THE EXCHANGE RATE AND THE RESERVE BANK'S ROLE IN THE FOREIGN EXCHANGE MARKET

The Australian economy operates with a floating exchange rate. This paper discusses the Australian dollar exchange rate within the context of the Reserve Bank's monetary policy framework, and the role of the Reserve Bank in the foreign exchange market.

1. What is the Exchange Rate and Why is it Important?

The exchange rate is the price of an Australian dollar expressed in terms of another currency. The two most common measures of the exchange rate used in Australia are:

• the exchange rate against the US dollar. Trading of Australian dollars on the foreign exchange market is predominantly against the US dollar. The US dollar is also the dominant international medium of exchange.
• the trade-weighted index or TWI. This is, in fact, not a price in terms of a single overseas currency, but a price in terms of a basket of currencies. This is often a better measure of general trends in the exchange rate than any one bilateral exchange rate, such as that against the US dollar, since the Australian dollar could be rising against the US dollar but falling against other currencies – in such circumstances, the TWI will give a measure of whether the Australian dollar is rising or falling on average.

The make-up of the TWI basket is determined by the relative shares of different countries in Australia's trade. It is reviewed annually to reflect changes in the composition of Australia's trade. The current composition of the TWI is shown in Table 1.

While these two measures of the Australian dollar exchange rate often move together, they have diverged at times (Graph 1). The main divergence occurred during the Asian crisis in 1997, when the exchange rate against the US dollar depreciated much more than the trade-weighted index because the Australian dollar appreciated against most Asian currencies.

Graph 1

Click to view larger

There are many alternative exchange rate indices which can be calculated, and which are relevant for different purposes. For instance, rather than using the conventional TWI based on trade weights, indices weighted by export shares or import shares separately might be more appropriate. Alternatively, a third-country export-weighted exchange rate index might be appropriate, if one wants to take into account competition that home-country exports experience in foreign markets from other countries, even if little trade occurs between the two competing countries. Another problem with trade weights is that they only cover goods and services that are actually traded. This does not necessarily respond to countries' shares of world production, and hence their influence on world prices. Therefore in some circumstances a GDP-weighted index may be appropriate.

Broad effective exchange rates like the TWI can sometimes mask important developments in the component bilateral exchange rates or groups of bilateral rates. For example, there has been a marked divergence in the trend movements of the Australian dollar against the exchange rates of the G10 and non-Japan Asian currencies, the two main groups making up the TWI. Against the former, the Australian dollar has depreciated in the post-float period, while against the latter the Australian dollar has appreciated significantly.

Graph 2

Click to view larger

Related RBA reading:

• Developments in the Trade-Weighted Index (October 2002)
• Measuring the Real Exchange Rate: Pitfalls and Practicalities (Ellis, 2001)

2. Why Does Australia have a Floating Exchange Rate?

Between the early 1970s and 1983, exchange rate policy in Australia moved through several regimes. The first major shift occurred in 1971, when exchange rate policy shifted from pegging the Australian dollar to the UK pound to pegging to the US dollar. This was followed by a period of pegging to the TWI, from 1974 to 1976. During each of these regimes, there were occasional revaluations and devaluations. A crawling peg against the TWI was subsequently adopted, until the Australian dollar was eventually floated in 1983. The history is shown in Graph 3.

Graph 3

Click to view larger

Under the fixed-rate and crawling peg arrangements which existed before 1983, the Reserve Bank cleared the foreign exchange market each day of any excess demand for or supply of Australian dollars. If, say, banks in aggregate had acquired US dollar balances in the course of their business and sold Australian dollars, the Reserve Bank would acquire those US dollars from the banks in exchange for Australian dollars at the prevailing exchange rate. As a result, volatile flows across the foreign exchanges would be reflected in matching purchases or sales of Australian dollars by the Reserve Bank. One problem with this was that, through these purchases and sales of Australian dollars, the Reserve Bank was subtracting from or adding to liquidity in the Australian money market. Domestic monetary conditions were therefore affected by the behaviour of international trade and capital flows.

