New Zealand's Current Account Deficit: Analysis based on the Intertemporal Optimisation Approach

Kim, Kunhong; Hall, Viv B; Buckle, Robert A Working Paper New Zealand's Current Account Deficit: Analysis based on the Intertemporal Optimisation Approach New Zealand Treasury Working Paper, No. 01/02 Provided in Cooperation with: The Treasury, New Zealand Government Suggested Citation: Kim, Kunhong; Hall, Viv B; Buckle, Robert A (2001) : New Zealand's Current Account Deficit: Analysis based on the Intertemporal Optimisation Approach, New Zealand Treasury Working Paper, No. 01/02, New Zealand Government, The Treasury, Wellington This Version is available at: https://hdl.handle.net/10419/205443 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Disclaimer: The views expressed are those of the author(s) and do not necessarily reflect the views of the New Zealand Treasury. The Treasury takes no responsibility for any errors or omissions in, or for the correctness of, the information contained in these working papers. TREASURY WORKING PAPER 01/02 New Zealand’s Current Account Deficit: Analysis based on the Intertemporal Optimisation Approach Kunhong Kim, Viv B. Hall and Robert A. Buckle∗ Abstract New Zealand’s Current Account of the Balance of Payments has been persistently in deficit since the early 1970s and increased markedly during the late 1990s. Is this a cause for significant concern? This paper tackles this question by evaluating New Zealand’s external solvency, the degree of optimality of the intertemporal consumption smoothing through its current account, and whether its international financial capital flows have been used in an optimal (consumption-smoothing) fashion. We carry out statistical tests in relation to external solvency. We also estimate a “benchmark” consumption-smoothing component for its current account based on an intertemporal optimisation model in order to carry out tests of the optimality of the size and volatility of the current account. We could not reject the hypotheses that New Zealand’s current account was consistent with optimal smoothing, that the external solvency condition has been satisfied, and that there is “no excess volatility” in international financial capital flows. JEL Classifications: E21, F32 Keywords: Current account; intertemporal; consumption-smoothing; New Zealand ∗Kim and Hall: School of Economics and Finance, Victoria University of Wellington, P. O. Box 600, Wellington, New Zealand (email: [email protected],[email protected]); Buckle: The Treasury, P. O. Box 3724, Wellington, New Zealand (email: [email protected] ). Kunhong Kim wishes to thank the Internal Grants Committee of the Victoria University of Wellington (VUW) for the Special Research Grant awarded for this research. We acknowledge without implication, very helpful suggestions from John Carlson, Paul Cashin and John McDermott, and comments received during presentations to VUW’s Research in Economics and Finance (REF) Workshop, the New Zealand Treasury’s Macroeconomic Issues Group, Osaka University’s Macroeconomics Workshop, the Departments of Economics at Purdue University and the University of Auckland, the Federal Reserve Bank of St. Louis, Research Instutute for Economics and Business Administration of Kobe University, and the July 2000 Conference of the New Zealand Association of Economists. Valuable advice and assistance on data matters has been provided by Vicki Plater. 3 1. Introduction New Zealand’s Current Account of the Balance of Payments has been persistently in deficit since the early 1970s (see Figure 1). Two further notable features of recent history have been: the current account deficit to GDP ratio deteriorated during the period 1984 to 1986 from around 4% into the 79% range; and during the latter part of the 1990s moved from around 1% into a 5 to 7% band (see Figure 2). During this latter period the current account deficit to GDP ratios of New Zealand’s two main trading partners, Australia and the US, also moved to historically quite high figures of around 6% and 4% respectively. A further noticeable aspect of Figure 2 is that while the deficit ratios for Australia and New Zealand have averaged around 5% during the 1980s and 1990s, New Zealand’s ratio has been more volatile than Australia’s. Is this persistent and recent sharp deterioration in the current account deficit a cause for significant concern for New Zealand and lenders of international financial capital? The theoretical literature, empirical findings and policy