Tuesday, January 27, 2015


The last two weeks have been full of monetary news with the Swiss Franc peg, and the ECB's announcement of Quantitative Easing (QE). A few thoughts.

As you have probably heard by now, the Swiss Central Bank removed the 1.20 cap vs. the euro, and the franc promptly shot up 20%.

To defend the peg, the Swiss central bank had bought close to a year's Swiss GDP of euros (short-term euro debt really) to issue similar amounts of Swiss Franc denominated debt.

This is a QE -- a big QE. Buy assets, print money (again, really interest-paying reserves). So to some extent the news items are related. And, it's pretty clear why the SNB abandoned the peg. If the ECB started essentially the opposite transaction -- buying debt and selling euros -- the SNB would soon be awash.

A few lessons:

A peg depends on credibility. The dollar is pegged to 4 quarters. The Fed is not racking up huge dollar for quarters QE, because everyone knows it will always be thus. The fact that the SNB had to buy euros at all is a great signal that everyone knew the peg was temporary. As, in fact, the SNB had made pretty clear. Sometime or other, probably when it's most important, investors thought, Swiss Francs will shoot up again. Might as well buy more of them.

An exchange rate peg is fiscal policy.  Really, the "credibility" a country needs is fiscal credibility.

The peg fell apart because the SNB was trying to do it alone. On the day of abandonment, the SNB lost about 20% of its balance sheet, since it owns Euros and owes Swiss Francs. Had things gone on any more before the plunge, they would have had to go begging to the Treasury for a recapitalization. "We just lost 20% of GDP, could you please send us some fresh government bonds to back our CHF debt issues?" That works seamlessly in economic models, but would be a political nightmare for a central bank.

So, if you want to run a peg, it should be done jointly with the Treasury. The central bank buys euros, sells francs, but immediately swaps the euro debt to the treasury for CHF debt. That at least is removes the first fragility, by taking the fiscal risk off the central bank balance sheet.

From the point of view of the nation as a whole, a strong demand for your government debt (that's what this is) is an invitation to profligacy, not a fiscal danger. That's why pegs usually break in the other direction: The central bank tries to peg a currency, let's call them pesos, against another, let's call them dollars. (Or gold.) People start to ask for dollars in exchange for pesos, the central bank starts to run out of reserves. At this point the treasury has to either tax, reduce spending, or credibly promise future taxes or spending reductions to borrow some dollars, and given them to the central bank in exchange for the bank's government debt. When that can't happen, the peg breaks. The essential problem is fiscal.

Switzerland had this in reverse: The Swiss were too darn thrifty.  Americans and Greeks know what to do if world capital markets come knocking and want to buy boatloads of your government debt. Print debt, give it to them, and send us Walmarts full of goods, or driveways full of Porsches.  Norway had a similar issue, with the world wanting to buy its oil. Norway decided not to go on a consumption binge, so their sovereign wealth fund buys equities; rights to future consumption.

Switzerland could have done the same: sell CHF bonds, use the proceeds to go on a consumption binge or buy about a year's GDP of foreign stocks. Instead, a referendum threatened a return to the gold standard.

Or, they could have said, "and by the way, we declare that we have the right to pay off our government debt in euros at 1.20, or to swap CHF debt for euro debt at that rate." Now that would have really enforced the peg. Devaluing the currency means engineering a partial default on government debt. Its fiscal policy and can't be done by the central bank alone.

QE and the ECB

Ben Bernanke famously said that QE works in practice but not in theory.  What that means, of course, is that the standard theory is wrong, and to the extent it "works" at all, it works by some other mechanism or theory. Permanent price impact by changing the private sector portfolio composition is the "theory" that Bernanke acknowledges really makes no sense. So why might a QE work?

In the US case, QE was arguably a signal of Fed intentions. Buying a trillion dollars of bonds and issuing a trillion dollars of, er... bonds (reserves are floating-rate debt) is a way for the Fed to tell markets that it will be years and years before interest rates go up. As I chat about QE with economists, this pretty much surfaces as the most plausible story for QE effects (along with, there weren't any long lasting effects.) Greenwood,  Hanson, Rudolph, and Summers make this point nicely, showing that Fed-induced changes in maturity structure have about twice the effect that Treasury selling more bonds does -- though exactly the same portfolio effect.

