Felix Salmon

Why Richard Koo’s idea won’t save the Eurozone

By Felix Salmon
April 16, 2012

A lot of people were looking forward to Richard Koo deliver his paper at INET last week; it comes with associated slides, here. Koo is the intellectual father of the idea of the balance-sheet recession — an idea which was born in Japan, has increasingly been adopted in the US and the UK, and which is now gaining traction in the Eurozone.

Koo’s diagnosis that Europe is in a balance-sheet recession, which he defines as a post-bubble-bursting state of affairs where individuals and companies choose to pay down their debts rather than borrow money, even when interest rates are at zero. That certainly seems to be the problem in Spain; Koo’s charts can be hard to read, but what you’re seeing here is a massive borrowing binge by the Spanish corporate sector — the dark-blue line — suddenly turning into net savings after the crisis hits. And to make matters worse, Spanish households — the red line — did exactly the same thing. As a result, the government had to run a massive deficit after the crisis; the four lines always have to sum to zero.


In this kind of a recession, monetary policy — reducing rates to zero — doesn’t work. And tax cuts don’t work either: they just increase household savings. You need government spending, at least until the economy has warmed up to the point at which companies and individuals start borrowing again. And the good news is that in a balance sheet recession, government spending is pretty much cost-free, since interest rates are at zero.

That’s the good news in Japan and the US and the UK, anyway. But it’s not the case in Spain. This is a big problem with the single currency, as Koo explains: it encourages capital to flow in exactly the wrong direction.

When presented with a deleveraging private sector, fund managers in non- eurozone countries can place their money only in their own government’s bonds if constraints prevent them from taking on more currency risk or principle risk.

In contrast, eurozone fund managers who are not allowed to take on more principle risk or currency risk are not required to buy their own country’s bonds: they can also buy bonds issued by other eurozone governments because they all share the same currency. Thus, fund managers at French and German banks were busily moving funds into Spanish and Greek bonds a number of years ago in search of higher yields, and Spanish and Portuguese fund managers are now buying German and Dutch government bonds for added safety, all without incurring foreign exchange risk.

The former capital flow aggravated real estate bubbles in many peripheral countries prior to 2008, while the latter flow triggered a sovereign debt crisis in the same countries after 2008. Indeed, the excess domestic savings of Spain and Portugal fled to Germany and Holland just when the Spanish and Portuguese governments needed them to fight balance sheet recessions. That capital flight pushed bond yields higher in peripheral countries and forced their governments into austerity when their private sectors were also deleveraging. With both public and private sectors saving money, the economies fell into deflationary spirals which took the unemployment rate to 23 percent in Spain and to nearly 15 percent in Ireland and Portugal.

With the recipients, Germany and Holland, also aiming for fiscal austerity, the savings that flowed into these countries remained unborrowed and became a deflationary gap for the entire eurozone. It will be difficult to expect stable economic growth until something is done about these highly pro-cyclical and destabilizing capital flows unique to the eurozone.

The solution to this problem is eurobonds. If all the eurozone countries funded themselves jointly and severally, then the yields on European government debt would be very low, and there would be no fiscal crisis in Spain.

But that’s not the solution that Koo provides to the problem that he quite correctly diagnoses. Instead, he said at the INET conference that the way to bring Spanish government bond yields down would be to massively decrease the number of people allowed to buy Spanish bonds.

This doesn’t make any sense to me at all. After all, in any market, the greater the number of potential buyers, the higher the price. But I’ll let Koo try to explain:

Eurozone governments should limit the sale of their government bonds to their own citizens. In other words, only German citizens should be allowed to purchase Bunds, and only Spanish citizens should be able to buy Spanish government bonds. If this rule had been in place from the outset of the euro, none of the problems affecting the single currency today would have happened.

The Greek government could not have pursued a profligate fiscal policy for so long if only Greeks had been allowed to buy its bonds, since private-sector savings in the country was nowhere near enough to finance the government’s spending spree…

The new rule will also resolve the capital flight problem by preventing Spanish savings from flowing into German Bunds. This will prompt Spanish fund managers facing private sector deleveraging to invest in Spanish government bonds, just as their counterparts in the US, UK or Japan must buy bonds issued by their own governments. That would allow Spanish government bond yields to come down to UK—if not to US—levels. With low bond yields and high bond prices, talk of a sovereign debt crisis would also disappear.

