Saturday, May 14, 2011


. Saturday, May 14, 2011

(click here for animation)

I agree with Krugman again:

How bad will it be if we don’t manage to raise the debt ceiling? ...

First, US government debt plays a special role in the financial system: T-bills are the universal safe asset, the ultimate collateral. That’s why, during moments of financial stress, the interest rate on T-bills has actually gone negative. Make that safe asset suddenly unsafe, and it might cause vast disruption.

Krugman should make the point much stronger. It will cause vast disruption. And not just for the U.S. financial system, although that would be bad enough. The effect on the global financial system would be far worse. There's a reason why, following the biggest financial crisis since the Great Depression, capital actually flowed into the U.S. banking system, which was the epicenter of the storm. It's because the U.S. was at the epicenter of the storm. Stabilizing the U.S. was necessary to stabilize the global financial and economic system. If you click on the link above and watch the animation, you'll notice that the U.S. becomes more central to the global banking system over the period 1999-2010. And you'll even notice that, while the network wobbles after the shock in late-2008, it reinforces itself pretty quickly. Why? Because if the U.S. went under, then all those thick black lines would disappear, and everyone else would go under too.

What does this mean? It means that the rest of the world is heavily invested in the U.S. Which means that if the U.S. has a downturn, the global economic system has a downtown. If the U.S. refuses to raise the debt ceiling, there are two possible outcomes, which are not mutually exclusive: quick, harsh austerity that will crush what little recovery we've had; some form of default on debt. In truth, one implies the other. Both of them involve a massive collapse in global aggregate demand, as well as the mother of all bank runs. This is important because U.S. Treasury bills are considered "riskless" by financial institutions around the world. If the "riskless" assets become risky overnight, the effects on bank balance sheets will be catastrophic. Global regulatory structures will effectively cease to exist (since enforcing capital requirements would make every bank insolvent), and runs on financial institutions will occur almost immediately. The U.S. won't be able to intervene to stabilize the banking sector since its debt-spending capacity has been eliminated, so the banks all melt down. If that happens, not only would the U.S. economy crash the world economy would crash as well. There are few countries immune from an American virus. This would be a pandemic.

Moreover, it's just not necessary. The U.S. can currently borrow for five years at negative real interest rates (i.e. adjusted for future inflation). That means that other people will currently pay us to take their money. Let me repeat: Other people will pay us to take their money for the next five years. Instead, we're considering blowing up the global economy. Bad deal.

What happens when global economies blow up? Well, all of the examples we have are pretty bad. They tend to lead to long depressions, world wars, nuclear bombs being dropped, that sort of thing. I'm trying not to be too hyperbolic, but the major thing separating this crisis from the 1930s is a series of global institutions that are buttressed by economic stability in the U.S., integrated Europe, and Asia. If we blow that apart, then things can get ugly very quickly. Quite frankly, I'd rather not conduct a natural experiment to see how well various IR theories hold up.

For these reasons, I don't think it will happen. I don't want to put anything past the current Congress, which is as petty and short-sighted as every other Congress, but the stakes here are just too high. Then again, we've been in similar situations before, and they didn't always end well. If Congress is engaged in a game of chicken, then I hope they're ready to jump. This isn't worth toeing the line.


Emmanuel said...

Having no Krugman fixation of my own, there is another beginner's error here:

Treasury yields are not a good indicator of foreign demand because the game is fixed. At regular intervals, the Fed has a programme of buying Treasuries. There's even a schedule for it last I checked.

Unless you can differentiate real demand from artificial demand in these yields, I'd not toot this "fact" too much (like a certain NYT columnist, for that matter).

It's like buying lots of drinks from your kid's lemonade stand and calling him a great salesman.

But on the larger point, don't fear--American debt addicts will never refuse a bigger fix. Nobody doubts this.

Dan Nexon said...

E: The yields don't appear to correlate with the size of FOMC purchases. The banking network is also pretty telling.

Kindred: I think the visualization would be better with centrality placement and arrows for net flows. Or something.

Kindred Winecoff said...

Dan -

We played around with that some. A few problems:

1. We don't want net flows (it's actually stocks at the end of each quarter), we want gross and we want it to be bidirectional (i.e. to show US holdings in UK, and UK holding in US). There's a lot more information included that way. If we just did net, then we'd (potentially) lose a lot of information about magnitudes. There are infinitely many ways to get the same net number, and we're really interested in exactly what those ties look like.