judgements about the implications of persistent and rising current account deficits have evolved considerably over the past decade or so. For example, recent research by Milesi-Ferretti and Razin (1996, p. 161) commenced by suggesting the conventional wisdom to be that “…current account deficits above 5% of GDP flash a red light, in particular if the deficit is financed with short-term debt or foreign exchange reserves, and if it reflects high consumption spending”. They concluded, however, (p. 178) that “… a specific threshold on persistent current account deficits (such as 5% of GDP for 3-4 years) is not per se a sufficiently informative indicator of sustainability. The size of current account imbalances should be considered in conjunction with exchange rate policy and structural factors, …”. A recent analysis for New Zealand by Collins et al. (1998) makes a similar judgement. They concluded (p. 30), after judging that the strengths of New Zealand’s wider sustainability indicators “…considerably outweigh her weaknesses.”, that “…although New Zealand’s current account deficit is sizeable, and will undoubtedly not remain at such an elevated level in the long-run, there are few reasons to believe that the transition to lower current account deficits will be disruptive to the economy.” The work reported in this paper takes a different approach. We carry out statistical tests in relation to external solvency. We also estimate a “benchmark” consumption-smoothing component for New Zealand’s current account based on an intertemporal optimisation model and use it to test the optimality of the size and volatility of the current account.1 Our analytical modelling and testing follows in the tradition of the intertemporal theoretic and empirical work developed in Sachs (1982), Campbell (1987), Campbell and Shiller (1987), Sheffrin and Woo (1990), Trehan and Walsh (1991), and Ghosh (1995). Major aspects of this literature have been summarised comprehensively in Obstfeld and Rogoff (1995) (OR). We specify an intertemporal optimisation model of the current account suitable for a small open economy, estimate the consumption-smoothing current account path using vector autoregression (VAR) methodology2and establish it as a “benchmark” current account path, and conduct a range of statistical tests to assist in forming judgements on a number of key empirical questions. 1The concepts of external solvency, sustainability, and optimality have recently been defined and addressed in MilesiFerretti and Razin (1996) and Cashin and McDermott (1998b). External (intertemporal) solvency is satisfied when a country fully meets its external obligations, in the sense of the present discounted value of its net external liabilities, i.e. its intertemporal budget constraint (ibc) is satisfied. Sustainability, in essence, requires that a country not be subject to ‘liquidity constraints’ imposed by foreign lenders. i.e. in addition to the ibc having to be satisfied, factors influencing (1) willingness (as well as ability) to pay, and (2) willingness to lend, should be taken into account. Intertemporal optimality for the purposes of this paper is as explained in section 2. 2Milesi-Ferretti and Razin (1996, p. 163) categorise the two main approaches to empirical current account modelling as: structural estimation with focus on the degree of persistence of responses to specific shocks; and VAR estimation and analysis of “benchmark” consumption-smoothed current accounts. 4 Key aims of this study are therefore (1) to establish an illustrative intertemporally optimal or “benchmark” path for New Zealand’s current account, and to identify the extent to which actual current account movements have deviated over time from the consumption-smoothed optimal path and whether international financial flows have been excessively volatile; and (2) to establish preliminary empirical conclusions relating to external solvency. Similar work has been reported for a number of countries. For example, Ghosh and Ostry (1995) have concluded that for a majority of developing countries, the hypothesis of full consumptionsmoothing could not be rejected. The hypothesis could also not be rejected for the US (Ghosh, 1995). However, for Canada, Otto (1992) found virtually no support for smoothing, and suggested this might have been due to Canada’s current account being “… more affected by temporary changes in the resource prices and terms of trade effects”. Ghosh (1995) also found that consumption-smoothing restrictions