Floating the exchange rate addressed this problem. After the float, the exchange rate was able to adjust to clear any excess demand or supply. The Reserve Bank no longer cleared the market, so foreign exchange flows ceased to have such a large effect on domestic liquidity. One of the final prerequisites for effective domestic monetary policy was achieved (the other, namely that the Government fully finance its budget surplus or deficit in the market at market interest rates, had been achieved in Australia in the early 1980s when the Australian government adopted a tender system for issuing bonds).

But this is not the only reason why there was a case for the floating of the Australian dollar. Economic theory tells us that the choice of exchange rate regime can also influence the way in which economies cope with external shocks. Countries which are susceptible to real external shocks will generally fare better with a floating exchange rate. Take the example, say, of a sharp drop in the terms of trade (the ratio of export prices to import prices). With a flexible exchange rate, the automatic response would be a depreciation, which would tend to cushion the impact of reduced incomes on the domestic economy. This market-driven reaction means, in principle, that there is less pressure to adjust monetary and fiscal policies in response to the initial shock. By contrast, a country with a fixed exchange rate would need to ease fiscal policy or adjust the exchange rate peg to avoid the contractionary effect of such a shock. This is a difficult problem for policymakers and, over the longer term, it would result in domestic policy settings being more volatile.

In summary, the floating exchange rate regime that has been in place since 1983 is widely accepted as having been beneficial for Australia. Such a regime is particularly well suited for the Australian economy, given it is relatively small, reasonably open and subject to sizeable shifts in its terms of trade. The floating exchange rate has acted as a buffer to external shocks, particularly shifts in the terms of trade, allowing the economy to absorb them without generating the large inflationary or deflationary pressures that tended to result under the previous fixed exchange rate regimes. While discretionary changes were made to the value of the Australian dollar in these regimes in response to developing pressures, it was extremely difficult to calibrate the adjustments either rapidly enough or accurately enough to provide an effective buffer against the shocks. Further, the floating exchange rate has contributed to the reduction in volatility of output that has occurred over the past two decades.

3. What Determines the Behaviour of the Exchange Rate?

One important determinant of a country's exchange rate over the long run is whether it has a higher or lower inflation rate than its trading partners. The theory of purchasing power parity suggests that the exchange rate between two countries will adjust such that a unit of each currency has the same purchasing power in both countries, once the exchange rate is taken into account. If a country's inflation rate is persistently higher than that of its trading partners, its exchange rate will tend to depreciate to prevent a progressive loss of competitiveness over time. Graph 4 shows the ratio of the average price level of Australia's trading partners relative to that in Australia and the TWI for the Australian dollar. From the mid 1970s through to the end of the 1980s, prices in Australia rose more quickly than prices overseas. The TWI depreciated over the same period, but a large part of this was doing no more than offsetting the cumulatively higher inflation Australia was experiencing. Much of what appears to be a potential gain in competitiveness due to the lower exchange rate was offset by a poor performance on inflation.

Estimates of real exchange rates adjust for this difference in inflation rates. Between 1970 and 1990, when the nominal TWI fell by about 40 per cent and Australia's inflation exceeded that of its trading partners, the real TWI depreciated by about 20 per cent. The movement in the real exchange rate is a better measure of changes in competitiveness than the movement in the nominal exchange rate, but it is still subject to considerable fluctuations.

Graph 4

Click to view larger

One of the strongest influences on the Australian dollar has been the terms of trade. For example, a rise in the terms of trade as a result of an increase in the prices of commodities (which are an important component of Australia's exports) provides an expansionary impulse to the economy through an increase in income. The increased demand for inputs from the export sector also creates inflationary pressure. An appreciation of the exchange rate counteracts these influences to some extent by inducing a substitution of domestic demand towards imported goods and services. However, there is evidence to suggest that this relationship has weakened since the late 1990s/early 2000s (Graph 5). The strong increase in commodity prices that began in 2002 boosted the terms of trade and was associated with an increase in the real TWI, although the rise has been less than what might have been expected given the historical relationship.

Graph 5

Click to view larger

A third major influence on the exchange rate is factors that affect capital transactions, such as relative rates of return on Australian dollar assets and changes in investor confidence in Australian dollar assets. Anecdotally, there have been three periods since the currency was floated when relative interest rates were seen as being a major influence. One was in the late 1980s, a period when Australian interest rates were much higher than overseas rates and the exchange rate rose sharply; the second was in the late 1990s, when Australian interest rates fell below US rates and the exchange rate depreciated; the third was in the first half of the 2000s, when Australian interest rates were again higher than overseas rates, as the major economies experienced a downturn.

