Tightening money and widening Eurozone spreads

How do we know the Eurozone crisis has been getting worse over the last few days? The indicator that I watch most closely, and I think others watch most closely, is the yield spread between German government bonds and the bonds of other Eurozone governments. If those spreads widen, it means the crisis is getting worse.

One big difference between Market Monetarists and Keynesian (or New Keynesian) macroeconomists is how they see the relationship between monetary policy and interest rates. Especially real interest rates.

Keynesians tend to think of monetary policy as working through its effects on real interest rates. High real interest rates are a sign of tight money, and low real interest rates as a sign of loose money. Though whether monetary policy is loose enough depends on whether real interest rates are above or below the natural rate of interest. So if real interest rates are low, but still above the even lower natural rate, monetary policy may be loose, but still not loose enough.

Monetarists tend to think of monetary policy as tight or loose depending on expectations of nominal GDP. Low interest rates, both nominal and real, can be a symptom of an overly tight monetary policy. When a tightening of monetary policy means a fall in expected future NGDP, desired investment may fall, and desired saving may rise, and interest rates, both nominal and real, may fall.

As I have explained in the past, the idea that a tighter monetary policy could cause lower real interest rates can be explained in Keynesian terms as an upward-sloping IS curve. Or, equivalently, as a leftward shift in a downward-sloping IS curve when expected future real demand falls because of tighter monetary policy.

All this refers to the risk-free rate of interest. What has been left out of this argument is the effect of tight money on the yield spread between safe and risky assets.

If you accept the Monetarist premise that tighter money means a fall in expected NGDP, then it is very easy to understand a widening of yield spreads as yet one more symptom of tight money.

Tight money may reduce interest rates on safe assets. At least, it reduces interest rates on safe assets that remain safe assets when monetary policy is tightened. But many previously safe assets will become risky assets when monetary policy is tightened, precisely because tight money means a fall in expected NGDP, and a fall in expected NGDP means less nominal income to pay fixed nominal liabilities.

For a given total quantity of assets, a tightening of monetary policy reduces the supply of safe assets, and increases the supply of risky assets. Because some previously safe assets now become risky. Since safe and risky assets are imperfect substitutes on the demand-side, this change in relative supply of safe and risky assets changes their relative price. The price of safe assets rises relative to the price of risky assets. So the yield on safe assets falls relative to the yield on risky assets. Tight money causes an increase in the yield spread.

Looked at in this way, the widening yield spreads between Eurozone government bonds is simply one more symptom of a tightening monetary policy.

Yes, this is an oversimplified picture. The rollover risk from short term bonds means there can be a run on government bonds not dissimilar to a run on a bank without an effective lender of last resort. And there can be multiple equilibria, because a high/low interest rate makes it harder/easier for a government to finance its debt, which increases/reduces the riskiness of that debt, and so increases/reduces the rate of interest at which people are willing to hold that debt.

But let's start with the simple stuff first. Widening yield spreads are a symptom of tightening monetary policy. Depressingly simple, really.

I'm not sure how much longer this particular yield spread will remain a useful indicator. So far, people seem to be assuming that German government bonds will remain safe. I wouldn't count on them remaining safe, if monetary policy continues to tighten.

Comments

[Ramanan]

Nick,

There are some Euro Area nations with high yields and pressure on yields and there are ones with low yields.

As some Economists such as Gavin Davies of FT and Paul Krugman have realized, the Euro Area problem is a balance of payments crisis. It is made worse by the fact that governments cannot take advances from the central bank.

A highly indebted Euro Area nation (with indebtedness measured as the negative of net international investment position relative to national income) has to net borrow to finance its current account deficit and refinance its existing liabilities from international banks, money markets and capital markets by hook or crook.

Nations with bad foreign sector have run into problems and Germany on the other hand is safer because it is a creditor nation.

There is an FT Alphaville on this as well. http://ftalphaville.ft.com/blog/2011/11/18/754341/looking-at-the-eurozone-through-a-niip-prism/ which quotes Goldman Sachs European Research.

In my view the spread is simply due to the external sector and market perception.

[Max]

"I'm not sure how much longer this particular yield spread will remain a useful indicator. So far, people seem to be assuming that German government bonds will remain safe. I wouldn't count on them remaining safe, if monetary policy continues to tighten."

Exactly why the ECB should explicitly guarantee German bonds. It would have no effect on yields (so not giving anyone a windfall), but would give the German government confidence that it won't lose its benchmark status if it does the right thing.

[Kosta]

A very nice story. Especially how it explains why the bonds of previously "safe" nations like France and Austria have separated from German Bunds as the European expected NGDP has progressively deteriorated.

