« The rebound in federal government revenues | Main | The Eurozone lender of last resort? »

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

Game theory is not my area, so I'll just ask a question on the monetary angle. The ECB prints money to bail out Greece. If this reduces the risk spread on Greek bonds then the speculators who sold Greece short get hurt. But if the action raises inflation, this may raise the inflation premium faster than the risk premium falls. Or am I completely missing the point here?

Greece is helped either way, what I am wondering is whether the wolves get hurt.

If the wolves are as clever as you suggest, I think they will figure out that you can't fire fast enough to kill them all if they all attack at once, and that in that case you won't fire at all because it wouldn't be logical. Your only hope is to convince them credibly that you aren't rational, which I think implies shooting one of the sheep now and then.

Take the mangiest-looking sheep, and in a Darwinian manner, cull the flock. Lace it with poison pills. Then cut up the carrion into smaller pieces and spread it around - so that a maximum number of wolves get sick or lose their appetite for more sheep. Repeat one or more times until the odds reverse.

Not sure if this has macro analogies - my knowledge of wolves and sheep economies is limited to Merry Melodies and Looney Tunes.

"Morning Ralph". "Morning Sam."

Scott: I don't think you are missing the point. Wouldn't it depend on the nature of the short? If you think of a short as a bet, if the bet were made in nominal terms the short-seller would still lose. But if the bet were on (say) the spread between nominal bonds and some real asset? Dunno. I need to think about that one.

Jim: But can the wolves coordinate their actions to all attack at once? If they can't, then which ever wolf attacks first will get shot. So all wolves wait for the other wolves to go first. Think of attacks over WWI trenches. No soldier wants to be the first to leave the trench. That's why armies spend so much time getting the soldiers to march in step, an otherwise totally useless activity. That's why they have a command structure. An army can always beat an equivalent number of individual soldiers, on level ground.

Predators like wolves (and bond market vigilantes?) tend to be lazy, opportunistic, and risk averse (unless they are desperate). Preemptively shooting a wolf or two will probably drive them off.

"I think they will figure out that you can't fire fast enough to kill them all"

There was a really bad Steven Segal movie (a redundant statement, I know) from the late 80's/early 1990s that might shed some insight on this. Segal was being set upon by a group of thugs (I know, that doesn't really narrow it down, that could be any Segal movie). He pulls his gun. The first thug says: "You can't shoot us all, officer". "You're righ", Segal says, then he shoots the first thug, "But I can get an A for effort". I seem to recall that that deterred the other thugs. I suppose that makes sense, if the rational thing is not to shoot the wolves and to just let them do their thing (because you'll eventually run out of bullets and they'll kill the sheep anyhow), you can convince the wolves that you're irrational by shooting one. In that scenario, the rational answer is to shoot a wolf.

"Your only hope is to convince them credibly that you aren't rational, which I think implies shooting one of the sheep now and then."

But then the wolves are likely to stick around until you leave and eat the dead sheep (because you're not likely to carry it off). That'll just keep them coming back for more, since they'll know that you'll shoot a sheep from time to time for them. If you really want to convince the wolves that you're irational, you'll shoot the biggest wolf, skin it, take off all your clothes, put on the dead wolf skin, start howling at the moon, dancing with the corpse of the now skinned dead wolf and chasing after the other wolves. After a display like that no wolf (or anyone else) will go anywhere near you.

In that case, the shepherd should shoot the first wolf that attacks, to make it clear early on that he is ready and willing to use his bullets. Wolves will die, and some sheep will die too, because there aren't enough bullets to kill all the wolves. But at least, less sheep die because some wolves are dead.

But what if using up the bullets actually angers some of the sheep?:/
Maybe because the bullets are made out their own wool ;0 The sheep might turn against their own shepherd, and both the sheep and wolves will feast on him instead.Wouldn't that make him sheepish (pun intended) about doing anything?

rogue: "Maybe because the bullets are made out their own wool ;0 The sheep might turn against their own shepherd, ...."

That is indeed part of the problem. And where my "bullet" metaphor breaks down. What the European shepherd is doing is to take away food from the stronger sheep (Germany and France), to feed to the weaker sheep. This not only annoys the stronger sheep, but makes the stronger sheep weaker. So ki goes up for Greece, but down for Germany.

But if the food comes from outside Europe (the IMF) it's different. And if the ECB can print bullets, it's different again.

"If you really want to convince the wolves that you're irational, you'll shoot the biggest wolf, skin it, take off all your clothes, put on the dead wolf skin, start howling at the moon, dancing with the corpse of the now skinned dead wolf and chasing after the other wolves. After a display like that no wolf (or anyone else) will go anywhere near you."