The importance of relative rates of return is difficult to demonstrate graphically, because it is often expectations about rates of return, rather than current rates, which are important and these expectations are difficult to measure. However, the general tendency is for higher domestic interest rates to be associated with a higher exchange rate when other things are equal.

Empirical models of the exchange rate

While it has been demonstrated that forecasting exchange rates is fraught with difficulty1, even attempts to model historical movements in exchange rates have met with mixed success. Efforts to model the Australian dollar exchange rate in the post-float era have been reasonably successful compared to other currencies, given its correlation with the terms of trade. These models are traditionally based on quarterly data, and have been relatively successful in explaining movements in the Australian dollar exchange rate over the short to medium term.2

While it is possible to identify a number of determinants of the exchange rate, it is important to note that their impact can vary over time. While the terms of trade has displayed a strong correlation with the exchange rate in the post-float era, there is evidence to suggest that this relationship weakened in the late 1990s and early 2000s. Over this period, Australia's terms of trade was rising but the nominal and real exchange rates both declined substantially. Some part of this decline reflected the large rise in the US dollar at the time, as there was a portfolio shift towards investment in technology stocks at the expense of so-called 'old economy' stocks prevalent in Australia.

Variables other than the terms of trade have sometimes helped explain movements in the Australian dollar exchange rate. At times, interest differentials have had an important role, at other times, the stock of foreign liabilities, the current account balance or growth rate differentials have been found to have an influence. In part, the changing influence of some of these variables reflects the varying focus of financial market participants.

Movements in the exchange rate over the very short run (e.g. intraday) have proved difficult to explain in a modelling framework. Research conducted in recent years using 'order flow' data, which report the size, price and direction of individual foreign exchange market transactions, has made some progress in modelling intraday exchange rate movements.3

Related RBA reading:

• Major Influences on the Australian Dollar Exchange Rate (Blundell-Wignall, Fahrer and Heath, 1993)
• A Small Model of the Australian Macroeconomy (Beechey, Bharucha, Cagliarini, Gruen and Thompson, 2000)
• Some Possible Long-term Trends in the Australian Dollar (Gruen, 2001)
• A Small Model of the Australian Macroeconomy: An Update (Stone, Wheatley and Wilkinson, 2005)
• Commodity Prices and the Terms of Trade (April 2005)
• The Impact of Monetary Policy on the Exchange Rate: A Study Using Intraday Data (Kearns and Manners, 2005)

4. The Exchange Rate and Monetary Policy

The exchange rate plays an important part in considerations of monetary policy in all countries. However, the exchange rate has not served as either a target or an instrument of monetary policy in Australia since the currency was floated. The exchange rate is best considered as part of the transmission mechanism. It serves to buffer the economy from external shocks, such that monetary policy can be directed towards achieving domestic price stability and growth.

Since the early 1990s, monetary policy has been conducted under an inflation targeting framework. The inflation target has replaced the exchange rate as the nominal anchor in the economy. Under the inflation targeting regime, monetary policy no longer targets any particular level of the exchange rate. Nor indeed has the Reserve Bank used intervention to defend any level of the exchange rate (see Sections 6, 7 and 8).

Various measures suggest that exchange rate volatility has been higher in the post-float period. However, output volatility has declined (Simon 2001). While this development is not unique to Australia, it is likely that the floating exchange rate has contributed to this decline in output volatility, together with a number of other economic reforms that have occurred in recent decades, including in both the product and labour markets as well as the reforms to the policy frameworks for both fiscal and monetary policy (Gruen and Stevens, 2000). In particular, exchange rate fluctuations have played a particularly important role in smoothing the influence of terms of trade shocks (Gruen and Wilkinson, 1991). Chen and Rogoff (2002) find a similar relationship in other commodity producing countries, but note that the relationship in Australia has been particularly robust.4

In addition to counterbalancing the influence of external shocks, particularly those related to movements in the terms of the trade, the other important role of the exchange rate in the transmission mechanism has been in its influence on inflation. Under the fixed exchange rate regimes, the Australian economy directly 'imported' the inflation rate of the country (or group of trading partners) to which the exchange rate was pegged. With the floating of the exchange rate, this was no longer the case. Instead, movements in the exchange rate itself became a direct influence on inflation. However, the extent of this influence has changed over the period since the float. The pass-through of exchange rate changes to consumer price inflation, through changes in the price of tradable goods and services has become more protracted (Heath et al, 2004). This phenomenon is not unique to Australia, having been also found in the United Kingdom, Brazil, Chile and the US inter alia.

