From China Financial Markets Newsletter
Michael Pettis
Seven steps to crisis
Step 1. The German savings rate rises.
Thanks to the labor reforms, German wage growth dropped sharply, and with it the household income share of Germany’s GDP. As the household income share declined, and because the household savings rate remained flat, the household consumption share also declined.
Because most consumption is household consumption, and Berlin did nothing to counteract the decline in household consumption, for example by increasing social spending, the total consumption share of GDP in Germany also declined during this period
All goods and services produced by an economy (i.e. GDP) are by definition either consumed or saved. As Germans consumed a smaller share of all the goods and services they produced, by definition they saved a larger share. Germany’s saving rate, in other words, rose.
Note that Germany’s savings rate rose not because German households decided to become thriftier but rather because the growth in wages and overall income, and therefore in consumption, was suppressed by the Hartz reforms.
Step 2. Weak consumption growth in Germany lowers private sector investment.
Businesses expand production capacity usually because they expect demand for their product to grow. As consumption growth lagged, however, German businesses responded by reducing their investment growth rate to below the GDP growth rate.
When investment is equal to savings, everything that is produced is purchased either for consumption or investment purposes. The demand generated by investment absorbs all the goods and services that an economy has produced but has not consumed. In Germany however savings rose faster than planned investment.
Step 3. As German savings began to grow faster than GDP, and planned German investment more slowly, there were four different ways in which the German economy could adjust: it could run a current account surplus, unemployment could rise, public sector investment could rise, or something could set off a rise in consumption faster than the rise in GDP.
It could run a current account surplus. Germany could sell to foreigners the excess of goods and services it produced at home but could not sell in Germany. In that case Germany would also export the excess of savings over investment, and so run a capital account deficit. The value of the current account surplus would be identical to the value of the capital account deficit, with excess German saving financing the purchase by foreigners of excess German production.
Unemployment could rise. German businesses could respond to growing inventories by cutting back production and firing workers. This would cause the total production of goods and services to drop. Total consumption would also drop as unemployed workers reduce their consumption, but because unemployed workers must still consume, the total production of goods and services would drop faster than total consumption, or, put differently, the savings share of GDP would drop. Rising unemployment, in other words, reduces the savings rate, not because consumption rises faster than GDP but rather because consumption declines more slowly than GDP.
Public sector investment could rise. The German government could borrow the excess savings and invest it in ways that increase German productivity. Higher government investment would counterbalance higher savings and the decline in private sector investment so that German savings would no longer exceed German investment.
A consumption boom could be set off. German investors could use excess German savings to buy assets like stocks or real estate. As asset prices rise, and the owners of these assets see their wealth rise, two things can happen. First, the “wealth effect” can cause Germans to increase their consumption faster than their income increases. Second, rising asset prices can cause them to increase their investment, especially in real estate. The first effect causes the savings share of GDP to drop and the second causes the investment share to rise, either of which retains the balance between savings and investment.
Step 4. In fact Germany ran current account surpluses and capital account deficits, mainly with other European countries.
For policy reasons Berlin and local German governments did not absorb the excess savings by borrowing and investing them locally.
During this time European financial authorities were encouraging a convergence in interest rates among very different European countries with very different interest-rate histories, creating a sharp division between some countries, such as Germany, in which inflation had been low and interest rates historically low, and other countries in which inflation had been high and interest rates historically high, too. In the former the real cost of this available German capital was high, while in the latter, which we will call “Spain” in the rest of this essay, the real cost of this available German capital was low.
Every economy consists of business and household entities with a wide range of risk appetites and optimism about the future, ranging from recklessness to prudence. When real interest rates are very low or negative in a growing economy, because of the diversity among businesses and households in their appetite for risk, inevitably some of them will borrow more than they otherwise would have in order to consume or invest more than might have been prudent. If we assume that Spain has exactly the same distribution of risk-seeking borrowers as Germany, and exactly the same distribution of reckless agents as Germany, and we offer Spanish and German borrowers equal access to capital, but at high real rates to the Germans and at low or negative real rates to the Spanish, most of the excess German savings will flow to Spanish borrowers and very little will flow to German borrowers. The difference between the two countries, in other words, is not in the distribution of reckless behavior but in the divergence in real interest costs.