Does the tighter monetary policy meaning lower expected NGDP causality run both ways? That is, does a drop in NGDP necessarily mean that monetary policy tightened (perhaps passively, or unintentionally on the part of the central bank, or that it became inappropriately too tight)?

Do we require another "safe" asset for German Bunds to become risky? I suppose that other asset has to be U.S. Treasuries? Or maybe Japanese Gov't Bonds? In the limit as European Expected NGDP decreases, German Bunds have to become risky, but it is still are relative game. Won't a drop in European NGDP also push down the U.S. expected NGDP? In other words, what will it take for German Bunds to appear risky compared to U.S. Tresuries?

[Doctor Why]

Maybe widening spreads (esp. between Germany and France) also reflect growing expectations that the Eurozone will disintegrate and the new German currency will appreciate against the currencies of the other former members.

[JKH]

In a normal tightening cycle at positive interest rate levels, the risk free yield curve eventually flattens/inverts as the CB pushes the policy rate to a peak. Then an inverted yield curve tends to lead the lowering of short rates back down into recovery.

A version of this curve flattening effect has been happening now at the zero bound. The inability of the CB to lower the nominal policy rate is a form of tightening, similar to keeping the policy rate too high for too long in a normal positive rate cycle.

So this yield curve effect holds for all rate levels – positive or zero bound. In all cases, it’s the market saying that expected policy rate increases will be delayed longer than previously expected, which is saying that money has been tight, zero bound or otherwise.

This is entirely a policy rate based interpretation. It’s consistent with your NGDP interpretation, but doesn’t depend on it.

[JKH]

My comment above applies roughly to widening of risk spreads as well, I guess.

The ECB is not yet understood to be a government LLR, so all sovereign debt is risky, including German.

The only "risk free rate" benchmark in the EZ is that which the ECB sets on its own liabilities. Until it is understood to be a government LLR, there really is no risk free yield curve to speak of.

To get a risk free yield curve, the EZ needs implicit/explicit ECB to government LLR plus Eurobonds. That would make it comparable to the Fed/Treasury case.

Eurobonds alone, without implicit/explicit LLR support, would not necessarily do it.

[Nick Rowe]

Ramanan: If all 17 countries were exactly the same, and had exactly the same risk under all monetary policies, then there wouldn't be a risk spread to widen when monetary policy tightened. My simplest story cannot be told without assuming some underlying differences in risk, whether due to differing debt/GDP ratios, different expected growth rates, or different capital account surpluses/deficits.

But none of those differences suddenly appeared overnight, in the last few days, weeks, months, or even years. (With the possible exception of Greece, which had not been open about its fiscal position.) So they don't explain why the spreads should vary so much so quickly.

So it's perhaps better to think of the two stories (mine and yours) as complements, rather than substitutes. Your story talks about those underlying differences between the 17 countries; mine talks about why those underlying differences should suddenly matter so much more very quickly.

I should maybe have talked about this in the post itself. Ah well, that's why God invented comments.

[JKH]

Ramanan,

Stumbled on your blog the other day - hadn't been aware of it.

Congrats. Looks good.

[Nick Rowe]

Max: there are lots of things like that the ECB could do to save the Euro, even at this very late stage. I'm not really sure it's worth saving, even though the breakup will be ghastly.

Kosta: the debate about whether tight money caused expected NGDP to fall, or whether money was too tight because it failed to prevent a fall in expected NGDP, is one that probably cannot be resolved. It's all in how you define "doing nothing". I'm not sure if it really matters. The difference between sins of commission and sins of ommission doesn't fit well into the economist's perspective.

"Do we require another "safe" asset for German Bunds to become risky?"

I was thinking about that, and never came to a clear answer. There's cash, of course, because a bond is a promise to pay cash, and that promise can't normally be safer than the cash itself. And recessions are caused by an excess demand for money, the medium of exchange, so a demand to switch from risky bonds into safer cash (the medium of exchange) is what we worry about. Conceivably there may be private bonds that remain safer than any government bond, including German, though that's perhaps unlikely. If Euro cash itself became seen as risky, so people switched into buying real assets, especially newly-produced real assets, that would not be a bad thing.

Doctor Why: my guess is that that must be part of it. The "risk" is not only a risk of failure to pay, it's a risk of paying in a new currency worth less than the Euro. For Germany, would the new German currency be likely to be worth more than the Euro? If not, it's best to hold Euro notes rather than German bonds. But if the best you could do with a Euro note is convert it into new German notes, then German bonds could be safe relative to Euro notes.

JKH: I had to read your comments twice, but I think I understand and agree.