It’s absolutely true that a lot of Spanish fund managers, in a flight-to-quality trade, are buying up German and Dutch government bonds. It’s also true that Koo’s proposed rule would prevent them from doing that. But it’s emphatically not true that if they couldn’t buy German or Dutch government bonds, they would buy Spanish government bonds instead.

I asked Koo about this, in Berlin, and the conceptual problem quickly emerged. On the one hand, Koo is careful not to say that he wants fully-fledged capital controls preventing Spanish fund managers from investing abroad at all. He’s confining his proposal just to government bonds, and is not including corporate bonds, equities, structured credit, or anything else that Spaniards can currently invest in. But, he still thinks that this one rule would, single-handedly, take the 6% of GDP that the Spanish private sector is currently saving, and divert it directly into the Spanish government bond market, sending yields there plunging.

It wouldn’t.

The fact is that when money flows into US Treasuries or JGBs or Gilts during a balance-sheet recession, that has absolutely nothing to do with the fact that they’re government bonds, and absolutely everything to do with the fact that they’re the dollar- or yen- or pound-based securities with the lowest perceived credit risk. If you ban Spanish institutional investors from investing in Bunds, then they’ll just buy something else with extremely low credit risk instead — AAA-rated corporate bonds, perhaps, or covered mortgage bonds from somewhere in the north, or some other kind of highly collateralized structured credit instrument. None of those things might have quite the degree of liquidity that Bunds can offer, but they’re still safer than Spanish government debt right now.

Koo is absolutely right that the flow of savings out of Spain is doing absolutely gruesome things to the Spanish economy: you can’t possibly grow when your companies and households are paying down debt, and all your national savings are fleeing the country. So maybe there’s a case for fully-fledged capital controls. But Koo’s weak-tea version would only serve to decrease, rather than increase, demand for Spanish government bonds. Their price would go down, their yields would go up, and Spain would be in an even worse position than it’s in now.

19 comments so far | RSS Comments RSS

Perhaps it is a dumb question, but what would happen if the government REFUSED to run a deficit? It seems Felix is treating corporate and household savings as the independent variables, and government savings (spending) as the dependent result. What happens if they don’t cooperate? Deflation?

And why can this argument not be turned around? Perhaps the CAUSE of the household/corporate savings is that the government is engaged in a massive borrowing/spending spree? After all, the lines must necessarily sum to zero…

Posted by TFF | Report as abusive

TFF, absent of private investment (corporations with record cash hoards have demonstrated no fear of doing nothing with their cash), if the government stopped running a deficit, unemployment would spike. the idea that having a huge supply of capital available will spur investment has been disproven. Supply side economics is voodoo.

Posted by KenG_CA | Report as abusive

@TFF – Spain’s government was running a surplus, which is easy to do when a real estate bubble creates employment, profits, and transactions to tax. From 2007 to 2009 Spain’s tax revenue fell by 16%. Their spending increased by 17% – higher unemployment benefits and such are probably part of that.

If the government had instead cut spending by 16%, without the private sector growing an equal amount (in $ terms), then the economy and tax base would merely have shrunk more. It seems highly unlikely that the private sector would have grown proportionately in the midst of their real estate crash, because they had started to run a surplus in reaction to losing a decade worth of investment in a speculative boondoggle.

But what if the government ran a balanced budget? Maybe the domestic savings would leave Spain to find borrowers (even more so than it recently has). Maybe the risk free interest rate would fall, if someone could find risk free savings in Spain. Since the private sector was already over indebted and their credit shot, it is unlikely that they would have reverted to zero savings no matter what the government did.

Savings without investment reduces economic activity. Eventually the savings are lost to default – a cruel resolution to the paradox of thrift.

Posted by kentwillard | Report as abusive

@kentwillard Nice explanation. You nicely illustrated what the ‘simplemongers’ ignore, debt as a percentage of GDP rises and falls no matter how prudent or profligate you are.

Posted by FifthDecade | Report as abusive

A Eurobond’s needed because
It’s one of the euro zone’s flaws
That investment was hot
Where the wanting was not
And decidedly cool where it was.


Posted by DrGoose | Report as abusive

“And the good news is that in a balance sheet recession, government spending is pretty much cost-free, since interest rates are at zero.” (FS)

Oh congratulations and well done, boss – you’ve discovered there really is a “free lunch” after all.

What’s you’re next miracle? – gonna pull a unicorn out off an empty hat?

Posted by MrRFox | Report as abusive

I don’t see how Koo thinks you could stop the banks from thinking up a trick to go around his rule that only those living on a country can own its sovereign debt.

