Australia Investment Report 2010 — Three Booms in a Trench Coat

Australia came through the crisis without a technical recession, and the explanation is usually given as "China." That is directionally right and analytically useless, because it obscures a sequence with very different implications at each stage.

A resource cycle has three phases:

Phase one is price. Demand rises and, because new supply takes years to build, prices rise sharply. This phase requires no investment and no additional output — the same volume of ore earns far more. It is pure income, arriving immediately, concentrated in existing producers and in government revenue.

Phase two is investment. High prices justify building new capacity. This phase is enormously labour- and capital-intensive — construction, engineering, logistics, accommodation. It is where the employment boom happens, and it is spread far more widely through the economy than the price phase.

Phase three is production. The new capacity comes online and volumes rise. This phase employs relatively few people, because modern extraction is capital-intensive. And it puts downward pressure on prices, since the whole industry's new capacity arrives at once.

The critical feature is that these do not coincide. Prices can peak while investment is still rising; investment can peak while production is still climbing. So the economy experiences a sequence of transitions, each with different winners.

The employment consequence is the one that matters most and was least anticipated: the labour-intensive phase is the middle one. When investment peaks, employment in the sector falls sharply even as output and export volumes are still rising. An economy can be setting export records while shedding resource jobs, which reads as a contradiction and is simply the sequence.

The second theme is that the real external vulnerability was not the commodity exposure. It was that the banking system funded a large domestic mortgage book partly through offshore wholesale markets — the structure that failed in the UK Investment Report 2008, in an economy with high household debt and expensive housing.

Three booms, three sets of beneficiaries

The phase distinction is worth developing because it determines who gains and when.

The price phase benefits:

• Existing producers, whose revenue rises with no additional cost.
• Government revenue, through resource taxes and company tax.
• The currency, which appreciates on the improved trade position.
• And national income broadly, through the terms-of-trade effect discussed below.

It does not create many jobs, because the same mines and the same workers produce the same volume at a higher price.

The investment phase benefits:

• Construction, engineering and services firms, often far from the resource regions.
• Labour, intensely — wages in relevant trades rise sharply and workers relocate.
• Equipment suppliers and importers.
• And regional economies hosting the construction.

This phase is where the boom is felt by most people, and it is the phase most easily mistaken for the whole cycle.

The production phase benefits:

• Producers, through volume.
• Export statistics, substantially.
• And relatively few workers, since operating a completed mine requires a fraction of the labour that building it did.

It also puts downward pressure on prices, because every producer's expansion completes in a similar window — the supply response the Global Investment Outlook 2014 describes.

The boom that creates jobs and the boom that creates export volumes are different booms, several years apart. An economy can be in the second while commentary is still describing the first.

The policy difficulty is that each transition requires a different response, and the transitions are visible only in arrears — investment intentions data exists, but the peak is identified after it has passed.

Income without output

The terms-of-trade effect is the least-understood element of the period, and it explains why headline growth statistics missed the scale of what was happening.

GDP measures the volume of output produced. It does not measure what that output is worth in terms of what it can buy.

A terms-of-trade gain — export prices rising relative to import prices — increases purchasing power without increasing output volume. The same iron ore now buys more imported goods.

So the economy's real income rises faster than its real output. The gap is genuine and can be very large during a price boom.

Where the additional income goes:

• Company profits, largely in the resource sector.
• Government revenue, through profit-based taxes.
• The currency, which appreciates — which is itself a mechanism for distributing the gain, since a stronger currency makes imports cheaper for every household.
• And wages, in sectors competing for the same labour.

The measurement point is practical: real gross domestic income is the right statistic for a commodity exporter's living standards, and it diverges substantially from GDP during price booms and busts. Both are published free by the national statistical agency, and the divergence is the terms-of-trade effect made visible.

The corollary is uncomfortable. A terms-of-trade reversal reduces income without reducing measured output, so an economy can experience a genuine decline in living standards while GDP growth remains positive. The statistics that people follow will say nothing is wrong. The Asia-Pacific Investment Report 2014 covers the reversal.

Two speeds, one currency

The "two-speed economy" was the period's dominant domestic debate, and its mechanism is a relative price adjustment rather than a policy failure.

The mechanism:

1. Resource exports boom, so the currency appreciates.
2. A stronger currency makes all other exports less competitive and all imports cheaper.
3. So manufacturing, tourism, education exports and import-competing sectors all face worse conditions — not because anything changed in those sectors, but because the exchange rate changed.
4. Meanwhile resources and related services boom.

This is the standard resource-driven adjustment, and it is genuinely a reallocation rather than a loss: capital and labour move from the sectors facing worse relative prices to the sector facing better ones. That is what a price signal is for.

Why it is nonetheless painful and politically difficult:

• The reallocation is geographic. The booming sector is in different regions from the declining ones, so it requires people to move rather than merely change jobs.
• It is skill-specific. A manufacturing worker does not become a mining engineer.
• It may not reverse. A closed factory does not reopen when the currency falls back — industrial capability, supplier networks and skills are lost in a way that is expensive to rebuild. This is the hysteresis argument, and it is the strongest case for intervention.
• And the sectors losing out are frequently the ones with long-term growth prospects, which is what makes the loss potentially permanent rather than cyclical.

The policy bind:

A single interest rate and a single exchange rate cannot serve two sectors moving in opposite directions. Every instrument that helps one damages the other, which is why the debate was unresolvable rather than merely unresolved.

The instrument that partially escapes the bind is a sovereign fund receiving resource revenue and investing it offshore — which reduces the currency appreciation directly, since the revenue is not converted domestically. This is the stabilisation fund design the Asia-Pacific Investment Report 2014 describes, and its currency effect is often overlooked in favour of its fiscal one.