This reminds me of a story my mom told from her days as the nurse in the county jail.

An inmate came in and after some time there the psychiatrist was asked to come and evaluate him because his behavior was strange. He was urinating and defecating in his cell and would suck on his urine soaked clothes and rub the feces on himself.

The psychiatrist after a thorough evaluation determined he wasnt crazy he was just trying to keep the other inmates away from him..... it worked.

What is it with macroeconomics and bizarre analogies - Robinson Crusoe, coconuts, and now wolves? Would it ever be in their interest to appear in sheep's clothing?

Generally I enjoy the blog, though: keep it up, all of you!

Hi Linda! good to see you commenting here. Thanks.

Well, wolves are sometimes used as a metaphor for a speculative attack.

And then there was the classic: "Macroeconomic policy and the optimal destruction of vampires" by Dennis Snower.
http://www.google.ca/url?sa=t&source=web&ct=res&cd=4&ved=0CC8QFjAD&url=http%3A%2F%2Fwww.ihs.ac.at%2Fpublications%2Fihsfo%2Ffo162.pdf&rct=j&q=macroeconomic+policy+and+the+optimal+destruction+of+vampires&ei=M1HjS6S7C4WKlweh2-yXAg&usg=AFQjCNGq9H1O8HE_qSYT7ksUM8yN1hkDGg

Hi Linda! What Nick said.

And besides, macro is inherently boring. Can't blame a person for trying to suck people in with a provocative metaphor.

Ha! Was just searching the cable TV channels for the UK election coverage and coincidently came across a PBS show in progress with wolves eating sheep. A search of the TV guide revelaed it was a BBC production called Human Senses - Smell and Taste. It made me chuckle considering what I had written earlier here somewhat tongue-in-cheek. The bit about wolves/sheep appears at 17:00.

Human Senses - Smell and Taste

Too funny.

I agree that any nation that surrenders its printing capacity is a sheep, but who is the wolf?

I wonder if you were to assign a probability of default to Greece, what that would be, and then whether the current rates would be too different from your own rate?

It's difficult to understand a world in which, say, Paul Krugman is openly echoing conventional wisdom that Greece will default -- that default is *inevitable* -- and yet simultaneously there is outrage that credit markets are demanding a risk premium. That somehow it is the misperception of risk that is forcing the crisis, rather than actual risk -- i.e. Greece's legendary inability to collect taxes, crushing debt burdens, and lack of currency sovereignty.

Supposedly all the investors are supposed to sleepwalk into the default, and then be surprised when it happens. If they do not sleepwalk into the default, then they are wolves.

I say the wolf is Germany, as it managed to secure a nice set of captive export markets within the EMU, and will soon be able to buy olives for pennies in the ensuing deflation. In a flexible exchange rate system, the Germans would get only inflation for their export position, but now they get olives.

So how about a model in which there is one wolf -- Germany -- circled by a lot of sheep. The sheep are trapped in a fence, and may injure themselves if they try to leave. No sheep knows how painful it will be to escape the fence.

If they all coordinate and break through the fence, then they can escape from the wolf. But if the wolf only eats one or two sheep, it may be better for the majority of sheep to remain in the fence, and hide behind the weakest one.

The ECB is indeed the shepherd, trying to keep the sheep in the same fence as the wolf. The shepherd does this because he believes that "wolves are a veil". To him, in a competitive equilibrium, it should not matter if the wolf eats the sheep or not. The sheep is better off eaten, and in any case other sheep will appear ex-nihilo as supply creates its own demand. The most important thing for the shepherd is that everyone remain inside the fence, because the fence is a common market with no currency-conversion costs.

The optimal strategy is clear, all the sheep should leave the fence, as soon as possible, and the shepherd should shoot himself for his economic stupidity. Only a single bullet is needed. The wolf will be free to roam, in equilibrium expending as much energy in the form of domestic inflation as he obtains from eating the sheep, until he stops being a mercantilist wolf.

You can always find the optimum path without expending an infinite number of bullets, or shooting anyone other than economic decision makers.

RSJ: you make some good points there.

One complicating factor though, is that the probability of default is not independent of (actual and prospective bondholders') expected probability of default. If countries (or firms) issued only long-term debt, that they knew they could pay off, if they needed to, when it came due, this complicating factor (liquidity crises) would not exist. It is only because they issue short-term debt, hoping they can roll it over when the time comes (and they don't have a money of their own) that expectations of default affect the actual probability of default.

Put it another way: some of those sheep are going to die anyway.