Some commentators have argued, bizarrely enough, that the causality is backwards, and that in fact all of the countries that we represent by Spain embarked on ill-advised consumption binges that occurred only coincidentally at the same time as Germany implemented the Hartz reforms. It was the subsequent collapse in Spanish savings rates, these commentators say, that caused them to wrest capital away from German investors, and the surge in their consumption that caused them to scoop up German tradable goods. It was Spain “pulling” capital and tradable goods out of Germany, in other words, and not Germany who was “pushing” capital and tradable goods into Spain. In the former case, however, Spanish borrowers could only have pulled capital out of Germany by forcing up interest rates, and by pulling tradable goods out of Germany, Spain would have also forced up the growth in German wages. The fact that European interest rates declined sharply during this period, and German wage growth decelerated sharply, would be extraordinarily counter-intuitive at best.
Step 5. “Spain” must accommodate the German current account surplus and the corresponding capital account deficit.
Spain had been importing moderate amounts of foreign capital before all of this happened, but after the Hartz reforms, the distortions in Germany were so great that the amount of excess German savings pouring into Spain in a matter of 3-5 years increased by 20-30 percentage points of Spain’s GDP. Spain’s corresponding current account deficit also soared to among the highest levels in its history. Just as Germany’s soaring current account surpluses and capital account deficits reflected a rising excess of German savings over German investment, the soaring current account deficits and capital account surpluses in Spain had to accommodate a rising excess of Spanish investment over Spanish savings. These are automatic consequences of net capital inflows.
There are four ways this could happen: Spain’s productive investments could rise, its non-productive investments could rise, Spanish unemployment could rise, or the Spanish consumption share of GDP could rise.
Spain’s productive investments could rise. More available capital at lower interest rates increases the amount of productive investments in Spain, which causes Spanish demand for investment-related goods and services to rise, including for tradable and non-tradable goods and services. As unemployed workers are put to work, part of this demand is met by rising supply, but unless unemployment in Spain is extremely high, much of it must be supplied from abroad.
Its non-productive investments could rise. More available capital at lower interest rates also increases the amount of non-productive investments in Spain to the extent that investors are unable to distinguish between the two. There are three reasons why non-productive might rise.
First, the rush of capital into Spain can cause GDP growth to rise, which can cause investors, including local governments, to overestimate future demand for infrastructure, and this is all the more likely when widespread corruption creates strong incentives to over-estimate future demand.
Second, when cheap capital pours into a country it can cause soaring asset prices, and as real estate prices soar, they create speculative demand for real estate, which creates even more scope for non-productive investment. Third, as German tradable goods increasingly displace Spanish tradable goods, Spanish investment in capacity is increasingly unproductive. For all these reasons, as non-productive investment in Spain rises, as in the case of productive investment, it causes Spanish demand for investment-related goods and services to rise, including for tradable and nontradable goods and services.
Spanish unemployment could rise. As German tradable goods increasingly displace Spanish tradable goods, Spanish investment in capacity becomes increasingly unproductive and Spanish manufacturers have to fire workers. Rising unemployment causes the savings rate to decline (as explained above). Before the 2008-09 crisis, however, fired workers were quickly hired by the rapidly-growing non-tradable goods sector, so the effect of the German tradable goods sector on Spanish unemployment was minimal until the crisis, with its effect being a shift of workers out of the tradable goods sector into the services and non-tradable sectors.
The Spanish consumption share of GDP could rise. More available capital at lower interest rates increases the amount of speculative investments in existing Spanish assets, such as stocks, bonds, and real estate. As their prices rise, Spanish households, especially those with greater risk appetites, believe they have become permanently wealthier, and they respond by expanding the consumption share of their income. They fund this increased consumption either by reducing their savings or by borrowing against their assets. This causes Spanish demand for consumption related goods and services to rise. As unemployed workers are put to work, part of this demand is met by rising supply, but unless unemployment in Spain is extremely high, eventually much of it must be supplied from abroad.
In three of these cases (excluding the third, rising unemployment) as Spanish demand for investment-related and consumption-related goods and services rose faster than Spanish supply of either, it had to import them from abroad, but because it can only import tradable goods (and some tradable services), an automatic shift had to occur in Spanish employment. Workers were transferred from the tradable sector to the nontradable sector, so that demand for tradable goods rose even faster, and this was resolved effectively by importing these goods from Germany. This is a simplification, and technically incorrect, but we will continue to use it over the rest of this essay. The goods that Spain imports as a consequence of the German capital “push” don’t have to be imported only from Germany. They can be imported from other countries, which simultaneously increase their imports from Germany, so that the rise in Spain’s current account deficit will be exactly matched by a rise in Germany’s current account surplus even if the two do not occur bilaterally.