However: the ‘only’ thing needed is to have people believe that Italy and Spain have economies that don’t need to default, and that even the ECB won’t want to force them into default. The ECB can still buy up those bonds and push down the interest within reasonable bounds. Italy and Spain are clearly too big to fail? So are those italian and spanish bonds now the big opportunity .. if not: the whole Eurozone will fail.

It could also buy up a balanced basket from all the EURO zone and then sell its own Eurobonds to fund that basket. And since nobody but the ECB can change the mix in that basket, it could work: inside the basket, capital cannot flow from one component to the other. Felix is right about those Eurobonds.

Posted by Twundit | Report as abusive

Thanks, kentwillard. Nice explanation.

Posted by TFF | Report as abusive

@TFF: kentwillard’s explanation is good. My take: GDP is the total of economic activity, so the short answer to your question: if the Spanish government refuses to spend when the private sector isn’t spending, then Spanish GDP shrinks until internal devaluation causes greater exports, and the now-lower GDP stops falling.

You could also read some basics about about GDP calculations, and sectoral composition. Off the top of my head, Pragmatic Capitalism has some good stuff on this.

GDP = consumption + investment + government spending + (imports – exports). If a country is running a trade deficit (net negative country savings), and households and business are saving (fiscal surplus), then either the government runs a corresponding fiscal deficit, or GDP contracts to make that accounting identity balance. There can be no other outcome.

That’s why governmental austerity when private balance sheets are deleveraging is catastrophic – every component of economic activity is decreasing, which is the same as saying GDP is decreasing, which is the same as saying those people, on average, are getting poorer.

Posted by SteveHamlin | Report as abusive

@Steve Hamlin – you wrote – “That’s why governmental austerity when private balance sheets are deleveraging is catastrophic – every component of economic activity is decreasing, which is the same as saying GDP is decreasing, which is the same as saying those people, on average, are getting poorer.” (SH)

Yes, they would be getting poorer in the scenario you describe. But if government steps in with bigger deficits to stoke demand, aren’t the people getting equally poorer when the additional government indebtedness is included in the calculation? The government’s incremental debt is actually the people’s debt, isn’t it?

Government intervention as you describe, which is the current practice, amounts to kicking the can … and not more than that, doesn’t it?

Posted by MrRFox | Report as abusive

I agree with you Mr. Fox.
I never quite understand in these “accounting identities” that government is some magic entity separate from the rest of the economy. How exactly is government spending different from and not a part of “consumption?” Yes, government can “Print” but the relationship between that and actual production and consumption doesn’t appear so linear (if if is, why aren’t we doing better?)
Now I am not one who believes that government doesn’t produce ANYTHING of value (streets, traffic signals, police, courts, schools, water works, etcetera) – there is government expenditures and government income (taxes as well as fees and even patents) – but it would be clearer just to roll it into a simpler identity – GDP equals consumption and production. (oh yeah, and export and import but it still is buying and earning).
Long story short – prices and incomes have to match and we may very well be spending more than we are earning – I suspect economists don’t want to say the standard of lving is going down (in “real” terms)

But I would also note that as long as central banks create money and give it to bankers to loan to people who cannot afford to pay back the loans on their underwater homes, when the bankers have demonstratably proven that they don’t know how to judge credit risk or economic return accurately, we are gonna remain in a stagnant economy. I guess I don’t buy that the Wizard of Oz knows what he is doing anymore…

Posted by fresnodan | Report as abusive

@MrRFox: “But if government steps in with bigger deficits to stoke demand, aren’t the people getting equally poorer when the additional government indebtedness is included in the calculation?”

Not if the governmental spending causes productive economic returns in excess of the real debt rate, and those returns would not have otherwise happened (see the current output gap for permanently lost wealth creation opportunity) – in that case, then that debt creates a net value-add.

Does borrowing money to build a factory make the company poorer? Only if the investment does not generate enough economic return to pay off the loan.

Were the nation’s infrastructures & systems worth building, and did they generate positive economic returns for the nation? I’d say most of those expenditures did.

If the government spending generates a lower return than the real interest rat of that debt, then it is value destroying (however, I’m also thinking of Keynes’ aphorism that in times of high unemployment, then it is a positive result to even pay a man to dig a hole and pay another man to fill it up, so in an output gap environment, the spending might not even need to be that productive.)

Posted by SteveHamlin | Report as abusive

@fresnodan: “it would be clearer just to roll it into a simpler identity – GDP equals consumption and production” You just restated the starting definition of GDP: now you need some way to total up that consumption and production. The traditional GDP components help to do that.

