Felix Salmon

Chart of the day, sovereign precariousness edition

Felix Salmon
Nov 2, 2013 06:42 UTC


Bond spreads, along with their close cousin credit default swaps, are a beautifully linear measure of sovereign default risk. They go up in a straight and steady line: the higher the number, the riskier the country is perceived to be. And so they’re normally the first and last place that people look when they’re interested in the chances of any given country defaulting.

But of course the world isn’t quite as simple as that, and — as we have learned the hard way — it’s the unexpected defaults which are the most damaging. So I asked Ben Walsh to put together this chart for me, showing a different measure of risk. It’s not a better measure, by any means. Still, it’s an interesting measure all the same, and it’s almost entirely uncorrelated with credit spreads.

The red bars, in this chart, are the credit spreads we’re all familiar with. Countries like France, the UK, Canada, and Germany are considered effectively risk-free, while Spain and South Africa and Italy are riskier, India’s worse, Greece is very bad, and Argentina is truly dreadful.

And then there’s the blue bars, which are related to the concept of a “sudden stop”: when all the windows close, and a country simply can’t borrow money any more. Such things can arrive with astonishing speed: the markets, in general, will only lend if they think that everybody else is willing to lend. So it’s easy for them to shut down very quickly indeed.

The blue bars are also related to the fact that sovereign creditworthiness, at least as the markets understand it these days, is not really a measure of a country’s ability to pay off its debts in a timely fashion. Japan is a highly creditworthy country, but it also has public debt of a quadrillion yen. Which, it is safe to say, it will not be able to pay off at any point in the foreseeable future.

In order to be considered creditworthy, even to the point of being considered risk-free, all you need to be able to do is roll over your debts, as they come due. There’s something scarily self-fulfilling about this state of affairs: if it’s all one big confidence game, then the minute that a country for whatever reason becomes unable to continue to borrow money, then catastrophe is imminent. And, as we saw during the debt-ceiling debacle, there are a lot of reasons why a country might become unable to continue to borrow money.

In a forthcoming paper, Lee Buchheit, the Cleary Gottlieb partner and godfather of sovereign debt restructuring, writes this:

No purchaser of a sovereign debt instrument today does so in the hope and expectation that when the debt matures the borrower will have the money to repay it. The purchaser does so in the hope and expectation that when the instrument matures the borrower will be able to borrow the money from somebody else in order to repay it. This is a crucial distinction. If by sovereign creditworthiness we mean that a sovereign is expected to be able to generate enough revenue from taxes or other sources to repay its debts as they fall due, then most countries are utterly insolvent…

We have allowed a financial system to develop for sovereigns that assumes a more or less perpetual state of benignity — in the debtor country, in the region, in the global economic and political environment and even in the natural world. A disturbance in any of these areas, if it frightens investors sufficiently, risks interrupting the expectation of refinancing that defines sovereign creditworthiness. Many sovereign borrowers would last only a few months, some only a few weeks, if shut off from the ability to refinance their maturing debts. Denied continued market access, they would burn through their reserves with frightening speed.

This is the idea behind the “hang time” measure in the chart above. What we did was to take a country’s primary deficit — the amount it needs to borrow every year to finance its operations — and add on its total annual debt service. We then took that number and divided it into the country’s total foreign reserves, to get an idea for the length of time that sovereign reserves would be able to fund not only operations, but also all of the country’s debt service requirements.

The results are quite startling. Some countries do extremely well on this measure: Russia, for instance has a primary surplus which is higher than its annual debt service, so it could quite happily continue to service its debt in perpetuity, even in the event of a sudden stop. And Argentina looks very healthy on this measure as well: it has a hang time of 280 weeks, or well over five years. Germany and Brazil would both last about three years; even Portugal would last more than a year. But the UK would run out of money in six months, the US would go bust in 17 weeks, and France has just nine weeks’ worth of reserves. Japan would have only about 14 days.

Of course, countries like Japan and the UK borrow overwhelmingly in their own currencies, and can always, if push comes to shove, print more of it. The US is in the same boat — and, to boot, has the exorbitant privilege of printing the world’s reserve currency. Where fiscal solutions fail, monetary solutions are likely to look attractive. But the chart does show how incredibly fiscally precarious many of the world’s safest credits really are.


This is not really the right analysis. A primary deficit can usually mostly be funded locally since the bulk of the expenditures are denominated in local currency. Reserves need to be sufficient to fulfill the demand of foreign exchange (imports, external debt). Fx reserves do not represent the money that the government can go to to fund its local deficits.

Posted by plasmaj | Report as abusive

There’s no global wine shortage

Felix Salmon
Nov 1, 2013 07:04 UTC

Have you heard about the global wine shortage? Of course you have: it’s been covered in pretty much every media outlet imaginable, but Roberto Ferdman’s piece for Quartz (“A global wine shortage could soon be upon us”) was one of the first, and also one of the most detailed. Still, it was the classic single-source article: it basically took one Morgan Stanley report, reproduced a bunch of the key charts, and added a clickbaity headline.

