by Michael Pettis, China Financial Markets
It seems to me that one of the automatic, if not always intended, consequences of Abenomics is to force up Japan’s current account surplus, and in fact to force it up substantially. This will have to do at least in part with deciding how to manage the country’s enormous government debt burden, which easily exceeds 200% of the country’s GDP.
If I am right, this should create two concerns. First, in a world struggling with insufficient demand and excess capacity, and in which the growth strategies of too many countries implicitly involve a significant increase in exports relative to imports, a major increase in Japan’s current account surplus could easily derail growth recovery elsewhere. The US for example has to worry that policies aimed at increasing domestic demand don’t simply result in rising debt as US demand bleeds out through the current account, while both China and Europe need strong external sectors to make their own difficult domestic adjustments less painful.
Second, it is not obvious that the world will be able to absorb a significant increase in the Japanese exports, and if Abenomics implicitly forces up the Japanese savings rate relative to investment (which is all that we mean when we say that economic policies force up current account surpluses), these policies can resolve themselves either in the form of high growth and soaring exports, or much lower growth and slowing imports. The former implies that Abenomics will be successful, while the latter that it will fail. It is not obvious, in other words, that Abenomics can succeed in a world of weak demand, and its failure is likely to make Japan’s domestic imbalances worse, not better.
It may seem a little quixotic to worry about a surging Japanese current account surplus just now when in fact Japan’s external balance has declined substantially and is surprising analysts on the downside. According to an article in last week’s Financial Times:
Japan’s current account balance plummeted by nearly two-thirds in August from a year ago, surprising forecasters that had assumed it would grow nearly a fifth. The current account is a broad measure of trade. A fall indicates Japan is receiving less income from overseas investments, despite help from the falling yen.
The current account surplus fell nearly 64 per cent in August, versus forecasts expecting an 18 per cent gain. The unadjusted balance in the month was Y161.5bn, against forecasts at Y520bn and down from Y577.3bn in July. Within the data, trade of goods and services was in deficit of more than Y1tn for a second consecutive month, while income fell to Y1.253tn from Y1.794tn a month before.
My concern, however, is unlikely to be played out over the next few quarters but rather over the next few years as Abenomics is implemented, and so Japan’s external position in the immediate future doesn’t matter. What matters, I think, is that in order to generate growth Tokyo is planning to implement polices aimed at raising both inflation and real GDP, and these policies are likely to force up the national savings rate relative to investment.
What is more, to the extent that these policies are successful in generating higher nominal GDP growth, they create a problem for Tokyo in how it decides to set domestic interest rates. Japan has never really resolved the overinvestment orgy of the 1980s. Instead of writing down bad debt it effectively transferred much of it to the government balance sheet, and now this huge debt burden is itself becoming, I think, a constraint on the success of policies designed by Tokyo to spur growth.
Before addressing the debt constraint, let me start by listing the reasons why I think Abenomics is likely to affect the trade surplus. First is the impact of Abenomics on pushing down the value of the yen. As I discuss in the first two chapters of my January book, The Great Rebalancing, currency depreciation does not affect the trade balance directly by changing relative prices. It does so indirectly by changing the relationship between savings and investment (the difference between the two being the current account balance). A depreciating currency reduces the real value of household income by acting effectively as a consumption tax on imported items. This also reduces the real value of household consumption.
The proceeds of this tax are used implicitly to subsidize the tradable goods sector, which effectively increases production in that sector. Of course as production rises relative to consumption, the difference between the two – the national savings rate – must also rise.
This means that as the yen depreciates, the consequence is likely to be an increase in the Japanese savings rate. If there is no commensurate increase in investment (and I assume that with excess capacity Japan does not need to increase investment much in order to produce higher output), Japan’s current account surplus must automatically rise. In the near term the investment rate is likely to rise, largely in response to greater confidence, but over the longer term downward pressure on the consumption share of GDP (which is the likely consequence of downward pressure on the household income share) will also put downward pressure on investment growth.
Savings is the obverse of consumption
But it doesn’t end there. Japan seems to be taking other steps to force up its domestic savings rate. Here is last Tuesday’s Financial Times:
Shinzo Abe,Japan’sprime minister, pledged to press ahead with the first increase in sales tax for over 15 years despite objections from some of his closest advisers, gambling that measures to address the country’s massive debts would not hinder his attempts to jump-start the economy.
Mr Abe said on Tuesday he would couple the consumption tax hike with roughly Y5tn in new public works spending, cash grants and other stimulus in order to blunt any negative impact on the economy.