The exposure nobody was discussing

The commentary of the period focused almost entirely on commodity price risk. The more acute vulnerability was in the banking system's funding structure.

The structure:

• Australian banks held large mortgage books, reflecting high house prices and high household debt.
• Domestic deposits did not fully fund this lending, leaving a gap.
• The gap was funded partly in offshore wholesale markets, much of it at relatively short maturities.
• Which means regular refinancing in markets outside the country's control.

This is precisely the structure described in the UK Investment Report 2008, with the same properties:

• The assets were sound. Australian mortgages performed well throughout.
• The funding was the exposure, and it could close for reasons entirely unrelated to Australian housing.
• And the amounts were large relative to the economy, per the banking-assets-to-GDP framework in that report.

Why it did not break in 2008–2010:

• A government guarantee of bank wholesale funding was introduced during the crisis, which allowed banks to keep refinancing.
• The domestic economy did not deteriorate, so credit quality never became a question.
• And the guarantee was credible, because the sovereign's fiscal position was strong — the credibility condition from the Global Investment Outlook 2012.

The lasting point is what the guarantee revealed: the state stood behind the banking system's offshore funding. That commitment existed before it was stated, and stating it was what made it work.

The mitigation over subsequent years — longer funding maturities, higher deposit shares, liquidity requirements — addressed the structure directly, and is one of the clearer cases of a lesson being acted on rather than merely noted.

What the sequence means for positioning

Pulling the phases together produces a framework that generalises to any resource economy.

During the price phase:

• Producer margins expand fastest, since costs are fixed and revenue rises.
• Government revenue surges, creating the fiscal illusion the Asia-Pacific Investment Report 2014 describes.
• The currency appreciates, which hurts every other tradeable sector.
• Equity in producers outperforms, and the effect is levered because costs do not move.

During the investment phase:

• Services, engineering and equipment suppliers benefit, often more reliably than producers.
• Producer margins compress, because capital spending is consuming the cash flow.
• Labour costs rise economy-wide, squeezing every sector competing for the same workers.
• And the current account may deteriorate, since capital equipment is imported.

During the production phase:

• Volumes rise and prices fall, so revenue growth is far below volume growth.
• Employment in the sector declines even as output rises.
• The currency weakens, which finally relieves the other tradeable sectors — years after they needed it.
• And producer returns depend on position on the cost curve, since price is now set by the marginal producer, per the Global Investment Outlook 2014.

The single most useful indicator is capital expenditure intentions in the resource sector, published by the statistical agency ahead of the spending. It identifies the investment phase's turn before employment responds, which is the transition that surprised the economy most.

The housing position underneath it all

Running beneath the resource cycle was a housing market that had appreciated for two decades without a serious correction, and the structural reasons are worth separating from the cyclical commentary.

What made Australian housing structurally expensive:

• Population growth was high by developed-economy standards, driven substantially by immigration, and concentrated in a small number of coastal cities.
• Supply was constrained by planning, land release processes and the geography of those cities.
• Tax treatment favoured leveraged property investment relative to other assets, which raised demand from investors as well as occupiers.
• And credit was available and cheap, funded partly through the offshore wholesale markets described above.

Each of these is a demand-side or supply-side factor operating continuously, which is why the appreciation was sustained rather than a spike. A market where demand grows steadily and supply cannot respond produces rising prices as an equilibrium outcome, not as a bubble.

Why this matters for the resource analysis: the two are connected through the banking system and through household balance sheets.

• Mortgages were the banks' dominant asset, so bank health was a function of housing.
• Household debt to income rose to high levels internationally, which makes consumption sensitive to interest rates and to house prices.
• And the resource boom supported incomes and migration, which supported housing demand.

So a resource downturn transmits to housing through employment and migration, and housing transmits to banks through collateral — the loop the Asia-Pacific Investment Report 2010 describes.

The commodity exposure, the banking system and the housing market were not three separate positions. They were connected through household income at one end and bank collateral at the other.

Why the correction repeatedly did not arrive when widely predicted is instructive: the structural supports did not weaken. Population growth continued, supply remained constrained, and rates fell rather than rose. Predicting a correction requires identifying which support fails, not merely observing that prices are high — a distinction the archive returns to whenever valuation alone is offered as a thesis.

What an allocator could act on

Separate the three phases before taking a resource-economy position. Price, investment and production benefit different sectors, arrive years apart, and end at different times.

Expect the employment peak to precede the output peak. The construction phase is labour-intensive and the production phase is not, so job losses coincide with record export volumes.

Use real gross domestic income, not GDP, for a commodity exporter. The terms-of-trade effect changes purchasing power without changing output volume, and both series are published free.

Read the two-speed economy as a relative price adjustment. No single interest rate or exchange rate can serve sectors moving in opposite directions, which is why the debate had no resolution.

Watch bank funding structure over commodity exposure. The offshore wholesale funding of a domestic mortgage book was the more acute vulnerability and attracted almost no commentary.

Track capital expenditure intentions as the leading indicator. They identify the investment phase's turn ahead of the employment response, which is the transition that does the most damage.

What 2010 established

• A resource cycle has three sequential phases with different beneficiaries and different endings.
• The labour-intensive phase is the middle one, so employment peaks years before output does.
• A terms-of-trade gain raises income without raising output, which standard growth statistics do not capture.
• The two-speed economy was a relative price effect, with hysteresis as the strongest argument for intervention.
• Offshore wholesale bank funding was the real external vulnerability, and the crisis guarantee revealed a commitment that already existed.