Greece Spain Portugal Argentina Italy etc are the wolves. The sheep are regular people with pensions and retirement accounts. Please get the cognative biases of your story straight.

Hmmm ... Germany is the wolf, and the ECB is the shepherd, eh? Fox, hen house...

Valid point Jon, over the longer term.

Nick Rowe,

Interesting post. However, where is the restucturing in the story. Wounded only during an attack without deaths or wounded woolfs that stay away for another say.
However, you are giving me the opportunity to return to a past debate as a teaser of samle food for thought. Suppose the shepperd is in a state of uncertainty (pure) of what to do and is frozen and paralysed, meaning that he holds the bullets without firing (real effect). He is hesitant to do so. He prefers to hold because he does not know what the firing will do. He is not holding bullets because he is in between shots or opportunities to shoot. He is not holding firing power but only bullets (nominal value)as he is not anticipating to use them. If he only knew what to do! If an outside force gives him more bullets (CB excessive reserves) he will do only one thing hold then as bullets not as firing power.He will not even trade them with other shepherds since he is uncertain if anybody else is out there!

In my own work determining interest rates there is a probability of default that weighs the expected shortfall effect induced by austerity measures that reduce income and raise the deficit as anticipated my market participants. This probability of default is not only a function of time but of pure uncertainty but also of the animal spirits of speculators. There is also of course an inflation expectation term which austerity measures magnify as they contain VAT and other sales tax measures. The result from austerity measures is not only to raise spreads against tradional theory analysis (austerity will lower rates) but also it will invert the yield curve. Both, of course, are happening.

Panayotis: in your own work, do you find that austerity measures actually increase the probability of default? (Or the amount of the default?)

I mean over some plausible range. Because I would guess that it's highly non-linear, right?

The first outstanding feature of real wolves is the cooperative pack behaviour. The second is availability of alternative prey. If not sheep, rodents. In this model, sheep are the only prey in town. Extremely bad pun intended.

These wolves only care about self interest and cannot communicate but are able to coordinate on extraneous information, then attack regardless of the reasonable expectation that a number of individual wolves will die. A rational sheep herder would credibly precommit to emptying the gun.

I get the metaphor but suspect that a few details of the game need to be worked out.

Maybe some wolves are familiar with fire arms and others are not? I suggest this because some of the punditry in recent days makes me wonder if many North Americans understand Europe about as well as they understand some small, obscure country in Africa.

Nick Rowe,

You are asking an interesting question that I am still laboring with. My current answer is that austerity measures do increase the probability because participants in the secondary markets evaluate these measures and their explosive feedback loop. They also consider the horizon of the program and this is what causes the yield curve to invert because they assume it to be short. There is a constant, a mean and a curvature term, so its non linear in this sense. Iam still debating the specification. Its the behavioral aspects, however, that are more important. The expected shortfall is the difference between the present value of public debt and the present value of the fiscal balance projected for the period of the program which can reflect an estimated insolvency.

Uncertainty and and the volatile and unstable nature of speculation ( animal spirits) are also important. They can overwhelm the inflationary expectations that everybody talks about.These inflationary expectations are specified to include an addition to current rate of inflation which is an expectation change with a duration effect. The point that I am trying to make is that interest rates and spreads are most probably rising for the horizon of decision making and this can have secondary negative eefcts on the outcome of these austerity measures. There is also a risk free rate term involved. I will stop because I am taking space in your blog.

I agree that borrowing costs affect default costs, but in both directions. Where was the complaining that Greek debt costs were "too cheap" prior to this crisis? If you want to blame this crisis on an expectations failure, then I would agree, but it was a general failure by the European Governments, european public, financial markets, and even most observers. They were all too optimistic before and there is consensus that default is inevitable now. I think the shift in opinion happened at about the same time in all these sectors.

" If countries (or firms) issued only long-term debt, that they knew they could pay off, if they needed to, when it came due, this complicating factor (liquidity crises) would not exist. "

Macro vs. micro. Debts are never paid off in the macro -- they are always rolled over. Even in the micro ("firm") case, it is rare to reduce debt -- think about what that would mean!

Rather think of all capital as incurring a liability. The liability never goes away, as it represents the discounted future profits of the firm. A firm has zero liabilities only when it dies. From that point, you decide on your optimal capital structure -- do you want to 50% equity and 50% debt, or 80% equity and 20% debt, etc. The decision will be based on things like tax breaks, changing borrowing costs in one instrument versus the other, etc.

Then you keep rolling everything over, making a stream of dividend and interest payments that only stops when your ability to make money stops.