Related RBA reading:

• Australia's Real Exchange Rate - is it Explained by the Terms of Trade or by Real Interest Differentials? (Gruen and Wilkinson, 1991)
• The Evolution of Monetary Policy: From Money Targets to Inflation Targets (Grenville, 1997)
• Pitfalls in the Use of Monetary Conditions Indexes (Stevens, 1998)
• Australian Macroeconomic Performance and Policies in the 1990s (Gruen and Stevens, 2000)
• The Decline in Australian Output Volatility (Simon, 2001)
• Inflation in Australia: Measurement and Modelling (Heath, Roberts and Bulman, 2004)

5. The Foreign Exchange Market

Foreign exchange turnover in Australia is currently around A$143 billion a day. This makes the Australian market the seventh largest in the world, though the two largest – London and New York – are much larger than the remainder. About half the turnover in the Australian foreign exchange market is against the Australian dollar. The remaining half is largely made up of trade in major currencies against the US dollar.

Activity in the Australian market has been very strong during recent times. In particular, trade in the euro has increased, in part due to some large European and US investment banks moving their Asia-Pacific trading desks to Sydney. Much of the pick-up has come from transactions executed with overseas banks, and trade in less traditional currencies has begun to expand.

Graph 6

Click to view larger

In addition to the traditional market segment (comprising turnover in spot, outright forwards and foreign exchange swaps), other foreign exchange derivatives are also traded in the Australian market. Derivatives constitute an important tool for many Australian companies, because they can be used to provide protection against adverse exchange rate movements. The Australian market processes around A$5 billion of these transactions every day, and contains a wide variety of products, ranging from very simple to more complex designs. As such, while trade in the traditional marketplace remains strong, the overall level of sophistication in Australia's foreign exchange market continues to increase.

As well as trading in Australia, there is considerable trading of the Australian dollar in other markets. Global trade in the Australian dollar averaged US$103 billion per day in April 2004 (the date of a recent global survey). This makes the Australian dollar the sixth most traded currency in the world, and the AUD/USD the fourth most traded currency pair. The size of the market indicates that the exchange rate is being determined in a liquid, active and competitive market place.

Graph 7

Click to view larger

Related RBA reading:

• The Australian Foreign Exchange and Derivatives Markets (June 2005).

6. Why Does the Reserve Bank Intervene in the Foreign Exchange Market?

The decision to float the Australian dollar allowed market forces to determine the value of the currency. However, this did not mean that the RBA had become indifferent to either the level of, or movement in, the exchange rate, since these can have a powerful influence on important aspects of the economy, particularly economic growth and inflation. For this reason, the RBA has occasionally intervened in the foreign exchange market. This approach to intervention stemmed from the Campbell Committee finding that a 'clean' float was unrealistic and that occasional intervention by authorities may be desirable.

When the exchange rate was first floated, the academic literature was strongly in favour of the view that financial markets would keep asset prices, exchange rates included, moving in line with changing economic fundamentals. This view was based on the well-known efficient markets hypothesis (EMH), the notion that the activities of rational traders will ensure that information relevant to determining the equilibrium for an asset price will be used in forming the actual market price. Under this view, there would be no role for intervention. However, over the past couple of decades, the academic literature has come to acknowledge that financial markets can overshoot – i.e. asset prices can move to levels that do not seem reasonable in the context of a range of economic and financial developments. There is an extensive literature, for example, on speculative bubbles, herding, fads, and other behaviour which can drive market prices away from their equilibrium values, even in a market which is deep and liquid. When such overshooting occurs, intervention may help in limiting the move or returning the exchange rate towards its equilibrium level, thus obviating the need for costly adjustment by the real economy to the incorrect signals which the exchange rate would otherwise give.