The charts, on their face, tell a pretty clear story, especially this one:

But if you look closely at the Morgan Stanley report, it starts to look less like a dispassionate analysis of supply and demand dynamics in the wine world, and more like an aggressively-argued attempt to put forward one particular investment thesis as strongly as possible. What’s more, the investment thesis is not, particularly, based on the existence of any present or future wine shortage; it’s simply trying to present the idea that demand for Australian wine exports is likely to rise, and to justify the fact that  a company called Treasury Wine Estates is the bank’s “top Australian consumer pick”. (The report was written by Morgan Stanley Australia.)

For instance, the scary chart above is actually this rather less scary chart, tweaked a little:

To create the first chart, Morgan Stanley just took the second chart, added 300 million cases to the red line, and then — this is pretty cunning — simply deleted 2013 altogether, so that the uptick at the end disappears. (The 300 million number is Morgan Stanley’s estimate of the annual demand for “non-wine uses” of wine.)

Although the first chart is scarier than the second chart, even the second chart does a little bait-and-switch, which you can only find by looking at the sourcing note at the bottom of the page. The numbers for the charts come from OIV, the Organisation Internationale de la Vigne et du Vin, including the estimate for 2012 production and consumption. But the 2013 estimate, showing a modest increase in production, is not the OIV estimate; it’s the Morgan Stanley estimate. And what Morgan Stanley doesn’t tell you is that the OIV estimate for 2013 is much higher. Here are the OIV charts:

These charts are less polished, but are actually much more useful. (They also have different units from the Morgan Stanley charts: they’re measured in million hectoliters, which is 100 million liters, while Morgan Stanley uses million unit cases, which is 9 million liters. So when Morgan Stanley says that 300 million unit cases are used for non-wine consumption, that works out at about 33 million hectoliters.)

For one thing, the OIV charts draw sensible straight lines between points, instead of turning them into elegant curves which make the trend seem continuous. The trend is not continuous: these are annual figures per vintage, and each vintage is a unique, separate event. What’s more, while the amount of wine that will be drunk and produced in 2013 is not yet entirely clear. So OIV gives a range of possibilities, while being reasonably certain that wine production is going to increase substantially this year, by somewhere between 7.1% and 10.5%. Morgan Stanley, by contrast, gives no rationale at all for the fact that it has a forecast which is much lower than OIV’s; indeed, nowhere in the Morgan Stanley report is its 2013 forecast ever even quantified.

Add it all up, and the OIV actually concludes, quite explicitly, that the production-consumption difference for wine will “be higher than the estimated industrial needs” in 2013, for the first time since 2007. In other words, far from entering a period of global wine shortage, it looks like the 2008-2012 period of shortage is actually ending.

This global wine shortage, then, just simply isn’t real. Don’t take my word for it: ask the wine trade. Stacy Finz of the San Francisco Chronicle asked a bunch of industry types about the Morgan Stanley report, and none of them took it seriously; Victoria Moore of the Telegraph conducted a similar operation in Europe, and came to much the same conclusion.

My wine-making contacts raised more than an eyebrow at the ready, steady, panic news.

“Tell them to come to the Languedoc if they are worried,” said one. “I think I can help them out.”

Another noted that it is still possible to buy hectares of good vineyard in parts of France and Spain for less than the cost of planting one. In other words, the price of some wine is still lower than its true cost of production, an indication that the balance of supply and demand is still favouring the demanders, not the suppliers.

The Morgan Stanley report paints a picture of a long-term secular downward trend in area under vine, which is running straight into a long-term secular upward trend in global demand for wine. But reality is more complicated than that: thanks to a combination of technology and global warming, an acre of vines can reliably produce more wine, and better wine, than it ever did in the 1970s. And of course if demand for wine really does start consistently exceeding supply, then there’s no reason why area under vine can’t stop going down and start going up.

But never mind all that: the Morgan Stanley report has numbers and charts, and journalists are very bad at being skeptical when faced with such things. Even Finz’s Chronicle article, which sensibly poured cold water on the report, ends with a “Wine by the numbers” box which simply reproduces all of Morgan Stanley’s flawed figures. And besides, the debunkings are never going to go viral in the way that the original “wine shortage!” articles did.

As analysts have known since long before Henry Blodget was covering Amazon, the way to make a splash is to come out with a bold, headline-worthy thesis. Morgan Stanley did exactly that with this report, and I’m sure succeeded beyond their wildest dreams. I’m sure they’ve been celebrating their PR coup all week — probably with sparkling wine of some description.


A splendid insight-Interestingly the link to Morgan- Stanley does not go through – Felix can you provide the reference ?- I am unable to get to the report and look at it. The fact that Frederico Castellucci affirmed your post must make you feel good!

Posted by Oddsonwine | Report as abusive