One quibble: "This is entirely a policy rate based interpretation. It’s consistent with your NGDP interpretation, but doesn’t depend on it."

Agreed (I think) on it being consistent with my story. Not sure I agree on whether it doesn't depend on it. What I mean by that is that we might have *observed* these correlations between risk spreads and the policy rate in the past, when monetary policy tightens, but if we want to *explain* those correlations, don't we need to say something about expected NGDP?

[JKH]

Ramanan,

Interestingly, TARGET2 is in effect an LLR function from the ECB to the individual NCB's, with NCB specific credit risk to be sorted out through ECB capital share exposure, etc.

So the various EZ nations don't ultimately rely on those other markets in the sense that the TARGET mechanism automatically fills in any private sector liquidity difference shortfall through the banking system.

BoP imbalances are symptomatic of differential risk, just as bond spreads are.

Was Krugman more focused on CA imbalances? There's more to it than that now.

[Nick Rowe]

Ramanan: a metaphor. Some countries are closer to the cliff than others. But if monetary policy is good, the cliff is a long way away, so those differences hardly matter at all. Tight monetary policy brings the cliff closer, so now some countries are very close to the cliff. Those differences now matter a lot.

[JKH]

Nick,

Agreed. A complete explanation of monetary policy should always reference NGDP, in context. And perhaps in the past it hasn't. That would be my thought in the most general terms as it relates to NGDPT. I'd put it in the mix of things to be considered in monetary policy, but with a soft T rather than a hard T.

[Phil Koop]

Do we require another "safe" asset for German Bunds to become risky?

The point is that it has to be a liquid asset denominated in euros - operators with euro liabilities need to balance them with euro assets. Isn't cash an alternative? Sure, but remember, the banking system itself is at risk. To hold cash in volume risklessly in practice means holding a deposit at the ECB. That possibility is not open to all market participants; many are obliged to make do with Bunds.

Although French and Austrian spreads have widened, note that their levels have hardly changed; the widening spreads are due to plunging Bund yields. That suggests a flight to "cash" (in the form of Bunds) rather than increase in the perceived riskyness or price of risk in French and Austrian bonds.

[Nick Rowe]

Phil: yep. In practice, safe Euro cash might mean a safety deposit box?

The fact that an increase in spreads is sometimes associated with a fall in German bond yields rather than always an increase in other yields does seem to support the Monetarist view that tight money can cause lower interest rates on safe assets. (Provided we assume German bonds are in fact safe). So I would say you need both bits together. The flight to cash that lowers German yields, plus the increased riskiness that widens spreads.

[David Pearson]

Nick,
"many previously safe assets will become risky assets when monetary policy is tightened"

Gorton, Shin and many others have described how the problems in the shadow banking system pre-dated the "tight money" of the fall of 2008. Specifically, 2007 through the summer of 2008 saw by far the largest number of AAA-rated bond downgrades in history, as well as multiple localized "runs" on shadow banking system liabilities. Was this because of "tight money"? If so, why did the stock market wait until October 2008 to crash?

Shin has a recent paper out on how European banks are at the core of the global shadow banking system, and how they will serve as a contagion vector to the U.S. and elsewhere. Just as in the 2007-2008 crisis, causality runs from the undercapitalization of towering shadow bank balance sheets; to their attempts to de-lever; to deflationary tendencies; to "tight money".

[David Pearson]

I imagine you could build a model in which the realized promise of stable NGDP leads to higher and higher financial system leverage. In this model, smaller and smaller disturbances cause larger potential deviations from NGDP trend, such that the "car" becomes impossible to drive. Financial frictions exist in this model: undercapitalized financial intermediaries fail/de-lever with no new intermediary to take their place.

[Kosta]

In practice, safe Euro cash might mean a safety deposit box?

That might work for individuals but companies, institutional investors, pension funds and even the really wealthy probably would find cash in a safety deposit box unwieldy and risky in of itself, n'est ce pas.

As Phil Koop stated, the safe asset has to be liquid. Gold fits the bill for the most part, and central banks are once again buying gold in size, so maybe Germany is becoming risky?

[Ramanan]

JKH,

Thanks.

I don't know how to put my thoughts ... but an attempt .. the TARGET2 is a like a cushion - helping the whole system but nothing more than that.

