adopted to look into non merely the short-term relation between the nominal and existent effectual alteration rates, but besides their long-term relationship. Consequences of this survey reveal that nominal depreciation leads to existent depreciation in the long-term but deviates in short-run which is in line with the findings of the earlier surveies ( for case by Bahmani-Oskooee and Miteza ( 2002 ) ) for less developed economic systems including India. Thus, the consequences imply nominal exchange rate can be used as a policy tool to stabilise the REER whenever necessary.

Cite this The Evidence From The ARDL Bounds Tests

The Evidence From The ARDL Bounds Tests. (2018, Apr 10). Retrieved from https://graduateway.com/the-evidence-from-the-ardl-bounds-tests/

Show less
  • Use multiple resourses when assembling your essay
  • Get help form professional writers when not sure you can do it yourself
  • Use Plagiarism Checker to double check your essay
  • Do not copy and paste free to download essays
Get plagiarism free essay

Search for essay samples now

Haven't found the Essay You Want?

Get my paper now

For Only $13.90/page