…The plan to increase the tax from 5 to 8 per cent next April had been approved by a previous government with the support of Mr Abe’s Liberal Democratic Party. But it was opposed by economists who had helped the premier draft his Abenomics strategy, as well as by some LDP politicians. The last time Japan increased the levy, in 1997, a deep recession followed that shook the party’s grip on power.
The increase in the consumption tax, part of the proceeds of which will be used to increase infrastructure investment, will accomplish many of the same results as the deprecation of the yen. A consumption tax, like a tariff, is effectively a kind of back-door currency devaluation, with a slightly different mix of losers among the household sector and winners among the producing sector.
By boosting production and reducing consumption, however, it automatically forces up the national savings rate in the same way as does currency depreciation. Even if 100% of the proceeds of the tax were used to fund increased infrastructure investment (and the article suggests that part, but not all, of the consumption taxes will be directed towards higher investment), because at least some of the investment spending will go to workers in the form of wages, who will save part of those wages, the net result will be that total savings will rise faster than total investment. Once again this must force up Japan’s current account surplus even further.
So far this all looks like an attempt by Abe to increase Japanese competitiveness and so increase its total share of global demand, but not by increasing Japanese productivity, which is the high road to growth, but rather by reducing the real Japanese household income share of what is produced. Japan (like Germany and China have done over the past decade) is attempting to increase employment by reducing wages, and this means that its workers will be able to purchase a declining share of what they produce. This effectively means Japan will be growing at the expense of its trading partners. As the Japanese become less able to consume all they produce, the excess must be exported abroad.
If the world were in ruddy good health, we might not worry too much about policies aimed at Japan’s pulling itself out of the mess created in the 1980s, but with the whole world struggling with weak demand and with country after country trying to reduce domestic unemployment by selling more abroad – effectively exporting unemployment (with Germany in particular hoping to resolve the European crisis not by increasing its net domestic demand, as it should, but rather by forcing German surpluses outside Europe) – there is a real question in my mind as to how successful the Japanese program of Abenomics is likely to be if it implicitly requires a burgeoning trade surplus.
Remember that if one country increases its savings rate, unless there is a net increase in global investment there must be a commensurate reduction in the savings rate of the rest of the world so that savings and investment always balance globally. There are broadly speaking two ways this can happen. In the pre-crisis days this reduction in the savings rate of the rest of the world occurred mainly in the form of soaring consumption fueled by credit, and in this way unemployment stayed low. Since the crisis – which because of the negative wealth effect saw credit-fueled consumption drop – foreign savings have been reduced by a rise in foreign unemployment
This means that if Japan forces up its savings rate, and assuming that we are unlikely to return in the next few years to a credit-fueled consumption binge, the only way the world can respond to a structural forcing up of the Japanese savings rate is either by higher unemployment outside Japan or, if Japan’s trade partners take steps to protect themselves from higher Japanese trade surpluses, higher unemployment inside Japan.
The debt-servicing cost of nominal GDP growth
But there is more, perhaps much more. Japan is struggling with an enormous debt burden, and perhaps this explains why Tokyo is so eager to engage in policies that force up the Japanese savings rate. As long as more than 100% of Japanese borrowing is funded by domestic savings (if Japan runs a current account surplus is must be a net exporter, not importer, of capital), it doesn’t have to rely on fickle foreigners, who might not be satisfied with coupons close to zero, to fund its enormous debt burden.
But the debt burden creates its own very dangerous source of trade instability. To understand why, we need to consider what happens to interest rates in Japan if nominal growth rates rise.
In Japan interest rates are currently very low, close to zero. With total government debt amounting to more than twice the country’s GDP – which puts it among the most heavily indebted governments in the world – it is not hard to see how low nominal interest rates benefit Japan. With interest rates close to zero, there is very little cashflow pressure on the government from servicing its debt.
Some people might argue that nominal interest rates do not matter. We should be looking at real interest rates, they would argue, and with Japan’s having experienced deflation for much of the past two decades, real interest rates in Japan are high and the nominal rate is largely irrelevant.
This is true, real interest rates do matter, but it doesn’t mean that nominal interest rates do not. In fact both real and nominal interest rates matter, albeit for different reasons. Real rates matter for all the obvious reasons – they represent the real cost to the borrower in terms of a transfer of resources from the borrower to the lender. But nominal rates also matter because they effectively determine the implicit amortization schedule of principal payments.
When the nominal rate is zero or close to zero in a deflationary environment, in other words, interest is effectively capitalized in real terms. In fact whenever the real rate exceeds the nominal rate, as it has in Japan for much of the past two decades, the cashflow cost of servicing the debt is lower than the real cost, and the difference is effectively converted into real principal and deferred. In real terms, in other words, Japanese debt is growing by the difference between the real rate and the nominal rate, and this effectively represents a reduction in the cashflow cost of servicing its debt.
