Pettis: Feedback Loops

By Michael Pettis. Cross posted from Michael Pettis’ China Financial Markets

Early this month Martin Wolf had another of his very interesting articles, this time on China, which I think suggests some of the concerns we must have about the upcoming adjustment. Wolf argues that it may be useful to think about Japan as a model for understanding the adjustment process in China since the Japanese model shows how risky it is to shift to a slow-growth model. I of course agree.

Over the next decade, China’s growth will slow, probably sharply. That is not the view of malevolent outsiders. It is the view of the Chinese government. The question is whether it will do so smoothly or abruptly. On the answer depends not only China’s own future, but also that of much of the world.

As an aside it is funny to me (and probably no surprise to any Chinese or foreign economist who writes about China) that Wolf had to stress that it is not just “malevolent outsiders” who predict slower Chinese growth. You would think that identifying risks in the Chinese growth model and suggesting ways to minimize or hedge them would be considered a service to China.

But for some reason in China there is a very vocal minority that considers any skepticism about the sustainability of the Chinese growth model to be either an insult to the Chinese people or a malevolent foreign conspiracy. The extent of this rather surprising rage may suggest a level of fragility in China’s social fabric and its self-confidence that could make any slowdown more difficult to manage, but this is a digression. My point in bringing up Wolf’s article is not to play amateur national psychologist but rather to suggest something about the process of shifting to a new growth model. Wolf says:

As the experience of Japan has shown, managing a shift from a high-investment, high-growth economy to a lower-investment, lower-growth economy is very tricky. I can envisage at least three risks. First, if expected growth falls from over 10 to, say, 6 per cent, the needed rate of investment in productive capital will collapse: under a constant incremental capital output ratio the fall would be from 50 per cent to, say, 30 per cent of GDP. If swift, such a decline would cause a depression, all on its own.

Second, a big jump in credit has gone together with reliance on real estate and other investments with falling marginal returns. Partly for this reason, the decline in growth is likely to mean a rise in bad debts, not least on the investments made on the assumption that past growth would continue. The fragility of the financial system could increase very sharply, not least in the rapidly expanding “shadow banking” sector.

Third, since there is little reason to expect a decline in the household savings rate, sustaining the envisaged rise in consumption, relative to investment, demands a matching shift in incomes towards households and away from corporations, including state enterprises. This can happen: the growing labour shortage and a move towards higher interest rates might deliver it smoothly. But, even so, there is also a clear risk that the resulting decline in profits would accelerate a collapse in investment.

The government’s plan is, of course, to make the transition to a better balanced and slower-growing economy smoothly. This is far from impossible. The government has all the levers it needs. Moreover, the economy continues to have much potential. But managing a decline in the growth rate without an investment collapse and financial disruption is far trickier than any general equilibrium model suggests.

In his article Wolf implicitly refers to a process on which he doesn’t actually dwell much, but which I think is very important. A lot of economies, and especially developing economies with distorted balance sheets and one or two major drivers of growth, can have embedded in their economic institutions self-reinforcing mechanisms that can be very powerful.

I discuss this a great deal in my bookThe Volatility Machine. As a consequence of these self-reinforcing mechanisms, I argue, movements in any direction can be sharply magnified, so that positive shocks will often result in much faster growth than anyone expected. But this comes at a cost. The reversal of these shocks often can result in much slower growth than anyone expected, or even in a wholly unexpected collapse into crisis.

Wolf mentions in his second point the relationship between slower growth and rising bad debt, for example, and he is absolutely correct, but I would add that rising bad debt itself puts pressure on the financial system in a way that creates at least two additional problems. First, it makes banks reluctant to increase credit further. Second, the rising bad debt increases the hidden transfer from the household sector needed to resolve the debt, which then puts downward pressure on household consumption.

This of course is self-reinforcing. Why? Because slowing growth caused an increase in bad debts, but an increase in bad debts will cause further slowing in growth.

Feedback loops

We speak of the case in which positive shocks are self-reinforcing, as a virtuous circle, and the case in which negative shocks are self-reinforcing as a vicious circle, but the important point is that these processes are part of the same system and are very common. It is usually a pretty safe bet, for example, that when an economy is surging forward at astonishing growth rates – rates which far exceeded anyone’s prior expectations – it has powerful positive feedback loops embedded within its economic institutions.