The RBA's approach to intervention has evolved over the past two decades. Broadly, there has been a shift away from concern about short-term volatility in the early post-float period to a focus on episodes where the exchange rate has clearly overshot. This change in emphasis has resulted in intervention strategy moving from generally small daily interventions with frequent changes in direction (often described as ‘testing and smoothing') to less frequent but larger scale intervention once the exchange rate had moved a long way.

Of course, the important issue is to identify in practice when the exchange rate has in fact overshot. Typically, the RBA has come to regard overshooting as unlikely to be occurring unless the exchange rate has moved a long way and, as noted, the move does not appear to be supported by economic and financial factors. This approach effectively means that the bulk of the RBA's intervention takes place around the cyclical highs and lows in the exchange rate.

In addition to circumstances where there appears to be misalignment, the RBA will also consider intervening in the market when conditions threaten to become disorderly. Persistent volatility, a sharp widening in bid-ask spreads or erratic movements of the exchange rate (especially at times of uncertainty about macroeconomic policy) may result in intervention to help restore order. Having said this, the RBA has become more comfortable with the ability of the market to cope with shocks of various types, so episodes when intervention is motivated by the desire to avoid disorderly conditions have become much less frequent.

Neither of the two reasons for intervention discussed above suggests that intervention could be used as an effective instrument of policy for achieving a particular level for the exchange rate. Nor does it imply the use of intervention to correct a monetary policy imbalance or to resist changes in the exchange rate which are in line with broader economic or financial developments.

In addition to intervention (i.e. transactions aimed purely at influencing the exchange rate) the RBA also undertakes more routine operations in the foreign exchange market, such as covering Government foreign exchange needs and rebuilding reserve holdings after periods of intervention. These give the RBA a fairly regular presence in the foreign exchange market, which allows it to gather information about the functioning of the market .

Related RBA reading:

• Profitability of Reserve Bank Foreign Exchange Operations: Twenty Years After The Float (Becker and Sinclair, 2004).

7. How Does the Reserve Bank Intervene?

When the RBA intervenes, it buys or sells Australian dollars against another currency, almost always the US dollar.5 To support the exchange rate at a time when it is depreciating, the RBA would sell foreign exchange and buy Australian dollars. If the RBA wanted to resist an appreciating exchange rate, it would buy foreign exchange and sell Australian dollars. The RBA has the capacity to deal in Australian dollars around the world, 24 hours a day.

As well as decisions about whether, and by how much, to intervene, the RBA also has discretion to vary the way the intervention is conducted and therefore the impact a given amount of intervention will have on the market. For example, if the RBA was intervening with the intention of influencing the exchange rate, it could enter the broker market directly through the electronic broker market. Because the broker market is the main mechanism used by interbank market participants to trade among themselves, knowledge of the RBA's presence in the market is immediately available to all active interbank players. They typically also inform their clients very quickly. This 'announcement effect' can itself have a significant impact on the exchange rate. Alternatively, if the RBA was intending to replenish foreign exchange reserves after a period of intervention, the aim might be to rebuild reserve holdings without having a significant impact on the exchange rate. Under these circumstances, the RBA might use an agent bank, so that the market as a whole is not aware of the RBA's presence. These are only two examples of the way in which the RBA can intervene in the foreign exchange market.

Intervention operations have implications for domestic liquidity. When the RBA buys Australian dollars, for example, there is a fall in the banking system's holdings of Australian dollars, thereby draining cash from the domestic money market. If the RBA took no further action, the market would be short of cash and domestic money market interest rates would tend to rise. This would be an example of unsterilised intervention. In effect, it would be a tightening of monetary policy since it leads to a rise in the cash rate.

The RBA can, of course, act in the domestic money market to replenish the banking system's liquidity by buying securities. This cancels, or 'sterilises', the liquidity effect of the intervention and leaves domestic interest rates unchanged. This is called sterilised intervention, and is the routine practice for central banks, unless they specifically set out to achieve a change in monetary policy. By using its domestic operations to keep cash rates around a target level, the RBA offsets excess demand for, or supply of, cash in the banking system whether it arises from intervention or from any other source.

