Financial Times has been publishing pieces around the theme of whether the markets’ view of sovereign risk of developed nations vis-à-vis emerging ones has changed.
If you’ve read my opinions as early as a year ago on this issue, specifically in relation to Turkish sovereign debt as well as its investment rating, you already know what I think. I simply cannot fathom why certain countries are entitled to the risk-free treatment they enjoy without any questioning or dissent. I was also troubled with assigning risk weights according to investment grades provided by credit rating agencies.
In the aftermath of Greece and Dubai, it turns out that banks, credit rating institutions and journalists have started entertaining the supposedly “radical” idea that AAA rated securities may not be risk-free after all.
To give you an example of my thought process, I had provided my reaction to a Moody's comment that “Turkey’s economic resilience is only medium, its debt affordability is low, and its susceptibility to political and financial event risk is assessed at high” in the following fashion:
Wow. Am I under some delusion or could this sentence be equally valid for many “developed” world economies as well? The sentence above basically says nothing and everything at the same time.”
At least my sanity remains intact. Furthermore, Tracy Alloway at Financial Times has recently quoted Fitch Solutions reiterating the markets’ increasing concern about developed economies based on future budget deficits, debt ratios, as well as decreasing tax revenues.
Dare I say that credit rating agencies have woken up from the cognitive dissonance they had been displaying regarding the major inconsistencies in their valuation and judgment of sovereign debt?
Of course, Fitch Solutions’ evaluation has no value to people who have already observed data relating to pertinent credit metrics of the countries in question. What is newsworthy is that sovereign CDS markets seem to be paying more attention to this discrepancy lately, and that credit rating agencies actually seem to acknowledge this situation.
Just three months before its collapse, high-level Lehman Brothers officials held a press conference in Turkey in which they reassured the entire Turkish investment community that they would remain in Turkey even if a disastrous political event hit the country.
I know it sounds funny now, but nobody thought this was remotely outrageous coming from a sub-prime laden entity. The fact that a titan like Lehman Brothers would not leave Turkey under such a circumstance was supposed to provide us all with a sense of comfort, and make no mistake, it did.
Guess who is still standing now? The markets indeed have an uncanny way of manifesting the unpredictable. Yet how it all seems to make perfect sense after the fact.
Sovereign Risk and CDS Spreads
How should we value sovereign risk and how should we approach the sovereign CDS market? A lot of late analysis has focused on debt to GDP metric, while recognizing that some “developed” nations look troublingly unsafe when evaluated with this ratio.
Ideally, you should have various metrics when measuring the financial "health" of these countries. Aside from debt to GDP ratio, significant metrics that come to my mind include external debt to GDP ratio, debt service ratio (as measured against net exports), government deficit ratio, current account balance to GDP, foreign reserve coverage, and various ratios that include real interest rates, real GDP growth, unemployment, exchange rates, and so on.
Because some are more important than others, it may be wise to assign different weights to each if you wish to come up with a scoring system. Unfortunately I don’t have access to data required to undertake such a project for many countries.
There are various studies in the academic literature that either predict default or propose valuation methodologies involving such metrics. Of course, these metrics are an indicator of ability to pay but they are not necessarily good predictors of the willingness to do so. Credit rating agencies claim that they consider all these metrics when coming up with their own investment grades.
Credit Rating Agencies: Do They Matter?
If objective quantitative methodologies were indeed being carried out consistently and fairly by credit rating agencies, assigned ratings for various countries would look different from what they do now. Therefore, it’s safe to conclude that a lot of subjectivity goes into their judgment of countries. For instance, they may decide to assign zero weight to a bad-looking metric (high debt to GDP ratios or enormous government deficits) for an AAA rated country. On the other hand, another country gets hammered for one single metric that looks somewhat unsafe.
So rating agencies have a way of magically shifting their criteria when dealing with different countries. If all else fails, they can always fall back on “intangibles”, in other words, those cookie-cutter statements and clichés that appeal to general investment prejudices when justifying credit decisions and ratings.
