Speech Recent Influences on the Exchange Rate
Address to CEDA Annual General Meeting Dinner
I am very pleased to have been asked to address CEDA's Annual General Meeting for the third time in Melbourne. When deciding on a topic for tonight, I realised that it would be very difficult to discuss the Australian economy without spending some time on the exchange rate and the various reasons advanced for its recent fall. So I have decided to plunge in and devote the whole speech to this topic. Obviously, we have been doing a lot of thinking about it, and I will see if I can say something useful about it in the 20 minutes I have available tonight.
What has Happened?
The exchange rate has behaved during 2000 in a way that no-one predicted. There used to be some elements of predictability for the Australian dollar in that its broad movements could usually be explained by changes in our terms of trade (or its close relative, commodity prices) and the difference between domestic and foreign interest rates. The relationship was not close on a month-to-month or quarter-to-quarter basis, but it could explain The exchange rate has behaved during 2000 in a way that no-one predicted. There used to be some elements of predictability for the Australian dollar in that its broad movements could usually be explained by changes in our terms of trade (or its close relative, commodity prices) and the difference between domestic and foreign interest rates. The relationship was not close on a month-to-month or quarter-to-quarter basis, but it could explain the large swings. This broad relationship is shown in Graph 1 which compares the predicted exchange rate and the actual exchange rate using an equation which has been used in the Bank over the years. (By the way, there is nothing unique about this equation as several of the large banks active in the foreign exchange market have also used a similar formulation.) What Graph 1 shows is that, on the basis of previous experience during the floating rate period, we could have expected a rise in the exchange rate in 2000. Instead, we have seen a fall and the gap between actual and predicted is over three standard deviations – the largest in the past fifteen years. Something therefore has changed, at least for the present.
Another way of making the same point is to say that the Australian dollar has never previously shown a significant fall with economic conditions as they presently are. Over the course of 2000, we have had a strong world economy, improving Australian terms of trade (rising commodity prices), a buoyant domestic economy, a declining current account deficit, a fiscal surplus and rising domestic interest rates. In the past, a significant fall in the Australian dollar has coincided with the opposite – a weak domestic economy and world economy, and falling commodity prices. It has also usually occurred against a background where there has been widespread public criticism of fiscal and monetary policy for being too expansionary. None of these elements is present on this occasion.
The comments I have made so far have treated the year as a single development, but I think it is possible to distinguish two phases. For the first seven months or so, the biggest influence on the exchange rate was the expectation of what was going to happen to Australian and US interest rates. The predominant view was that the strong US economy would mean that US interest rates would rise more than those in other countries (including Australia), and this contributed to a strong US dollar and falling Australian dollar. Any suggestion that the Australian economy was weakening (or the US economy strengthening) led to a fall in the Australian dollar. In the first seven months of the year, virtually all of the large movements in the Australian dollar could be explained this way. For example, on two occasions, a weak monthly retail trade figure caused the Australian dollar to fall by more than one US cent in a day. Our forthcoming Statement on Monetary Policy (due next Monday) examines this in detail.
The irony of this situation was that this process actually involved a misinterpretation of the Australian economy. The signs of weakness that were seized upon were anomalies; the underlying economy remained very strong, incipient inflationary pressures were starting to appear, and during this period our interest rates were raised, as it turned out, more or less by the same amount as those in the United States. Nevertheless, the misinterpretation prevailed and the currency fell. Over the six months from late January, the Australian dollar fell by 11 per cent against the US dollar and 8 per cent in trade-weighted terms.
The second phase occurred from around mid year, when it became widely expected by the market that the US economy was going to have a soft landing, and that no further increases in US interest rates were likely. As a result, the market's expectation of future interest differentials moved in a direction favourable to Australia. Had the foreign exchange market continued to behave as it did in the first half of the year, the Australian dollar would have risen, but it did not. It steadied for a while, and then fell appreciably from mid August. It was this period that convinced a lot of people that some new factor must be at work, which was overwhelming the normal determinants of the exchange rate. It was also in this period that the alternative explanations were put forward most volubly.