In my book I focus mostly on balance sheet feedback loops, but they also exist just as powerfully in the underlying economy. Urbanization, for example, can create very strong feedback loops. How does it work? In the early stages of growth, productive jobs are created in the urban areas, for example as factories are built, and workers very quickly leave the countryside to take these jobs. They are also willing to move quickly around the country, so that migration is very sensitive to the perception of demand.

As workers move to the cities, their need for housing and services immediately rises, and their expenditures create additional jobs. What’s more, their remittances create capital accumulation and higher expenditures in their home areas that cause rural growth to increase by more than it otherwise would have. In the end the original investment in the factory is quickly multiplied throughout the economy, so that growth creates urbanization and urbanization creates more growth in a virtuous circle.

This process is reinforced by the impact of the original investment on the financial sector. As workers get jobs, part of their income is consumed, creating more demand and more jobs, and part of it is saved, which allows banks to direct the additional savings into higher investment. Since the growth impact of this process in the early stages of industrialization can be very high, this creates stronger growth expectations, which then justify even higher investment in capacity and infrastructure.

It is probably not a coincidence that in such developing countries that are growing quickly we almost always see credit growth far surpass anything we might have expected. The impact of financial deepening can be extremely strong in a country that starts out with a very weak and underdeveloped financial system, and as growth exceeds expectations year after year, perhaps not surprisingly, credit standards are weakened and money pours into projects that might have otherwise been considered risky

But what happens when these productive jobs dry up and the economy starts to slow, especially if it slows after credit has been too liberally extended? For one thing, either very quickly the workers go home, or they remain in the cities as unemployed workers without savings or social safety nets. In the former case the goods and services they demanded also disappear quickly and unemployment rises by even more than the direct impact of the reduction in jobs. If they remain in the city they create a drag on social expenditure (and perhaps a rise in crime) that transfers spending from more productive to less productive sectors.

As urbanization reverses, or even as it simply slows, it becomes self-reinforcing in the wrong direction. These kinds of feedback loops exist in every economy, but for a variety of reasons they seem much stronger in developing countries with weak institutional structures, in countries that are undergoing rapid social and economic change, and in countries with rigid and unsophisticated financial systems.

This feedback process may explain one of the puzzles typical of developing countries, and especially developing countries undergoing an investment boom. The historical precedents suggest that in the early stages of a growth miracle we are always surprised by the extent of growth – growth far exceeds even our wildest expectations.

Once the economy begins to slow, however, we have also been – in every case that I can identify – shocked by how vicious the slowdown turned out to be. This would not be a surprise if indeed these economies are caught up in very powerful feedback loops. On the contrary, this would be normal.

What does all of this have to do with Martin Wolf’s article? Wolf suggest plausible reasons for expecting a slowdown in Chinese growth, but his idea of a slowdown is relatively moderate and would, in fact, be considered rapid growth in most economies. But if part of the explanation for China’s spectacular – and spectacularly unexpected – growth of the past three decades has to do with the positive feedback loops that are so typical of developing countries with fragile and unsophisticated financial systems, then a moderate slowdown in growth may be an impossible target to achieve. Once growth starts to slow, the self-reinforcing impact on urbanization, on credit growth, on financial distress, and on expectations may force growth rates to drop far more sharply than any “plausible” analysis would suggest.

Back to Europe

Speaking about feedback loops, Europe continues to bounce violently up and down in the markets. In my January newsletter I expressed my surprise that the market was acting as if it thought the bulk of Europe’s problems was behind us, but since then the market has changed its beliefs. The events in Cyprus have reminded us of just how volatile the situation is and just how pro-cyclical it can be.

As I see it however events in Europe are still unfolding just as they are supposed to. Last week I saw a worrying article in the Financial Times.

Spain’s economic crisis and the near-collapse of its banking sector last year have conspired to choke off the flow of bank loans – threatening to dry out the vast and versatile pool that dominates Spain’s private sector.

In the five years since the crisis started, no fewer than 450,000 small and medium-sized enterprises have gone under, says Jesús Terciado, the president of Cepyme, the Spanish SME association. “But it is not just about staying in business – it’s also about growth. There is no way you can grow your business at the moment,” he says.

I have previously discussed the financial distress process – which affects companies as well as countries – as consisting of a highly pro-cyclical change in the behavior of major stakeholders in response to deteriorating credibility, which itself causes credibility to deteriorate further. In other words the more we worry about the government’s ability to repay debt, the more we behave in ways that have the effect of further reducing the government’s ability to repay the debt. We do this by acting in ways that automatically increase the amount of the debt, make the balance sheet more fragile, or reduce debt-servicing capacity.