At times of heavy intervention, this has the potential to cause substantial changes in the RBA's balance sheet as, for example, it sells US dollars in the foreign exchange market and sterilises this by buying domestic securities. To avoid the costs that can arise from this, the RBA uses foreign exchange swaps as the main vehicle for sterilising its intervention. In a situation where the RBA has bought Australian dollars and sold US dollars in its intervention operations, it subsequently undertakes a swap in which it lends Australian dollars and borrows US dollars. The settlement flows from the first leg of the swap offset those arising from the intervention transaction, and therefore remove the need for further operations to control liquidity. As each swap consists of a spot with an offsetting forward transaction, it does not alter the net balance of demand and supply for Australian dollars in the foreign exchange market, and therefore does not cancel out the effect on the exchange rate of the original intervention.

Related RBA reading:

• Profitability of Reserve Bank Foreign Exchange Operations: Twenty Years After The Float (Becker and Sinclair, 2004).

8. Intervention Since the Float

Over the post-float period, there have been three broad cycles of foreign exchange intervention reflecting the three cycles in the exchange rate: one in the second half of the 1980s; one in the first half of the 1990s; and one since 1997 (Graph 8). Each cycle began with the Australian dollar falling and the RBA selling foreign exchange (i.e. buying Australian dollars), with the position reversed during the second phase of the cycle as the exchange rate rose.

The timing of each cycle shown in the graph below is defined in terms of the major turning points in the RBA's cumulative foreign exchange position resulting from intervention.6 For instance, the initial phase of the first cycle extended from December 1983 to September 1986 during which the RBA was a net seller of foreign exchange with cumulative net sales peaking at US$6.0 billion in September 1986. From there through to September 1991, the RBA was a net buyer. Total purchases over that period amounted to US$11.7 billion, so that the cumulative position since the float moved from a net short position of US$6.0 billion in September 1986 to a net long position of US$5.7 billion in September 1991. The subsequent cycle saw net sales of foreign exchange between October 1991 and October 1993, and net purchases from November 1993 to September 1997, while the final cycle saw net sales from October 1997 to February 2002 and net purchases from March 2002 onwards.7

Graph 8

Click to view larger

The table below provides some summary statistics on the RBA's foreign exchange transactions in the market, for the entire period since the float, and in each of the three cycles of intervention outlined earlier.

Overall, in the 20 years since the float of the Australian dollar, the RBA has transacted with the market on 40 per cent of trading days. A large proportion of these transactions represent cover for Government business and, reflecting this, the RBA has transacted to purchase foreign exchange on about three times as many days as it has to sell foreign exchange. For the same reason, the average size of purchases is around half the average size of sales. Because of the change in the style of intervention after the 1980s, the figures for the latter period are quite different from those in the earlier period:

• the proportion of days on which the RBA intervened in the market declined after the 1980s. In the episode between 1983 and 1986, as the exchange rate fell, the RBA intervened in the market to buy Australian dollars on 40 per cent of days. In the subsequent two episodes when the exchange rate fell, the RBA intervened in the market on only about 21 per cent and 4 per cent of days, respectively;
• the average transaction size increased. The size of average daily sales of foreign exchange to support the Australian dollar in the 1983 to 1986 episode was about A$20 million, but rose to around A$155 million in the subsequent two episodes; and
• the maximum size of daily intervention increased. For example, the largest daily sale of foreign exchange was equivalent to about A$250 million in the 1980s episode, but about A$1,200 million in each of the subsequent two cycles.8

Since the float in 1983, the net position resulting from sales and purchases of foreign exchange by the RBA (including transactions with the Australian Government) has fluctuated around zero. This provides an approximate indication that the RBA has not systematically tried to support or weaken the Australian dollar over the floating period as a whole.

Related RBA reading:

• Profitability of Reserve Bank Foreign Exchange Operations: Twenty Years After The Float ( Becker and Sinclair, 2004).

9. Has Intervention Been Effective?

Evaluating the effectiveness of intervention is a difficult process because it cannot be known how the exchange rate would have behaved in the absence of intervention. In addition, in estimating the contemporaneous effect of intervention there is an endogeneity problem; that is, intervention is largely triggered by exchange rate movements, yet intervention can cause exchange rate movements. These difficulties have led to the development of a number of different methods of evaluating the effectiveness of intervention, two of which have been employed by Reserve Bank staff in recent years to evaluate the effectiveness of RBA intervention.

