Economics Letters North-Holland
OFFICIAL AUSTRALIAN 1977 TO 1986 Sandra
26 (1988) 73-75
IN FOREIGN EXCHANGE
University of Wollongong, Woilongong, N.S. W. 2500, Australia Received
The intervention behaviour of the Reserve Bank of Australia for a period both before and after the float of the Australian currency is analysed. It is found that ‘leaning against the wind’ is the intervention policy used by the Bank.
1. Introduction The official policy of the Reserve Bank of Australia in the foreign exchange market since the float of the Australian dollar in December 1983 was initially one of ‘testing and smoothing’. Since August 1986, the policy of the Reserve Bank has been more explicit, namely it has ‘traded in the market on its own account to enhance market efficiency and to promote stability [Marsden and Jones (1987)]. The objective of central bank intervention under a floating exchange rate regime is to reduce instability in exchange markets by reducing the magnitude of random fluctuations away from the equilibrium rate and by smoothing out swings in the exchange rate movement. A distinction is made here between two intervention strategies - reducing and countering erratic fluctuations away from an equilibrium rate and consequently smoothing the path of the exchange rate through time, and moderating the speed of exchange rate movements toward what is considered a new equilibrium. The former strategy is more closely identified with intervention on a very short term basis (that is, daily or weekly). The latter strategy implies intervention into the market of considerably greater magnitude and is consequently more controversial and more long term than the first strategy. This latter policy is commonly referred to as ‘leaning against the wind’. Both Quirk (1977) and Hut&son (1984) have empirically tested and analysed the use of this policy by the Japanese central bank. Intervention by use of a ‘leaning against the wind’ policy, moreover, is in accordance with the International Monetary Fund guidelines for management of members’ exchange rate policies [Hut&son (1984)].
2. A model of intervention behaviour The analysis here considers the type and extent of the intervention undertaken by the Reserve Bank of Australia over a time period from January 1977 to December 1986. In choosing a time period which crosses a change in management of the exchange rate, some implications of structural change have to be accommodated and tested. The period since the floating of the Australian dollar * I am grateful
0 1988, Elsevier Science Publishers
alone, however, was considered to be an insufficient time period for comprehensive data analysis, and some comparisons of the type and level of intervention since the float may be instructive. The model employed here postulates simple leaning against the wind behaviour with lagged response variables [Hutchison (1984)],
where I is an intervention proxy: the change in gold and foreign exchange held by the Australian Reserve Bank from the end of the preceding month is converted into U.S. dollars at the average of that month’s exchange rate, and 2 is the percent change of the Australian dollar/U.S. dollar spot exchange rate from the end of the preceding month. It is hypothesised that evidence of leaning against the wind would yield a negative sign on the percentage change of the exchange rate value and the lagged exchange rate value. The lagged exchange rate recognises that policies are modified gradually and that inte~ention reacts to the rate of exchange rate change in previous months. The lagged intervention variable recognises that the intervention policy this month is in part determined by the intervention policy in previous months. The lagged intervention variable would thus be positively related to the dependent variable.
3. Empirical results The results in table 1 indicate that the Reserve Bank of Australia does use a leaning against the wind policy in its intervention strategy. That is, if the Australian dollar/U.S. dollar exchange rate (e) is appreciating, the value of e will fall. The Reserve Bank moderates the speed of the exchange rate movement by purchasing foreign financial assets. A positive sign on the dependent variable represents accumulation of foreign reserves. The results in table 1 for the entire sample indicate that every time the dollar appreciates (depreciates) by a cent the Reserve Bank authorities buy (sell) $7.6 million of foreign exchange reserves. The equivalent amounts for the subsamples are $4.7 million and $8.8 million, respectively. This indicates that inte~ention activity increased rather than decreased in the period after the float of the Australian dollar. Moreover, the coefficient for the percentage change
Table 1 Regression Independent variables R-squared DWChow test constant e e-1 I-1
a r-statistics b Significant
0.0498 1.8873 0.0113 (1.336) - 0.0047 (- 1.07) 0.0026 (0.582) 0.2737 (2.493) b in parentheses. at the 5% level of confidence.
0.3587 1.8476 0.0039 (0.476) -0.0088 (- 4.72) b 0.0029 (1.237) 0.0769 (0.456)
3 sample 1986
Jan. 1977-Dec. 0.1282 1.9148 0.3874 0.0090 (1.433) -0.0076 (-3.544) b 0.3591 (1.587) 0.2496 (2.759) b
in foreign exchange markets
in the exchange rate is not significant for the first subsample. Leaning against the wind was not a significant policy in the pre-float period. The lagged variables in table 1 have good explanatory power. The lagged intervention variable is of the correct sign for all sample periods, but insignificant for the second subsample. This result is intuitively appealing as it indicates that the intervention policy altered after floating of the Australian dollar in December 1983. As mentioned previously, however, intervention activity increased rather than decreased after the dollar was floated. For all sample periods, the lagged exchange rate variable is insignificant and of incorrect sign. Thus, the lagged exchange rate variable is contrary to the results of both Hutchison and Quirk, positively related to the dependent variable. This suggests that the authorities have followed a policy of leaning against the wind in the i~ediate month, but have made a compensating adjustment of leaning into the wind in the preceding month. The evidence of a different intervention management policy since December 1983 is a pleasing result. The F-statistic for the Chow test of structural stability of the parameters between the two time periods, however, is not significant at the five percent level. The hypothesis of a stable relations~p between the parameters is therefore not rejected and the subsample disturbance variances can be considered equal. The Durbin-Watson statistic is used in spite of the presence of a lagged dependent variable as the Durbin’s h-statistic could not be calculated in this case.
4. Conclusions This paper provides evidence of the official Reserve Bank policy on intervention in the foreign exchange market. The bank apparently does follow an intervention policy of moderate ‘testing and smoothing’ of the exchange rate. Moreover, there is evidence that the policy of the bank has altered since the floating of the currency in December 1983. Leaning against the wind inte~ention behaviour is more prevalent after than before the float.
References Hutchison, M.M., 1984, Official Japanese intervention in foreign exchange markets: Leaning against the wind?, Economics Letters 15, 115-120. Intemational Monetary Fund, Inte~ational Financial Statistics, various issues (IMF: Washington, DC). Marsden, J.S. and P.L. Jones, 1987, Monetary and exchange rate policy: A volatility based perspective, Conference on Australian Monetary Policy Post Campbell, Victoria. Quirk, P.J., 1977, Exchange rate policy in Japan: Leaning against the wind, IMF Staff Papers, Nov., 642-664. Reserve Bank of Australia, Bulletin, various issues.