One of the key modifications in the way stakeholders respond to a crisis is in the behavior of small and medium enterprises. Business owners respond to deteriorating economic conditions by disinvesting and firing workers, while lenders respond to worsening credit risks by tightening credit conditions. Both responses only worsen the underlying debt problem, and cause the economy constantly to underperform expectations, until at some point conditions spiral out of control.

No matter what Madrid does, in other words, the debt crisis will ensure that the Spanish economy continues to deteriorate as stakeholders respond to the deterioration. Without a major positive shock that can permanently reverse the underlying process of financial distress, Spain cannot help but continue its slow slide into crisis, and should any other country leave the euro, the pressures on Spain will automatically intensify as the burden of adjustment is forced onto a smaller group of countries.

Meanwhile in what also shouldn’t be a surprise, Spain’s savings rate continues to drop. Here is an article in El País:

Spain’s household savings rate declined to a new record low last year as rampant unemployment and the loss of spending power due to inflation and wage cuts forced families to dip into their piggy banks to make ends meet. According to figures released Tuesday by the National Statistics Institute (INE), households set aside 8.2 percent of their disposal income last year, the lowest rate since the INE began compiling the series in 2000. The figure was down 2.8 points from a year earlier.

The INE said disposable income in the fourth quarter declined 4.2 percent from the same period a year earlier, a fall of 7.835 billion euros to 176.766 billion. Wages in the period declined 8.5 percent, with was barely offset by slight increases in income from other sources such as interest rates and deposits.

Why are Spanish savings declining?

The article explains the falling savings rate as reflecting rising unemployment (fired workers, after all, still have to consume), but it is perhaps easier to understand the process, and how difficult it is to unwind, by working through the balance of payments. If Germany is still exporting savings (i.e. running a current account surplus), the rest of the world has no choice but to import savings, and because of monetary constraints the rest of the world in this case is peripheral Europe.

Why is Germany exporting savings? This, as I have pointed out many times, has nothing to do with German thrift. It has to do with wage repression. From 1991 to 2000, according to an interesting recent paper by Thomas Palley, nominal compensation per employee in Germany rose by 3.2% a year on average, whereas it rose a little more quickly in Europe overall (3.5% a year). From 2001 to 2010, while European-wide nominal compensation growth dropped to 2.4% annually, however, it dropped to 1.1% in Germany.

The sharp slowdown in German wage growth was a direct consequence of agreements hammered out between the government, large businesses and labor unions to repress wages in order to make German goods more competitive in international markets. It might seem surprising that slower wage growth would create a surge in savings, but since savings is simply total production less total consumption, and wage repression had a secondary consequence of repressing consumption growth, by definition it puts upward pressure on German savings.

In the 1990s Germans saved less than they invested domestically, so they imported foreign capital and ran current account deficits, but after the wage-repressing agreements, German’s savings rate was forced up to the point where it so far exceeded investment that Germany began exporting capital in 2002. The amount of net exported capital surged to 7.5% of GDP by 2007 – or, to put it another way, German’s current account surplus was 7.5% of GDP. Because the structure of the euro ensured that the corresponding deficits to German’s forced surpluses were likely to be the countries of peripheral Europe, German’s net capital exports had to result in net capital imports for the peripheral Europe.

There is only two ways a country like Spain could increase its net imports of foreign savings. Its investment could rise faster than its savings, or its savings could fall faster than its investment. In either case, as savings exceeded investment, Spain would import the difference, and so run a current account deficit equal to its net capital imports.

Of course with the economy in such terrible shape, there is no chance that investment is rising. In fact it is probably falling. This means that Spanish savings must also fall, and there are practically speaking two ways it can do so. First, the Spanish can go on a credit-fueled consumption boom that drives down the savings rate while keeping unemployment low, as they did before 2008-09. Second unemployment can surge, as it has since then.

Because a return to the consumption binge is pretty unlikely any time soon, is it such a big surprise that we are seeing savings adjust through high unemployment? Without a significant reversal of Germany’s too-high savings rate, Spain really doesn’t have much choice. The only ways Spain can reduce unemployment is either with a surge in domestic investment, or with a reversal of the trade deficit, which probably requires that it leave the euro and devalue. This is mostly just arithmetic.

The exorbitant privilege

For one last quick point I want to return to some of the absurdities of the currency war. During the 1980s Japan was fairly certain that its large current account surpluses had nothing to do with monetary policy and everything to do with management and business techniques.