2. All the plots that placed nodes according to centrality were very muddy. It was nodes stacked on top of nodes. We could have made node size fixed for ease of visualization but again, we'd be throwing out information.

3. This is because many traditional centrality measures aren't all that interesting in this network, since most nodes are tied to most other nodes at least a little bit. What's interesting is the patterns in the strength of ties rather than centrality per se. That said, the size of nodes in our depiction is in-degree, and from that you can easily see that the US and UK are most central (have the highest in-degree), while still seeing where the ties are coming from pretty clearly.

In the paper this comes from, we use a data set with more countries (but only one time period) and use arbitrary cut-offs to establish links. We lose a decent amount of info that way, but it allows us to plot with centrality placement. It looks like what you'd expect: US and UK at the center, Germany/Luxembourg important regional sub-hubs in Europe, and everyone else basically in the periphery. (Surprisingly, Japan isn't very central in any of the data we have.)

Emmanuel, I didn't say anything about foreign demand. I said the US can borrow for 5 years at negative interest rates. "Other people" was meant to be "not the government".

On the schedule you posted, there are two TIPS purchases that total $2-4bn. That's not enough to move anything around that much. Moreover, it's fairly well agreed that TIPS yields provide the best (or one of best) pictures of the real cost of sovereign borrowing, as it internalizes inflation expectations as well as credit risk (which is basically zero). I'm not sure why you think otherwise.

Dan Nexon said...

Kindred: thanks for the explanation. Is the paper available for download?

Emmanuel said...

Everyone--I'm more interested in stocks than flows. The Fed schedule merely hints that government purchases are nothing to sneeze at. F'rinstance, PIMCO calculates them as representing over two-thirds of Treasury buying since QE2 started.

We won't have to wait for long to see how much Fed buying has depressed mid-to-long duration Treasury yields as QE2 wraps up at the end of June.

Also: let's not be Amerocentric here. The relevant yield comparison here for foreign buyers isn't the real yield but, roughly speaking, the difference between what they can get at home and what they can get abroad.

Kindred Winecoff said...

Dan, Final revisions should be done early this week. I'll let you know when it's done. I think there's a decent chance you'll like it, as it gets to some of the "IPE is messed up" discussion that you and Henry were driving at, although it focus more on ways to conceptualize the international system in useful ways than more of the micro-level elite politics that Henry was driving at.

Emmanuel, points taken. But expectations are important too, and expectations are already priced in to some extent at least.

Dan Nexon said...

Let me guess - international structure as networks (cf. Hafner-Burton and Montgomery)? I advocate that in my last and only book--and have been saying we should be doing this for ten years--so I bet I'd be happy with it....

Kindred Winecoff said...

Yeah, the point is to complicate H-B/Montgomery by focusing on "complex" networks, ie those where the intensity of ties, not just their presence or absence, is considered an important variable. Particularly for issues in IPE where most states have some tie to most other states, but the size/strength of those ties varies fairly dramatically. As I said, I suspected you might be intrigued by the general drift, at least.

I'm sorry to say I haven't read your book yet. It's in my pile, and I'm hoping to get to it in a month or two. I'm interested in it on a number of levels, not least because I was raised in a family that took quite a lot of pride in the Reformation tradition, certain strands of it at least, and the effect it had in transforming European society. I'm interested to see how well the narratives I was raised on hold up. I'm also interested, more and more all the time, in how periods of instability shape future orders. So I'm expecting to enjoy it a lot.

Thomas Oatley said...

Complex networks are not about the intensity of ties. Complex networks are “networks whose structure is irregular, complex and dynamically evolving in time.” 

The approach we try to develop in the paper thus shares the motivation you elaborate on page 26 of your book--to model the structure of relationships as they exist and explain how the evolve over time. To conceptualize these relationships as economic, social, informational, and political.

We'll send you the paper this week--we are writing the conclusion now--and would love to have comments.

Dan Nexon said...

Cool. As you know, I don't have the technical chops to say much of interest about that side of it, so I probably won't be much good to you. But I'd love to see it. We'll correspond, but it sounds like something I'd like to cite.

Thomas Oatley said...

Any feedback helps. It is also clear that we need to cite your book. We'll be in touch. Thanks, Thomas

Dan Nexon said...

Wasn't fishing for a cite. While it's nice, I'm sure it isn't required.

Thomas Oatley said...

Sorry--I didn't mean to imply that. I hadn't read the book until about 25 minutes ago. Then I read what I could via google books after I read your description of your argument. It is pretty evident that we are making very similar arguments. So, we should cite your book.





Add to Technorati Favorites