Before getting on to some of the newer explanations for the currency's movements, it is worth looking at some more conventional explanations.
The first of these is the strength of the US dollar itself, which has been the biggest reason behind our recent experience. This has affected us in a mechanical sense in that if the US dollar is rising, the currency on the other side of the bilateral exchange rate must fall. We have been part of this, as is shown by the fact that the Australian dollar has fallen more against the US dollar than in trade-weighted terms. There has also been an indirect effect in that the rising US dollar has affected the investment preferences of many market participants in an extrapolative way which I will discuss later. The US dollar in 2000 has risen against the currencies of all other industrial countries, although its rise against the yen has been relatively small. These two currencies – the US dollar and the yen – have been the strongest currencies in 2000, and they happen to be our largest trading partners and so have the largest weights in our trade-weighted index.
Interestingly, the newer explanations for currency movements are largely based on the assumption that countries with strong currencies will have characteristics similar to the United States (see later). But Japan is almost the opposite to the United States – weak economy, barely positive interest rates, massive fiscal deficit, trade surplus and an economic culture based on conventional manufacturing. I do not wish to dwell on these differences, other than to note that they show the dangers of coming up with a general explanation of exchange rate movements which turns out to be only applicable to one country or one period of time.
The second explanation relies on the well-known tendency for exchange rates to be driven by momentum. That is – once a trend has been established, for example, by the events in the first half of the year – it tends to continue and thereby to overshoot the level implied by fundamentals. There have been many documented examples of this, and market participants often place weight on this factor when explaining movements in exchange rates. I have no doubt that this has been an important factor over recent months, as it has been on several earlier occasions in our past. A related market rule of thumb that has appeared this year, but was entirely absent last year, is the one which relates movements in the Australian dollar to movements in the euro. The good thing is that these trends and rules of thumb do not continue indefinitely, and the further the exchange rate departs from fundamentals, the more likely it is that some unforeseen event will come along to jolt the rate into a new direction.
The third conventional explanation gives prominence to Australia's external position – the current account deficit and the accumulation of foreign debt and other liabilities. This explanation emerges whenever the Australian dollar is falling, and then goes into hibernation whenever it is rising. The problem with this explanation is that our external position has been relatively stable over the past decade or more, and so it is difficult to use it as the explanation for an event that only occurred as recently as this year. Over the past decade, the current account deficit has cycled around an average of about 4½ per cent of GDP, and our foreign debt has stayed a little over 40 per cent of GDP. It is true that total external liabilities to GDP has risen, but against this, the ratio of debt servicing to exports has halved over the past decade, and the current account deficit itself has been declining during the period of the falling Australian dollar. I think it is very difficult to make the case that our external position has recently deteriorated, or that it is the explanation for the falling Australian dollar in 2000.
While I am generally sceptical that this year's events can be explained by a sudden focus on trends in the current account position, it may be more promising to think about developments from the perspective of capital flows.
If, for one reason or another, investors found Australia a relatively less attractive destination for their funds, compared with some other destination – e.g. the United States – we would expect that this will have implications for the exchange rate. Some will want to assert that this occurred because people suddenly focused on Australia's current account, but this seems unsatisfactory because the recipient of the flows – the United States – is seeing its current account deficit expand to unprecedented size. In fact, as a share of GDP, Australia's current account deficit is well on the way to being smaller than its US counterpart, yet the US dollar has been rising, and the Australian dollar falling. So to make sense of what has been happening, it is necessary to think in terms of capital flows being driven by changes in the perceived returns in the different countries. In the remainder of this talk, I would like to examine this apparent shift in capital flows, and to think about whether it is temporary or permanent.
Attitudes of International Investors
There are two reasons put forward for why overseas investors may now have a smaller appetite for investing in Australian assets – the first applies to debt and the second to equity.