In the past decade China has also argued the same thing. When trade surpluses are extraordinarily high, it cannot possibly be the consequence of distorted monetary policy, it seems. This makes the recent dispute between Japan and China a little surprising. According to an article in the South China Morning Post:

Many of China’s top economists are livid at what they view as an effective currency devaluation by Japan and are calling on the People’s Bank of China to retaliate by weakening the yuan to defend itself in what they see as a new currency war. These economists, including Tsinghua University professor Li Daokui and ANZ Bank’s Liu Ligang, see Japan’s plan to double its monetary base within two years as “blackmail” and have criticised the Japanese central bank’s decision to open the liquidity floodgates to bump up the economy.

Liu said Japan’s unprecedented easing programme, aimed at ending more than two decades of deflation, was “a monetary blackmail” targeted at other export-driven Asian countries such as China and that the central bank should sell more yuan and buy the US dollar to push down the yuan. He also called on authorities to guard against a fresh wave of hot money into China’s fragile financial markets, warning that Japan’s move would reignite the so-called carry trade, under which investors borrow in low-interest yen and invest in high- interest markets.

It is interesting that while Beijing has complained bitterly about the exorbitant privilege that allows the US to exploit foreign purchase of dollars for reserves, when Tokyo acts to push down the yen, many of the same people in Beijing argue that the Chinese response should be to transfer to the US even more of this exorbitant privilege. It seems that none of the world’s economies really want to take any of the exorbitant privilege for themselves.

And why should they? Buying US dollars in order to push down the value of their currencies – and this is all the US exorbitant privilege consists of – allows other countries to increase domestic employment by turbo-charging exports growth at the expense of domestic consumption (and imports). Although countries that do this usually insist that this higher domestic employment does not come at the expense of US employment, no one is eager to carry the current account deficit that comes with the exorbitant privilege.

My guess is that until the US takes steps to prevent foreign accumulation of US government bonds, or, if other countries want to retain the use of an international traded currency, they agree not to game the reserve system, the US will always be faced with the choice either of absorbing demand deficiency from every part of the world that wants more growth or of engaging in its own version of currency war, QE. The very same people who complain about the currency consequences of QE, in other words, are arguing that while it makes sense for them to expand aggressively, it is morally wrong when other countries do it.

That sounds a lot to me like old-fashioned beggar-thy-neighbor politics. And there is no reason we should expect this to end soon.

Print Friendly
Tweet about this on Twitter0Digg thisShare on Reddit0Share on StumbleUpon0Share on Facebook0Share on LinkedIn0Share on Google+1Buffer this pageEmail this to someone
This entry was posted in Guest Post on by .

About Lambert Strether

Lambert Strether has been blogging, managing online communities, and doing system administration 24/7 since 2003, in Drupal and WordPress. Besides political economy and the political scene, he blogs about rhetoric, software engineering, permaculture, history, literature, local politics, international travel, food, and fixing stuff around the house. The nom de plume “Lambert Strether” comes from Henry James’s The Ambassadors: “Live all you can. It’s a mistake not to.” You can follow him on Twitter at @lambertstrether.


  1. vlade

    “The very same people who complain about the currency consequences of QE, in other words, are arguing that while it makes sense for them to expand aggressively, it is morally wrong when other countries do it.”

    How human!

  2. EenAnderGeluid

    A good analysis of the origins and effects of the German current account surplus, and the damaging effects of these imbalances !

    Much of it also applies to my country – the Netherlands – which is also running a huge surplus, mainly by wage suppression. Recently yet another “deal” was made between the union leadership and government. But Holland differs from Germany in being much less industrial and much more FIRE-oriented. That’s why Holland is now in a deeper crisis then Germany.

    A second good part of this analysis is the beggar-thy-neighbor politics.
    Here it would be interesting to explore ways to an exit of the crisis WITHOUT resorting to more export and endless growth. The whole neoliberal scheme in Europe is based on export-competition and a race to the bottom. Obviously it is a deadend policy.

    1. Ignacio

      FIRE-Oriented? I suppose it means FInantial & REal estate oriented. I also guess that The Netherlands is exporting finance services (Real Estate services?) in vast amounts to run such surplus. Please, can you expand your explanation?

      1. EenAnderGeluid

        FIRE = Finance, Insurance, Real Estate.