Let me try to elaborate: Whenever you read an analysis that appeals to the western ego about strong education systems, superior institutions, supreme rule of law; or, alternatively resorts to general third-world stereotyping about uneducated commons, poor infrastructure, terrorism, scare-mongering in the form of religious coups, unreligious coups, or military coups (all of which I’ve read in published “research” about the Turkish economy), you should know that you have fallen victim to such a bias (or hedonic editing) in investment evaluation and justification.
I fail to see how that is any different from judging different types of people via different standards in the same court of law.
Should credit rating agencies’ ratings be taken seriously in any way? If you read the stuff I wrote in the past regarding the inequalities manifested by these agencies, you may think that my answer is a firm “no."
Unfortunately, it does not work that way. Indeed, we’ve seen during the credit crisis that many institutions depended upon their AAA rating to gain access to cheap financing. In fact, it’s fair to say that certain institutions made (and continue to make) money based on a business model enabled by credit rating agencies’ (and hence the system’s) stamp of approval.
The Sovereign CDS Market
The sovereign CDS market is a definite improvement over credit rating agency grades.
For instance, Financial Times states that UK’s CDS implied trading band is around AA, and even “inching towards an implied rating of AA-“. This could give you an example of the gap between the AAA credit rating of UK and its perceived rating in the marketplace.
So should investors start soaking up those CDSs in the expectation of the ultimate Black Swan event?
Better yet, do we believe that CDS sellers will be able to make good on their obligations if and when such a blow-up occurs? Do CDS buyers really expect anyone to shoulder these insurance burdens the same way the U.S. taxpayer did when AIG failed to make good on the insurance contracts it sold?
Gillian Tett has written an extensive piece
on the issue of banks’ perception of western sovereign risks, among other things. The article confirms that sovereign CDS market is a good indicator of how institutions view tail risks, and promises to be a busy place in the coming year. (It may require registration.)
Profit Opportunities in the Sovereign CDS Market – Yet Another Proposition
Trading desks have been trying to “arb” the CDS market, but this type of trading strategy is outside the realm of possibility for the retail investor. Yet because the matter has hit the blogosphere, I feel the urge to comment.
Tyler Durden has proposed a one-variable (debt to GDP ratio) system somewhat tongue in cheek, while Shishir Nigam at Seeking Alpha elaborated on the idea further.
Tyler Durden has apparently allowed the software to come up with its own function to increase data’s goodness of fit. So if you look carefully at his graph, you’ll notice that what came out of the black box is not a linear or log function but more like a cubic function indeed.
I hereby take Tyler Durden’s (and Shishir Nigam’s) pieces as a formal challenge to propose an alternative model:
If it’s all about increasing R-squared value, I got a couple of ideas for those who have access to data. Increase the number of variables to include the following factors that are known to be of importance when evaluating sovereign risk:
Apart from debt to GDP ratio, you may want to consider external debt to GDP ratio, government deficit ratio, average maturity of debt, current account balance to GDP, debt service ratio, net FDI to GDP, and coverage ratio involving foreign reserves. As far as economic and market indicators go, you may add real GDP growth, real interest rate, real effective exchange rate, unemployment rate and currency volatility to the mix.
If that’s not good enough, include the investment rating of the sovereign nation provided by credit rating agencies as a determinant (but not an indicator) of market’s perception of sovereign risk.
Feel free to include the kitchen sink and see if that helps in increasing the goodness of fit. If you get meaningful coefficients and a higher R-squared value, then we are talking real algo trading strategy.
Is the CDS trading lower or higher than implied by the model? That does not necessarily mean under or overvaluation, especially if you are aware of explanatory reasons outside of model’s variables that may account for the differential. However, if your research shows that one of the explanatory variables is subject to a significant change in the future, then the CDS price will adjust accordingly as soon as that change is publicly known and discounted by the markets.
Note that this is a cross-sectional regression idea and does not include a time component. Therefore it does not provide an analysis of how sovereign CDS prices change during times of financial distress.
Disclaimer: The above is presented as a modeling idea and does not constitute investment or trading advice. Past goodness of fit does not guarantee future success.
Disclosure: Long Turkish sovereign debt