were rejected for Canada, as well as for Japan, Germany and the United Kingdom. He was, however, comfortable with the model’s ability to capture current account directions and turning points in all cases. Conclusions for Australia have varied by study, by sample period and by data source. For example, Milbourne and Otto (1992), utilising per capita data for the period 1959:3 to 1989:1, found that the consumption-smoothing hypothesis was rejected either for the full sample period or for the post 1983:4 floating exchange rate period3. Similarly, using the extended data period 1960-61 to 199495, an intertemporal optimisation model based on less restrictive assumptions, and individually deflated expenditure component series, Guest and McDonald (1998) rejected Australia’s having optimally smoothed consumption over their full sample period. Perhaps more importantly, though, they also reported (p. 213) “…there is less evidence for this since 1984-85, suggesting that deregulation of capital markets may have facilitated the optimal smoothing of consumption”. Two recent studies by Cashin and McDermott (1998a, 1998b) also suggest key conclusions can vary over time. Utilising annual data for the period 1954-94, they concluded that “…the Australian current account was not used to smooth consumption optimally in the period prior to the relaxation of capital controls in the early 1980s…” and that “…in the period since the mid-1980s [i.e. following the move to a fully flexible nominal exchange rate regime for the Australian dollar in December 1983 and, at the same time, the complete removal of capital and exchange controls], Australia’s current account deficits have become excessive…”. Subsequently, however, utilising quarterly data for the period 1984:1 – 1998:2, Cashin and McDermott suggested that despite their having found international capital flows to be larger than optimal during the 1980s, in the 1990s “such flows have been broadly consistent with those predicted by the consumption-smoothing approach”. More specifically, they identified a structural break at 1990:4, and their overall conclusion was that “…it appears that, over time, Australia’s international borrowing decisions have been increasingly determined by changes in economic fundamentals.” It is evident from this brief review that results from applying the intertemporal optimisation approach have varied by country and by time period, and that in some cases the degree of financial market regulation can influence results. The application of the intertemporal optimisation approach to New Zealand therefore complements earlier studies in several respects. The sample used for this study covers a smaller and more volatile open economy than has previously been examined, which was initially characterised by pervasive financial market regulation that was removed during the second half of the sample period. Furthermore, the latter part of the sample period includes the period of the Asian financial crisis when New Zealand’s current account moved further into deficit. The structure for this paper is as follows. Section 2 introduces and explains key economic and 3One suggestion they made for further work (1992, p. 383) was for “relaxing the single commodity assumption…, since it introduces a role for relative prices in explaining consumption and current account behaviour. ” See also Sheffrin and Woo, p. 252, and OR pp. 1755-59. 5 econometric methodology. Major empirical results are presented in Section 3. Conclusions appear in Section 4. 2. Methodology The economic model utilised is a basic intertemporal optimisation model of the current account, of the type developed and explained in Sachs (1982), Sheffrin and Woo (1990), Ghosh (1995), Ghosh and Ostry (1995), Obstfeld and Rogoff (1995), Cashin and McDermott (1998a, 1998b), and Agénor et al. (1999). The model reflects the permanent income theory of consumption and saving4.It therefore implies that temporary shocks could to a large extent be smoothed in the short term, and be reflected instead in substantial short term fluctuations in national saving and the current account. We consider a small open economy that consumes a single good. It is inhabited by a large number of like individuals with infinite planning horizons. The economy is small in the sense that it takes the path of world real interest rates as exogenous. We assume that only riskless bonds are traded in the international capital market and that the world real interest rate on bonds is fixed5. There