        The Netherlands lost large part of their industrial base (mining, shipbuilding, textiles, cars, aviation etc) in the 70’s and 80’s. This was partially neutralized by rising natural gas export. The dutch economy depends for a large part on transit trade to and from Germany (Rotterdam) with petrochemical industry as an important sector linked to this. In the 80’s the banking sector (Amsterdam) began to expand on the real estate bubble en participating in the new unchained global financial markets. Dutch banking grew to 5 times GNP. The Netherlands are also an important link in the financial offshore-business.

        The country recieved a double whammy of a popping real estate bubble and a financial crisis which made the nationalisation of 2 of the 4 systemic banks necessary, and a thrird on life support. Natural gas passed its peak already – and when international trade (Germany) slows down we’re in real deep trouble.

    2. financial matters

      “”A second good part of this analysis is the beggar-thy-neighbor politics.
      Here it would be interesting to explore ways to an exit of the crisis WITHOUT resorting to more export and endless growth. The whole neoliberal scheme in Europe is based on export-competition and a race to the bottom. Obviously it is a deadend policy.””

      Yes, I agree. One way to stablize the swings would be with a JG/ELR Job Guarantee/Employer of last resort. When private sector jobs are lost, this labor can be put to work on publicly oriented projects which would save on unemployment benefits/food stamps etc and lead to useful work such as infrastructure repair. When the economy picks up employers can hire from this low wage pool by offering higher wages.

      This can ease credit tightening problems also as the govt is spending money into the economy.

      This works best in countries with a sovereign currency but still runs into problems in the US as many view the US budget as similar to a household budget. It’s also not impossilbe for the euro but would take similar mind set changes such as Germany recognizing that a strong Spain is useful for its exports and that maybe their labor should be better paid which would help domestic consumption.

  3. russell1200

    Realizing there are some apples and oranges, but two cases come to mind where the positive feedback loop seems to have been shut down, or at least transferred somewhat gracefully.

    The first is Greenspan’s mid-90s interest rate hikes. Granted it was temporary, but it seems to have worked for a little while.

    You also have a number of times when wartime spending was stopped, and you didn’t have the post war crash. They always seem to be special cases though where either the country was on the flank of the main war (Great Britain in Napoleonic Wars, Japan in WW1, or U.S. in WW2)and was in a position of being able to take advantage of new markets, or was able to shift its over expansion into rebuilding the other participants, or achieved similar advantages in victory (German unification).

  4. F. Beard

    and part of it is saved, which allows banks to direct the additional savings into higher investment. Michael Pettis

    This sounds like the “loanable funds model” which, to my meager knowledge has been discredited.

    With the banks, “loans create deposits” and conversely “loan repayment destroys deposits.” So it seems to me that additional investment (presumably new bank loans/deposits) allows additional savings and not the other way around.

  5. burnside

    In broad terms, Pettis seems to say that development is a hand played once, the chief variable being how rapidly it’s deployed, and that subsequent events are determined largely by the velocity or aggressiveness of the development phase.

  6. Wat Tyler

    “The government has all the levers it needs”

    Contrast this with “government is the problem” ideology and the two are opposed and mutually exclusive. Like “Thunder Dome”, two go in but only one comes out. The winning system will dominate the remainder of the 21st century and perhaps beyond.

    The rumble is on!


    1. F. Beard

      To the extent that “government is the problem” is it not because some chose merely to ameliorate injustice (on the cheap, mind you) rather than abolish it? Rather than a “New Deal”, should we not have been given an “Honest Deal” plus reparations for past injustice? Especially when the deflation of the times could have been cured with massive amounts of new fiat just given to the population and which was cured instead later by the massive waste of lives and material of WWII?

      1. Wat Tyler

        Yes but giving money to citizens for food and shelter is a moral hazzard and communism whereas giving money to the military to blow up stuff in a distant land is “truth, justice, and the American way” (picture Reagan with tears welling up). It is not the size of deficits but the beneficiaries that matters – or it seems that way to me.


  7. Systemic Disorder

    Excellent point on German wage cuts — indeed, this is the secret of Germany’s past decade. German worker productivity has been increasing by more than one percent a year while real wages have been slowly declining.

    Germany is increasingly dependent on exports. If exports decline because of lower wages in export markets, German workers aren’t going to be able to make up the difference.

  8. allcoppedout

    Most management theory is fit for ridicule rather than analysis. Bias is usually evident in the first paragraph. The problem is an outdated theory of comparative advantage and management fear it is too useless to be able to manage people not threatened by poverty.

Comments are closed.