IMF stress tests China/Australia bust


With the Gittins! campaign for national ignorance going swimmingly, it is not at all surprising that we must look offshore for a sensible risk assessment of Australia vis-a-vis China.

On cue, as a part of its annual report on the Australian economy, the IMF has produced a three part stress test of the Australian fallout in the event of a China bust. So far as I can tell it has gone completely unreported.

The first of the scenarios examines the outcome if China is able to do what so far it has been unable to do, rebalance towards internal demand drivers for growth:


…we assume that a successful rebalancing of China’s growth toward private consumption comes about by a comprehensive set of
policy measures that includes exchange rate appreciation, financial sector reform, structural reforms, and a stronger social safety net.

4 All these measures support private consumption and shift resources from the tradables to the nontradables sector while slowing the build-up of export capacity through private investment.
Exchange rate appreciation: The nominal exchange rate appreciates against the U.S. dollar by 20 percent over five years. Because prices in China are sticky, the nominal appreciation translates into a persistent real effective appreciation of about 16 percent over five years.

  • Financial sector reform: Interest rate liberalization will result in a higher cost of capital. At the same time households and SMEs are assumed to get increased access to financial services.
  • Fiscal reform: Government spending is re-oriented toward the social safety net, affordable housing and the provision of health and education services in a way that is neutral for the budget balance. As a result of fiscal and financial sector reform, households save less.
  • Structural reforms: Reforms facilitate a reallocation of capital and labor from the tradables to the nontradables sector.

5. As a result, the model suggests that China’s annual GDP growth rate would fall by 1½ percent and the level of GDP would be about 7 percent below baseline after five years. However, Chinese consumers enjoy improved terms of trade, which raises their welfare. Domestic demand increases by about 8 percent of GDP while net exports fall. The current account surplus narrows by 4 percent of GDP over five years.

6. Rebalancing toward domestic demand in China is mirrored by rebalancing toward external demand in the rest of the world. Nevertheless, global GDP remains about ¾ percent below baseline after five years, as structural adjustment in the global economy takes more than five years to complete. Commodity revenues in Australia remain little changed as the lower demand from China is largely compensated for by higher demand elsewhere.

7. The gain in competitiveness leads to an increase in noncommodity exports of goods and services and the current account improves by about ¾ percent of GDP after five years. Real GDP is modestly higher, however, nominal GDP falls ¾ percent below baseline because of lower terms of trade. Government revenues from commodities are lower while government expenditures change little. As a result, the public debt to GDP ratio increases, though not by much.

As regular readers will know, this is the scenario presented often by Michael Pettis. Only on his reading China slows to an annual growth rate between 3% and 5% as the subsidies that support high growth via fixed investment unwind. There are two other large unknowns in this modeling too. The first is the assumption that the growth that is redistributed to deficit nations is as commodity-centric as that in China. This seems a bold assertion to me given Western nations have a much more mature infrastructure matrix than China does. Second, any major reduction in the Chinese surplus necessarily means a major reduction in the availability of capital to deficit nations. Which means interest rates will rise. Not necessarily bad but one wonders what the impact would be on the still wildly indebted West.

Given the likelihood that commodity supply will have caught up with bubble expectations in five years as well, my view is that the IMF overestimates the likely prices of commodities in the scenario. Still, a slow wrestle with the current account deficit is something we can handle.


Scenario two is a more severe a one year slowdown in Chinese growth arising from a real estate bust:

…we assume a decline in investment drives a slowdown in China, possibly because of problems in the real estate market or some financial market disturbances.

6. GDP growth falls to around 6 percent for one year and the level of real GDP returns to baseline after five years, with 1 percent higher-thanbaseline growth in each of the four years following the growth shock.

8. We assume no large discretionary fiscal policy response from the Chinese authorities. If the Chinese authorities were to respond as they did following the Global Financial Crisis, by increasing public spending on infrastructure, it would likely cushion the impact of the slowdown on demand for commodities.

9. Lower growth in China leads to a persistent fall in global commodity prices by about 13 percent. While this hurts commodity exporters, it benefits commodity importers and reduces the impact of slower Chinese growth on the global economy.

10. The impact on the global economy is mitigated by automatic fiscal stabilizers in Australia and other countries. However, with limited fiscal space, many advanced countries may not have the policy flexibility to let automatic stabilizers work fully. To illustrate the role of fiscal stabilizers, two scenarios were run: one in which automatic stabilizers are allowed to work fully (full policy flexibility) and another where automatic stabilizers are reduced in Europe, Japan, and the United States (limited policy flexibility).

7 In Japan and the United States, policy interest rates are at the lower bound and cannot be reduced. China’s renminbi is assumed to be pegged to the U.S. dollar.

11. Australia suffers a terms of trade shock. The size of the impact, however, depends significantly on the degree of policy flexibility elsewhere.

  • With limited policy flexibility elsewhere, Australia’s terms of trade falls by about 10 percent in the first year, relative to baseline. The impact of the fall in global commodity prices on the terms of trade is partly offset by a fall in the price of commodity imports, including oil. With full policy flexibility in advanced countries, the fall in Australia’s terms of trade would be cut by two-thirds.
  • Real GDP falls by about ¾ percent relative to baseline, because of lower demand from China and the negative impact of the Chinese shock on global demand. However, with policy flexibility in advanced countries the fall would be less than ¼ percent. The reduction in commodity prices amplifies the negative impact on nominal incomes, with nominal GDP falling by 2½ percent in the case of limited policy flexibility (1½ percent with full policy flexibility).
  • Government revenue in Australia falls directly, through a decline in resource taxes and company income taxes, and indirectly
  • through lower economic activity. We assume public expenditure is reduced gradually to balance the budget. As a result, transitory deficits add to public net debt which rises by about 3 percentage points of GDP over two years.