Imagine if a Euro Area nation was disconnected from the rest of the world. As the government runs deficits and hence is increasing its debt, the private sector is building surpluses and hence increasing its assets. The government bonds are absorbed without there being a pressure on yields. Of course, the bond markets can at some point take the government for a ride, given that there is no chance that governments can get advances from their home NCB but let us ignore that for now. Now bring in the foreign sector and assume the nation runs continuous current account deficits. While it is doing so, the whole region becomes indebted and this debt has to be financed/refinanced from abroad. Lenders have a home bias and they are not going to easily finance a deficit nation causing borrowing to become progressively expensive. This is even more true if the rest of the world thinks that the indebtedness is getting out of control.

All this was made complicated by the fact that lenders continued to finance the deficit regions pre-crisis as if the whole process can continue forever except suddenly realizing that it cannot continue forever.

Its a bit like the US private sector deficit pre-crisis when the private sector was happily borrowing at a large scale till it things started to get out of control.

Similarly for many years, the differential in the Euro Area nations' government debt were close but then during the crisis suddenly started spiking.

I wouldn't think that external imbalances are symptoms - they were important in leading to imbalances. Its the debtor nations which are facing more problems now than creditor nations. That much is clear.

The TARGET2 mechanism creates just creates more credit - as in it increases assets and liabilities of the NCBs. However the nation as a whole (domestic private sector + government + NCB) still has to borrow from the rest of the world (with an easy game for the NCB but not the other two).

For example if the government is able to finance its deficit and roll its existing debt from domestic lenders, some other sector will be left requiring financing from abroad because there aren't even funds locally.

The TARGET2/Eurosystem makes the analysis difficult because in some sense the NCB is domestic but from a different point of view, it is foreign.

[Ramanan]

Nick,

I agree everyone is near a cliff. I am just trying to give an argument on why some are slightly better off than others.

"But none of those differences suddenly appeared overnight, in the last few days, weeks, months, or even years. (With the possible exception of Greece, which had not been open about its fiscal position.) So they don't explain why the spreads should vary so much so quickly."

Yes it is completely true that none of the difference appeared overnight or months or years. But the markets continued to behave as if these imbalances did not matter until the participants realized that something is wrong.

(There was a paper by Willem Buiter from 2006 where he blamed the ECB for treating all government bonds as equivalent even though they are not.)

To me it looks like a general rule that markets completely ignore differences for a while and then suddenly start to look at differences.

[Ramanan]

"because there aren't even funds locally." @ November 22, 2011 at 12:55 PM

should be .. because there aren't enough funds locally.

[JKH]

Ramanan,

I should have said that TARGET2 reconciles intra-EZ balance of payments in Euros.

And it does so automatically.

E.g. Euro bank deposit flight from Greece to Germany is reconciled this way, as is the bilateral CA deficit of Greece with Germany in Euros, to the extent they aren't funded by other means.

Payment in Euros is made to Germany automatically through the banking system and TARGET, if Greece fails to attract Euro denominated funding otherwise.

This is separate from BoP of the EZ as a whole with the rest of the world.

[JKH]

The Fed has rejected NGDPT, as per the minutes of the last FOMC meeting, released today:

http://www.federalreserve.gov/monetarypolicy/fomcminutes20111102.htm

"Several participants observed that the efficacy of nominal GDP targeting depended crucially on some strong assumptions, including the premise that the Committee could make a credible commitment to maintaining such a strategy over a long time horizon and that policymakers would continue adhering to that strategy even in the face of a significant increase in inflation."

Translated - it won't work

[Nick Rowe]

David: Tight money is one thing that can go wrong with an economy and financial system, but I agree it's not the only thing that can go wrong. Increasing spreads aren't going to be a perfect indicator of tight money, just because other things can cause increased spreads too.

I never know what policy conclusion to draw from the Minsky hypothesis. Suppose we believe Minsky is right. Should we actively use monetary policy to destabilise the system, to stop it getting too fragile? At first sight, it seems to me, that won't help. The system will always adjust to the same level of risk regardless of how big or how small the shocks are.

Ramanan: "To me it looks like a general rule that markets completely ignore differences for a while and then suddenly start to look at differences."

Agreed, it does look like that, which is a puzzle. Maybe monetary policy could be part of the explanation of that puzzle? None of those differences really mattered very much, as long as there was little risk of a serious and long recession.

JKH: I don't think they are being very consistent. Half the time they say they can't increase AD, and the other half the time they say if they can increase AD but it will only cause inflation. If the latter, then why? Do they think the US economy is already at the NAIRU? In which case, why is the Fed asking for fiscal policy to increase AD.

[JKH]

Well, Scott S. is declaring victory, just like after the Bernanke press conference.

My goodness, such delusion.

But he does admit NGDPT is just a version of flexible inflation targeting:

"NGDP targeting is basically flexible inflation targeting, where you let the inflation rate deviate above or below 2% as output growth deviates from 3%. If there is any fundamental shift in framework, it would not be going from (flexible) inflation to NGDP growth rate targeting, it would be going from flexible inflation targeting to flexible price level targeting, or NGDP level targeting."

