Residents of the United States, please read this important information before proceeding
Please read this important information before proceeding.
The Australian dollar has been a star performer among the world’s ten big currencies (G10) over the past decade. Since 2003, it has risen around 71% against USD and 49% in trade-weighted terms. However, we believe its fortunes are likely to change. We see a weak AUD as the flipside to a strong USD: while the USD may gain pro-cyclical attributes and benefit from positive local economic news, the reverse should be true for AUD, which could lose its status as a “go-to” cyclical currency as markets become more concerned about the Australian economy. Figure 1 shows that AUD’s historically stretched correlation with global equity markets has already started to normalise. Australian assets generally may decouple from the ongoing increase in risk appetite.
Fundamentals? What fundamentals?
In our view, fundamentals suggest that the AUD is living on borrowed time. Investment in the country’s mining sector is likely to peak soon (and decline thereafter), while public demand is likely to slow due to planned fiscal consolidation and the exact path to a rebalanced economy remains uncertain. Our economists expect the economy to grow by 3% and 2.3% this year and next. While these rates are respectable compared to some G10 countries, they are not particularly high historically. The growth differential with the US is likely to fall to zero by 2014, below its 20-year average (Figure 2), calling to mind the late 1990s, when robust US relative economic performance drove the USD to outperform AUD (and most other G10 currencies). Probably the most striking part of Australia’s outlook is its persistent current-account deficit. Despite being the beneficiary of China’s strong growth this last decade, Australia still manages to import more than it exports. The deficit has averaged 4.3% of GDP since 2000.
As a result, Australia has become a net debtor to the rest of the world. Its net liabilities, at 58% of GDP, are the second largest in the G10 FX space, after New Zealand at 67%.
Figure 1: AUD correlation with risk has been normalizing
Figure 2: Australia-US growth differential to fall to zero
Figure 3: Portfolio flows form a bulk of foreign liabilities
Figure 4: 2-year rate spread suggests weaker AUD/USD
Admittedly, the third worst is the US, but its net liabilities are just 27%, and the USD is of course the world’s reserve currency (and will be for the foreseeable future).
Furthermore, the bulk of Australia’s net foreign liabilities (Figure 3) have been accumulated via portfolio flows (specifically debt), which tend to be more volatile and less sticky than other types of financing such as direct investment (which forms only a small part of its Net International Investment position). Foreign investors own around 70% of all government securities.1 This dependence on potentially volatile flows poses a risk to AUD over the medium to long term. Should investors turn bearish on AUD, a fall in the currency could be further magnified if these existing positions are liquidated.
“Australian stocks and bonds also look relatively unattractive”
What would change market views on AUD?
Most market participants appear well aware of AUD’s shaky fundamentals, but selling AUD has proved to be one of the ‘pain trades’ of recent years. Attractive rates, flows chasing remaining AAA assets (Australia still has its rating), underpinned commodity prices, and central-bank liquidity have all supported the currency. However, three of these four factors now look exhausted.
First, as Figure 4 shows, AUD/USD has massively decoupled from levels suggested by the relative rate spread. Second, foreign investors as noted may now be heavily exposed to AUD bonds, making any material increase in positions less likely. Third, we now see AUD decoupling from commodity prices, the outlook for which is in any case now less optimistic. The fourth factor, central bank liquidity, is the only one likely to remain intact.
Global liquidity, however, on its own is not enough to support assets/currencies that look fundamentally unattractive. The experience of the South African rand (ZAR) – formally a high beta currency – is a prime example. It is, therefore, not surprising that AUD correlation with risk has started to fall from overstretched levels.
Where does AUD fair value lie? And what drives it?
Our bilateral AUD/USD Behavioural Equilibrium Exchange Rate model (BEER) suggests a medium-term fair value of around AUD/USD 0.90 (Figure 5), a 7% overvaluation from current levels. Moreover, the details show that fair value has been, by and large, driven by an increase in Australia’s terms of trade, the ratio of export to import prices, relative to the US.2 This is depicted by the blue-shaded area in Figure 6. Australia’s exports are heavily skewed to commodities, and to commodities that we are not particularly upbeat on (see below). Hence, its terms of trade are likely to deteriorate, pushing estimated fair value down further, below its current AUD/USD 0.90, making AUD even less attractive.