Under your view, instead of private entities producing and consuming less and the government producing and consuming less (GDP goes down), you have all entities producing and consuming less (GDP goes down). It doesn’t change anything.

The GDP accounting identity components are merely a deconstruction of how you’re wanting to look at it, and that is useful because it allows the available data sources to map to the separate components.

I suspect you’re thinking of some sort of full Ricardian Equivalence, which as applied here might argue that any governmental debt-financed spending cancels out any aggregate demand increase from that spending, because rational taxpayers know they’ll have to pay for that debt via higher taxes in the future, so as a result they’ll immediately save for those future taxes, and those increased savings contemporaneously and completely offset any increased spending. It’s a ridiculous concept that only applies in theory, with rigid assumptions about human behavior that do not have basis in the real world. Homoeconomis Rationalis is a mythical creature akin to Bigfoot.

Posted by SteveHamlin | Report as abusive

This is why I enjoy Felix’s blog so much. I think I’ve learned more about economics just from reading the comments here than I did in college.

Posted by spectre855 | Report as abusive

KenG_CA, kentwillard, DrGoose, SteveHamlin -

Nice comments to accompany this article!

Posted by TobyONottoby | Report as abusive

@TFF if the government refused to run a deficit, that would set up a battle royale between the various interest groups which benefit from government spending, especially retirees, versus the taxpayers. Ultimately, both would lose. Retirement ages would go up substantially, other entitlements would be cut substantially, taxes which have little or no immediate effect on economic activity (property taxes and especially taxes on the land rather than the improvements, inheritance taxes) would go up dramatically, other taxes might rise, depending on what side of the Laffer curve they are on now. For parts of Europe, raising income and payroll taxes is going to be counter-productive, given how high these taxes are already. But even the taxes were set optimally, balancing the budget would lead to short-term massive deflation. Everyone would go bankrupt and the debt overhand would be cleared up. The owners of debt (the rich) would be devastated like they were back in the Great Depression. The poor would do okay because modern societies won’t tolerate actual starvation. The middle-class would be wiped out in the short-run, but they’d bounce back once the deflation was over. In short, a balanced budget and massive deflation and liquidation would be wonderful for Main Street and a disaster for Wall Street. Anyone arguing that a balanced budget is bad for the common people is really trying to keep Wall Street fat and happy.

Posted by revelo | Report as abusive

Koo’s idea seems to attack the symptom, not the cause. Having more bond holders might help bond prices but the real problem is that there is too much debt.

have a look at this:

http://thatretiredguy.blogspot.pt/2012/0 4/no-crisis-is-not-over.html

Posted by FinanceChicken | Report as abusive

@revelo You’ve missed out a vital descriptor – the time frame over which the budget should be balanced… every quarter? Each year? Why stick at an arbitrary 12 month period, why not go for 15 months? What about an electoral cycle of 4/5 years (depending on country)? What about balancing the budget over the business cycle? If you stick rigidly to too short a period of time for the budget to be balanced over, you get an artificial imbalance which skews policy and accentuates boom and bust; in good years, government spending is low, in bad years the same dollar spend is high as a percentage.

So you either allow for variability in government spending (no more political promises can be kept, if they are now) or you match the budget balancing over the time frame of the variability – which is itself a variable (the length of the business cycle is not related to astrophysical events ie the calendar year). But if you do that, you have to explain why the Govt keeps surpluses in the good times and doesn’t pay them back to taxpayers as tax rebates. If you pay them out as tax rebates to win votes, you’re going to end up with a deficit later when the cycle moves out of boom into bust. Because one of the main roles of government is to build and run infrastructure, annual spending is far less changeable than the budget balancers would like it to be.

This isn’t rocket science, heck, even that history book, the Bible, mentions Pharaoh being advised to keep seven years of grain as a reserve against times of famine; whatever happened to prudently keeping money in reserve?

Posted by FifthDecade | Report as abusive

Euro or dollar system doesn’t require the four figures to add up to zero. Under these currencies new money is created when new credit is created. This is called credit expansion. Conversely, under credit contraction the total amount of credit and therefore the total amount of money in the currency system decreases.

Therefore, when private sector chooses to pay back their debt ie. deleverage, there is no need for Gvt to step in and start to borrow. Unless of cource the policy makers don’t want the credit contraction to happen.

Posted by Eskola | Report as abusive

Post Your Comment

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/