That's progress. Now just adjust for the choice of an explicit or an implicit inflation target, with flexibility, and there's no issue of substance - other than price level versus growth, which is probably the only issue in either flexible or inflexible form.

[Ramanan]

JKH,

Yes, different from the BoP with the rest of the world.

What I was trying to get at was BoP within the Euro Area. A good proportion of many nations' BoP is internal, i.e., within the Euro Area.

Normally one thinks of BoP as between countries and since countries have different currencies, one tends to think of it as between currencies. I believe the intuition of most economists has been that since the Euro Zone has eliminated currencies, it has eliminated the balance of payments problems within the Euro Area itself. It is for this reason that the discussion on this has been limited.

[Ramanan]

Nick,

"Agreed, it does look like that, which is a puzzle. Maybe monetary policy could be part of the explanation of that puzzle? None of those differences really mattered very much, as long as there was little risk of a serious and long recession."

My only explanation for that is what Keynes said - the markets can remain irrational more than you can remain solvent.

Willem Buiter was writing in 2006 - I guess his solution was a bit of monetary policy - differential treatment of haircuts on collateral for government debt - as in Greek government bonds should have different haircut than German bonds when banks borrow from the Eurosystem.

Had that been implemented, it could have brought a smoother fall in the markets and probably policy makers would have started noticing the problems earlier.

[Toni Vallen]

"This is entirely a policy rate based interpretation. It’s consistent with your NGDP interpretation
Normally one thinks of BoP as between countries and since countries have different currencies, one tends to think of it as between currencies

[TMDB (123)]

"So far, people seem to be assuming that German government bonds will remain safe. I wouldn't count on them remaining safe, if monetary policy continues to tighten."

This is the Lake Wobegon theory in reverse - all the Eurozone members are below average. All the Eurozone members are likely to depreciate against EUR if they quit. All the sovereign collateral accepted by the ECB has below average quality. Potential seigniorage per capita is below average in all the Eurozone countries.

Interestingly, today's events partially support the Lake Wobegon thesis - see the bund auction failure. On the other hand, DAX has rallied after the failure.

[Jacques René Giguère]

Even with the same currency, BoP problems exist.
That's why federation have mechanism such as equalization payments to balance things out.And unitary state, by enforcing common standards of expenditures , do the same trick.
If we applied forensic accounting technique to internal BoP in Canada, we could prove that QC and ON buy their oil from AB even if not a one drop flows east. And since it generates rent ( that is AB doesn't buy as much as it sell) we must correct that by either federal spending or equalization payments. Oil sheikdoms solve the problem by arms purchases and other no-productive non-consummable goods.
The internal BoP is another problem the euro-designers never thought about. By the minute we plumb the depths of their idicy...

[Kosta]

From Reuters "Disastrous" bond sale shakes confidence in Germany

The German debt agency was forced to retain almost half of a sale of 6 billion euros due to a shortage of bids by investors. The result pushed the cost of borrowing over 10 years for the bloc's paymaster above those for the United States for the first time since October.

The move in yields is small (~10 basis points) but are German Bunds about to get risky?

[Alex]

Cash is one option (Germans, in general, hold far more cash than any other Europeans). Another one, for big players, is to get a bank licence or control of something that does. Then you can park your cash with the central bank. I'm aware of a couple of major corporates (Siemens is one) that have done exactly that.

[Sergei]

"Interestingly, TARGET2 is in effect an LLR function from the ECB to the individual NCB's"

TARGET2 is a LLR for NCB's but NCB's are rather irrelevant. They lend to banks only against collateral which is limited. Yes, banking system used to able to produce ABS, etc. and is still somewhat able to produce covered bonds both of which are eligible. However its main source of LLR funds is obviously government bonds. So this LLR function can easily collapse because there will be no collateral left which Italian banks could post to get funds to settle Italian CA deficit against Germany. And it is already getting out of control as Italian banks have already asked ECB to extend the collateral rules.

TARGET2 is the ultimate time-bomb of the euro banking system.

[from Italy]

Well, someone commented about World of Warcraft's kind of economies in a past comment.
What about if the guys at Facebook or Google+ or Twitter would introduce a virtual currency in their system which could be used not only for buying services IN the system, but also OUTSIDE of it, I do mean that you can pay your employee both at FarmVille or at your actual farm with FB money and he can use that money with whomever accept them... will this solve the tight money problem? :D

[it's a joke, it's a joke :D]