is no restriction on international borrowing and lending. Population size is normalized to one so that we can identify per capita quantity variables with national aggregate quantities. The representative agent of this economy maximizes lifetime utility () [] ∑ ∞ =+ 0 E jjtt jCu β (1) where β is the subjective discount factor with 10 << β and β β /)1( −is the subjective rate of time preference, t E is the conditional expectations operator based on the information set of the representative agent at period t,andCis private consumption. The period utility function )(Cu is strictly increasing in consumption and strictly concave: 0)( > ′Cu and 0)( < ′′ Cu . The series of budget constraints faced by the representative agent is captured by the current account identity tttttttt GICrBYBBCA −−−+=−≡ +1(2) 4In this basic form, it does not treat demographic influences explicitly, has no explicit terms of trade variables, has no explicit (time varying) risk premium, and has not utilised alternative forms of utility function. Studies which have considered these issues include those of : Guest and McDonald (1998, p. 217), who utilise an additive, constant elasticity form of utility function; and OR (1995), who have postulated CES composite consumption (p 1752) and also an absolute risk aversion utility function ( p. 1791), have given preliminary consideration (as have Milbourne and Otto, 1992, p. 383) to terms of trade related specifications (pp 1755-59), and have explicitly considered demographic factors (pp 1759-64). 5Bergin and Sheffrin (2000) present and test an intertemporal model which allows for variable interest rates and exchange rates. They find that for Australia and Canada, but not for the United Kingdom, including the variable interest and exchange rate influences significantly improves the fit of the model relative to a benchmark model that excludes them. The improvement is attributed primarily to the intratemporal elements of the theory (i.e. allowing for substitution between internationally-traded goods and nontraded goods). The basic model specified and tested in this paper does not allow for variable interest and exchange rate influences. This is partly because the Bergin and Sheffrin paper came to our attention after the empirical work reported here was completed, and partly because (as shown in section 3) our first order, two equation (unrestricted) VAR model and benchmark consumption-smooothed current account path provides robust results and are unlikely to be improved significantly by the additional equation and variables. This could be because intratemporal substitution between internationally-traded and nontraded goods has not been empirically significant for New Zealand. The latter can, however, be tested in follow-up research. 6 where Yis the economy’s real GDP, Bis the beginning of period real net stock of outstanding foreign assets (debts if negative), rBY +is real GNP (defined as real GDP plus interest income on the outstanding stock of net foreign assets), Iis real investment, Gis real government consumption, and CA is the real current account balance (defined as real GNP minus real private and public expenditure, GIC ++ ). Taking expectations of (2) conditional on the information set, and recursively eliminating future values of the stock of foreign assets, yields the intertemporal budget constraint: ∑ ∞ =++ ∞→ ++++ −       + +−−−       + =+− 01)(E 1 1 )(E 1 1 )1( lim jTtt T T jtjtjtjtt j tB r GICY r Br .(3) Requiring that the country’s budget processes be externally solvent rules out Ponzi schemes in which debt is continually rolled over. External solvency requires that the last term in (3) must equal zero: 0)(E 1 11 lim =−       +++ ∞→ Ttt T T B r.(4) If condition (4) is satisfied, the discounted value of the expected future stock of debt converges to zero as the time horizon goes to infinity. Equation (3) then implies that ∑∑ ∞ =+ ∞ =++++       + ≡−−−       + =+− 00 E 1 1 )(E 1 1 )1( jjtt j jjtjtjtjtt j tTB r GICY r Br .