Only in the case of an individual household, say buying a mortgage, can you talk of the debt being "paid off". But even then, households in their productive years often elect to roll the debt over, going from house to house, or car to car. They eventually become debt free, at least for retirement, but in aggregate households do not become debt free.

The "representative" household is never debt free.

IIRC, the maturity of Greek Debt is not different from the maturity of Canadian or U.S. debt -- about 5-6 years, but a steady flow of it needs to be rolled over.

This Bloomberg article was far too appropriately title not to be posted here EU Preps Euro Fund to Fight ‘Wolfpack,’ Debt Crisis

From the contents of the article, it's unclear if the ECB is going to get into the fray.

Yep. The sunspot has appeared over the weaker sheep, so the wolves can use it to coordinate their actions, and form a pack.

So all the Finance ministers have been meeting, though not just the Euro ones. UK is in there too. But until the ECB joins in, and starts printing bullets, all this does is re-distribute "strength" around between the sheep. It does not strengthen the sheep as a whole. Like mountain climbers roping themselves together. Makes it harder for one to fall; but if one falls they all fall.

Well, it looks like the ECB is being cajoled into action. From Charlemagne's Notebook in The Economist
Firstly, the European Central Bank president, Jean-Claude Trichet, has flown off to meet the members of his governing body to discuss the ECB buying government debt on the secondary markets, in large amounts. Mr Trichet is reportedly furious that member countries of the eurozone have been putting it about that he is doing their bidding in this. The independence of the ECB is already looking threadbare as it is, and he feels it is astonishingly unhelpful for countries like France to limit his room for manoeuvre, by appearing to issue him with orders. That said, an awful lot of diplomats wandering round this EU complex seem remarkably confident that the ECB will do what the political leaders of the eurozone hope.

Poses some interesting questions about the independence of the ECB.

There's another line in the blog which fits this post
[The EU's] gamble, or hope if you want to be more optimistic, is that if big shots in the eurozone put unlimited guarantees on the table, the markets will understand that this is enough to shield Spain, for instance. And as one well-informed person I was talking to put it, the markets cannot get to Italy if they cannot get to Spain.

Going off of your last analogy Nick, the EU does stand a fair chance of making Spain invunerable to the wolves, and if they succeed, then sacrficing Greece and Portugal might be worth it. Of course, it might be better if the ECB just starts shooting.

What is the "secondary market for securities"? Because whatever it is, the ECB has just decided to intervene in it: http://www.bloomberg.com/apps/news?pid=20601087&sid=apehB.EV_9Kw&pos=4
Does it mean buying Greek bonds in the open market?

Yep. Looks like the ECB will be buying Greek bonds. But it's going to be "sterilised" purchases. http://www.bloomberg.com/apps/news?pid=20601087&sid=a0RoeU1FECGY

Why sterilised? With what will it sterilise (what will it sell)?

Good question about why they are bothering with sterilization, but I guess the German influence in the ECB is hard to shake. I am very curious about what they plan to sell to sterilize the purchases. I don't know what they can sell, but in effect, won't whatever they do sell be the equivalent of a Euroland-wide debt security? This will basically mean they will be taking Greek (or Portugese or whomever's) bonds off the market and replacing them with much higher quality debt. Which should be a net positive

There's no story yet, but this Bloomberg headline explains part of the reason why the purchases will be sterilized Weber Says ECB Bonds Purchases Pose Significant Stability Risks

The Germans (or at least Axel Weber, president of the German Bundesbank) are worried about price stability.

Here's Krugman arguing that the most recent EU rescue package is not enough to save Greece and may not be enough to save Spain Shock and Uh? Here's a relevant snippet:

When the first announcement came, my reaction was to say that the EU was making the classic mistake, treating a solvency problem as if it were a liquidity problem.

Announcement #2, from the ECB, changes things somewhat. It now seems that Trichet has been dragged kicking and screaming into becoming at least a semi-Bernanke, engaging in much more expansionary policies than before. (Yes, the ECB says that they’re only liquidity operations, and will be sterilized, yada yada — we can only hope that they don’t really mean it.)

A more expansionary monetary policy could make a real difference — especially if the ECB ends up accepting somewhat higher inflation.

With regards to the shepherd guarding the sheep, does it matter what form the expansionary monetary policy takes? Is the unsterilized purchase of sovereign bonds preferable to holding interest rates inordinately low while allowing inflation?

"The Germans (or at least Axel Weber, president of the German Bundesbank) are worried about price stability."

Another central banker who would intellectually benefit from being unemployed. I am starting a list.

The comments to this entry are closed.

Search this site

  • Google

    WWW
    worthwhile.typepad.com
Blog powered by Typepad