When nominal interest rates are positive and higher than the real rate, however, there is effectively an acceleration of real principal payments. This means that as long as nominal rates are very low, the real cost of servicing the debt is low and the principal payments are postponed, with some of the interest even being capitalized. As nominal rates rise, however, the real cost of servicing the debt during each payment period consists of interest plus some real principal.
This is just a long, perhaps pedantic, way of pointing out that even if the real interest rate in Japan declines, debt servicing is likely to be much more difficult as the nominal rate rises. Japan might be paying a lower real rate, but it is also implicitly paying down principle, instead of capitalizing it. Tokyo would need a significant increase in revenues, or a significant decrease in expenditures, to cover the cost.
So what would force Japan to raise its nominal interest rate? In principle the nominal interest rate should be more or less in line with the nominal GDP growth rate. If it is higher, growth generated by investing capital is disproportionately retained by net savers (including mainly the household sector). There is, in other words, a hidden transfer of resources from net borrowers to net savers.
If the nominal lending rate is lower than the nominal GDP growth rate, as is the case in China today and Japan during the 1980s, the opposite occurs. There is a hidden transfer from net savers to net borrowers, and because net savers are mainly the household sector, this will put downward pressure on the household share of income even as it gooses investment growth. This hidden transfer has been at the heart of the rapid economic growth that typically occurs in financially repressed economies during the earlier stages, and is also at the heart of the investment misallocation process that typically occurs during the later stages. We have seen this very clearly in China.
Will Tokyo raise interest rates?
Japan is trying to generate both positive inflation and real GDP growth, so that it is trying urgently to raise the growth rate of nominal GDP. What happens if and when it is successful? For example let us assume that Japan’s GDP is able to grow nominally by 4-5% a year – what will happen to the nominal Japanese interest rate?
Tokyo can either raise interest rates in line with nominal GDP growth rates or it can keep them repressed. In the former case, debt-servicing costs would soar, ultimately to 8% of GDP or more. This would create a problem for Tokyo in its ability to service its tremendous debt burden. It would need a primary surplus of around 8% of GDP just to keep debt levels constant, and it is hard to imagine how such a huge surplus would be consistent with nominal GDP growth rates of 4-5%.
If it were to raise income taxes it would create a huge burden for the household sector and almost certainly force up the national savings rate by forcing down the household share of GDP. Remember that during the 1980s Japan, like China today, generated rapid growth in part through financial repression, and one of the consequences of that rapid growth was an extraordinarily high savings rate along with a huge current account surplus, both of which were ultimately unsustainable. Japan has spent much of the past twenty years rebalancing GDP back in favor of the household sector, and to reverse this process may provide relief in the short term, but it is hard to see how I can be helpful in the medium term.
On the other hand if, in order to make its debt burden manageable Tokyo represses interest rates to well below the nominal GDP growth rate, it is effectively transferring a significant share of GDP from the household sector to the government in the form of the hidden financial repression tax. This is what Japan was doing in the 1980s, with all of the now-obvious consequences.
Japan’s enormous debt burden was manageable as long as GDP growth rates were close to zero because this allowed both for the country to rebalance its economy and for Tokyo to make the negligible debt servicing payments even as it was effectively capitalizing part of its debt servicing cost. If Japan starts to grow, however, it can no longer do so. Unless it is willing to privatize assets and pay down the debt, or to impose very heavy taxes of the business sector, one way or the other it will either face serious debt constraints or it will begin to rebalance the economy once again away from consumption.
As this happens Japan’s saving rate will inexorably creep up, and unless investment can grow just as consistently, Japan will require ever larger current account surpluses in order to resolve the excess of its production over its domestic demand. If it has trouble running large current account surpluses, as I expect in a world struggling with too much capacity and too little demand, Abenomics is likely to fail in the medium term.
Perhaps all I am saying with this analysis is that debt matters, even if it is possible to pretend for many years that it doesn’t (and this pretense was made possible by the implicit capitalization of debt-servicing costs). Japan never really wrote down all or even most of its investment misallocation of the 1980s and simply rolled it forward in the form of rising government debt. For a long time it was able to service this growing debt burden by keeping interest rates very low as a response to very slow growth and by effectively capitalizing interest payments, but if Abenomics is “successful”, ironically, it will no longer be able to play this game. Unless Japan moves quickly to pay down debt, perhaps by privatizing government assets, Abenomics, in that case, will be derailed by its own success.
This is an abbreviated version of the newsletter that went out three weeks ago. Academics, journalists, and government and NGO officials who want to subscribe to the newsletter should write to me at firstname.lastname@example.org, stating your affiliation, please. Investors who want to buy a subscription should write to me, also at that address.