(5) Current outstanding real stock of debt, t Br)1( +− , must be equal to the present discounted value of current and expected future trade balance surpluses,TB(defined as real GDP minus real private and public expenditure, GIC ++ ). Proposition 1 in Trehan and Walsh (1991) provides the necessary and sufficient condition for satisfying the solvency condition (4) when the real interest rate is constant. That proposition applied to our context implies that, if TB)L1( λ −is a mean zero stationary stochastic process with r+<≤ 10 λ , then the solvency condition (4) holds if and only if there is a linear combination of TB and Bthat is stationary. Therefore if the current account balance CA (which is the linear combination of TB and Bby the definition of rBTBCA +≡ ) is stationary, then we can say that the solvency condition is satisfied. Hakkio and Rush (1991) discuss the condition required for solvency when the real interest rate is not constant but stationary. They show that, if revenue and expenditure processes are I(1), solvency requires that the inclusive of interest revenues be cointegrated with expenditures. Proposition 2 of Trehan and Walsh (1991) applies to the more general case when the real interest rate is allowed to vary and is not necessarily stationary. In this case, stationarity of the current account deficit is sufficient to imply intertemporal solvency condition holds, as long as the expected real interest rate is positive. With perfect capital mobility, Fisherian separability holds in this model. Facing an exogenously given world real interest rate, the representative agent of the small open economy determines investment and output independently of the level of consumption. We assume that government expenditure is exogenous. Therefore output, investment, and government consumption may all be treated as exogenous when choosing the optimal path for consumption. 7 Necessary conditions for the representative agent’s optimal consumption decision problem include L,1,0)]([E)1()]([E 1= ′ += ′+++ jCurCu jttjtt β ,(6) which implies for 0=jthat )]([E)1()( 1+ ′ += ′ttt CurCu β .(7) With a view to empirical implementation, we consider the case in which period utility is quadratic,6 2 0 2 )( C a CCu −= with 0 0>a. With quadratic period utility function, equation (6) becomes L,1,0]E1)[1(E1 100 =−+=− +++ jCarCa jttjtt β .(8) If the subjective discount factor β and the market discount factor 1/(1+r) are equal so that 1)1( =+ r β , (8) implies LL ===== ++++ 11 EEE jttjttttt CCCC .(9) Equation (9) represents the representative agent’s consumption smoothing motive. When the subjective discount factor is different from the market discount factor the representative agent has a consumption tilting motive as well as a consumption smoothing motive. For example, if β is smaller than 1/(1+r)sothat 1)1( <+ r β , (8) implies LL >>>>> ++++ 11 EEE jttjttttt CCCC ; and the representative agent wants to have consumption tilted towards the present. Equation (8) can be written as L,1,0E )1( 1 E1=+ + =+++ jC r Cjttjtt α β (10) 6With a quadratic utility function, the certainty equivalence principle holds, which implies that the representative agent’s forecasting and optimisation problems separate. The representative agent makes its decisions under uncertainty by acting as if future stochastic variables were sure to turn out equal to their expected values. This separation of forecasting from optimisation considerations is computationally very convenient and explains why quadratic functions are assumed in much applied work. For more general functional forms, the certainty equivalence principle does not hold. With quadratic utility, 0)( = ′′′ Cu so that variability of future net output does not affect consumption. When 0)( > ′′′ Cu , agents engage in precautionary saving that depends on the variability of future net output and not just expected values. Ghosh and Ostry (1994) used constant absolute risk aversion utility function and added a precautionary effect to the kind of intertemporal optimisation current account model utilised here. The key parameter appearing in their extended model is the lifetime innovation in net output. But the length of the data series required to measure this parameter accurately is such that the extended model cannot be utilised for our study. 8 where       + −= )1( 1 1 1 0ra β α . Recursions on equation (10) imply that L,2,1 )]1([1 )]1([1 )1( 1 E1= +− +− ⋅+       + =− − +j r r C r Cj t j jtt β β α β (11) Substituting (11) into (5) and solving for t Cgives optimal consumption      − +                  −−       + ++       + −= ∑ ∞ =+++ r GIY r Br r C jjtjtjt j ttt α β 0 2 *)( 1 1 E)1( )1( 1 1, which can be rewritten as      − +                  −−       ++ += ∑ ∞ =+++ r GIY rr B r C jjtjtjt j ttt α θ 0 *)( 1 1 E 1 1(12) where 1)1( )1( 2−+ + =r rr β β θ .7 Since 1)1()1( )1( 1)1( )1( 2−+++ + = −+ + =rrr rr r rr ββ β β β θ , (13) it is very clear that 1< θ if and only if 1)1( >+ r β . The representative consumer wants to tilt consumption towards the