Figure 5: AUD/USD trading above its BEER fair value
Figure 6: Commodities by and large caused AUD’s rise
Looking at even longer-term valuations based on inflation differentials only (such as purchasing power parity), AUD looks even more overstretched. Most valuation measures suggest downside risks. The OECD estimates AUD/USD fair value around 0.66.
Outlook for iron ore and China demand
Iron ore is Australia’s key commodity, accounting for more than 20% of its total exports. With China being the world’s biggest iron-ore importer (and Australia’s main export market), the outlook for Chinese demand remains pivotal for iron ore supply-demand dynamics (Figure 7). While we expect demand growth from China to remain healthy, we do not believe that it will be able to exceed the rate of supply growth in the market, which has recently increased in response to the higher prices that have followed decades of underinvestment in the sector.
The ongoing industrialisation and urbanisation process in China will continue to support demand for iron ore, but the government’s shift of focus from investment in infrastructure and fixed assets to domestic consumption should translate into a more moderate rate of demand growth for raw materials, including iron ore. The expectation that robust supply growth will not be completely absorbed by moderating demand growth suggests a more challenging outlook for prices. We expect that the market will remain oversupplied and prices are likely to trend lower, from the current value of about $120/tonne to perhaps around $100/tonne in coming years. With the outlook for
Australia’s most important commodity doubtful, the key factor that has supported AUD over the past decade is likely to diminish, leaving AUD vulnerable, we think.
Equity and bond markets
The Australian stock market is up by about 10% in local currency terms this year, almost as much as the other developed markets, but it has underperformed markedly in dollars, in which terms its price relative is currently at the lowest level since mid-2009. Although the resource sector accounts for a relatively large portion of the index compared to other stock markets, the sector’s recent underperformance (due to falling commodity prices) has been muted by a strong rise in financials, which form around half of the index. While the overall index valuation does not look overstretched (it is mildly expensive on a price-to-earnings basis, but mildly undervalued on a price-to-book basis – Figure 8), this large financial sector (it accounts for “only” 21% of the MSCI World index, and 17% of the S&P500), banking in particular, adds to our wariness of the equity market. Our equity views are unhedged, and so the likely fall in the AUD also counts against it. Property prices have looked frothy for a long time, and with other countries coming out of the other side of a real estate crisis we think there are more attractive returns to be had from banking sectors elsewhere – for example, in the US.
Australian bonds look historically expensive. The 10-year yield, at 3.3%, is higher than in the US or Europe, and is actually positive in real terms, but the spread to US bonds is the lowest since 2007. As noted above the investor base is largely international, and a fragile currency could hit demand. We do often hedge bond positions, but Australia’s higher interest rates make that expensive, effectively cancelling the yield attraction.
Conclusion: investment implications
We expect the decade-long AUD bull trend to come to an end. AUD materially overshot its fundamentals and with the last supporting factors easing the clock is ticking. We expect AUD/USD to fall towards the 0.90s levels on a one- to two-year time horizon.
But we recognize that it is expensive to short AUD against USD (it has the highest interest rates in the G10 FX space). Rather, we prefer doing so against the Canadian dollar (CAD). CAD’s higher rates make long positions less expensive. Moreover, for clients wishing to express the view via options, implied volatility on AUD/CAD is lower than on AUD/USD, and that implementation is cheaper too. We retain a constructive view on CAD and expect the currency to benefit from its exposure to the US economy and underpinned oil prices.
Generally, we would be underweight Australia relative to benchmarks in investment portfolios generally.
Figure 7: Australia’s ore exports to China
Figure 8: Local stock market: not a valuation call
1 The Reserve Bank of Australia, Statement on Monetary Policy, 10 May 2013
2 We do not include a relative productivity variable in the model (which we do for other currencies) as it lacks any explanatory power. This is because Australia’s labour productivity was virtually stagnant between 2003 and 2011, lagging the rise of AUD. This reflects a number of factors: investment in mining and utilities takes time to feed through to increasing output; the extraction of lower grade mineral deposits increases the difficulty of mining; environmental targets need to be met. In this respect Australia’s poor productivity performance differs from “Dutch disease,” because it was widespread, and included the mining sector itself.