12. A depreciation of the Australian dollar helps buffer the shock, as do cuts in the policy interest rate. The exchange rate depreciation redirects exports to non-commodities and reduces imports. The contribution of net exports to GDP growth in real terms improves. However, the depreciation is not strong enough to fully offset the impact of lower commodity prices on the nominal trade balance, which worsens by about 1½ percent of GDP. Because about half of private earnings from commodities accrue to foreign shareholders, the drop in the current account balance is smaller than the decline in the trade balance.

Clearly, we live in the policy constrained world, so that’s the scenario we need to consider. The scenario as painted looks OK, although a 13% delcline in commodity prices looks awfully sanguine given the degree of perfection still priced into the bulks, iron ore and coal. Moreover, the IMF has not factored in any second round effects. For instance, in the event of a terms of trade shock, there will be a major blow to confidence, consumption will retrench, house prices come under pressure and bank funding costs rise. The IMF is assuming a 3% of GDP fiscal response (somewhere around $35 billion) which looks like the outer limit to me, leaving the nation’s public debt above 50% of GDP, and making ratings agencies very nervous. It doesn’t mention interest rates but it’s safe to say, they’d be at record lows.


Still, a rapid bounce back in China on stimulus (which would come) may bail our long term prospects out before we got the ratings cut.

The final scenario modeled by the IMF is a Chinese export bust, based upon a Western recession:

This scenario represents the tail risk of a confidence crisis triggered by high sovereign debt in advanced countries, particularly Europe. It assumes higher risk premiums on government bond rates (75 basis point area wide) and a risk premium on the exchange rate, leading to a real effective exchange rate depreciation in the Remaining Countries block. A permanent deleveraging by private households and a two-year discretionary reduction in the fiscal deficit are also assumed (1¾ percent of GDP). In the United States, households deleverage as well, but no discretionary fiscal tightening was assumed. Government bond rates rise in all countries, except China.

14. As a result, advanced economies enter into a recession led by a decline in investment. Global GDP falls by about 3 percent. China’s output declines most, hit by the reduction in imports in advanced countries directly and a related fall in investment. International commodity prices fall by about 25 percent. The terms of trade in commodity export countries worsen, and improve in those countries which are net commodity importers. Lower commodity prices support consumption in the latter countries.

15. Macroeconomic policy flexibility is assumed to be limited in most advanced countries. Automatic fiscal stabilizers in Europe and the United States are reduced to ⅓ of their OECD estimates, and no fiscal space is assumed to be left in Japan. Again, in Japan and the United States, policy interest rates are at the lower bound and cannot be reduced and China’s renminbi is assumed to be pegged to the U.S. dollar.

16. GDP in Australia falls too, but by less than in the other regions, despite a large decline in commodity prices. This is largely due to a strong policy reaction and a flexible exchange rate and sizeable foreign ownership in mining companies:

  • Policy interest rates are reduced by more and earlier than elsewhere and the exchange rate depreciates against all other countries except the Remaining Countries block (Europe). Real net noncommodity exports improve and mitigate the impact of the
  • worsening commodity balance on the wider economy.
  • Automatic fiscal stabilizers in Australia are allowed to work fully by keeping expenditures on the planned path while temporarily absorbing shortfalls in government revenues, including from commodities, in a larger fiscal deficit. Net debt rises by about 10 percent of GDP after five years. This cushions the blow to domestic incomes and demand.
  • Mining profits are squeezed, as commodity prices fall. However, a large share of foreign ownership limits the impact on the domestic economy and the current account. Nevertheless, lower profits depress investment in Australia.

Hmmm, no wonder it hasn’t been reported. Given the near universal garbage floating around that China is decoupled from a Western slowdown, there’d hardly be enough egg to go around.

Much the same comments about the national debt made above apply here to.

I don’t wish to be too alarming. These are stress tests and scenarios not yet reality. But, there is logic in the thought that we currently face the possibility of the final two scenarios happening simultaneously. That is, a Western recession triggered by European and US austerity (not to mention financial tumult) and a Chinese real estate pop.


Finally, the IMF admits itself that its commodity models are not a perfect fit for Australia:

The commodity basket in GIMF includes energy, minerals, and agriculture commodities, while Australian commodity exports to China are mainly iron ore and coal, which represent a larger input share in Chinese output than elsewhere. The simulations are therefore likely to underestimate the negative impact of rebalancing in China on the demand for Australia’s key commodities.

I suggest you read the document in full.

About the author
David Llewellyn-Smith is Chief Strategist at the MB Fund and MB Super. David is the founding publisher and editor of MacroBusiness and was the founding publisher and global economy editor of The Diplomat, the Asia Pacific’s leading geo-politics and economics portal. He is also a former gold trader and economic commentator at The Sydney Morning Herald, The Age, the ABC and Business Spectator. He is the co-author of The Great Crash of 2008 with Ross Garnaut and was the editor of the second Garnaut Climate Change Review.