future if 1< θ . Ghosh (1995, p.113), Ghosh and Ostry (1995, p.309), Cashin and McDermott (1998a, p.351), Cashin and McDermott (1998b, p.10), and Agénor et al. (1999, p.4) all take an interpretation for θ diametrically opposite to the above, stating that consumption is tilted towards the present when 1< θ . Also, unlike the above analysis, they seem to have taken no explicit account of the existence of the constant term in the optimal consumption when the utility function is quadratic. The optimal consumption level can be decomposed into the consumption smoothing part and the consumption-tilting part by noting that when 1)1( =+ r β , there is no consumption tilting. The optimal consumption level then becomes                  −−       ++ +∑ ∞ =+++ 0 )( 1 1 E 1 1 jjtjtjt j tt GIY rr Br . We use SM t Cto denote consumption-smoothing component of the optimal consumption. 7A similar result is shown in Sargent (1987), p. 365. 9                  −−       ++ += ∑ ∞ =+++ 0 )( 1 1 E 1 1 jjtjtjt j tt SM tGIY rr BrC . (14) It is the annuity value of the representative consumer’s total discounted wealth net of investment and government consumption. The consumption-tilting component is the difference between the optimal level of consumption * t Cand its smoothing component SM t C. Equations (12) and (14) imply the following relationship between the optimal consumption level and its consumptionsmoothing component. r CC t SM t θα θ += * We define the consumption-smoothing component of the current account as r CGIrBYCGIrBYCA ttttt SM ttttt SM t θα θ −−−−+=−−−+= ** . (15) Substituting (14) into (15) implies that the consumption-smoothing component of the current account can be represented as ∑ ∑ ∑ ∞ =+ ∞ =+++ ∞ =+++       ++ −=        −−       ++ −−−=                  −−       ++ +−−−+= 0 0 0 * E r1 1 1 )( 1 1 E 1 )( 1 1 E 1 1 jjtt j t jjtjtjt j tttt jjtjtjt j tttttt SM t Z r r Z GIY rr r GIY GIY rr BrGIrBYCA (16) where GIYZ −−= has been termed in the literature net output or national cash flow. Rearranging terms in the right-hand side of the last equality in (16) yields ∑ ∞ =+ ∆ + −= 1 *E )1( 1 jjtt j SM tZ r CA . (17) Equation (17) shows that the consumption-smoothing component of the current account is in deficit when the present discounted value of future net output changes is positive, and it is in surplus in the opposite case. The consumption-smoothing component of the current account deficit is the predictor of future increases in net output. According to equation (17), permanent shocks, which have no effect on ∆Z, leave the consumption-smoothing component of the current account unaffected, whereas temporary shocks to Z(e.g. an unexpected temporary increase in Gor I) would lead the current account to act as a buffer to smooth consumption. Equation (17) shows that creating the model implied consumption-smoothing component of the current account series requires estimating the present value of expected changes in net output, where expectation is conditional on the information set used by the representative agent. As shown 16 4. Conclusion There has been a long-standing debate concerning the implications and appropriate policy response to New Zealand’s persistent current account deficit. This debate has been heightened by the substantial increase in the current account deficit during the 1990s, especially following the Asian financial crisis in the late 1990s. This paper contributes to this debate by evaluating New Zealand’s external solvency, the degree of optimality of the intertemporal consumption smoothing through its current account, and whether its international financial capital flows have been used in an optimal (consumption-smoothing) fashion. We carried out statistical tests in relation to external solvency. We also estimated a “benchmark” consumption-smoothing component for its current account based on an intertemporal optimisation model and used it to test the optimality of the size and volatility of the current account. Specific results are: (1) Despite substantial deterioration in New Zealand’s current account deficits during the late 1990s, its current account movements over our sample period as a whole have been consistent with its intertemporal budget constraint and hence its formal external solvency condition has been satisfied; (2) The current account balance predicted by the simple intertemporal optimisation model used in this paper has satisfactorily reflected the actual directions and turning points for the consumption smoothing component of