But what is the signal in ECB QE? Well, a decidedly different one. The signal is, I think, not about interest rates, but that the ECB will buy government debt. "What it takes" is now taken. Yes, there is this lovely pretense that national central banks buy the bonds, so the ECB doesn't hold credit risk. But if a country defaults, where is the national central bank going to come up with funds to pay the ECB?

So, when we think of what expectations people derive from ECB QE, and with that how it might or might not "work," the obvious conclusion is that the Eurobonds are now being printed. Like all bonds, they will either be repaid, inflate, or default.

Torsten Slok sends on this interesting graph. 80% of Greek debt is now in the hands of "foreign official." Now you know why nobody is worrying about "contagion" anymore. The negotiation is entirely which government will pay.


  1. It makes perfect sense to me why pegs usually break. I can't understand why this one needs to break. Can you explain it to me like I'm a finance student :-)?

    I'm the SNB. I can print as much paper as I want with CHF on it to buy euros. People want my paper a lot, so I get a lot of euros (maybe I swap these for eurodenominated assets). Some day in the future I stop buying euros, the price of the euro falls and now all the stuff I've bought in the past is worth less. But I bought all that stuff with little pieces of paper, so why do I care?

    Probably there is something deep here that I don't understand about why you call helicopter drops fiscal policy and not monetary policy. Maybe I am stuck in the model world and not the real world, but I have never understood this point (which I realize is at the heart of a bunch of your work -- sorry :-( ).

    1. I'm no economist, but I think inflation expectations may be at the heart of the issue. SNB is privately owned - taking on ECB QE would have been a mammoth task and would have inflated SNB balance sheet to gargantuan proportions. Irrational maybe, but there nevertheless.

      Re helicopter drops, think of it as a huge cut to income tax. Fiscal!

    2. So I get how a company goes bankrupt:

      A = L + E

      You take some cash out of A and you buy some bonds/stocks, then those stocks loose value and E goes down.

      But this is like a company that takes a piece of paper writes 1000 on it and it's worth 1000 dollars (CHF). It's like making assets go up and equity go up at the same time, then if you buy some different assets with the cash and they lose all their value, they just go back to where they started. How do you lose in a real economic sense? I mean, we could call that an accounting loss, but why does it matter?

      John once wrote: "However, our Federal Reserve can create as much more money as anyone might desire and more. There is about $10 trillion of Treasury debt still outstanding. The Fed can buy it."

      Why can't the SNB just buy all the outstanding swiss debt and if that doesn't work buy other stuff? Not being able to inflate should be wonderful, right? All the goods in the world are free.

    3. Charlie: from one finance student to another, I think the most useful analogy is the pricing of IPOs. Companies typically sell equity for below the market price during the IPO placement, and those who get these artificially cheap shares benefit from the pop when the shares start trading. Likewise, there was a pop when the Swiss abandoned the peg and allowed the Franc to trade at market prices. The size of the jump tells us about the resources that were used to maintain the peg this long. Now that the Franc is free to float, any claims issued by the Swiss central bank will be priced more efficiently.

      One difference with the IPO analogy is that we know exactly when an IPO will happen, but not when a currency peg will be altered or abandoned altogether.

    4. I can see that real resources are used when a peg breaks in the other direction, when a country is using its (dollar reserves) to buy its own currency. But here the Swiss are just printing little pieces of paper and trading it for valuable stuff. They aren't giving up future cash flows, like an IPO. I mean, they should be. If you print a lot of money, your supposed to get inflation, but the Swiss were begging for inflation.

      If the problem is: "We just lost 20% of GDP, could you please send us some fresh government bonds to back our CHF debt issues?" Isn't the easy fix just to buy government debt instead of euros? I mean that's kinda what John is saying about swapping with the treasury, but they could just as easily say, we are going to buy back Franc debt until we hit our peg, and it they don't hit their peg, they just buy back all Swiss debt. If they still haven't hit the peg, they can then start buying euros with no worries, right?

      It's more like a country that just decided to cut the money supply dramatically. It looks like a huge deflationary shock.

    5. By trading the Franc for less than it is really worth, they were indeed giving up future cash flows, by foregoing some of the assets they could have bought if they were trading at a freely-floating rate. There was no immediate need to fix the large loss of 20%, because it was not a bill that suddenly became due. But it does show how costly the peg was in the long term, and I think we should stop writing contracts in nominal terms so that we no longer have a need for such costly policies.