On Australian debt, whether at the short or long end or whether private or public, current interest rates are now lower than formerly was the case. For a good part of the past two decades, Australian interest rates were well above US rates, as we struggled to get Australian inflation back to a respectably low rate. Even when inflation fell, a number of years passed before the fall was finally reflected in inflationary expectations and interest rates. As a result, it was only over the past three or so years that Australian interest rates lost their significant premium over US rates (Graph 2).
The argument essentially says that, since Australian interest-bearing securities no longer contain a premium, investors' appetite for them has declined, and so the demand for Australian dollars to buy them has fallen. On the surface, this sounds plausible, but we should ask why have rates fallen. Surely they fell because Australian debt became more attractive to investors. Whereas in 1994 rates of 300 basis points above US rates were required to get investors to hold our bonds, they now do so willingly at rates only slightly above US rates.
Clearly, the simple form of this argument has a few logical gaps. However, there does seem to be something at work here – perhaps due to differing appetites for Australian bonds on behalf of local compared with foreign investors – because there has been a reversal in recent years of the former high levels of capital inflow into Australian bonds. In the five years between June 1992 and June 1997, these inflows averaged $4.3 billion per year, whereas over the past two years there have been outflows of $5.2 billion per year.
The second argument, namely that there has been a change in preference by equity investors, is a relatively recent one, but it has been put forward very forcefully. It essentially says that equity investment (direct and portfolio) is less attracted to Australia than formerly since we are viewed as an ‘old economy’ because we do not have a big enough exposure to the new growth areas, particularly the information and communications technology sector (ICT). The contrast is drawn between the United States, which is seen as the ‘new economy’, and other countries, including Australia and Europe, which are seen as the ‘old economy’.
Let me say at the outset, there are two questions at issue – first, do a substantial number of providers of capital think this way? – and second, is there economic substance to the contrast between the two types of economy? There can be little doubt that the answer to the first question is in the affirmative – this line of thinking is very widespread at present, and therefore has no doubt been an important recent influence on our exchange rate. The answer to the second is, in my view, in the negative – the differences between the two types of economy – ‘old’ as applied to Australia and ‘new’ as applied to the United States – have been greatly exaggerated and the similarities ignored. This is important because if there is not fundamental economic substance to the view, it will not last – it will be seen as a phase through which capital markets passed. I would like to spend the remainder of my time tonight explaining why I believe this is so.
The thing that stands out about the economic performance of Australia and the United States over the past decade or the past few years is not the contrast, but the remarkable similarity. For an investor, the most likely place to make money is where there is strong growth and a rapid increase in productivity. The United States is justly proud of its performance in these two areas, but the one comparable country in the OECD area that can match it – in fact, exceed it – has been Australia. I have shown the growth and productivity figures before, so I will not repeat them again, but will rely on a US source to make my point.
An article in the US Federal Reserve Board's October Bulletin sets out to explain the remarkable recent pick-up in US productivity and to contrast it with 16 other OECD countries. It is an excellent and thoroughly researched article which broadly achieves its objectives, but it keeps finding an irritating exception. Whether it is growth of labour productivity, growth of multi-factor productivity or the extent of capital deepening, Australia matches or exceeds the excellent US performance, rather than sharing the lacklustre performance of the majority of other countries.
But proponents of the ‘new economy’ view would reply that they are not interested in the whole economy, they are interested primarily in the ICT part where the United States unquestionably is the world leader. But only 3.9 per cent of the US workforce is employed in the ICT sector (the comparable figure for Australia is 2.6 per cent). Why you should make an assessment of a country by concentrating on the 3.9 per cent and ignoring the other 96.1 per cent is a question I cannot answer. On a similar note, the willingness to adopt new technology is surely the key to improving productivity, and here there are a number of indicators that place Australia very high in the world rankings. A recent OECD publication on this subject finds that, among the 27 or so OECD countries, Australia is:
- third highest in ICT expenditure as a percentage of GDP;
- sixth highest in PC penetration;
- eighth highest in internet hosts per 1000 population;
- third best in internet access cost; and
- third best in secure web-servers for electronic commerce.