the current account. The null of the NoGranger causality hypothesis that the current account has not signalled subsequent changes in net output has been rejected. Furthermore, a Wald test of nonlinear restrictions implied by the model has not been rejected. All of these results are consistent with optimal smoothing having been achieved; (3) We also examined the sensitivity of the results to the decomposition between the consumption-tilting and consumption-smoothing components, by obtaining empirical results without stochastic detrending. This is equivalent to imposing no consumption-tilting. This did not alter our results in any material way; (4) Finally, it can be noted that the variance ratio of our actual and model implied current account series is consistent with “no excess volatility” in international financial capital movements for consumptionsmoothing purposes. 17 APPENDIX: DATA SOURCES Two basic data sets were constructed for the period 1982:2 to 1999:3. One set was converted from nominal to real terms, by using the implicit price deflator for GDP; the other utilises series for the individual components of GDE published directly in real terms. The latter are therefore the standard national system of accounts constant price measures. Seasonally adjusted series for private final consumption expenditure (C), gross fixed capital formation and increase in stocks (I), general government final consumption expenditure (G), and GDP (gross domestic expenditure, Y), in current and constant 1991-92 prices were taken from Statistics New Zealand’s (SNZ) September 1999 quarter release of Gross Domestic Product data. The implicit price deflator for GDP series was computed as the ratio of our current price and constant price GDP series, and is the same (after converting to base 1991-92 =100) as SNZ’s published series. The Gross National Product (i.e. Y+rB) series are obtained by adding to GDP SNZ’s Balance of Payments’ (BoP) series “Balance on (International) Investment Income”. The rB series in current price form was seasonally adjusted using X11 (and deflated by the GDP deflator). The nominal series for our sample period was taken primarily from the recently released BoP statistics compiled using the IMF’s BoP Manual, 5th edition (BPM5), and for observations prior to 1986:4 from BPM4. No official series exists in real seasonally adjusted form. The rB series have negative values for all observations in our sample period. Our current account series (CA) in current and constant price terms were computed (in residual fashion) from (Y+rB)–(C+I+G), and when converted to year ended current account to GDP ratios at quarterly intervals, follow very closely the corresponding ratios published by SNZ using BPM4. “National Cash Flow”/”Net Output”, Z, was calculated from Y-I-G; and the population series used to convert our data to per capita form was obtained by linking (at 1991:2) SNZ’s series for de facto mean population (SBEC) and resident mean population (SEIC). The value imposed for the real world interest rate, r, was the conventionally used 4% per annum. 18 REFERENCES Agénor, Pierre-Richard, Claude Bismut, Paul Cashin, and C. John McDermott (1999), “Consumption Smoothing and the Current Account: Evidence for France”, Journal of International Money and Finance,18, 1-12. Bergin, Paul R. and Steven M. Sheffrin (2000), “Interest Rates, Exchange Rates and Present Value Models of the Current Account”, The Economic Journal,110, April, 535-558. Campbell, John (1987), “Does Saving Anticipate Declining Labor Income? An Alternative Test of the Permanent Income Hypothesis”, Econometrica,55, November, 1249-1273. Campbell, John and Robert Shiller (1987), “Cointegration and Tests of Present Value Models”, Journal of Political Economy,95, October, 1062-1088. Cashin, Paul and C. John McDermott (1998a), “Are Australia’s Current Account Deficits Excessive?”, The Economic Record,74 (227), December, 346-361. Cashin, Paul and C. John McDermott (1998b), “International Capital flows and National Creditworthiness: Do the Fundamental Things Apply as Time Goes By?”, IMF Working Paper, WP/98/172, December. Collins, Sean, Francisco Nadal De Simone, and David Hargreaves (1998), “The Current Account Balance: An Analysis of the Issues”, Reserve Bank of New Zealand Bulletin,61(1), March, 15-34. Ghosh, Artish R. (1995), “International Capital Mobility Amongst the Major Industrialised Countries: Too Little or Too Much?”, The Economic Journal,105, 107-128. Ghosh, Artish R. and Jonathan D. Ostry (1994), “Export Instability and the External Balance in Developing Countries”, IMF Staff Papers,41(2), 214-235. Ghosh, Artish R. and Jonathan D. Ostry (1995), “The Current Account in Developing Countries: A Perspective from the Consumption-Smoothing Approach”, The World Bank Economic Review,9(1), 305-333. Guest, Ross S. and Ian M. McDonald (1998), “The Socially Optimal Level of Saving in Australia, 1960-61 to 1994-95”, Australian Economic Papers,37(3), September, 213-235. Hakkio, Craig S. and Mark Rush (1991), “Is the Budget Deficit Too Large?”, Economic Inquiry, 29, 429-445. Milbourne, Ross and Glenn Otto (1992), “Consumption Smoothing and the Current Account”, Australian Economic Papers,31, December, 369-384. Milesi-Ferretti, Gian Maria and Assaf Razin (1996), “Persistent Current Account Deficits: a Warning Signal?”, International Journal of Finance and Economics,1, 161-181. Obstfeld, Maurice and Kenneth Rogoff (1995), “The Intertemporal Approach to the Current Account”, Chapter 34 in G. Grossman and K. Rogoff (eds.)Handbook of International Economics, Volume 3, 1731-1799. 19 Otto, Glenn (1992), “Testing a Present-value Model of the Current Account: Evidence from US and Canadian Time Series”, Journal of International Money and Finance,11, 414-430. Sachs, Jeffrey (1982), “The Current Account in the Macroeconomic Adjustment Process”, Scandinavian Journal of Economics,84, 147-159. Sargent, Thomas J. (1987), Macroeconomic Theory, Academic Press. Sheffrin, Steven M. and Wing Thye Woo (1990), “Present Value Tests of an Intertemporal Model of the Current Account”, Journal of International Economics,29, 237-253. Trehan, Bharat and Carl E. Walsh (1991), “Testing Intertemporal Constraints: Theory and Applications to U.S. Federal Budget and Current Account Deficits”, Journal of Money, Credit, and Banking,23(2), 206-223. 20 Figure 1: Balance on Current Account, New Zealand Nominal NZ$m, March Years 1950/51 to 1998/99 Figure 2: Current Account to GDP (%) New Zealand (IMF 4th ed.), Australia 21 Figure 3: Real Current Account Balance Based on Nominal Expenditure Series Deflated by the GDP Deflator BasedonRealExpenditureSeries 22 Figure 4: Actual and Predicted Current Account Demeaned and Detrended, 1991/92 NZ$m, r= 0.04 p.a. Figure 5: Actual and Predicted Current Account Demeaned but Not Detrended, 1991/92 NZ$m, r= 0.04 p.a. 23 Figure 6: Actual vs. Predicted Current Account with 95% Confidence Band in percent of GDP, r= 0.04 p.a. 24 Table 1: Tests for Unit Roots Series Deflated by GDP Deflator, 1982:3 – 1999:3 (Augmented Dickey-Fuller t-statistic) Non-per capita data Per capita data Level 1st difference Level 1st difference C0.51 -12.15** -0.39 -12.41** Z-1.44 -9.39** -2.34 -9.40** Z+rB -1.63 -9.29** -2.57 -9.28** TB♠-3.92** -10.44** -4.00** -10.37** CA -3.07*-10.58** -3.32*-10.60** Asymptotic critical values are: 1%, -3.51; 5%, -2.89; 10%, -2.58 * Null hypothesis of unit root rejected at 5% level of significance, in favour of stationarity. ** Null hypothesis of unit root rejected at 1% level of significance, in favour of stationarity. ♠Means of series are all positive at $122.26m, $17.10m, $32.60, and $5.31, respectively. Table 2: Cointegration Regressions and Estimates of θ θθ θ Series Deflated by GDP Deflator, 1982:2 – 1999:3 Non-per capita data Per capita data Constant Term No Constant Term Constant Term No Constant Term Cointegration Regression Constant 191.82 49.72 (t-stat) (0.40) (0.21) θ 0.90 0.92 0.90 0.92 (t-stat) (22.01) (189.92) (12.91) (185.69) ADF t-stat♣-3.53*-3.47*-3.56*-3.53* ♣Asymptotic critical values are: 1%, -3.96; 5%, -3.37; 10%, -3.07. * Null hypothesis of unit root rejected at 5% level of significance, in favour of stationarity. 25 Table 3: VAR Parameters, 1982:4 – 1999:3 Series Deflated by GDP Deflator, Cointegration Regression Used Non-per capita data Per capita data Const.Term No Const. Term Const. Term No Const.Term ∆ZtSM t CA ∆ZtSM t CA ∆Zt SM t CA ∆ZtSM t CA ∆Zt-1 -0.20 -0.08 -0.20 -0.08 -0.19 -0.08 -0.20 -0.08 (t-stat) (-1.73) (-0.74) (-1.73) (-0.73) (-1.67) (-0.75) (-1.67) (-0.74) SM t CA 1−-0.20 0.74 -0.20 0.75 -0.22 0.74 -0.21 0.75 (t-stat) (-2.23) (8.61) (-2.19) (8.64) (-2.33) (8.66) (-2.30) (8.67) Table 4: Tests based on Rt, 1982:4 – 1999:3 Series Deflated by GDP Deflator, Cointegration Regression Used Non-per capita data Per capita data Const. Term No Const. Term Const. Term No Const. Term ∆Zt-1 0.12 0.12 0.11 0.12 (t-stat) (1.21) (1.22) (1.18) (1.18) SM t CA 1−-0.06 -0.06 -0.05 -0.05 (t-stat) (-0.79) (-0.79) (-0.66) (-0.67)