    6. How would you differentiate your story from my story, which is that the Swiss just had a large tightening of monetary policy?

    7. I agree with you on that. And this spills over into other countries where people take out home loans denominated in Swiss francs. But those are external costs, and the internal costs to the Swiss of the peg apparently were not worth the benefits.

  2. Great post. The Eurozone's unique institutional features (read: monetary union w/o fiscal union) raises further interesting points regarding the fiscal nature of ECB QE relative to Fed or SNB QE. The NCB purchases are indeed a fig leaf, as risk will be shared through the Target2 system. And this is where the analogy to the SNB is most interesting it seems. Capital outflows from the periphery in the event of a crisis will be effectively financed by the Bundesbanks growing T2 credit. This credit is in essence an FX "floor" for the Lira/Franc/Drachma/Peseta/Escuda. To be truly ex post "unlimited" (not just ex ante unlimited as in the SNB case) the German polity must give it's full backing to such mutualization of risk. Given that seems a clear violation of the Treaties, we may wake up to find out that the peripheral FX floor (or equally the Deutschmark ceiling) is not actually there. If the fiscal backing isn't credible, which one could argue is the case without further explicit fiscal integration at the EMU level involving true loss of sovereignty and joint/several liability issuance, the how can ECB QE credibly boost inflation expectations?

    1. "Given that seems a clear violation of the Treaties"

      As we saw with the violation of the Stability and Growth Pact (which Germany had forced on the EZ), Germany is unencumbered by minor inconveniences like the Maastricht Treaty. If they choose not to mutualize risk they'll wrap themselves in the Treaty as the reason but if forced by circumstances to ignore the Treaty they will. "Realpolitik" rules the EZ.

  3. The long run policy that is consistent with an exchange rate ceiling is, "we'll be at least as inflationary as the reference central bank". No amount of fiscal backing allows you to deviate from that.

    Probably the SNB never intended to adopt such a policy, though. Rather, the ceiling was just a tactical measure to deal with an overvalued (in the opinion of the SNB) currency.

    Incidentally, central banks recapitalize themselves by withholding dividends. That's a lot easier than going hat in hand to governments.

  4. The Swiss tried pegs twice before in the 70s and each time ended up with inflation that forced them to abandon the peg. I think they saw the writing on the wall and bailed before Draghi pulled them down the same path again. Good for them. They swallowed a bitter pill now rather than 2 bitter pills later.

    The EZ should copy a few pages from the SNB playbook about not taking the most expedient route that gets you through for the moment. Unfortunately that will never happen with weasels like Juncker running the show.

  5. .....where is the national central bank going to come up with funds to pay the ECB?....
    Where ? The same place where the ECB is going, the liability side of its balance sheer, where it will be digit some zeros.... ECB QE means that some (45 billions a month) european bonds will disappear for good, never to come back. Germany will buy 215 billions of its bonds against a NET issuance expected to be of 4 (yes) billions, so it will reduce ist 2,000 something Bund Bobll Schatz market by 10%. Even Italy will buy 147 billions against a NET issuance of 62 billions and therefore will reduce its 1,700 billions BTP CCT market by something like 5%. Public Debt will not pile up any more in Europe. We are only at 20% of GDP in Eurozone in terms of buying back our gov bonds, let us hope we keep going and eventually get rid of, you know, 30% of it. As Mike Woodford said in London: "All this talk of exit strategies is deeply negative," at a London Business School seminar on the merits of Helicopter money, or "overt monetary financing". "If we are going to scare the horses, let's scare them properly. Let's go further and eliminate government debt on the bloated balance sheet of central banks," he said.