Step 6. As speculation and cheap capital causes markets to rise, the Spanish debt burden will automatically rise until speculative price increases and a rising debt burden reach their limits.
The different ways described above in which Spain could have responded to the increase in capital imports each has a different impact on the Spanish debt burden.
Capital that flows into productive investments does not result in a rising debt burden because the capacity to service debt rises as quickly, or more quickly, than the debt servicing cost.
Capital that flows into non-productive investments causes the debt burden to rise because there is no consequent increase in the capacity to service debt, even as there is an increase in the debt-servicing cost.
Capital that finances the increase in consumption also causes the debt burden to rise because there is no consequent increase in the capacity to service debt while there is an increase in the debt-servicing cost.
When debt-servicing costs rise faster than debt-servicing capacity, the growth in debt is unsustainable and it becomes increasingly necessary to roll debt over as principle and interest come due.
Rising asset prices and rising debt are self-reinforcing, but neither is infinite, and both will eventually be forced to stop rising. There are many triggers for the reversal, which we do not need to get into here, but once it happens, market prices fall, and it becomes more difficult to roll over the debt. This occurred in 2009, and the consequence was that German capital stopped flowing into Spain. As a result Spanish governments, businesses and households could not service existing debt to Germany and they could no longer fund the domestic consumption and investment that had previously been funded by German savings.
Step 7. When something is unsustainable, eventually it will stop.
As Spanish and other peripheral European debt rose to levels at which creditors were increasingly worried about funding further debt increases, and as the underlying economies were damaged as resources were misallocated as a result of asset price distortions, all it took was a trigger to set off a system breakdown. The US sub-prime crisis served as that trigger. Once creditors became sufficiently concerned about the adverse impact of US events on underlying liquidity, they began to raise interest rates, reduce maturities and otherwise transform liabilities in ways that were highly selfreinforcing, and in a very short period of time deficit countries found themselves unable to refinance their external borrowings. Without balancing inflows on the capital account, their deficits had to collapse.
By definition a collapse in the current account deficit can occur either in the form of a drop in investment or a surge in savings. Investment dropped when creditors became unwilling to fund new investment and as stock and real estate prices collapsed, and the savings rate rose as Spanish households cut back on consumption. The net result was a sharp drop in demand and a corresponding surge in unemployment.
After the crisis there were a limited number of ways Germany and Spain could adjust. Once the rest of Europe was no longer able to absorb the German surplus, and because Germany took no steps to rebalance domestic demand but instead maintained the existing distortions in which German workers and German households retained a small share of GDP, the export of German savings abroad meant that the full European surplus, perhaps the largest ever recorded as a share of global GDP, had to be absorbed abroad.
In light of the June 23 Brexit vote, it might be worth noting that the British current account deficit, which had ranged between 1% and 3% for most of the previous decade, jumped 3.6% in 2008 and dropped back to its normal range over the next two years. Since then, as the European surplus has surged, the British deficit has climbed steadily to over 5% by 2015 and substantially higher in the first quarter of 2016. This requires one or both of two British consequences, rising unemployment or rising inequality.
Lessons
It's not just the outflows we need to worry about but the initial inflows.
Capital flows require more management (i.e. restrictions and regulations) if we want to avoid the negative consequences of unproductive capital inflows. This in turn requires a realisation that the 'market' cares about profit not productivity. Profit can be made in remarkably unproductive ways.
It's not cultural propensities that should be the major explanatory variable for economic outcomes. Rather it is macroeconomic flows shaped by policy (ie. institutions) that are most important. Culture and path dependency will shape what are considered to be appropriate policies. In sum, institutions matter more than culture, although culture may shape what institutions are chosen and developed in the first place.
For Australia, a negative spiral might begin with an external reassessment of investment opportunities or a decline in employment. Given the highly indebted status of Australian households this would reduce consumption and lead to a reassessment of the housing market.
A housing price collapse in Australia would lead to a reassessment of debt, a collapse in consumption and rising unemployment, which would lead to further house price falls and so on.
Australia has avoided falling into this negative spiral through a combination of good luck and some good policy. However the rise in debt and the failure of authorities to regulate credit will soon be seen as remarkably poor policy.