The first ( Kearns and Rigobon, 2003), which attempted to capture the interdependence of exchange rate movements and central bank intervention, supported the description of RBA intervention as 'leaning against the wind'. i.e. acting to slow or correct excessive trends in the exchange rate. Intervention was found to have a significant effect on the exchange rate, particularly on the day of intervention.

The second (Becker and Sinclair, 2004) used the 'profits test' to evaluate the effectiveness of intervention, as advocated by Friedman (1953).9 The application of the profits test relies on the central bank acting as a stabilising long-term speculator. If the objective of the central bank is to limit fluctuations in the exchange rate, this will tend to involve the purchase of the local currency (sale of foreign exchange) when the exchange rate is relatively low, and the sale of the local currency (purchase of foreign exchange) when the exchange rate is high. If the central bank is successful in buying low and selling high, its intervention should yield a profit. It follows from this that if a central bank has been profitable in its intervention, it must have bought low and sold high, therefore contributing to the stabilisation of the exchange rate.

The graph below demonstrates that during each cycle the RBA has been successful in buying low and selling high in its intervention activities. The horizontal lines represent the average exchange rate at which net transactions were undertaken for particular periods. The fact that the RBA's intervention activities since the float have been profitable provides an indication that the RBA's intervention has therefore worked to stabilise the exchange rate during each cycle. The wider gap between average buy and sell rates in the latter period on the graph also shows the change in the RBA's intervention stance, which now focuses on larger movements in the exchange rate.

Graph 9

Click to view larger

Related RBA reading:

• Profitability of Reserve Bank Foreign Exchange Operations: Twenty Years After The Float (Becker and Sinclair, 2004)
• Identifying the Efficacy of Central Bank Interventions: Evidence from Australia (Kearns and Rigobon, 2003)

Footnotes

1. See Meese and Rogoff (1983) ‘Empirical Exchange Rate Models of the Seventies: do they fit out of sample?', Journal of International Economics, 14(1/2), pp3–24. (back to text)
2. See, for example, related reading and references therein: Blundell-Wignall, Fahrer and Heath (1993), Beechey et al (2000), Stone et al (2005). (back to text)
3. See, for example, Evans and Lyons (2006) ‘Understanding Order Flow', International Journal of Finance and Economics, 11, pp 3–23. (back to text)
4. Chen and Rogoff (2003) ‘Commodity currencies', Journal of International Economics, vol. 60(1), pp 133–160. (back to text)
5. It will usually subsequently re-balance the various currencies it holds in order to restore the proportions in line with its foreign currency benchmark. For example, a sale of US dollars for Australian dollars will require a subsequent round of transactions to sell some euros and yen (the two other foreign currencies held) and buy US dollars so that the proportions of each currency held are restored to benchmark. (back to text)
6. Changes in the RBA's net foreign exchange position are measured as the net of transactions with the market, the Government, and all other counterparties. Foreign currency received as earnings on foreign assets has been included as a 'purchase' of foreign exchange. The RBA takes into account such earnings in calculating how much foreign exchange to buy in the market to cover sales to the Government. Earnings therefore influence the discretionary actions of the RBA. (back to text)
7. The exact dates of the cycles are as follows: Cycle 1 – sales from 12 December 1983 to 25 September 1986 and purchases from 26 September 1986 to 23 September 1991; Cycle 2 – sales from 24 September 1991 to 27 October 1993 and purchases from 28 October 1993 to 2 September 1997; Cycle 3 – sales from 3 September 1997 to 7 February 2002 and purchases from 8 February 2002 to 30 June 2004. (back to text)
8. Note from Table 1 that in the second part of the first cycle the appreciation in the exchange rate was temporarily disrupted by the October 1987 stock market crash, with an ensuing bout of intervention and a maximum daily sale of foreign exchange of A$1,025 million. (back to text)
9. Friedman, M (1953), ‘The case for flexible exchange rates', Essays on Positive Economics, University of Chicago Press, Chicago, pp157–203. (back to text)