    This could done with a flick of the fingers. The debt would vanish. This is what our Mario Draghi is doing.
    BTW Target2 does not play a role here because austerity is still in place, there is no capital flight (outside Greece) and Target2 balances are getting smaller

  6. John,
    I can’t help but think that the SNB’s only real power is the provision of SFs – the liability side. Trying to influence EUR-SF by buying EUR is slightly crazy, because it’s ineffective on the EUR side (no chance of moving this large market). So the SNB provides SF reserves to its banking system – handing it an effective asset capital infusion – and creates SF demand deposits with whomever sold the SF bond to the bank. These SF deposits can then be used in much more fungible fashion – consumption, investment, taxed – than the original CH government bonds, and so ought to weaken the currency more effectively against the EUR.
    So why bother buying EUR at all? Perhaps this could fine-tune daily EUR-SF demand at 1.20, but really all that the SNB controls is SF provision. Credibility failed because the SNB impaled itself on holding EUR assets, and so tied its balance sheet to whatever the ECB decided to do – why not buy gold or equities instead, and keep credible degrees of freedom?
    In a similar vein, why does the composition of assets matter at all? “High quality” CB assets are usually required by law. CBs also usually don’t want to distort markets, so large-and-liquid is desirable. (And CBs need income to keep up operations.) The SNB would have been much more credible if it bought Edelweiss flowers from Swiss citizens for 1,000SF each.
    The only CB asset composition argument I can see is that CBs ought to want reasonable liquid assets that will gain in value with inflation – assets matched with liabilities, that is. Gold is phenomenal by this measure, and fixed income bonds are the worst possible holding: if CB’s were to actually stimulate the economy via asset purchases, and lift interest rates and nominal GDP, they automatically impale their balance sheets, and make asset sales more difficult – madness!

  7. I'm puzzled by your claim that an exchange rate peg is fiscal policy. It seems to hinge on the idea that eventually all of those Euros the SNB was holding would fall in value, SNB's equity would be wiped out, and they would have to go "begging the treasury for recapitalization". Is there a law that prohibits the SNB from operating with negative capital? I don't see why the SNB would need to recapitalize to continue about the business of defending the exchange rate floor. Inflation could not have been a serious concern. The price level has fallen since 2011 and nominal interest rates indicated extremely low inflation for the foreseeable future.

    I have yet to hear a reasonable explanation for how this policy change is consistent with an inflation target of 2%. A far better explanation is that the SNB wants continued deflation.

    1. The Czech CB has a negative capital of about 5% of GDP and has had for years. The Slovak CB joined the eurozone while it had negative equity and still operates with negative equity. I too don't understand why anyone would need to recapitalize a CB.

  8. I don't get why an exchange rate is "fiscal" policy. And I don't understand why SNB would hold euro-assets just because their intervention rule was expressed in terms of a Euro-Franc rate. As for being forced to go off the peg because, with ECB's tentative excursion into QE, SNB would be "awash" with Euros, so what? Even if they had to buy more, they would not have to hold more.

    1. Monetary and fiscal policy are linked via the government intertemporal budget constraint, The underlying assumption is that ultimately the central bank answers to the fiscal policy makers and so monetary policy is subservient to the fiscal authority.

  9. " The central bank buys euros, sells francs, but immediately swaps the euro debt to the treasury for CHF debt."

    Doesn't that leave the government with foreign currency denominated debt?

  10. "The fact that the SNB had to buy euros at all is a great signal that everyone knew the peg was temporary....Sometime or other,...investors thought,...Swiss Francs will shoot up again. Might as well buy more of them."

    If the SNB had said that they were willing to raise the peg in the future, to 1.30 or 1.40 for example, if Swiss inflation didn't approach their target, then wouldn't that have made it easier to defend the peg at 1.20? If investors believed that there was significant risk that the franc would fall from 1.20 to 1.30, then they would presumably be less willing to buy franc at 1.20 even if they believed that, eventually, its value would rise once the peg was lifted. The SNB can maintain a peg longer than investors can remain solvent. Of course, if the SNB was going to do that, they might as well have gone directly to an inflation target rather than using the exchange rate as an intermediate target.

  11. CHF increased it's value because of increased number of CHF loans in the last years, as debt gives value to money.

    In the future, very few will take loans anymore in CHF; the CHF value should go to sustainable value for the Swiss economy.

  12. "But here the Swiss are just printing little pieces of paper and trading it for valuable stuff. They aren't giving up future cash flows, like an IPO. I mean, they should be."

    Aren't they giving those pieces of paper, that they can exchange for valuable stuff (in Euro countries -sounds weird geographically differentiating between Swiss and Euro) to someone who can use those pieces of paper to buy valuable stuff located in Switzerland?

    By conversion appears to be a lean on my stuff vs a lean on your stuff. Also appears currencies do have a determinable reference value.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.