Not surprisingly, we do not match the United States in most of these measures, but we are almost always in the top quartile or higher, suggesting that among OECD countries we should be regarded as being towards the top end of the range in willingness to embrace new technology.
Again, the proponents of the ‘new economy’ view find this reasoning unconvincing. They tend to concentrate on the production of the ICT sector, and give Australia low marks for not having more resources in this area. While there is no doubt that the very sophisticated fringe of ICT, as exemplified by Silicon Valley, would be an asset to any country, most of the production side of ICT is much more humdrum. In fact, the country with the highest share of GDP derived from ICT is not the United States; Korea and Hungary, for example, both have higher shares. The United States has outsourced a good deal of the production to less sophisticated countries with lower cost structures, with the result that the United States now has the largest deficit in trade in ICT of all countries. This, I think, helps make the point that a policy of trying to direct resources into ICT, particularly its production at the expense of other aspects of the economy, could be self-defeating. It would be all the more disappointing if such a policy were sold on the grounds that it would be a constructive response to the current fall in the exchange rate.
Perhaps the biggest weakness with the ‘new economy’ argument as an explanation of exchange rate movements can be found in the recent behaviour of the US dollar and US investments. It is frequently asserted that the reason the US dollar is rising is because everyone wants to buy US equities, particularly tech stocks, because that is where the high returns are. In fact, anyone who followed this approach this year would have been disappointed. Not only were they buying into something with an already high price and low yield, they would have seen the prices fall noticeably over the year – the Dow by 5 per cent and the NASDAQ by 16 per cent. (In Australia, the All Ordinaries has risen over the same period.) To the extent that the US investments made positive returns, it would have been only due to the appreciation of the US dollar itself, not to the assets purchased, which turns the underlying reasoning on its head.
The Australian dollar has shown some big swings in both directions since it was floated in 1983. This is the fourth time in that era that the Australian dollar has fallen by more than 10 per cent in trade-weighted terms in a year. On each of the earlier occasions, it seemed at the time that the fall would go on forever, and there were plenty of people who had explanations to support its continuation. However, on each occasion, it recovered and it will do so again this time.
The principal reason for the fall over recent months has been the rising US dollar, which has affected us directly and indirectly. Another reason is to be found in the interaction between the early fall based on expectations of interest differentials plus the intrinsic tendency for foreign exchange markets to continue to go in one direction, i.e. to overshoot, and to come up with explanations to justify the overshooting. Finally, over recent months, there has been a tendency for markets to judge countries and their exchange rates by the ‘new economy model’. Ironically, this is exactly the period when US technology stocks have languished.
I should now conclude by saying something about the implications for monetary policy. Monetary policy will continue to be conducted according to the medium-term principles contained in our inflation-targeting approach. This means that the exchange rate will be taken into account, along with the other variables that contribute to our inflation outlook (where this is appropriately defined to exclude once-off factors such as petroleum prices and the GST). On the other hand, as can be seen from our current behaviour, we have no intention of departing from our medium-term approach in an ad hoc attempt to push up the exchange rate for its own sake.
The period ahead will contain many uncertainties, both here and abroad, but we have coped with uncertainty before. Over the past year, we have had the unusual combination of strong domestic growth, a falling current account deficit and a declining exchange rate, all of which are conducive to future growth. We still have to ensure that the medium-term outlook for inflation stays on track – which means making sure that several temporary factors have only a ‘once-off’ effect and are not reflected in the ongoing inflation rate. If we succeed in doing that, we will be in a good position to continue the expansion we have had since 1991.
For example, the US dollar in 1989, the Japanese yen in 1995, in a strengthening direction, and then in 1998 in a weakening direction. 
Net equity flows into Australia in 1999/2000 (only data for the first three quarters are available) appear to have been less than in the two preceding years. 
CJ Gust and JR Marquez, ‘Productivity Developments Abroad’, Federal Reserve Bulletin, October 2000. 
The Knowledge-based Economy: A Set of Facts and Figures, OECD, Paris, 1999. 
Measuring the ICT Sector, OECD, Paris, 2000.